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.
30 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
32 PP/02 Make -n do something, by making it not do something.
33 When combined with -bP, the name of an option is not output.
35 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
38 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
39 $sender_host_name and config options to manage this, and basic check
42 PP/05 DSCP support for outbound connections and control modifier for inbound.
44 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
45 (Only plugin which currently uses this is kerberos4, which nobody should
46 be using, but we should make it available and other future plugins might
47 conceivably use it, even though it would break NAT; stuff *should* be
48 using channel bindings instead).
50 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
51 name; added for Sendmail compatibility; requires admin caller.
52 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
53 just ignore it); requires trusted caller.
54 Also parse but ignore: -Ac -Am -X<logfile>
57 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
59 TL/02 Add +smtp_confirmation as a default logging option.
61 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
62 Patch by Magnus Holmgren from 2007-02-20.
64 TL/04 Bugzilla 1281 - Spec typo.
65 Bugzilla 1283 - Spec typo.
66 Bugzilla 1290 - Spec grammar fixes.
68 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
70 TL/06 Add Experimental DMARC support using libopendmarc libraries.
72 TL/07 Fix an out of order global option causing a segfault. Reported to dev
73 mailing list by by Dmitry Isaikin.
75 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
77 JH/02 Support "G" suffix to numbers in ${if comparisons.
79 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
81 NM/01 Bugzilla 1197 - Spec typo
82 Bugzilla 1196 - Spec examples corrections
84 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
86 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
87 gnutls_enable_pkcs11, but renamed to more accurately indicate its
90 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
91 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
93 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
94 "acl {{name}{arg}...}", and optional args on acl condition
97 JH/05 Permit multiple router/transport headers_add/remove lines.
99 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
101 JH/07 Avoid using a waiting database for a single-message-only transport.
102 Performance patch from Paul Fisher. Bugzilla 1262.
104 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
107 JH/09 Add $headers_added variable, with content from use of ACL modifier
108 add_header (but not yet added to the message). Bugzilla 199.
110 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
111 Pulled from Bugzilla 817 by Wolfgang Breyha.
113 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
115 (nb: this is the same fix as in Exim 4.80.1)
117 JH/11 Add A= logging on delivery lines, and a client_set_id option on
120 JH/12 Add optional authenticated_sender logging to A= and a log_selector
123 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
125 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
126 advertise SMTP AUTH mechanism to us, instead of a generic
127 protocol violation error. Also, make Exim more robust to bad
128 data from the Dovecot auth socket.
130 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
132 When a queue runner is handling a message, Exim first routes the
133 recipient addresses, during which it prunes them based on the retry
134 hints database. After that it attempts to deliver the message to
135 any remaining recipients. It then updates the hints database using
138 So if a recipient address works intermittently, it can get repeatedly
139 deferred at routing time. The retry hints record remains fresh so the
140 address never reaches the final cutoff time.
142 This is a fairly common occurrence when a user is bumping up against
143 their storage quota. Exim had some logic in its local delivery code
144 to deal with this. However it did not apply to per-recipient defers
145 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
147 This change adds a proper retry rule check during routing so that the
148 final cutoff time is checked against the message's age. We only do
149 this check if there is an address retry record and there is not a
150 domain retry record; this implies that previous attempts to handle
151 the address had the retry_use_local_parts option turned on. We use
152 this as an approximation for the destination being like a local
153 delivery, as in LMTP.
155 I suspect this new check makes the old local delivery cutoff check
156 redundant, but I have not verified this so I left the code in place.
158 TF/02 Correct gecos expansion when From: is a prefix of the username.
160 Test 0254 submits a message to Exim with the header
164 When I ran the test suite under the user fanf2, Exim expanded
165 the header to contain my full name, whereas it should have added
166 a Resent-Sender: header. It erroneously treats any prefix of the
167 username as equal to the username.
169 This change corrects that bug.
171 GF/01 DCC debug and logging tidyup
172 Error conditions log to paniclog rather than rejectlog.
173 Debug lines prefixed by "DCC: " to remove any ambiguity.
175 TF/03 Avoid unnecessary rebuilds of lookup-related code.
177 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
178 Bug spotted by Jeremy Harris; was flawed since initial commit.
179 Would have resulted in OCSP responses post-SNI triggering an Exim
180 NULL dereference and crash.
182 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
184 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
185 Bug detection, analysis and fix by Samuel Thibault.
186 Bugzilla 1331, Debian bug #698092.
188 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
190 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
191 Server implementation by Todd Lyons, client by JH.
192 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
193 config variable "prdr_enable" controls whether the server
194 advertises the facility. If the client requests PRDR a new
195 acl_data_smtp_prdr ACL is called once for each recipient, after
196 the body content is received and before the acl_smtp_data ACL.
197 The client is controlled by bolth of: a hosts_try_prdr option
198 on the smtp transport, and the server advertisement.
199 Default client logging of deliveries and rejections involving
200 PRDR are flagged with the string "PRDR".
202 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
203 fclose(). Diagnosis by Todd Lyons.
205 PP/17 Update configure.default to handle IPv6 localhost better.
206 Patch by Alain Williams (plus minor tweaks).
209 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
210 This is now consistent with GnuTLS, and is now documented: the
211 previous undocumented portable approach to treating the option as
212 unset was to force an expansion failure. That still works, and
213 an empty string is now equivalent.
215 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
216 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
217 not performing validation itself.
219 PP/20 Added force_command boolean option to pipe transport.
220 Patch from Nick Koston, of cPanel Inc.
222 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
225 TF/04 Added udpsend ACL modifer and hexquote expansion operator
227 PP/21 Fix eximon continuous updating with timestamped log-files.
228 Broken in a format-string cleanup in 4.80, missed when I repaired the
229 other false fix of the same issue.
230 Report and fix from Heiko Schlichting.
233 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
234 Report from Prashanth Katuri.
236 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
237 It's SecureTransport, so affects any MacOS clients which use the
238 system-integrated TLS libraries, including email clients.
240 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
241 using a MIME ACL for non-SMTP local injection.
242 Report and assistance in diagnosis by Warren Baker.
244 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
246 JH/16 Fix comparisons for 64b. Bugzilla 1385.
248 TL/09 Add expansion variable $authenticated_fail_id to keep track of
249 last id that failed so it may be referenced in subsequent ACL's.
251 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
254 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
255 ldap library initialization, allowing self-signed CA's to be
256 used. Also properly sets require_cert option later in code by
257 using NULL (global ldap config) instead of ldap handle (per
258 session). Bug diagnosis and testing by alxgomz.
260 TL/12 Enhanced documentation in the ratelimit.pl script provided in
261 the src/util/ subdirectory.
263 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
264 renamed to Transport Post Delivery Action by Jeremy Harris, as
267 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
268 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
269 redis_servers = needs to be configured which will be used by the redis
270 lookup. Patch from Warren Baker, of The Packet Hub.
272 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
274 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
275 hostname or reverse DNS when processing a host list. Used suggestions
276 from multiple comments on this bug.
278 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
280 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
281 Missed a few lines, added it to make the runtest require no keyboard
284 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
285 contains upper case chars. Make router use caseful_local_part.
287 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
288 support when GnuTLS has been built with p11-kit.
294 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
296 This, or similar/improved, will also be change PP/11 of 4.82.
302 PP/01 Handle short writes when writing local log-files.
303 In practice, only affects FreeBSD (8 onwards).
304 Bugzilla 1053, with thanks to Dmitry Isaikin.
306 NM/01 Bugzilla 949 - Documentation tweak
308 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
311 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
313 PP/02 Implemented gsasl authenticator.
315 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
317 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
318 `pkg-config foo` for cflags/libs.
320 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
321 with rest of GSASL and with heimdal_gssapi.
323 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
324 `pkg-config foo` for cflags/libs for the TLS implementation.
326 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
327 properties get this fed in as external SSF. A number of robustness
328 and debugging improvements to the cyrus_sasl authenticator.
330 PP/08 cyrus_sasl server now expands the server_realm option.
332 PP/09 Bugzilla 1214 - Log authentication information in reject log.
333 Patch by Jeremy Harris.
335 PP/10 Added dbmjz lookup type.
337 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
339 PP/12 MAIL args handles TAB as well as SP, for better interop with
340 non-compliant senders.
341 Analysis and variant patch by Todd Lyons.
343 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
344 Bug report from Lars Müller <lars@samba.org> (via SUSE),
345 Patch from Dirk Mueller <dmueller@suse.com>
347 PP/13 tls_peerdn now print-escaped for spool files.
348 Observed some $tls_peerdn in wild which contained \n, which resulted
349 in spool file corruption.
351 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
352 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
353 or write after TLS renegotiation, which otherwise led to messages
356 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
357 as a tracking header (ie: a signed header comes before the signature).
358 Patch from Wolfgang Breyha.
360 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
361 comma-sep list; embedded commas doubled.
363 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
365 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
367 Report and patch from Dmitry Banschikov.
369 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
370 Removed SSL_clear() after SSL_new() which led to protocol negotiation
371 failures. We appear to now support TLS1.1+ with Exim.
373 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
374 lets Exim select keys and certificates based upon TLS SNI from client.
375 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
376 before an outbound SMTP session. New log_selector, +tls_sni.
378 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
379 NULL dereference. Report and patch from Alun Jones.
381 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
382 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
383 Not seeing resolver debug output on NetBSD, but suspect this is a
384 resolver implementation change.
386 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
387 Left warnings. Added "eximon gdb" invocation mode.
389 PP/21 Defaulting "accept_8bitmime" to true, not false.
391 PP/22 Added -bw for inetd wait mode support.
393 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
394 locate the relevant includes and libraries. Made this the default.
396 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
397 Bugzilla 1246, report and most of solution from Tomasz Kusy.
399 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
400 This may cause build issues on older platforms.
402 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
403 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
404 gnutls_require_mac & gnutls_require_protocols (no longer supported).
405 Added SNI support via GnuTLS too.
406 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
408 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
410 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
411 Applied second patch from Janne, implementing suggestion to default
412 multiple-strings-in-record handling to match SPF spec.
414 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
416 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
417 read-only, out of scope).
418 Patch from Wolfgang Breyha, report from Stuart Northfield.
420 PP/29 Fix three issues highlighted by clang analyser static analysis.
421 Only crash-plausible issue would require the Cambridge-specific
422 iplookup router and a misconfiguration.
423 Report from Marcin Mirosław.
425 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
427 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
428 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
429 As part of this, removing so much warning spew let me fix some minor
430 real issues in debug logging.
432 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
433 assignment on my part. Fixed.
435 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
436 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
437 Janne Snabb (who went above and beyond: thank you).
439 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
440 string otherwise requires a connection and a bunch more work and it's
441 relatively easy to get wrong. Should also expose TLS library linkage
444 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
445 64-bit ${eval} (JH/03).
447 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
448 GNU libc to support some of the 64-bit stuff, should not lead to
449 conflicts. Defined before os.h is pulled in, so if a given platform
450 needs to override this, it can.
452 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
453 protection layer was required, which is not implemented.
454 Bugzilla 1254, patch from Wolfgang Breyha.
456 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
457 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
458 tls_dhparam take prime identifiers. Also unbreak combination of
459 OpenSSL+DH_params+TLSSNI.
461 PP/39 Disable SSLv2 by default in OpenSSL support.
467 PP/01 Solaris build fix for Oracle's LDAP libraries.
468 Bugzilla 1109, patch from Stephen Usher.
470 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
472 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
475 TF/02 Fix a couple more cases where we did not log the error message
476 when unlink() failed. See also change 4.74-TF/03.
478 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
479 lock up or crash if it happened to be inside a call to libc when it
480 got a SIGUSR1 from exiwhat.
482 The SIGUSR1 handler appends the current process status to the process
483 log which is later printed by exiwhat. It used to use the general
484 purpose logging code to do this, but several functions it calls are
485 not safe for signals.
487 The new output code in the SIGUSR1 handler is specific to the process
488 log, and simple enough that it's easy to inspect for signal safety.
489 Removing some special cases also simplifies the general logging code.
490 Removing the spurious timestamps from the process log simplifies
493 TF/04 Improved ratelimit ACL condition.
495 The /noupdate option has been deprecated in favour of /readonly which
496 has clearer semantics. The /leaky, /strict, and /readonly update modes
497 are mutually exclusive. The update mode is no longer included in the
498 database key; it just determines when the database is updated. (This
499 means that when you upgrde Exim will forget old rate measurements.)
501 Exim now checks that the per_* options are used with an update mode that
502 makes sense for the current ACL. For example, when Exim is processing a
503 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
504 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
505 must specify per_mail/readonly. If you omit the update mode it defaults to
506 /leaky where that makes sense (as before) or /readonly where required.
508 The /noupdate option is now undocumented but still supported for
509 backwards compatibility. It is equivalent to /readonly except that in
510 ACLs where /readonly is required you may specify /leaky/noupdate or
511 /strict/noupdate which are treated the same as /readonly.
513 A useful new feature is the /count= option. This is a generalization
514 of the per_byte option, so that you can measure the throughput of other
515 aggregate values. For example, the per_byte option is now equivalent
516 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
518 The per_rcpt option has been generalized using the /count= mechanism
519 (though it's more complicated than the per_byte equivalence). When it is
520 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
521 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
522 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
523 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
524 Note that using per_rcpt with a non-readonly update mode in more than
525 one ACL will cause the recipients to be double-counted. (The per_mail
526 and per_byte options don't have this problem.)
528 The handling of very low rates has changed slightly. If the computed rate
529 is less than the event's count (usually one) then this event is the first
530 after a long gap. In this case the rate is set to the same as this event's
531 count, so that the first message of a spam run is counted properly.
533 The major new feature is a mechanism for counting the rate of unique
534 events. The new per_addr option counts the number of different
535 recipients that someone has sent messages to in the last time period. It
536 behaves like per_rcpt if all the recipient addresses are different, but
537 duplicate recipient addresses do not increase the measured rate. Like
538 the /count= option this is a general mechanism, so the per_addr option
539 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
540 example, measure the rate that a client uses different sender addresses
541 with the options per_mail/unique=$sender_address. There are further
542 details in the main documentation.
544 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
546 TF/06 Removed a few PCRE remnants.
548 TF/07 Automatically extract Exim's version number from tags in the git
549 repository when doing development or release builds.
551 PP/02 Raise smtp_cmd_buffer_size to 16kB.
552 Bugzilla 879. Patch from Paul Fisher.
554 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
555 Heavily based on revision 40f9a89a from Simon Arlott's tree.
558 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
560 PP/05 Variable $av_failed, true if the AV scanner deferred.
561 Bugzilla 1078. Patch from John Horne.
563 PP/06 Stop make process more reliably on build failure.
564 Bugzilla 1087. Patch from Heiko Schlittermann.
566 PP/07 Make maildir_use_size_file an _expandable_ boolean.
567 Bugzilla 1089. Patch from Heiko Schlittermann.
569 PP/08 Handle ${run} returning more data than OS pipe buffer size.
570 Bugzilla 1131. Patch from Holger Weiß.
572 PP/09 Handle IPv6 addresses with SPF.
573 Bugzilla 860. Patch from Wolfgang Breyha.
575 PP/10 GnuTLS: support TLS 1.2 & 1.1.
577 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
580 PP/11 match_* no longer expand right-hand-side by default.
581 New compile-time build option, EXPAND_LISTMATCH_RHS.
582 New expansion conditions, "inlist", "inlisti".
584 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
586 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
588 PP/14 fix log_write() format string regression from TF/03.
589 Bugzilla 1152. Patch from Dmitry Isaikin.
595 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
597 PP/02 Harmonised TLS library version reporting; only show if debugging.
598 Layout now matches that introduced for other libraries in 4.74 PP/03.
600 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
602 PP/04 New "dns_use_edns0" global option.
604 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
607 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
608 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
610 TK/01 Updated PolarSSL code to 0.14.2.
611 Bugzilla 1097. Patch from Andreas Metzler.
613 PP/07 Catch divide-by-zero in ${eval:...}.
616 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
619 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
620 format-string attack -- SECURITY: remote arbitrary code execution.
622 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
623 time unintentionally subject to list matching rules, letting the header
624 cause arbitrary Exim lookups (of items which can occur in lists, *not*
625 arbitrary string expansion). This allowed for information disclosure.
627 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
628 INT_MIN/-1 -- value coerced to INT_MAX.
634 NM/01 Workround for PCRE version dependency in version reporting
637 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
638 This fixes portability to compilers other than gcc, notably
639 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
641 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
642 makefiles for portability to HP-UX and POSIX correctness.
644 PP/01 Permit LOOKUP_foo enabling on the make command-line.
645 Also via indented variable definition in the Makefile.
646 (Debugging by Oliver Heesakkers).
648 PP/02 Restore caching of spamd results with expanded spamd_address.
649 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
651 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
652 Improves build reliability. Fix from: Frank Elsner
654 NM/02 Fix wide character breakage in the rfc2047 coding
655 Fixes bug 1064. Patch from Andrey N. Oktyabrski
657 NM/03 Allow underscore in dnslist lookups
658 Fixes bug 1026. Patch from Graeme Fowler
660 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
661 Code patches from Adam Ciarcinski of NetBSD.
663 NM/04 Fixed exiqgrep to cope with mailq missing size issue
666 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
667 is logged, to avoid truncation. Patch from John Horne.
669 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
670 Patch from Jakob Hirsch.
672 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
673 SQL string expansion failure details.
674 Patch from Andrey Oktyabrski.
676 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
677 Patch from Simon Arlott.
679 PP/09 New lookups functionality failed to compile on old gcc which rejects
680 extern declarations in function scope.
681 Patch from Oliver Fleischmann
683 PP/10 Use sig_atomic_t for flags set from signal handlers.
684 Check getgroups() return and improve debugging.
685 Fixed developed for diagnosis in bug 927 (which turned out to be
688 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
689 Patch from Mark Zealey.
691 PP/12 Bugzilla 1056: Improved spamd server selection.
692 Patch from Mark Zealey.
694 PP/13 Bugzilla 1086: Deal with maildir quota file races.
695 Based on patch from Heiko Schlittermann.
697 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
698 Patch from Uwe Doering, sign-off by Michael Haardt.
700 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
701 variable declaration deep within a block. Bug and patch from
704 PP/15 lookups-Makefile IRIX compatibilty coercion.
706 PP/16 Make DISABLE_DKIM build knob functional.
708 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
709 Patch by Simon Arlott
711 TF/03 Fix valgrind.h portability to C89 compilers that do not support
712 variable argument macros. Our copy now differs from upstream.
718 TF/01 Failure to get a lock on a hints database can have serious
719 consequences so log it to the panic log.
721 TF/02 Log LMTP confirmation messages in the same way as SMTP,
722 controlled using the smtp_confirmation log selector.
724 TF/03 Include the error message when we fail to unlink a spool file.
726 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
727 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
728 for maintaining out-of-tree patches for some time.
730 PP/01 Bugzilla 139: Documentation and portability issues.
731 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
732 Handle per-OS dynamic-module compilation flags.
734 PP/02 Let /dev/null have normal permissions.
735 The 4.73 fixes were a little too stringent and complained about the
736 permissions on /dev/null. Exempt it from some checks.
737 Reported by Andreas M. Kirchwitz.
739 PP/03 Report version information for many libraries, including
740 Exim version information for dynamically loaded libraries. Created
741 version.h, now support a version extension string for distributors
742 who patch heavily. Dynamic module ABI change.
744 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
745 privilege escalation vulnerability whereby the Exim run-time user
746 can cause root to append content of the attacker's choosing to
749 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
752 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
753 If dropping privileges for untrusted macros, we disabled normal logging
754 on the basis that it would fail; for the Exim run-time user, this is not
755 the case, and it resulted in successful deliveries going unlogged.
756 Fixed. Reported by Andreas Metzler.
762 PP/01 Date: & Message-Id: revert to normally being appended to a message,
763 only prepend for the Resent-* case. Fixes regression introduced in
764 Exim 4.70 by NM/22 for Bugzilla 607.
766 PP/02 Include check_rfc2047_length in configure.default because we're seeing
767 increasing numbers of administrators be bitten by this.
769 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
771 PP/03 Bugzilla 994: added openssl_options main configuration option.
773 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
775 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
777 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
779 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
780 without a peer certificate, leading to a segfault because of an
781 assumption that peers always have certificates. Be a little more
782 paranoid. Problem reported by Martin Tscholak.
784 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
785 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
786 NB: ClamAV planning to remove STREAM in "middle of 2010".
787 CL also introduces -bmalware, various -d+acl logging additions and
788 more caution in buffer sizes.
790 PP/09 Implemented reverse_ip expansion operator.
792 PP/10 Bugzilla 937: provide a "debug" ACL control.
794 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
796 PP/12 Bugzilla 973: Implement --version.
798 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
800 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
802 PP/15 Bugzilla 816: support multiple condition rules on Routers.
804 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
805 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
806 ignore trailing whitespace.
808 JJ/02 prevent non-panic DKIM error from being sent to paniclog
810 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
813 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
814 Notification from Dr Andrew Aitchison.
816 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
817 ExtendedDetectionInfo response format.
818 Notification from John Horne.
820 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
823 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
824 XSL and documented dependency on system catalogs, with examples of how
827 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
830 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
831 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
832 configuration file which is writeable by the Exim user or group.
834 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
835 of configuration files to cover files specified with the -C option if
836 they are going to be used with root privileges, not just the default
839 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
840 option (effectively making it always true).
842 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
843 files to be used while preserving root privileges.
845 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
846 that rogue child processes cannot use them.
848 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
849 run-time user, instead of root.
851 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
852 Exim run-time user without dropping privileges.
854 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
855 result string, instead of calling string_vformat() twice with the same
858 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
859 for other users. Others should always drop root privileges if they use
860 -C on the command line, even for a whitelisted configure file.
862 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
864 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
870 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
871 $data_path, and $header_path variables; fixed documentation bugs and
874 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
875 exipick to access non-standard spools, including the "frozen" queue
878 NM/01 Bugzilla 965: Support mysql stored procedures.
879 Patch from Alain Williams
881 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
883 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
884 Patch from Andreas Metzler
886 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
887 Patch from Kirill Miazine
889 NM/05 Bugzilla 671: Added umask to procmail example.
891 JJ/03 installed exipick 20100323.0, fixing doc bug
893 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
894 directory. Notification and patch from Dan Rosenberg.
896 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
898 TK/02 Improve log output when DKIM signing operation fails.
900 MH/01 Treat the transport option dkim_domain as a colon separated
901 list, not as a single string, and sign the message with each element,
902 omitting multiple occurences of the same signer.
904 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
905 Bugzilla 985, 986. Patch by Simon Arlott
907 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
908 Patch by Simon Arlott
910 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
911 MBX locking. Notification from Dan Rosenberg.
917 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
919 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
921 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
923 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
925 NM/04 Bugzilla 915: Flag broken perl installation during build.
931 TK/01 Added patch by Johannes Berg that expands the main option
932 "spamd_address" if it starts with a dollar sign.
934 TK/02 Write list of recipients to X-Envelope-Sender header when building
935 the mbox-format spool file for content scanning (suggested by Jakob
938 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
939 (http://www.dcc-servers.net/) support via dccifd. Activated by
940 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
942 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
943 by Mark Daniel Reidel <mr@df.eu>.
945 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
946 When building exim an external PCRE library is now needed -
947 PCRE is a system library on the majority of modern systems.
948 See entry on PCRE_LIBS in EDITME file.
950 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
951 conversation. Added nologin parameter to request.
952 Patch contributed by Kirill Miazine.
954 TF/01 Do not log submission mode rewrites if they do not change the address.
956 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
958 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
959 log files in place. Contributed by Roberto Lima.
961 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
963 TF/03 Bugzilla 615: When checking the local_parts router precondition
964 after a local_part_suffix or local_part_prefix option, Exim now
965 does not use the address's named list lookup cache, since this
966 contains cached lookups for the whole local part.
968 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
969 Robert Millan. Documentation is in experimental-spec.txt.
971 TF/04 Bugzilla 668: Fix parallel build (make -j).
973 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
975 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
976 Patch provided by Jan Srzednicki.
978 TF/05 Leading white space used to be stripped from $spam_report which
979 wrecked the formatting. Now it is preserved.
981 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
982 that they are available at delivery time.
984 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
986 TF/08 TLS error reporting now respects the incoming_interface and
987 incoming_port log selectors.
989 TF/09 Produce a more useful error message if an SMTP transport's hosts
990 setting expands to an empty string.
992 NM/06 Bugzilla 744: EXPN did not work under TLS.
993 Patch provided by Phil Pennock.
995 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
996 Patch provided by Richard Godbee.
998 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
999 acl_smtp_notquit, added index entry.
1001 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1002 Patch provided by Eugene Bujak.
1004 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1005 accept(). Patch provided by Maxim Dounin.
1007 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1008 Patch provided by Phil Pennock.
1010 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1012 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1013 Patch provided by Brad "anomie" Jorsch.
1015 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1016 Patch provided by Dean Brooks.
1018 TK/05 Add native DKIM support (does not depend on external libraries).
1020 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1021 Patch provided by Graeme Fowler.
1023 NM/16 Bugzilla 851: Documentation example syntax fix.
1025 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1027 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1030 NM/19 Bugzilla 745: TLS version reporting.
1031 Patch provided by Phil Pennock.
1033 NM/20 Bugzilla 167: bool: condition support.
1034 Patch provided by Phil Pennock.
1036 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1037 clients. Patch provided by Phil Pennock.
1039 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1040 Patch provided by Brad "anomie" Jorsch.
1042 NM/23 Bugzilla 687: Fix misparses in eximstats.
1043 Patch provided by Heiko Schlittermann.
1045 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1046 Patch provided by Heiko Schlittermann.
1048 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1049 plus update to original patch.
1051 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1053 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1054 Patch provided by David Brownlee.
1056 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1058 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1060 NM/30 Bugzilla 888: TLS documentation bugfixes.
1062 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1064 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1065 Unlike the original bugzilla I have changed all shell scripts in src tree.
1067 NM/33 Bugzilla 898: Transport filter timeout fix.
1068 Patch by Todd Rinaldo.
1070 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1071 Patch by Serge Demonchaux.
1073 NM/35 Bugzilla 39: Base64 decode bug fixes.
1074 Patch by Jakob Hirsch.
1076 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1078 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1080 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1082 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1088 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1089 ALT-N's libdkim that I have put here:
1090 http://duncanthrax.net/exim-experimental/
1092 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1093 were called 'true' and it seems that C99 defines that as a reserved
1094 keyword to be used with 'bool' variable types. That means you could
1095 not include C99-style headers which use bools without triggering
1096 build errors in sieve.c.
1098 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1099 as mailq or other aliases. Changed the --help handling significantly
1100 to do whats expected. exim_usage() emits usage/help information.
1102 SC/01 Added the -bylocaldomain option to eximstats.
1104 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1106 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1108 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1114 PH/01 Another patch from the Sieve maintainer.
1116 PH/02 When an IPv6 address is converted to a string for single-key lookup
1117 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1118 dots are used instead of colons so that keys in lsearch files need not
1119 contain colons. This was done some time before quoting was made available
1120 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1121 (notated using the quote facility) so as to distinguish them from IPv4
1122 keys. This meant that lookups for IP addresses in host lists did not work
1123 for iplsearch lookups.
1125 This has been fixed by arranging for IPv6 addresses to be expressed with
1126 colons if the lookup type is iplsearch. This is not incompatible, because
1127 previously such lookups could never work.
1129 The situation is now rather anomolous, since one *can* have colons in
1130 ordinary lsearch keys. However, making the change in all cases is
1131 incompatible and would probably break a number of configurations.
1133 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1136 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1137 conversion specification without a maximum field width, thereby enabling
1138 a rogue spamd server to cause a buffer overflow. While nobody in their
1139 right mind would setup Exim to query an untrusted spamd server, an
1140 attacker that gains access to a server running spamd could potentially
1141 exploit this vulnerability to run arbitrary code as the Exim user.
1143 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1144 $primary_hostname instead of what libspf2 thinks the hosts name is.
1146 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1147 a directory entry by the name of the lookup key. Previously, if a
1148 symlink pointed to a non-existing file or a file in a directory that
1149 Exim lacked permissions to read, a lookup for a key matching that
1150 symlink would fail. Now it is enough that a matching directory entry
1151 exists, symlink or not. (Bugzilla 503.)
1153 PH/03 The body_linecount and body_zerocount variables are now exported in the
1156 PH/04 Added the $dnslist_matched variable.
1158 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1159 This means they are set thereafter only if the connection becomes
1162 PH/06 Added the client_condition to authenticators so that some can be skipped
1163 by clients under certain conditions.
1165 PH/07 The error message for a badly-placed control=no_multiline_responses left
1166 "_responses" off the end of the name.
1168 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1170 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1171 (without spaces) instead of just copying the configuration text.
1173 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1175 PH/11 Added $max_received_linelength.
1177 PH/12 Added +ignore_defer and +include_defer to host lists.
1179 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1180 way in which PCRE > 7.0 is built.
1182 PH/14 Implemented queue_only_load_latch.
1184 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1185 MAIL command. The effect was to mangle the value on 64-bit systems.
1187 PH/16 Another patch from the Sieve maintainer.
1189 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1191 PH/18 If a system quota error occurred while trying to create the file for
1192 a maildir delivery, the message "Mailbox is full" was not appended to the
1193 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1194 only to a quota excession during the actual writing of the file.
1196 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1197 characters?) which causes problems in log lines. The DN values are now
1198 passed through string_printing() before being added to log lines.
1200 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1201 and InterBase are left for another time.)
1203 PH/21 Added message_body_newlines option.
1205 PH/22 Guard against possible overflow in moan_check_errorcopy().
1207 PH/23 POSIX allows open() to be a macro; guard against that.
1209 PH/24 If the recipient of an error message contained an @ in the local part
1210 (suitably quoted, of course), incorrect values were put in $domain and
1211 $local_part during the evaluation of errors_copy.
1217 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1218 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1221 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1222 issue a MAIL command.
1224 PH/02 In an ACL statement such as
1226 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1228 if a client was not listed at all, or was listed with a value other than
1229 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1230 the condition was not true (as it should be), so access was not denied.
1231 The bug was that the ! inversion was incorrectly passed on to the second
1232 item. This has been fixed.
1234 PH/03 Added additional dnslists conditions == and =& which are different from
1235 = and & when the dns lookup returns more than one IP address.
1237 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1238 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1240 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1241 FSYNC, which compiles an option called disable_fsync that allows for
1242 bypassing fsync(). The documentation is heavily laced with warnings.
1244 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1246 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1247 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1248 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1249 including adding "make clean"; (3) Added -fPIC when compiling the test
1250 dynamically loaded module, to get rid of a warning.
1252 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1253 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1254 = 0s. The bug is otherwise harmless.
1256 PH/07 There was a bug in the dovecot authenticator such that the value of
1257 $auth1 could be overwritten, and so not correctly preserved, after a
1258 successful authentication. This usually meant that the value preserved by
1259 the server_setid option was incorrect.
1261 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1263 PH/09 Installed PCRE release 7.0.
1265 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1266 run for batched SMTP input. It is now run at the start of every message
1267 in the batch. While fixing this I discovered that the process information
1268 (output by running exiwhat) was not always getting set for -bs and -bS
1269 input. This is fixed, and it now also says "batched" for BSMTP.
1271 PH/11 Added control=no_pipelining.
1273 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1274 patch, slightly modified), and move the expansion of helo_data till after
1275 the connection is made in the smtp transport (so it can use these
1278 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1280 PH/14 Added log_selector = +pid.
1282 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1284 PH/16 Add ${if forany and ${if forall.
1286 PH/17 Added dsn_from option to vary the From: line in DSNs.
1288 PH/18 Flush SMTP output before performing a callout, unless control =
1289 no_callout_flush is set.
1291 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1292 was true (the default) a successful delivery failed to delete the retry
1293 item, thus causing premature timeout of the address. The bug is now
1296 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1298 PH/21 Long custom messages for fakedefer and fakereject are now split up
1299 into multiline reponses in the same way that messages for "deny" and
1300 other ACL rejections are.
1302 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1303 with slight modification.
1305 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1306 draft, changing the syntax and factoring some duplicate code".
1308 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1309 for deliveries of the second and subsequent messages over the same SMTP
1312 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1313 ${reduce, with only minor "tidies".
1315 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1317 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1318 expansion side effects.
1320 PH/27 When a message times out after an over-quota error from an Exim-imposed
1321 quota, the bounce message says "mailbox is full". This message was not
1322 being given when it was a system quota that was exceeded. It now should
1325 MH/03 Made $recipients available in local_scan(). local_scan() already has
1326 better access to the recipient list through recipients_list[], but
1327 $recipients can be useful in postmaster-provided expansion strings.
1329 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1330 in the case of a MAIL command with additional options following the
1331 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1332 were accidentally chopped off.
1334 PH/29 SMTP synchronization checks are implemented when a command is read -
1335 there is a check that no more input is waiting when there shouldn't be
1336 any. However, for some commands, a delay in an ACL can mean that it is
1337 some time before the response is written. In this time, more input might
1338 arrive, invalidly. So now there are extra checks after an ACL has run for
1339 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1340 pipelining has not been advertised.
1342 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1344 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1345 returned in the SMTP response when smtp_return_error_details was false.
1346 This has been fixed.
1348 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1349 instead of the C I/O that was originally supplied, because problems were
1350 reported on Solaris.
1352 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1353 Exim which did not show up earlier: it was assuming that a call to
1354 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1355 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1356 was a macro that became an assignment, so it seemed to work. This has
1357 changed to a proper function call with a void return, hence the compile
1358 error. Exim's code has been fixed.
1360 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1363 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1365 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1367 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1368 because it turns out to be too big or there is a timeout) there is no
1369 "Completed" line in the log. When some messages of this type were
1370 selected by exigrep, they were listed as "not completed". Others were
1371 picked up by some special patterns. I have improved the selection
1372 criteria to be more general.
1374 PH/38 The host_find_failed option in the manualroute router can now be set
1375 to "ignore", to completely ignore a host whose IP address cannot be
1376 found. If all hosts are ignored, the behaviour is controlled by the new
1377 host_all_ignored option.
1379 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1380 homing or because of multiple MX records with /mx) generated more than
1381 one IP address, and the following item turned out to be the local host,
1382 all the secondary addresses of the first item were incorrectly removed
1383 from the list, along with the local host and any following hosts (which
1384 is what is supposed to happen).
1386 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1387 whoever called Exim into the -H file. In the case of the daemon it was
1388 behaving confusingly. When first started, it used values for whoever
1389 started the daemon, but after a SIGHUP it used the Exim user (because it
1390 calls itself on a restart). I have changed the code so that it now always
1393 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1394 message are rejected with the same error (e.g. no authentication or bad
1395 sender address), and a DATA command is nevertheless sent (as can happen
1396 with PIPELINING or a stupid MUA), the error message that was given to the
1397 RCPT commands is included in the rejection of the DATA command. This is
1398 intended to be helpful for MUAs that show only the final error to their
1401 PH/42 Another patch from the Sieve maintainer.
1403 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1404 Eximstats - Fixed some broken HTML links and added missing column headers
1406 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1407 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1409 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1411 PH/43 Yet another patch from the Sieve maintainer.
1413 PH/44 I found a way to check for a TCP/IP connection going away before sending
1414 the response to the final '.' that terminates a message, but only in the
1415 case where the client has not sent further data following the '.'
1416 (unfortunately, this is allowed). However, in many cases there won't be
1417 any further data because there won't be any more messages to send. A call
1418 to select() can be used: if it shows that the input is "ready", there is
1419 either input waiting, or the socket has been closed. An attempt to read
1420 the next input character can distinguish the two cases. Previously, Exim
1421 would have sent an OK response which the client would never have see.
1422 This could lead to message repetition. This fix should cure that, at
1423 least in a lot of common cases.
1425 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1426 advertised in response to EHLO.
1432 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1433 fixed by 4.65/MH/01 (is this a record?) are fixed:
1435 (i) An empty string was always treated as zero by the numeric comparison
1436 operators. This behaviour has been restored.
1438 (ii) It is documented that the numeric comparison operators always treat
1439 their arguments as decimal numbers. This was broken in that numbers
1440 starting with 0 were being interpreted as octal.
1442 While fixing these problems I realized that there was another issue that
1443 hadn't been noticed. Values of message_size_limit (both the global option
1444 and the transport option) were treated as octal if they started with 0.
1445 The documentation was vague. These values are now always treated as
1446 decimal, and I will make that clear in the documentation.
1452 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1453 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1456 MH/01 Don't check that the operands of numeric comparison operators are
1457 integers when their expansion is in "skipping" mode (fixes bug
1458 introduced by 4.64-PH/07).
1460 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1461 child addresses, Exim now panics and dies. Previously, because the count
1462 is held in a short int, deliveries were likely to be lost. As such a
1463 large number of recipients for a single message is ridiculous
1464 (performance will be very, very poor), I have chosen to impose a limit
1465 rather than extend the field.
1471 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1472 leftover -K file (the existence of which was triggered by #402).
1473 While we were at it, introduced process PID as part of the -K
1474 filename. This should rule out race conditions when creating
1477 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1478 processing considerably. Previous code took too long for large mails,
1479 triggering a timeout which in turn triggers #401.
1481 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1482 in the DK code in transports.c. sendfile() is not really portable,
1483 hence the _LINUX specificness.
1485 TF/01 In the add_headers option to the mail command in an Exim filter,
1486 there was a bug that Exim would claim a syntax error in any
1487 header after the first one which had an odd number of characters
1490 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1491 callout verification, Exim cached a "reject" for the entire domain. This
1492 is correct for most verifications, but it is not correct for a recipient
1493 verification with use_sender or use_postmaster set, because in that case
1494 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1495 case of MAIL FROM:<> rejection from other early rejections (e.g.
1496 rejection of HELO). When verifying a recipient using a non-null MAIL
1497 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1498 Whatever the result of the callout, the value of the domain cache is
1499 left unchanged (for any other kind of callout, getting as far as trying
1500 RCPT means that the domain itself is ok).
1502 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1505 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1506 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1507 session, the daemon ignores ECONNECT errors and logs others; it now
1508 ignores EPIPE as well.
1510 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1511 (quoted-printable decoding).
1513 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1514 later the small subsequent patch to fix an introduced bug.
1516 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1518 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1520 PH/08 An error is now given if message_size_limit is specified negative.
1522 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1523 to be given (somewhat) arbitrary names.
1525 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1528 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1529 miscellaneous code fixes
1531 PH/10 Added the log_reject_target ACL modifier to specify where to log
1534 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1535 hostname. This is wrong, because it relates to the incoming message (and
1536 probably the interface on which it is arriving) and not to the outgoing
1537 callout (which could be using a different interface). This has been
1538 changed to use the value of the helo_data option from the smtp transport
1539 instead - this is what is used when a message is actually being sent. If
1540 there is no remote transport (possible with a router that sets up host
1541 addresses), $smtp_active_hostname is used.
1543 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1544 tweaks were necessary in order to get it to work (see also 21 below):
1545 (a) The code assumed that strncpy() returns a negative number on buffer
1546 overflow, which isn't the case. Replaced with Exim's string_format()
1548 (b) There were several signed/unsigned issues. I just did the minimum
1549 hacking in of casts. There is scope for a larger refactoring.
1550 (c) The code used strcasecmp() which is not a standard C function.
1551 Replaced with Exim's strcmpic() function.
1552 (d) The code set only $1; it now sets $auth1 as well.
1553 (e) A simple test gave the error "authentication client didn't specify
1554 service in request". It would seem that Dovecot has changed its
1555 interface. Fortunately there's a specification; I followed it and
1556 changed what the client sends and it appears to be working now.
1558 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1561 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1562 address A is aliased to B and C, where B exists and C does not. Without
1563 -v the output is "A verified" because verification stops after a
1564 successful redirection if more than one address is generated. However,
1565 with -v the child addresses are also verified. Exim was outputting "A
1566 failed to verify" and then showing the successful verification for C,
1567 with its parentage. It now outputs "B failed to verify", showing B's
1568 parentage before showing the successful verification of C.
1570 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1571 look up a TXT record in a specific list after matching in a combined
1574 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1575 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1576 they consult the DNS. I had assumed they would set it the way they
1577 wanted; and indeed my experiments on Linux seem to show that in some
1578 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1579 To be on the safe side, however, I have now made the interface to
1580 host_find_byname() similar to host_find_bydns(), with an argument
1581 containing the DNS resolver options. The host_find_byname() function now
1582 sets these options at its start, just as host_find_bydns() does. The smtp
1583 transport options dns_qualify_single and dns_search_parents are passed to
1584 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1585 of host_find_byname() use the default settings of RES_DEFNAMES
1586 (qualify_single) but not RES_DNSRCH (search_parents).
1588 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1589 spool_read_header() do less string testing, by means of a preliminary
1590 switch on the second character of optional "-foo" lines. (This is
1591 overdue, caused by the large number of possibilities that now exist.
1592 Originally there were few.) While I was there, I also converted the
1593 str(n)cmp tests so they don't re-test the leading "-" and the first
1594 character, in the hope this might squeeze out yet more improvement.
1596 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1597 flag allowing group syntax was set by the header_syntax check but not
1598 turned off, possible causing trouble later; (2) The flag was not being
1599 set at all for the header_verify test, causing "group"-style headers to
1600 be rejected. I have now set it in this case, and also caused header_
1601 verify to ignore an empty address taken from a group. While doing this, I
1602 came across some other cases where the code for allowing group syntax
1603 while scanning a header line wasn't quite right (mostly, not resetting
1604 the flag correctly in the right place). These bugs could have caused
1605 trouble for malformed header lines. I hope it is now all correct.
1607 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1608 with the "reply" argument non-NULL. The code, however (which originally
1609 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1610 but it didn't always do it. This confused somebody who was copying the
1611 code for some other use. I have removed all the tests.
1613 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1614 feature that was used to support insecure browsers during the U.S. crypto
1615 embargo. It requires special client support, and Exim is probably the
1616 only MTA that supported it -- and would never use it because real RSA is
1617 always available. This code has been removed, because it had the bad
1618 effect of slowing Exim down by computing (never used) parameters for the
1619 RSA_EXPORT functionality.
1621 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1622 authenticator to fail if there's a tab character in the incoming data
1623 (there should never be unless someone is messing about, as it's supposed
1624 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1625 if the connection is using TLS or if the remote IP is the same as the
1626 local IP, and the "valid-client-cert option" if a client certificate has
1629 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1630 authenticators. This can be used for authorization after authentication
1631 succeeds. (In the case of plaintext, it servers for both authentication
1634 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1635 if any retry times were supplied.
1637 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1638 connection, where there is no client IP address to check. In this
1639 situation, the verify now always succeeds.
1641 PH/25 Applied John Jetmore's -Mset patch.
1643 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1645 PH/27 In a string expansion for a processed (not raw) header when multiple
1646 headers of the same name were present, leading whitespace was being
1647 removed from all of them, but trailing whitespace was being removed only
1648 from the last one. Now trailing whitespace is removed from each header
1649 before concatenation. Completely empty headers in a concatenation (as
1650 before) are ignored.
1652 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1653 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1655 PH/29 [Removed. This was a change that I later backed out, and forgot to
1656 correct the ChangeLog entry (that I had efficiently created) before
1657 committing the later change.]
1659 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1660 address errors (4xx response to RCPT) over the same connection as other
1661 messages routed to the same hosts. Such deliveries are always "forced",
1662 so retry times are not inspected. This resulted in far too many retries
1663 for the affected addresses. The effect occurred only when there were more
1664 hosts than the hosts_max_try setting in the smtp transport when it had
1665 the 4xx errors. Those hosts that it had tried were not added to the list
1666 of hosts for which the message was waiting, so if all were tried, there
1667 was no problem. Two fixes have been applied:
1669 (i) If there are any address or message errors in an SMTP delivery, none
1670 of the hosts (tried or untried) are now added to the list of hosts
1671 for which the message is waiting, so the message should not be a
1672 candidate for sending over the same connection that was used for a
1673 successful delivery of some other message. This seems entirely
1674 reasonable: after all the message is NOT "waiting for some host".
1675 This is so "obvious" that I'm not sure why it wasn't done
1676 previously. Hope I haven't missed anything, but it can't do any
1677 harm, as the worst effect is to miss an optimization.
1679 (ii) If, despite (i), such a delivery is accidentally attempted, the
1680 routing retry time is respected, so at least it doesn't keep
1681 hammering the server.
1683 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1684 in ${readsocket because some servers need this prod.
1686 PH/32 Added some extra debug output when updating a wait-xxx database.
1688 PH/33 The hint "could be header name not terminated by colon", which has been
1689 given for certain expansion errors for a long time, was not being given
1690 for the ${if def:h_colon_omitted{... case.
1692 PH/34 The spec says: "With one important exception, whenever a domain list is
1693 being scanned, $domain contains the subject domain." There was at least
1694 one case where this was not true.
1696 PH/35 The error "getsockname() failed: connection reset by peer" was being
1697 written to the panic log as well as the main log, but it isn't really
1698 panic-worthy as it just means the connection died rather early on. I have
1699 removed the panic log writing for the ECONNRESET error when getsockname()
1702 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1703 runs only) independently of the message's sender address. This meant
1704 that, if the 4xx error was in fact related to the sender, a different
1705 message to the same recipient with a different sender could confuse
1706 things. In particualar, this can happen when sending to a greylisting
1707 server, but other circumstances could also provoke similar problems.
1708 I have changed the default so that the retry time for these errors is now
1709 based a combination of the sender and recipient addresses. This change
1710 can be overridden by setting address_retry_include_sender=false in the
1713 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1714 remote server are returned as part of bounce messages. This was not
1715 happening for LMTP over a pipe (the lmtp transport), but now it is the
1716 same for both kinds of LMTP.
1718 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1719 in header lines that were in fact CNAMEs. This is no longer the case.
1721 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1722 and queue runs started by the daemon processed all messages. This has
1723 been fixed so that -R and -S can now usefully be given with -q<time>.
1725 PH/40 Import PCRE release 6.7 (fixes some bugs).
1727 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1729 PH/42 Give an error if -q is specified more than once.
1731 PH/43 Renamed the variables $interface_address and $interface_port as
1732 $received_ip_address and $received_port, to make it clear that these
1733 values apply to message reception, and not to the outgoing interface when
1734 a message is delivered. (The old names remain recognized, of course.)
1736 PH/44 There was no timeout on the connect() call when using a Unix domain
1737 socket in the ${readsocket expansion. There now is.
1739 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1740 be meaningful with "accept".
1742 SC/01 Eximstats V1.43
1743 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1745 SC/02 Eximstats V1.44
1746 Use a glob alias rather than an array ref in the generated
1747 parser. This improves both readability and performance.
1749 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1750 Collect SpamAssassin and rejection statistics.
1751 Don't display local sender or destination tables unless
1752 there is data to show.
1753 Added average volumes into the top table text output.
1755 SC/04 Eximstats V1.46
1756 Collect data on the number of addresses (recipients)
1757 as well as the number of messages.
1759 SC/05 Eximstats V1.47
1760 Added 'Message too big' to the list of mail rejection
1761 reasons (thanks to Marco Gaiarin).
1763 SC/06 Eximstats V1.48
1764 Mainlog lines which have GMT offsets and are too short to
1765 have a flag are now skipped.
1767 SC/07 Eximstats V1.49 (Alain Williams)
1768 Added the -emptyok flag.
1770 SC/08 Eximstats V1.50
1771 Fixes for obtaining the IP address from reject messages.
1773 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1774 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1775 whitesspace changes from 4.64-PH/27
1777 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1780 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1781 are found, allow negative numbers in numeric criteria)
1783 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1785 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1786 to match changes made in 4.64-PH/43
1788 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1790 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1791 "domain" is actually an IP address, and force a failure. This locks out
1792 those revolvers/nameservers that support "A-for-A" lookups, in
1793 contravention of the specifications.
1795 PH/48 When a host name was looked up from an IP address, and the subsequent
1796 forward lookup of the name timed out, the host name was left in
1797 $sender_host_name, contrary to the specification.
1799 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1800 restricted to single-key lookups, Exim was not diagnosing an error if
1801 * or *@ was used with a query-style lookup.
1803 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1805 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1806 long ago, but noone interested enough thought of it. Let's just say that
1807 the "1.1" means that there are some new functions that weren't there at
1808 some point in the past.
1810 PH/51 Error processing for expansion failure of helo_data from an smtp
1811 transport during callout processing was broken.
1813 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1814 tested/used via the -bh/-bhc/-bs options.
1816 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1817 bug, fixed in subsequent PCRE releases).
1819 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1820 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1822 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1828 SC/01 Use a glob alias rather than an array ref in eximstats generated
1829 parser. This improves both readability and performance.
1831 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1832 Don't display local sender or destination tables in eximstats unless
1833 there is data to show.
1834 Added average volumes into the eximstats top table text output.
1836 SC/03 Collect data on the number of addresses (recipients) as well
1837 as the number of messages in eximstats.
1839 TF/01 Correct an error in the documentation for the redirect router. Exim
1840 does (usually) call initgroups() when daemonizing.
1842 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1843 with consistent privilege compared to when running as a daemon.
1845 TF/03 Note in the spec that $authenticated_id is not set for local
1846 submissions from trusted users.
1848 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1849 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1851 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1852 by adding some example configuration directives to the default
1853 configuration file. A little bit of work is required to uncomment the
1854 directives and define how usernames and passwords are checked, but
1855 there is now a framework to start from.
1857 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1858 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1859 without this. I don't know how relevant this is to other LDAP libraries.
1861 PH/02 Add the verb name to the "unknown ACL verb" error.
1863 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1865 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1867 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1868 directory not expanded when it should be if an expanded home directory
1869 was set for the address (which is overridden by the transport).
1871 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1874 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1875 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1876 because it is too late at that time, and has no effect.
1878 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1879 security issue with \' (bugzilla #107). I could not use the
1880 PQescapeStringConn() function, because it needs a PGconn value as one of
1883 PH/08 When testing addresses using -bt, indicate those final addresses that
1884 are duplicates that would not cause an additional delivery. At least one
1885 person was confused, thinking that -bt output corresponded to deliveries.
1886 (Suppressing duplicates isn't a good idea as you lose the information
1887 about possibly different redirections that led to the duplicates.)
1889 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1890 systems where poll() doesn't work, in particular OS X.
1892 PH/10 Added more information to debugging output for retry time not reached.
1894 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1895 operations in malware.c.
1897 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1900 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1901 syslog_duplication set false, log lines that would normally be written
1902 both the the main log and to the reject log were not written to syslog at
1905 PH/14 In the default configuration, change the use of "message" in ACL warn
1906 statements to "add_header".
1908 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1909 not followed by a command (e.g. "seen endif").
1911 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1912 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1915 PH/17 Added extra conditions to the default value of delay_warning_condition
1919 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1920 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1921 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1924 The Auto-Submitted: and various List- headers are standardised, whereas I
1925 don't think Precedence: ever was.
1927 PH/18 Refactored debugging code in route_finduser() to show more information,
1928 in particular, the error code if getpwnam() issues one.
1930 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1931 This is apparently needed in addition to the PH/07 change above to avoid
1932 any possible encoding problems.
1934 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1935 but not after initializing Perl.
1937 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1938 output them only if debugging. By default they are written stderr,
1939 apparently, which is not desirable.
1941 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1944 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1947 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1949 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1950 authenticated or an ident call has been made. Suppress the default
1951 values for $authenticated_id and $authenticated_sender (but permit -oMai
1952 and -oMas) when testing with -bh.
1954 PH/24 Re-jigged the order of the tests in the default configuration so that the
1955 tests for valid domains and recipients precede the DNS black list and CSA
1956 tests, on the grounds that those ones are more expensive.
1958 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1959 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1960 This bug exists in every version of Exim that I still have, right back to
1963 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1964 However, an attempt to turn on case-sensitivity in a regex key by
1965 including (?-i) didn't work because the subject string was already
1966 lowercased, and the effects were non-intuitive. It turns out that a
1967 one-line patch can be used to allow (?-i) to work as expected.
1973 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1974 other effects) broke the use of negated acl sub-conditions.
1976 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1979 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1980 "Deny" causes Exim to reject the incoming connection with a 554 error.
1981 Unfortunately, if there is a major crisis, such as a disk failure,
1982 tcp-wrappers gives "deny", whereas what one would like would be some
1983 kind of temporary error. A kludge has been added to help with this.
1984 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1985 554 error is used if errno is still zero or contains ENOENT (which occurs
1986 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1989 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1991 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1992 errors. Otherwise a message that provokes a temporary error (when other
1993 messages do not) can cause a whole host to time out.
1995 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1996 addresses were routed directly to files or pipes from a redirect router.
1997 File deliveries just didn't batch; pipe deliveries might have suffered
2000 PH/06 A failure to get a lock for a hints database would erroneously always say
2001 "Failed to get write lock", even when it was really a read lock.
2003 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2004 of 0600. This has been changed to use the value of the lockfile_mode
2005 option (which defaults to 0600).
2007 PH/08 Applied small patch from the Sieve maintainer.
2009 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2010 folder from quota calculations, a direct delivery into this folder messed
2011 up the contents of the maildirsize file. This was because the regex was
2012 used only to exclude .Trash (or whatever) when the size of the mailbox
2013 was calculated. There was no check that a delivery was happening into an
2014 excluded directory. This bug has been fixed by ignoring all quota
2015 processing for deliveries into excluded directories.
2017 PH/10 Added the maildirfolder_create_regex option to appendfile.
2023 PH/01 The code for finding all the local interface addresses on a FreeBSD
2024 system running IPv6 was broken. This may well have applied to all BSD
2025 systems, as well as to others that have similar system calls. The broken
2026 code found IPv4 interfaces correctly, but gave incorrect values for the
2027 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2028 that it would not match correctly against @[] and not recognize the IPv6
2031 PH/02 The ipliteral router was not recognizing addresses of the form user@
2032 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2034 PH/03 Added disable_ipv6.
2036 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2037 decoded form, because it is most often used to construct To: headers
2038 lines in autoreplies, and the decoded form may well be syntactically
2039 invalid. However, $reply_address has leading white space removed, and all
2040 newlines turned into spaces so that the autoreply transport does not
2043 PH/05 If group was specified without a user on a router, and no group or user
2044 was specified on a transport, the group from the router was ignored.
2046 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2047 for visible compile-time settings that can be used to change these
2048 numbers, for those that want even more. Backwards compatibility with old
2049 spool files has been maintained. However, going back to a previous Exim
2050 release will lost any variables that are in spool files.
2052 PH/07 Two small changes when running in the test harness: increase delay when
2053 passing a TCP/IP connection to a new process, in case the original
2054 process has to generate a bounce, and remove special handling of
2055 127.0.0.2 (sic), which is no longer necessary.
2057 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2058 be the same on different OS.
2060 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2063 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2064 whether --show-vars was specified or not
2066 JJ/02 exipick: Added support for new ACL variable spool format introduced
2069 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2070 syntactically invalid From: or Reply-to: line, and a filter used this to
2071 generate an autoreply, and therefore failed to obtain an address for the
2072 autoreply, Exim could try to deliver to a non-existent relative file
2073 name, causing unrelated and misleading errors. What now happens is that
2074 it logs this as a hard delivery error, but does not attempt to create a
2077 PH/11 The exinext utility has a -C option for testing purposes, but although
2078 the given file was scanned by exinext itself; it wasn't being passed on
2079 when Exim was called.
2081 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2082 an end-of-file indication when reading a command response.
2084 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2085 compiled. In many other places in Exim, IPv6 addresses are always
2086 recognized, so I have changed this. It also means that IPv4 domain
2087 literals of the form [IPV4:n.n.n.n] are now always recognized.
2089 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2090 used if the router is not running as root, for example, when verifying at
2091 ACL time, or when using -bh. The debugging output from this situation was
2092 non-existent - all you got was a failure to exec. I have made two
2095 (a) Failures to set uid/gid, the current directory, or a process leader
2096 in a subprocess such as that created by queryprogram now generate
2097 suitable debugging ouput when -d is set.
2099 (b) The queryprogram router detects when it is not running as root,
2100 outputs suitable debugging information if -d is set, and then runs
2101 the subprocess without attempting to change uid/gid.
2103 PH/15 Minor change to Makefile for building test_host (undocumented testing
2106 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2107 additional section of a DNS packet that returns MX or SRV records.
2108 Instead, it always explicitly searches for A/AAAA records. This avoids
2109 major problems that occur when a DNS server includes only records of one
2110 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2111 fixed another bug: if SRV records were looked up and the corresponding
2112 address records were *not* found in the additional section, the port
2113 values from the SRV records were lost.
2115 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2116 using the correct key (the original address) when searching the retry
2117 rules in order to find which one to use for generating the retry hint.
2119 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2120 adding the default one. Similarly, if it contains a Reply-To: header, the
2121 errors_reply_to option, if set, is not used.
2123 PH/19 When calculating a retry time, Exim used to measure the "time since
2124 failure" by looking at the "first failed" field in the retry record. Now
2125 it does not use this if it is later than than the arrival time of the
2126 message. Instead it uses the arrival time. This makes for better
2127 behaviour in cases where some deliveries succeed, thus re-setting the
2128 "first failed" field. An example is a quota failure for a huge message
2129 when small messages continue to be delivered. Without this change, the
2130 "time since failure" will always be short, possible causing more frequent
2131 delivery attempts for the huge message than are intended.
2132 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2134 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2135 $1, $2, $3) because the numerical variables can be reset during some
2136 expansion items (e.g. "match"), thereby losing the authentication data.
2138 PH/21 Make -bV show the size of off_t variables so that the test suite can
2139 decide whether to run tests for quotas > 2G.
2141 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2142 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2143 filecount value is greater than 2G or if a quota value is greater than 2G
2144 on a system where the size of off_t is not greater than 4, a panic error
2147 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2148 never match. The debug and -bh output now contains an explicit error
2149 message indicating a malformed IPv4 address or mask.
2151 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2152 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2153 PH/23 above applies.
2155 PH/25 Do not write to syslog when running in the test harness. The only
2156 occasion when this arises is a failure to open the main or panic logs
2157 (for which there is an explicit test).
2159 PH/26 Added the /no_tell option to "control=freeze".
2161 PH/27 If a host name lookup failed very early in a connection, for example, if
2162 the IP address matched host_lookup and the reverse lookup yielded a name
2163 that did not have a forward lookup, an error message of the form "no IP
2164 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2165 could be logged. Now it outputs the IP address instead of "NULL".
2167 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2168 allows the sender and the authenticated sender to be set when
2169 submitting a message from within Exim. Since child_open_exim() is
2170 documented for local_scan(), the new function should be too.
2172 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2173 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2174 results in an empty string is now treated as unset.
2176 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2178 PH/31 Added sender_verify_fail logging option.
2180 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2181 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2182 tidied the source and removed it altogether.
2184 PH/33 When a queue run was abandoned because the load average was too high, a
2185 log line was always written; now it is written only if the queue_run log
2186 selector is set. In addition, the log line for abandonment now contains
2187 information about the queue run such as the pid. This is always present
2188 in "start" and "stop" lines but was omitted from the "abandon" line.
2190 PH/34 Omit spaces between a header name and the colon in the error message that
2191 is given when verify = headers_syntax fails (if there are lots of them,
2192 the message gets confusing).
2194 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2195 names, as there are now some PTR records that contain slashes. This check
2196 is only to protect against broken name servers that fall over on strange
2197 characters, so the fact that it applies to all lookups doesn't matter.
2199 PH/36 Now that the new test suite is complete, we can remove some of the
2200 special code in Exim that was needed for the old test suite. For example,
2201 sorting DNS records because real resolvers return them in an arbitrary
2202 order. The new test suite's fake resolver always returns records in the
2205 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2206 bounces) except when queue_only is set, to avoid logging races between
2207 the different processes.
2209 PH/38 Panic-die if .include specifies a non-absolute path.
2211 PH/39 A tweak to the "H" retry rule from its user.
2213 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2214 a label. They prevented compilation on older perls.
2216 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2217 a warning to be raised on newish perls.
2219 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2220 on queue. Changes to match documented behaviour of showing count of
2221 messages matching specified criteria.
2223 PH/40 Changed the default ident timeout from 30s to 5s.
2225 PH/41 Added support for the use of login_cap features, on those BSD systems
2226 that have them, for controlling the resources used by pipe deliveries.
2228 PH/42 The content-scanning code uses fopen() to create files in which to put
2229 message data. Previously it was not paying any attention to the mode of
2230 the files. Exim runs with umask(0) because the rest of the code creates
2231 files with open(), and sets the required mode explicitly. Thus, these
2232 files were ending up world-writeable. This was not a big issue, because,
2233 being within the spool directory, they were not world-accessible. I have
2234 created a function called modefopen, which takes an additional mode
2235 argument. It sets umask(777), creates the file, chmods it to the required
2236 mode, then resets the umask. All the relevant calls to fopen() in the
2237 content scanning code have been changed to use this function.
2239 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2240 to 24 hours. This avoids potential overflow problems when processing G
2241 and H retry rules. I suspect nobody ever tinkers with this value.
2243 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2245 PH/45 When the plaintext authenticator is running as a client, the server's
2246 challenges are checked to ensure they are valid base64 strings. By
2247 default, the authentication attempt is cancelled if an invalid string is
2248 received. Setting client_ignore_invalid_base64 true ignores these errors.
2249 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2250 they are received. Thus, the responses can be made to depend on the
2251 challenges. If an invalid string is ignored, an empty string is placed in
2254 PH/46 Messages that are created by the autoreply transport now contains a
2255 References: header, in accordance with RFCs 2822 and 3834.
2257 PH/47 Added authenticated_sender_force to the smtp transport.
2259 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2261 PH/49 Installed latest patch from the Sieve maintainer.
2263 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2264 was also set, the appendfile transport was unnecessarily scanning a
2265 directory of message files (e.g. for maildir delivery) to find the count
2266 of files (along with the size), even though it did not need this
2267 information. It now does the scan only if it needs to find either the
2268 size of the count of files.
2270 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2272 PH/52 Two bugs concerned with error handling when the smtp transport is
2275 (i) Exim was not creating retry information for temporary errors given
2276 for individual recipients after the DATA command when the smtp transport
2277 was used in LMTP mode. This meant that they could be retried too
2278 frequently, and not timed out correctly.
2280 (ii) Exim was setting the flag that allows error details to be returned
2281 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2282 recipients that were returned after the DATA command.
2284 PH/53 This is related to PH/52, but is more general: for any failing address,
2285 when detailed error information was permitted to be returned to the
2286 sender, but the error was temporary, then after the final timeout, only
2287 "retry timeout exceeded" was returned. Now it returns the full error as
2288 well as "retry timeout exceeded".
2290 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2291 do this, and (what is worse) MTAs that accept it.
2293 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2294 will now be deprecated.
2296 PH/56 New os.c-cygwin from the Cygwin maintainer.
2298 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2299 formats (previously only available in exim formats via -bpr, -bpru,
2300 and -bpra. Now also available in native and exiqgrep formats)
2302 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2303 with very large, slow to parse queues
2305 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2307 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2309 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2310 responses to authentication challenges, though it was showing the
2311 challenges; (ii) I've removed the CR characters from the debug output for
2314 PH/58 Allow for the insertion of a newline as well as a space when a string
2315 is turned into more than one encoded-word during RFC 2047 encoding. The
2316 Sieve code now uses this.
2318 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2319 data_4xx, lost_connection, tls_required.
2321 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2322 message was being sent as an SMTP response.
2324 PH/61 Add -l and -k options to exicyclog.
2326 PH/62 When verifying, if an address was redirected to one new address, so that
2327 verification continued, and the new address failed or deferred after
2328 having set something in $address_data, the value of $address_data was not
2329 passed back to the ACL. This was different to the case when no
2330 redirection occurred. The value is now passed back in both cases.
2332 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2333 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2334 use login_cap.h, so on its own it isn't the distinguishing feature. The
2335 new name refers directly to the setclassresources() function.
2337 PH/65 Added configuration files for NetBSD3.
2339 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2341 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2342 is preferred over IPv4.
2344 PH/68 The bounce_return_message and bounce_return_body options were not being
2345 honoured for bounces generated during the reception of non-SMTP messages.
2346 In particular, this applied to messages rejected by the ACL. This bug has
2347 been fixed. However, if bounce_return_message is true and bounce_return_
2348 body is false, the headers that are returned for a non-SMTP message
2349 include only those that have been read before the error was detected.
2350 (In the case of an ACL rejection, they have all been read.)
2352 PH/69 The HTML version of the specification is now built in a directory called
2353 spec_html instead of spec.html, because the latter looks like a path with
2354 a MIME-type, and this confuses some software.
2356 PH/70 Catch two compiler warnings in sieve.c.
2358 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2359 function verify_get_ident() calls ip_connect() to connect a socket, but
2360 if the "connect()" function timed out, ip_connect() used to close the
2361 socket. However, verify_get_ident() also closes the socket later, and in
2362 between Exim writes to the log, which may get opened at this point. When
2363 the socket was closed in ip_connect(), the log could get the same file
2364 descriptor number as the socket. This naturally causes chaos. The fix is
2365 not to close the socket in ip_connect(); the socket should be closed by
2366 the function that creates it. There was only one place in the code where
2367 this was missing, in the iplookup router, which I don't think anybody now
2368 uses, but I've fixed it anyway.
2370 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2371 well as to direct DNS lookups. Otherwise the handling of names in host
2372 lists is inconsistent and therefore confusing.
2378 PH/01 Two changes to the default runtime configuration:
2380 (1) Move the checks for relay_from_hosts and authenticated clients from
2381 after to before the (commented out) DNS black list checks.
2383 (2) Add control=submission to the relay_from_hosts and authenticated
2384 clients checks, on the grounds that messages accepted by these
2385 statements are most likely to be submissions.
2387 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2389 (1) Generate an error if the third argument for the ${prvs expansion is
2392 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2395 (3) Reset the variables that are obtained from the first argument of
2396 ${prvscheck and used in the second argument before leaving the code,
2397 because their memory is reclaimed, so using them afterwards may do
2400 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2401 one (it's much easier than Tom thought :-).
2403 (5) Because of (4), we can now allow for the use of $prvscheck_result
2404 inside the third argument.
2406 PH/03 For some reason, the default setting of PATH when running a command from
2407 a pipe transport was just "/usr/bin". I have changed it to
2410 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2411 anything to be listed in the output from -bV.
2413 PH/05 When a filter generated an autoreply, the entire To: header line was
2414 quoted in the delivery log line, like this:
2416 => >A.N.Other <ano@some.domain> <original@ddress> ...
2418 This has been changed so that it extracts the operative address. There
2419 may be more than one such address. If so, they are comma-separated, like
2422 => >ano@some.domain,ona@other.domain <original@ddress> ...
2424 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2425 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2426 not being looked up in the rDNS to get a host name, Exim was showing the
2427 IP address twice in Received: lines, even though the IP addresses were
2428 identical. For example:
2430 Received: from [1.2.3.4] (helo=[1.2.3.4])
2432 However, if the real host name was known, it was omitting the HELO data
2433 if it matched the actual IP address. This has been tidied up so that it
2434 doesn't show the same IP address twice.
2436 PH/07 When both +timestamp and +memory debugging was on, the value given by
2437 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2438 called by the debug printing, thereby overwriting the timestamp buffer.
2439 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2441 PH/08 When the original message was included in an autoreply transport, it
2442 always said "this is a copy of the message, including all the headers",
2443 even if body_only or headers_only was set. It now gives an appropriate
2446 PH/09 Applied a patch from the Sieve maintainer which:
2448 o fixes some comments
2449 o adds the (disabled) notify extension core
2450 o adds some debug output for the result of if/elsif tests
2451 o points to the current vacation draft in the documentation
2452 and documents the missing references header update
2456 o fixes a bug in processing the envelope test (when testing
2457 multiple envelope elements, the last element determinted the
2460 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2461 Electronic Mail") by including:
2463 Auto-submitted: auto-generated
2465 in the messages that it generates (bounce messages and others, such as
2466 warnings). In the case of bounce messages for non-SMTP mesages, there was
2467 also a typo: it was using "Auto_submitted" (underscore instead of
2468 hyphen). Since every message generated by Exim is necessarily in response
2469 to another message, thes have all been changed to:
2471 Auto-Submitted: auto-replied
2473 in accordance with these statements in the RFC:
2475 The auto-replied keyword:
2477 - SHOULD be used on messages sent in direct response to another
2478 message by an automatic process,
2480 - MUST NOT be used on manually-generated messages,
2482 - MAY be used on Delivery Status Notifications (DSNs) and Message
2483 Disposition Notifications (MDNs),
2485 - MUST NOT be used on messages generated by automatic or periodic
2486 processes, except for messages which are automatic responses to
2489 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2490 to the default Received: header definition.
2492 PH/12 Added log selector acl_warn_skipped (default on).
2494 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2495 variables because (a) they are in the wrong storage pool and (b) even if
2496 they were copied, it wouldn't work properly because of the caching.
2498 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2499 checking when decoding. Apparently there are clients that generate
2500 overlong encoded strings. Why am I not surprised?
2502 PH/15 If the first argument of "${if match_address" was not empty, but did not
2503 contain an "@" character, Exim crashed. Now it writes a panic log message
2504 and treats the condition as false.
2506 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2508 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2509 extension "envelope-auth". The code is finished and in agreement with
2510 other implementations, but there is no documentation so far and in fact,
2511 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2512 not changing the active code.
2514 Print executed "if" and "elsif" statements when debugging is used. This
2515 helps a great deal to understand what a filter does.
2517 Document more things not specified clearly in RFC3028. I had all this
2518 sorted out, when out of a sudden new issues came to my mind. Oops."
2520 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2523 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2524 canonical form (as documented). However, after a host name lookup from
2525 the IP address, check_host() was doing a simple string comparison with
2526 addresses acquired from the DNS when checking that the found name did
2527 have the original IP as one of its addresses. Since any found IPv6
2528 addresses are likely to be in abbreviated form, the comparison could
2529 fail. Luckily, there already exists a function for doing the comparison
2530 by converting both addresses to binary, so now that is used instead of
2531 the text comparison.
2533 PH/20 There was another similar case to PH/19, when a complete host name was
2534 given in a host list; looking up its IP address could give an abbreviated
2535 form, whereas the current host's name might or might not be abbreviated.
2536 The same fix has been applied.
2542 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2543 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2546 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2547 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2549 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2551 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2552 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2553 Also, for "Auto-Submitted", treat anything other than "no" as
2554 non-personal, in accordance with RFC 3834. (Previously it treated
2555 anything starting "auto-" as non-personal.)
2557 TF/01 The control=submission/name=... option had a problem with syntax
2558 errors if the name included a slash character. The /name= option
2559 now slurps the rest of the string, so it can include any characters
2560 but it must come last in the list of options (after /sender_retain
2563 PH/05 Some modifications to the interface to the fake nameserver for the new
2571 TK/01 Added the "success_on_redirect" address verification option. See
2572 NewStuff for rationale and an example.
2574 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2576 PH/02 Patch to exigrep to allow it to work on syslog lines.
2578 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2579 fread() to skip over the body file's header line, because in Cygwin the
2580 header line is locked and is inaccessible.
2582 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2583 co-exist for some time) to make it clear that it is the Exim ID that is
2584 referenced, not the Message-ID: header line.
2586 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2587 string_format() function, because snprintf() does not exist on all
2590 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2591 ${stat: expansion item.
2593 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2594 protocol synchronization error", to keep the pedants happy.
2596 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2597 well as for IRIX systems, when gcc is being used. See the host.c source
2600 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2602 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2605 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2606 transport and to the smtp transport in LMTP mode.
2608 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2610 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2611 run a filter in a subprocess. This could lead to confusion in subsequent
2612 lookups in the parent process. There should also be a search_tidyup() at
2613 the end of the subprocess.
2615 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2616 only if the host matched helo_try_verify_hosts, which caused the
2617 verification to occur when the EHLO/HELO command was issued. The ACL just
2618 tested the remembered result. Now, if a previous verification attempt has
2619 not happened, "verify = helo" does it there and then.
2621 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2623 TK/03 Fix log output including CR from clamd.
2625 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2626 exist provoked a memory error which could cause a segfault.
2628 PH/15 Installed PCRE 6.2
2630 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2632 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2633 of the problem. Specifically, suggested +O2 rather than +O1 for the
2636 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2638 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2639 record, the port was not correctly logged when the outgoing_port log
2640 selector was set (it logged the transort's default port).
2642 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2643 fallback_hosts, and "hosts" in the smtp transport.
2645 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2646 host errors such as "Connection refused".
2648 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2649 authentication with radiusclient 0.4.9:
2651 - Error returned from rc_read_config was caught wrongly
2652 - Username/password not passed on to radius server due to wrong length.
2654 The presumption is that some radiusclient API changes for 4.51/PH/17
2655 were not taken care of correctly. The code is still untested by me (my
2656 Linux distribution still has 0.3.2 of radiusclient), but it was
2657 contributed by a Radius user.
2659 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2660 expanding the "port" option of the smtp transport.
2662 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2663 while reading a MIME header. Thanks to Tom Hughes for a patch.
2665 PH/24 Include config.h inside local_scan.h so that configuration settings are
2668 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2669 that in an ACL for RCPT (for example), you can examine exactly what was
2672 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2673 commands, but it was not correctly comparing the address with the actual
2674 client host address. Thus, it would show the EHLO address in Received:
2675 header lines when this was not necessary.
2677 PH/27 Added the % operator to ${eval:}.
2679 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2680 it should be ignoring failures (because with -C, for example, it has lost
2681 privilege). It wasn't ignoring creation failures other than "already
2684 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2687 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2688 because an input line was too long, either on its own, or by virtue of
2689 too many continuations, the temporary file was not being removed, and the
2690 return code was incorrect.
2692 PH/31 Missing "BOOL" in function definition in filtertest.c.
2694 PH/32 Applied Sieve patches from the maintainer.
2696 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2698 PH/33 Added "verify = not_blind".
2700 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2701 Local/Makefile (with some defaults set). These are used in built scripts
2702 such as exicyclog, but they have never been used in the exim_install
2703 script (though there are many overriding facilities there). I have
2704 arranged that the exim_install script now takes note of these two
2707 PH/35 Installed configuration files for Dragonfly.
2709 PH/36 When a locally submitted message by a trusted user did not contain a
2710 From: header, and the sender address was obtained from -f or from an SMTP
2711 MAIL command, and the trusted user did not use -F to supply a sender
2712 name, $originator_name was incorrectly used when constructing a From:
2713 header. Furthermore, $originator_name was used for submission mode
2714 messages from external hosts without From: headers in a similar way,
2715 which is clearly wrong.
2717 PH/37 Added control=suppress_local_fixups.
2719 PH/38 When log_selector = +received_sender was set, and the addition of the
2720 sender made the log line's construction buffer exactly full, or one byte
2721 less than full, an overflow happened when the terminating "\n" was
2724 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2725 when the result of a list match is failure because a DNS lookup failed.
2727 PH/40 RM_COMMAND is now used in the building process.
2729 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2730 the "build-* directories that it finds.
2732 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2733 address in a domain literal was a prefix of an interface address.
2735 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2736 when verifying a sender address, unless rewrite_headers is false.
2738 PH/44 Wrote a long comment about why errors_to addresses are verified as
2739 recipients, not senders.
2741 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2742 the ratelimit ACL was added.
2744 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2746 PH/46 Added extra information about PostgreSQL errors to the error string.
2748 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2749 avoiding the need to install special zones in a real server. This is
2750 backwards compatible; if it can't find the fake resolver, it drops back.
2751 Thus, both old and new test suites can be run.
2753 TF/02 Added util/ratelimit.pl
2755 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2756 clock is set back in time.
2758 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2759 Candler <B.Candler@pobox.com>.
2761 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2762 OK for addresses that are the result of redirections.
2764 PH/48 A number of further additions for the benefit of the new test suite,
2765 including a fake gethostbyname() that interfaces to the fake DNS resolver
2768 TF/06 The fix for widen_domains has also been applied to qualify_single and
2769 search_parents which are the other dnslookup options that can cause
2772 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2775 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2777 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2778 service" if it encountered an MX record with an empty target hostname.
2779 The message is now "an MX or SRV record indicated no SMTP service".
2781 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2782 if the DNS of the sending site is misconfigured. This is quite a
2783 common situation. This change restores the behaviour of treating a
2784 helo verification defer as a failure.
2786 PH/51 If self=fail was set on a router, the bounce message did not include the
2787 actual error message.
2793 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2795 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2796 command itself ended in error, the underlying message about the transport
2797 filter timeout was being overwritten with the pipe command error. Now the
2798 underlying error message should be appended to the second error message.
2800 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2802 PH/02 Reduce the amount of output that "make" produces by default. Full output
2803 can still be requested.
2805 PH/03 The warning log line about a condition test deferring for a "warn" verb
2806 was being output only once per connection, rather than after each
2807 occurrence (because it was using the same function as for successful
2808 "warn" verbs). This seems wrong, so I have changed it.
2810 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2811 it should not have, which might have caused a crash in the right
2812 circumstances, but probably never did.
2814 PH/04 Installed a modified version of Tony Finch's patch to make submission
2815 mode fix the return path as well as the Sender: header line, and to
2816 add a /name= option so that you can make the user's friendly name appear
2819 TF/03 Added the control = fakedefer ACL modifier.
2821 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2822 Mark Lowes for thorough testing.
2824 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2826 TK/03 Merged latest SRS patch from Miles Wilton.
2828 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2829 with the definition in sysexits.h (which is #included earlier).
2830 Fortunately, Exim does not actually use EX_OK. The code used to try to
2831 preserve the sysexits.h value, by assumimg that macro definitions were
2832 scanned for macro replacements. I have been disabused of this notion,
2833 so now the code just undefines EX_OK before #including unistd.h.
2835 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2836 in the smtp transport. When a block could not be written in a single
2837 write() function, the timeout was being re-applied to each part-write.
2838 This seems wrong - if the receiver was accepting one byte at a time it
2839 would take for ever. The timeout is now adjusted when this happens. It
2840 doesn't have to be particularly precise.
2842 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2843 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2845 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2846 without a domain if the check to <postmaster@domain> fails.
2848 SC/01 Eximstats: added -xls and the ability to specify output files
2849 (patch written by Frank Heydlauf).
2851 SC/02 Eximstats: use FileHandles for outputing results.
2853 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2855 SC/04 Eximstats: fixed display of large numbers with -nvr option
2857 SC/05 Eximstats: fixed merging of reports with empty tables.
2859 SC/06 Eximstats: added the -include_original_destination flag
2861 SC/07 Eximstats: removed tabs and trailing whitespace.
2863 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2865 TK/06 MBOX spool code: Add real "From " MBOX separator line
2866 so the .eml file is really in mbox format (even though
2867 most programs do not really care). Patch from Alex Miller.
2869 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2870 The latter is generated from $received_to and is only set if the
2871 message has one envelope recipient. SA can use these headers,
2872 obviously out-of-the-box. Patch from Alex Miller.
2874 PH/08 The ${def test on a variable was returning false if the variable's
2875 value was "0", contrary to what the specification has always said!
2876 The result should be true unless the variable is empty.
2878 PH/09 The syntax error of a character other than { following "${if
2879 def:variable_name" (after optional whitespace) was not being diagnosed.
2880 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2881 accidental colon was present, for example, could give incorrect results.
2883 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2884 result is used (not including appendfile, where other changes are about
2887 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2888 This involved changing a lot of size variables from int to off_t. It
2889 should work with maildirs and everything.
2891 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2892 spamd dying while we are connected to it.
2894 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2897 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2898 function for BDB 4.3.
2900 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2902 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2903 the sources for PCRE have been re-arranged and more files are now
2906 PH/15 The code I had for printing potentially long long variables in PH/11
2907 above was not the best (it lost precision). The length of off_t variables
2908 is now inspected at build time, and an appropriate printing format (%ld
2909 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2910 to be "long long int" or "long int". This is needed for the internal
2911 formatting function string_vformat().
2913 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2914 the configuration file to be ":syslog", then the script "guesses" where
2915 the logs files are, rather than using the compiled in default. In our
2916 case the guess is not the same as the compiled default, so the script
2917 suddenly stopped working when I started to use syslog. The patch checks
2918 to see if log_file_path is "". If so, it attempts to read it from exim
2919 with no configuration file to get the compiled in version, before it
2920 falls back to the previous guessing code."
2922 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2923 implementing BATV in an Exim configuration. See NewStuff for the gory
2926 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2927 Makefile that are specific to HP-UX.
2929 PH/18 If the "use_postmaster" option was set for a recipient callout together
2930 with the "random" option, the postmaster address was used as the MAIL
2931 FROM address for the random test, but not for the subsequent recipient
2932 test. It is now used for both.
2934 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2935 patch removes a few documentation additions to RFC 3028, because the
2936 latest draft now contains them. It adds the new en;ascii-case comparator
2937 and a new error check for 8bit text in MIME parts. Comparator and
2938 require names are now matched exactly. I enabled the subaddress
2939 extension, but it is not well tested yet (read: it works for me)."
2941 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2942 rework some of the code of TK/09 above to avoid the hardwired use of
2943 "%lld" and "long long". Replaced the call to snprintf() with a call to
2946 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2947 records point to non-existent hosts", "retry timeout exceeded", and
2948 "retry time not reached for any host after a long failure period".
2950 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2951 experimental DomainKeys support:
2953 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2954 (2) On an error such as an illegally used "control", the wrong name for
2955 the control was given.
2957 These problems did NOT occur unless DomainKeys support was compiled.
2959 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2961 PH/24 Added ${if match_ip condition.
2963 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2964 databases so that it will be absolutely obvious if a crash occurs in the
2965 DB library. This is a regular occurrence (often caused by mis-matched
2968 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2969 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2970 picked up on a user's system that detects such things. There doesn't seem
2971 to be a gcc warning option for this - only an attribute that has to be
2972 put on the function's prototype. It seems that in Fedora Core 4 they have
2973 set this on a number of new functions. No doubt there will be more in due
2976 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2977 specify a transport. However, if an address that was verified by such a
2978 router was the subject of a callout, Exim crashed because it tried to
2979 read the rcpt_include_affixes from the non-existent transport. Now it
2980 just assumes that the setting of that option is false. This bug was
2981 introduced by 4.51/PH/31.
2983 PH/28 Changed -d+all to exclude +memory, because that information is very
2984 rarely of interest, but it makes the output a lot bigger. People tend to
2985 do -d+all out of habit.
2987 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2988 code in os-type was giving problems when libc.so lives in lib64, like on
2991 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2992 aren't the modern standard, and it seems that some systems' include files
2993 don't always have them. Exim was already checking for some of the newer
2994 ones like T_AAAA, and defining it itself. I've added checks for all the
2995 record types that Exim uses.
2997 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2998 not automatically generating a new one, as it is supposed to. This
2999 prevented TLS from working. If the file did exist, but contained invalid
3000 data, a new version was generated, as expected. It was only the case of a
3001 non-existent file that was broken.
3003 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3004 with a change in libdomainkeys > 0.64.
3006 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3007 from DNS. If the selector record carries the flag, it now has
3008 precedence over the domain-wide flag.
3010 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3012 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3013 the use of a port name that isn't defined in /etc/services) occurred, the
3014 message was deferred as in a normal delivery, and thus remained on the
3015 spool, instead of being failed because of the mua_wrapper setting. This
3016 is now fixed, and I tidied up some of the mua_wrapper messages at the
3019 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3020 the messages in a hash of arrays rather than using individual hashes.
3021 This is a bit cleaner and results in dramatic memory savings, albeit
3022 at a slight CPU cost.
3024 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3025 as requested by Marc Sherman.
3027 SC/10 Eximstats: added histograms for user specified patterns as requested
3030 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3032 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3033 fopen() in the content-scanning modules that did not already have it.
3039 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3040 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3042 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3044 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3046 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3047 to test Sieve filters that use "vacation".
3049 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3050 that changes the way the GnuTLS parameters are stored in the cache file.
3051 The new format can be generated externally. For backward compatibility,
3052 if the data in the cache doesn't make sense, Exim assumes it has read an
3053 old-format file, and it generates new data and writes a new file. This
3054 means that you can't go back to an older release without removing the
3057 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3058 work if there are any delivery delays because "one_time" forces the
3059 parent to be marked "delivered", so its unseen clone is never tried
3060 again. For this reason, Exim now forbids the simultaneous setting of
3063 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3064 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3065 entry if you want to know the details. The fix, however, neglected to
3066 consider the case when local delivery batching is involved. The test for
3067 "previously delivered" was not happening when checking to see if an
3068 address could be batched with a previous (undelivered) one; under
3069 certain circumstances this could lead to multiple deliveries to the same
3072 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3073 in its include files, and this causes problems building Exim.
3075 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3076 header_syntax) but Exim was just ignoring anything given after a slash.
3077 In particular, this caused confusion with an attempt to use "verify =
3078 reverse_host_lookup/defer_ok". An error is now given when options are
3079 supplied for verify items that do not have them. (Maybe reverse_host_
3080 lookup should have a defer_ok option, but that's a different point.)
3082 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3083 defined by RFC 821) to 2048, because there were problems with some AUTH
3084 commands, and RFC 1869 says the size should be increased for extended
3085 SMTP commands that take arguments.
3087 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3090 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3091 "unknown" error; now it says that the functionality isn't in the binary.
3093 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3094 an address' error message when a string expansion fails (syntax or
3095 whatever). Otherwise the password may appear in the log. Following change
3096 PH/42 below, there is no longer a chance of it appearing in a bounce
3099 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3101 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3102 of its addresses was ever tried. (Bugzilla bug #2.)
3104 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3105 the result incorrectly in the debug output. (It correctly added a newline
3106 to what was transported.)
3108 TF/01 Added $received_time.
3110 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3111 commented out examples of how to interface to a virus scanner and to
3112 SpamAssassin. Also added commented examples of av_scanner and
3113 spamd_address settings.
3115 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3116 and controls are allowed in which ACLs. There were a couple of minor
3117 errors. Some of the entries in the conditions table (which is a table of
3118 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3119 negation of where the condition IS allowed.
3121 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3123 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3124 header file does not have a version number, so I've had to invent a new
3125 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3126 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3127 radiusclient), but it was contributed by a Radius user.
3129 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3130 files or directories, for OpenSSL.
3132 PH/19 When an Exim process that is running as root has to create an Exim log
3133 file, it does so in a subprocess that runs as exim:exim so as to get the
3134 ownership right at creation (otherwise, other Exim processes might see
3135 the file with the wrong ownership). There was no test for failure of this
3136 fork() call, which would lead to the process getting stuck as it waited
3137 for a non-existent subprocess. Forks do occasionally fail when resources
3138 run out. I reviewed all the other calls to fork(); they all seem to check
3141 PH/20 When checking for unexpected SMTP input at connect time (before writing
3142 the banner), Exim was not dealing correctly with a non-positive return
3143 from the read() function. If the client had disconnected by this time,
3144 the result was a log entry for a synchronization error with an empty
3145 string after "input=" when read() returned zero. If read() returned -1
3146 (an event I could not check), uninitialized data bytes were printed.
3147 There were reports of junk text (parts of files, etc) appearing after
3150 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3152 PH/22 Added support for macro redefinition, and (re)definition in between
3153 driver and ACL definitions.
3155 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3156 forgetting to use the resulting value; it was using the unexpanded value.
3158 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3159 hadn't been configured. The fix is from Juergen Kreileder, who
3160 understands it better than I do:
3162 "Here's what I see happening with three configured cyrus_sasl
3163 authenticators configured (plain, login, cram-md5):
3165 On startup auth_cyrus_sasl_init() gets called for each of these.
3166 This means three calls to sasl_listmech() without a specified mech_list.
3167 => SASL tests which mechs of all available mechs actually work
3168 => three warnings about OTP not working
3169 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3171 With the patch, sasl_listmech() also gets called three times. But now
3172 SASL's mech_list option is set to the server_mech specified in the the
3173 authenticator. Or in other words, the answer from sasl_listmech()
3174 gets limited to just the mech you're testing for (which is different
3176 => the return list contains just 'plain' or 'login', 'cram-md5' or
3177 nothing depending on the value of ob->server_mech.
3179 I've just tested the patch: Authentication still works fine,
3180 unavailable mechs specified in the exim configuration are still
3181 caught, and the auth.log warnings about OTP are gone."
3183 PH/25 When debugging is enabled, the contents of the command line are added
3184 to the debugging output, even when log_selector=+arguments is not
3187 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3188 answer is "GNU", and only if the return is "GNU/something" is the answer
3191 PH/27 $acl_verify_message is now set immediately after the failure of a
3192 verification in an ACL, and so is available in subsequent modifiers. In
3193 particular, the message can be preserved by coding like this:
3195 warn !verify = sender
3196 set acl_m0 = $acl_verify_message
3198 Previously, $acl_verify_message was set only while expanding "message"
3199 and "log_message" when a very denied access.
3201 PH/28 Modified OS/os.c-Linux with
3203 -#ifndef OS_LOAD_AVERAGE
3204 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3206 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3207 nomenclature these days.)
3209 PH/29 Installed patch from the Sieve maintainer that adds the options
3210 sieve_useraddress and sieve_subaddress to the redirect router.
3212 PH/30 In these circumstances:
3213 . Two addresses routed to the same list of hosts;
3214 . First host does not offer TLS;
3215 . First host accepts first address;
3216 . First host gives temporary error to second address;
3217 . Second host offers TLS and a TLS session is established;
3218 . Second host accepts second address.
3219 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3220 and peerdn, if requested) that were in fact used only for the second
3223 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3224 attention to any local part prefix or suffix that was matched by the
3225 router that accepted the address. It now behaves in the same way as it
3226 does for delivery: the affixes are removed from the local part unless
3227 rcpt_include_affixes is set on the transport.
3229 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3230 timeout during the DATA phase of an incoming message.
3232 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3233 applied a patch sanctioned by the Sieve maintainer.
3235 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3236 the uid or gid is negative. A case of a negative gid caused this to be
3237 noticed. The fix allows for either to be negative.
3239 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3240 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3243 PH/36 The change PH/12 above was broken. Fixed it.
3245 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3246 the grounds that routing the same address twice would always produce the
3247 same answer. This might have been true once, but it is certainly no
3248 longer true now. Routing a child address may depend on the previous
3249 routing that produced that child. Some complicated redirection strategies
3250 went wrong when messages had multiple recipients, and made Exim's
3251 behaviour dependent on the order in which the addresses were given.
3253 I have moved the duplicate checking until after the routing is complete.
3254 Exim scans the addresses that are assigned to local and remote
3255 transports, and removes any duplicates. This means that more work will be
3256 done, as duplicates will always all be routed, but duplicates are
3257 presumably rare, so I don't expect this is of any significance.
3259 For deliveries to pipes, files, and autoreplies, the duplicate checking
3260 still happens during the routing process, since they are not going to be
3263 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3264 It corrects a timeout issue with spamd. This is Ian's comment: "The
3265 background is that sometimes spamd either never reads data from a
3266 connection it has accepted, or it never writes response data. The exiscan
3267 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3268 blindly assumes that writes won't block so it may never time out."
3270 PH/39 Allow G after quota size as well as K and M.
3272 PH/40 The value set for $authenticated_id in an authenticator may not contain
3273 binary zeroes or newlines because the value is written to log lines and
3274 to spool files. There was no check on this. Now the value is run through
3275 the string_printing() function so that such characters are converted to
3276 printable escape sequences.
3278 PH/41 $message_linecount is a new variable that contains the total number of
3279 lines in the message. Compare $body_linecount, which is the count for the
3282 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3283 bounce and delay warning messages, except in certain special cases, which
3286 (a) An SMTP error message from a remote host;
3287 (b) A message specified in a :fail: redirection;
3288 (c) A message specified in a "fail" command in a system filter;
3289 (d) A message specified in a FAIL return from the queryprogram router;
3290 (e) A message specified by the cannot_route_message router option.
3292 In these cases only, Exim does include the error details in bounce and
3293 warning messages. There are also a few cases where bland messages such
3294 as "unrouteable address" or "local delivery error" are given.
3296 PH/43 $value is now also set for the "else" part of a ${run expansion.
3298 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3299 being worked on, but at least Exim now implements the latest version to
3302 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3303 process to complete was treated as a delivery failure, a timeout while
3304 writing the message to the pipe was logged, but erroneously treated as a
3305 successful delivery. Such timeouts include transport filter timeouts. For
3306 consistency with the overall process timeout, these timeouts are now
3307 treated as errors, giving rise to delivery failures by default. However,
3308 there is now a new Boolean option for the pipe transport called
3309 timeout_defer, which, if set TRUE, converts the failures into defers for
3310 both kinds of timeout. A transport filter timeout is now identified in
3313 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3314 systems where "make" and "gmake" are different, calling "gmake" at top
3315 level broke things. I've arranged for the value of $(MAKE) to be passed
3316 from the Makefile to this script so that it can call the same version of
3320 A note about Exim versions 4.44 and 4.50
3321 ----------------------------------------
3323 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3324 changes of various kinds. As a consequence, a big documentation update was
3325 needed. This delayed the release for rather longer than seemed good, especially
3326 in the light of a couple of (minor) security issues. Therefore, the changes
3327 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3328 release. So 4.44 and 4.50 are in effect two different branches that both start
3331 I have left the 4.50 change log unchanged; it contains all the changes since
3332 4.43. The change log for 4.44 is below; many of its items are identical to
3333 those for 4.50. This seems to be the most sensible way to preserve the
3334 historical information.
3340 1. Minor wording change to the doc/README.SIEVE file.
3342 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3343 computation of the current number of files was incorrect.
3345 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3346 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3349 4. Give more explanation in the error message when the command for a transport
3350 filter fails to execute.
3352 5. There are several places where Exim runs a non-Exim command in a
3353 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3354 was being done only for the command run by the queryprogram router. It is
3355 now done for all such subprocesses. The other cases are: ${run, transport
3356 filters, and the commands run by the lmtp and pipe transports.
3358 6. Added CONFIGURE_GROUP build-time option.
3360 7. Some older OS have a limit of 256 on the maximum number of file
3361 descriptors. Exim was using setrlimit() to set 1000 as a large value
3362 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3363 systems. I've change it so that if it can't get 1000, it tries for 256.
3365 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3366 was an oversight, and furthermore, ever since the addition of extra
3367 controls (e.g. 4.43/32), the checks on when to allow different forms of
3368 "control" were broken. There should now be diagnostics for all cases when a
3369 control that does not make sense is encountered.
3371 9. Added the /retain_sender option to "control=submission".
3373 10. $recipients is now available in the predata ACL (oversight).
3375 11. Tidy the search cache before the fork to do a delivery from a message
3376 received from the command line. Otherwise the child will trigger a lookup
3377 failure and thereby defer the delivery if it tries to use (for example) a
3378 cached ldap connection that the parent has called unbind on.
3380 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3381 of $address_data from the recipient verification was clobbered by the
3382 sender verification.
3384 13. The value of address_data from a sender verification is now available in
3385 $sender_address_data in subsequent conditions in the ACL statement.
3387 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3389 15. Added a new option "connect=<time>" to callout options, to set a different
3392 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3393 was its contents. (It was OK if the option was not defined at all.)
3395 17. A "Completed" log line is now written for messages that are removed from
3396 the spool by the -Mrm option.
3398 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3399 information about exactly what failed.
3401 19. Added -dd to debug only the daemon process.
3403 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3404 handles timeouts, both on the server side and network timeouts. Renamed the
3405 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3407 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3408 It is now set to "smtps".
3410 22. $host_address is now set to the target address during the checking of
3411 ignore_target_hosts.
3413 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3414 being passed; this would have caused $sender_host_name to have been used if
3415 matching the list had actually called for a host name (not very likely,
3416 since this list is usually IP addresses). A host name is now passed as
3419 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3420 code that specifies a non-restarting handler (typically sigaction() in
3421 modern systems) in an attempt to fix a rare and obscure crash bug.
3423 25. Narrowed the window for a race in the daemon that could cause it to ignore
3424 SIGCHLD signals. This is not a major problem, because they are used only to
3425 wake it up if nothing else does.
3427 26. A malformed maildirsize file could cause Exim to calculate negative values
3428 for the mailbox size or file count. Odd effects could occur as a result.
3429 The maildirsize information is now recalculated if the size or filecount
3432 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3433 support for a long time. Removed HAVE_SYS_VFS_H.
3435 28. Installed the latest version of exipick from John Jetmore.
3437 29. In an address list, if the pattern was not a regular expression, an empty
3438 subject address (from a bounce message) matched only if the pattern was an
3439 empty string. Non-empty patterns were not even tested. This was the wrong
3440 because it is perfectly reasonable to use an empty address as part of a
3441 database query. An empty address is now tested by patterns that are
3442 lookups. However, all the other forms of pattern expect the subject to
3443 contain a local part and a domain, and therefore, for them, an empty
3444 address still always fails if the pattern is not itself empty.
3446 30. Exim went into a mad DNS loop when attempting to do a callout where the
3447 host was specified on an smtp transport, and looking it up yielded more
3448 than one IP address.
3450 31. Re-factored the code for checking spool and log partition space into a
3451 function that finds that data and another that does the check. The former
3452 is then used to implement four new variables: $spool_space, $log_space,
3453 $spool_inodes, and $log_inodes.
3455 32. The RFC2047 encoding function was originally intended for short strings
3456 such as real names; it was not keeping to the 75-character limit for
3457 encoded words that the RFC imposes. It now respects the limit, and
3458 generates multiple encoded words if necessary. To be on the safe side, I
3459 have increased the buffer size for the ${rfc2047: expansion operator from
3462 33. It is now permitted to omit both strings after an "if" condition; if the
3463 condition is true, the result is "true". As before, when the second string
3464 is omitted, a false condition yields an empty string. This makes it less
3465 cumbersome to write custom ACL and router conditions.
3467 34. Failure to deliver a bounce message always caused it to be frozen, even if
3468 there was an errors_to setting on the router. The errors_to setting is now
3471 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3472 canonical form (fully expanded) before being placed in
3473 $sender_host_address.
3475 36. The table in the code that translates DNS record types into text (T_A to
3476 "A" for instance) was missing entries for NS and CNAME. It is just possible
3477 that this could have caused confusion if both these types were looked up
3478 for the same domain, because the text type is used as part of Exim's
3479 per-process caching. But the chance of anyone hitting this buglet seems
3482 37. The dnsdb lookup has been extended in a number of ways.
3484 (1) There is a new type, "zns", which walks up the domain tree until it
3485 finds some nameserver records. It should be used with care.
3487 (2) There is a new type, "mxh", which is like "mx" except that it returns
3488 just the host names, not the priorities.
3490 (3) It is now possible to give a list of domains (or IP addresses) to be
3491 looked up. The behaviour when one of the lookups defers can be
3492 controlled by a keyword.
3494 (4) It is now possible to specify the separator character for use when
3495 multiple records are returned.
3497 38. The dnslists ACL condition has been extended: it is now possible to supply
3498 a list of IP addresses and/or domains to be looked up in a particular DNS
3501 39. Added log_selector=+queue_time_overall.
3503 40. When running the queue in the test harness, wait just a tad after forking a
3504 delivery process, to get repeatability of debugging output.
3506 41. Include certificate and key file names in error message when GnuTLS fails
3507 to set them up, because the GnuTLS error message doesn't include the name
3508 of the failing file when there is a problem reading it.
3510 42. Allow both -bf and -bF in the same test run.
3512 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3514 44. The "Exiscan patch" is now merged into the mainline Exim source.
3516 45. Sometimes the final signoff response after QUIT could fail to get
3517 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3518 < 0 before doing a fflush(). This bug looks as though it goes back to the
3519 introduction of TLS in release 3.20, but "sometimes" must have been rare
3520 because the tests only now provoked it.
3522 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3523 (this can affect the format of dates).
3525 47. exim_tidydb, when checking for the continued existence of a message for
3526 which it has found a message-specific retry record, was not finding
3527 messages that were in split spool directories. Consequently, it was
3528 deleting retry records that should have stayed in existence.
3530 48. Steve fixed some bugs in eximstats.
3532 49. The SPA authentication driver was not abandoning authentication and moving
3533 on to the next authenticator when an expansion was forced to fail,
3534 contradicting the general specification for all authenticators. Instead it
3535 was generating a temporary error. It now behaves as specified.
3537 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3538 (the order specifies the preference for clients). The order is now AES256,
3539 AES128, 3DES, ARCFOUR128.
3541 51. Small patch to Sieve code - explicitly set From: when generating an
3544 52. Exim crashed if a remote delivery caused a very long error message to be
3545 recorded - for instance if somebody sent an entire SpamAssassin report back
3546 as a large number of 550 error lines. This bug was coincidentally fixed by
3547 increasing the size of one of Exim's internal buffers (big_buffer) that
3548 happened as part of the Exiscan merge. However, to be on the safe side, I
3549 have made the code more robust (and fixed the comments that describe what
3552 53. Now that there can be additional text after "Completed" in log lines (if
3553 the queue_time_overall log selector is set), a one-byte patch to exigrep
3554 was needed to allow it to recognize "Completed" as not the last thing in
3557 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3558 patch that reportedly fixes this has been added. I am not expert enough to
3559 create a test for it. This is what the patch creator wrote:
3561 "I found a little strange behaviour of ldap code when working with
3562 Windows 2003 AD Domain, where users was placed in more than one
3563 Organization Units. When I tried to give exim partial DN, the exit code
3564 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3565 But simultaneously result of request was absolutely normal ldap result,
3566 so I produce this patch..."
3568 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3569 REFERENCE, so I have modified the code to exclude the patch when that macro
3572 55. Some experimental protocols are using DNS PTR records for new purposes. The
3573 keys for these records are domain names, not reversed IP addresses. The
3574 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3575 leaves it alone. Component reversal etc. now happens only for IP addresses.
3578 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3580 57. Double the size of the debug message buffer (to 2048) so that more of very
3581 long debug lines gets shown.
3583 58. The exicyclog utility now does better if the number of log files to keep
3584 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3586 59. Two changes related to the smtp_active_hostname option:
3588 (1) $smtp_active_hostname is now available as a variable.
3589 (2) The default for smtp_banner uses $smtp_active_hostname instead
3590 of $primary_hostname.
3592 60. The host_aton() function is supposed to be passed a string that is known
3593 to be a valid IP address. However, in the case of IPv6 addresses, it was
3594 not checking this. This is a hostage to fortune. Exim now panics and dies
3595 if the condition is not met. A case was found where this could be provoked
3596 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3597 components; fortuitously, this particular loophole had already been fixed
3598 by change 4.50/55 above.
3600 If there are any other similar loopholes, the new check in host_aton()
3601 itself should stop them being exploited. The report I received stated that
3602 data on the command line could provoke the exploit when Exim was running as
3603 exim, but did not say which command line option was involved. All I could
3604 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3605 running as the user.
3608 61. There was a buffer overflow vulnerability in the SPA authentication code
3609 (which came originally from the Samba project). I have added a test to the
3610 spa_base64_to_bits() function which I hope fixes it.
3613 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3614 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3616 63. The daemon start-up calls getloadavg() while still root for those OS that
3617 need the first call to be done as root, but it missed one case: when
3618 deliver_queue_load_max is set with deliver_drop_privilege. This is
3619 necessary for the benefit of the queue runner, because there is no re-exec
3620 when deliver_drop_privilege is set.
3622 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3623 This has been fixed.
3625 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3626 was in use, was not putting the data itself into the right store pool;
3627 consequently, it could be overwritten for a subsequent message in the same
3628 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3631 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3633 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3634 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3635 the function were treating the return as a boolean value, which happened to
3636 work because 0=false and not-0=true, but is not correct code.
3638 68. The host_aton() function was not handling scoped IPv6 addresses (those
3639 with, for example, "%eth0" on the end) correctly.
3641 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3642 negated items (that is, ~something) in unsigned ints. Some compilers
3643 apparently mutter when there is no cast.
3645 70. If an address verification called from an ACL failed, and did not produce a
3646 user-specific message (i.e. there was only a "system" message), nothing was
3647 put in $acl_verify_message. In this situation, it now puts the system
3650 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3651 session; change 4.31/43 added the unwanted input to the log line - except
3652 that it did not do this in the start of session case. It now does.
3654 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3655 This is wrong and can cause the other end to generate a synchronization
3656 error if it is another Exim or anything else that does the synchronization
3657 check. A QUIT command is no longer sent after a timeout.
3659 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3660 during host lookups.
3662 74. The defer_ok option of callout verification was not working if it was used
3663 when verifying addresses in header lines, that is, for this case:
3665 verify = header_sender/callout=defer_ok
3667 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3668 those file descriptors could be used for SMTP connections. If anything
3669 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3670 could be sent to the SMTP client, causing chaos. The daemon now opens
3671 stdin, stdout, and stderr to /dev/null when it puts itself into the
3674 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3675 log by default. The user can override this with suitable Perl magic.
3677 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3678 the log message when discard triggers, was not working for the DATA ACL or
3679 for the non-SMTP ACL.
3681 78. Error message wording change in sieve.c.
3683 79. If smtp_accept_max_per_host was set, the number of connections could be
3684 restricted to fewer than expected, because the daemon was trying to set up
3685 a new connection before checking whether the processes handling previous
3686 connections had finished. The check for completed processes is now done
3687 earlier. On busy systems, this bug wouldn't be noticed because something
3688 else would have woken the daemon, and it would have reaped the completed
3691 80. If a message was submitted locally by a user whose login name contained one
3692 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3693 It caused a spool format error. I have fixed the spool reading code. A
3694 related problem was that the "from" clause in the Received: line became
3695 illegal because of the space(s). It is now covered by ${quote_local_part.
3697 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3700 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3702 83. Patch from Sieve maintainer to fix unterminated string problem in
3703 "vacation" handling.
3705 84. Some minor changes to the Linux configuration files to help with other
3706 OS variants using glibc.
3708 85. One more patch for Sieve to update vacation handling to latest spec.
3711 ----------------------------------------------------
3712 See the note above about the 4.44 and 4.50 releases.
3713 ----------------------------------------------------
3719 1. Change 4.43/35 introduced a bug that caused file counts to be
3720 incorrectly computed when quota_filecount was set in an appendfile
3723 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3724 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3727 3. Give more explanation in the error message when the command for a transport
3728 filter fails to execute.
3730 4. There are several places where Exim runs a non-Exim command in a
3731 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3732 was being done only for the command run by the queryprogram router. It is
3733 now done for all such subprocesses. The other cases are: ${run, transport
3734 filters, and the commands run by the lmtp and pipe transports.
3736 5. Some older OS have a limit of 256 on the maximum number of file
3737 descriptors. Exim was using setrlimit() to set 1000 as a large value
3738 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3739 systems. I've change it so that if it can't get 1000, it tries for 256.
3741 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3742 was an oversight, and furthermore, ever since the addition of extra
3743 controls (e.g. 4.43/32), the checks on when to allow different forms of
3744 "control" were broken. There should now be diagnostics for all cases when a
3745 control that does not make sense is encountered.
3747 7. $recipients is now available in the predata ACL (oversight).
3749 8. Tidy the search cache before the fork to do a delivery from a message
3750 received from the command line. Otherwise the child will trigger a lookup
3751 failure and thereby defer the delivery if it tries to use (for example) a
3752 cached ldap connection that the parent has called unbind on.
3754 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3755 of $address_data from the recipient verification was clobbered by the
3756 sender verification.
3758 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3759 was its contents. (It was OK if the option was not defined at all.)
3761 11. A "Completed" log line is now written for messages that are removed from
3762 the spool by the -Mrm option.
3764 12. $host_address is now set to the target address during the checking of
3765 ignore_target_hosts.
3767 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3768 being passed; this would have caused $sender_host_name to have been used if
3769 matching the list had actually called for a host name (not very likely,
3770 since this list is usually IP addresses). A host name is now passed as
3773 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3774 code that specifies a non-restarting handler (typically sigaction() in
3775 modern systems) in an attempt to fix a rare and obscure crash bug.
3777 15. Narrowed the window for a race in the daemon that could cause it to ignore
3778 SIGCHLD signals. This is not a major problem, because they are used only to
3779 wake it up if nothing else does.
3781 16. A malformed maildirsize file could cause Exim to calculate negative values
3782 for the mailbox size or file count. Odd effects could occur as a result.
3783 The maildirsize information is now recalculated if the size or filecount
3786 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3787 support for a long time. Removed HAVE_SYS_VFS_H.
3789 18. Updated exipick to current release from John Jetmore.
3791 19. Allow an empty sender to be matched against a lookup in an address list.
3792 Previously the only cases considered were a regular expression, or an
3795 20. Exim went into a mad DNS lookup loop when doing a callout where the
3796 host was specified on the transport, if the DNS lookup yielded more than
3799 21. The RFC2047 encoding function was originally intended for short strings
3800 such as real names; it was not keeping to the 75-character limit for
3801 encoded words that the RFC imposes. It now respects the limit, and
3802 generates multiple encoded words if necessary. To be on the safe side, I
3803 have increased the buffer size for the ${rfc2047: expansion operator from
3806 22. Failure to deliver a bounce message always caused it to be frozen, even if
3807 there was an errors_to setting on the router. The errors_to setting is now
3810 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3811 canonical form (fully expanded) before being placed in
3812 $sender_host_address.
3814 24. Updated eximstats to version 1.33
3816 25. Include certificate and key file names in error message when GnuTLS fails
3817 to set them up, because the GnuTLS error message doesn't include the name
3818 of the failing file when there is a problem reading it.
3820 26. Expand error message when OpenSSL has problems setting up cert/key files.
3823 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3824 (this can affect the format of dates).
3826 28. exim_tidydb, when checking for the continued existence of a message for
3827 which it has found a message-specific retry record, was not finding
3828 messages that were in split spool directories. Consequently, it was
3829 deleting retry records that should have stayed in existence.
3831 29. eximstats updated to version 1.35
3832 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3833 1.35 - bugfix such that pie charts by volume are generated correctly
3835 30. The SPA authentication driver was not abandoning authentication and moving
3836 on to the next authenticator when an expansion was forced to fail,
3837 contradicting the general specification for all authenticators. Instead it
3838 was generating a temporary error. It now behaves as specified.
3840 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3841 (the order specifies the preference for clients). The order is now AES256,
3842 AES128, 3DES, ARCFOUR128.
3844 31. Small patch to Sieve code - explicitly set From: when generating an
3847 32. Exim crashed if a remote delivery caused a very long error message to be
3848 recorded - for instance if somebody sent an entire SpamAssassin report back
3849 as a large number of 550 error lines. This bug was coincidentally fixed by
3850 increasing the size of one of Exim's internal buffers (big_buffer) that
3851 happened as part of the Exiscan merge. However, to be on the safe side, I
3852 have made the code more robust (and fixed the comments that describe what
3855 33. Some experimental protocols are using DNS PTR records for new purposes. The
3856 keys for these records are domain names, not reversed IP addresses. The
3857 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3858 leaves it alone. Component reversal etc. now happens only for IP addresses.
3861 34. The host_aton() function is supposed to be passed a string that is known
3862 to be a valid IP address. However, in the case of IPv6 addresses, it was
3863 not checking this. This is a hostage to fortune. Exim now panics and dies
3864 if the condition is not met. A case was found where this could be provoked
3865 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3866 components; fortuitously, this particular loophole had already been fixed
3867 by change 4.50/55 or 4.44/33 above.
3869 If there are any other similar loopholes, the new check in host_aton()
3870 itself should stop them being exploited. The report I received stated that
3871 data on the command line could provoke the exploit when Exim was running as
3872 exim, but did not say which command line option was involved. All I could
3873 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3874 running as the user.
3877 35. There was a buffer overflow vulnerability in the SPA authentication code
3878 (which came originally from the Samba project). I have added a test to the
3879 spa_base64_to_bits() function which I hope fixes it.
3882 36. The daemon start-up calls getloadavg() while still root for those OS that
3883 need the first call to be done as root, but it missed one case: when
3884 deliver_queue_load_max is set with deliver_drop_privilege. This is
3885 necessary for the benefit of the queue runner, because there is no re-exec
3886 when deliver_drop_privilege is set.
3888 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3889 was in use, was not putting the data itself into the right store pool;
3890 consequently, it could be overwritten for a subsequent message in the same
3891 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3894 38. Sometimes the final signoff response after QUIT could fail to get
3895 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3896 < 0 before doing a fflush(). This bug looks as though it goes back to the
3897 introduction of TLS in release 3.20, but "sometimes" must have been rare
3898 because the tests only now provoked it.
3904 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3905 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3906 Now it returns an integer. A number of calls were still expecting a T/F
3907 return. Fortuitously, in all cases, the tests worked in OK situations,
3908 which is the norm. However, things would have gone wrong on any write
3909 failures on the smtp file descriptor. This function is used when sending
3910 messages over SMTP and also when doing verify callouts.
3912 2. When Exim is called to do synchronous delivery of a locally submitted
3913 message (the -odf or -odi options), it no longer closes stderr before doing
3916 3. Implemented the mua_wrapper option.
3918 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3920 5. Implemented the functions header_remove(), header_testname(),
3921 header_add_at_position(), and receive_remove_recipient(), and exported them
3924 6. If an ACL "warn" statement specified the addition of headers, Exim already
3925 inserted X-ACL-Warn: at the start if there was no header name. However, it
3926 was not making this test for the second and subsequent header lines if
3927 there were newlines in the string. This meant that an invalid header could
3928 be inserted if Exim was badly configured.
3930 7. Allow an ACL "warn" statement to add header lines at the start or after all
3931 the Received: headers, as well as at the end.
3933 8. Added the rcpt_4xx retry error code.
3935 9. Added postmaster_mailfrom=xxx to callout verification option.
3937 10. Added mailfrom=xxxx to the callout verification option, for verify=
3940 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3941 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3943 12. Inserted some casts to stop certain compilers warning when using pointer
3944 differences as field lengths or precisions in printf-type calls (mostly
3945 affecting debugging statements).
3947 13. Added optional readline() support for -be (dynamically loaded).
3949 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3950 same clock tick as a message's arrival, so that its received time was the
3951 same as the "first fail" time on the retry record, and that message
3952 remained on the queue past the ultimate address timeout, every queue runner
3953 would try a delivery (because it was past the ultimate address timeout) but
3954 after another failure, the ultimate address timeout, which should have then
3955 bounced the address, did not kick in. This was a "< instead of <=" error;
3956 in most cases the first failure would have been in the next clock tick
3957 after the received time, and all would be well.
3959 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3960 being recognized when the domain list was tested by the match_domain
3961 condition in an expansion string.
3963 16. Added the ${str2b64: operator.
3965 17. Exim was always calling setrlimit() to set a large limit for the number of
3966 processes, without checking whether the existing limit was already
3967 adequate. (It did check for the limit on file descriptors.) Furthermore,
3968 errors from getrlimit() and setrlimit() were being ignored. Now they are
3969 logged to the main and panic logs, but Exim does carry on, to try to do its
3970 job under whatever limits there are.
3972 18. Imported PCRE 5.0.
3974 19. Trivial typo in log message " temporarily refused connection" (the leading
3977 20. If the log selector return_path_on_delivery was set and an address was
3978 redirected to /dev/null, the delivery process crashed because it assumed
3979 that a return path would always be set for a "successful" delivery. In this
3980 case, the whole delivery is bypassed as an optimization, and therefore no
3983 21. Internal re-arrangement: the function for sending a challenge and reading
3984 a response while authentication was assuming a zero-terminated challenge
3985 string. It's now changed to take a pointer and a length, to allow for
3986 binary data in such strings.
3988 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3990 23. Exim was not respecting finduser_retries when seeking the login of the
3991 uid under which it was called; it was always trying 10 times. (The default
3992 setting of finduser_retries is zero.) Also, it was sleeping after the final
3993 failure, which is pointless.
3995 24. Implemented tls_on_connect_ports.
3997 25. Implemented acl_smtp_predata.
3999 26. If the domain in control=submission is set empty, Exim assumes that the
4000 authenticated id is a complete email address when it generates From: or
4001 Sender: header lines.
4003 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4004 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4005 chown and chgrp in /bin and hostname in /usr/bin.
4007 28. Exim was keeping the "process log" file open after each use, just as it
4008 does for the main log. This opens the possibility of it remaining open for
4009 long periods when the USR1 signal hits a daemon. Occasional processlog
4010 errors were reported, that could have been caused by this. Anyway, it seems
4011 much more sensible not to leave this file open at all, so that is what now
4014 29. The long-running daemon process does not normally write to the log once it
4015 has entered its main loop, and it closes the log before doing so. This is
4016 so that log files can straightforwardly be renamed and moved. However,
4017 there are a couple of unusual error situations where the daemon does write
4018 log entries, and I had neglected to close the log afterwards.
4020 30. The text of an SMTP error response that was received during a remote
4021 delivery was being truncated at 512 bytes. This is too short for some of
4022 the long messages that one sometimes sees. I've increased the limit to
4025 31. It is now possible to make retry rules that apply only when a message has a
4026 specific sender, in particular, an empty sender.
4028 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4029 it possible to be selective about when SMTP synchronization is enforced.
4031 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4033 32. Implemented hosts_connection_nolog.
4035 33. Added an ACL for QUIT.
4037 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4040 35. Added mailbox_size and mailbox_filecount to appendfile.
4042 36. Added control = no_multiline_responses to ACLs.
4044 37. There was a bug in the logic of the code that waits for the clock to tick
4045 in the case where the clock went backwards by a substantial amount such
4046 that the microsecond fraction of "now" was more than the microsecond
4047 fraction of "then" (but the whole seconds number was less).
4049 38. Added support for the libradius Radius client library this is found on
4050 FreeBSD (previously only the radiusclient library was supported).
4056 1. When certain lookups returned multiple values in the form name=value, the
4057 quoting of the values was not always being done properly. Specifically:
4058 (a) If the value started with a double quote, but contained no whitespace,
4060 (b) If the value contained whitespace other than a space character (i.e.
4061 tabs or newlines or carriage returns) it was not quoted.
4062 This fix has been applied to the mysql and pgsql lookups by writing a
4063 separate quoting function and calling it from the lookup code. The fix
4064 should probably also be applied to nisplus, ibase and oracle lookups, but
4065 since I cannot test any of those, I have not disturbed their existing code.
4067 2. A hit in the callout cache for a specific address caused a log line with no
4068 reason for rejecting RCPT. Now it says "Previous (cached) callout
4069 verification failure".
4071 3. There was an off-by-one bug in the queryprogram router. An over-long
4072 return line was truncated at 256 instead of 255 characters, thereby
4073 overflowing its buffer with the terminating zero. As well as fixing this, I
4074 have increased the buffer size to 1024 (and made a note to document this).
4076 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4077 when Exim is waiting for an SMTP response from a remote server, Exim
4078 restarts its select() call on the socket, thereby resetting its timeout.
4079 This is not a problem when such interrupts are rare. Somebody set up a cron
4080 job to run exiwhat every 2 minutes, which is less than the normal select()
4081 timeout (5 or 10 minutes). This meant that the select() timeout never
4082 kicked in because it was always reset. I have fixed this by comparing the
4083 time when an interrupt arrives with the time at the start of the first call
4084 to select(). If more time than the timeout has elapsed, the interrupt is
4085 treated as a timeout.
4087 5. Some internal re-factoring in preparation for the addition of Sieve
4088 extensions (by MH). In particular, the "personal" test is moved to a
4089 separate function, and given an option for scanning Cc: and Bcc: (which is
4090 not set for Exim filters).
4092 6. When Exim created an email address using the login of the caller as the
4093 local part (e.g. when creating a From: or Sender: header line), it was not
4094 quoting the local part when it contained special characters such as @.
4096 7. Installed new OpenBSD configuration files.
4098 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4099 try to make them clearer.
4101 9. Callout options, other than the timeout value, were being ignored when
4102 verifying sender addresses in header lines. For example, when using
4104 verify = header_sender/callout=no_cache
4106 the cache was (incorrectly) being used.
4108 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4109 only the Cygwin environment.
4111 11. When return_path_on_delivery was set as a log selector, if different remote
4112 addresses in the same message used different return paths and parallel
4113 remote delivery occurred, the wrong values would sometimes be logged.
4114 (Whenever a remote delivery process finished, the return path value from
4115 the most recently started remote delivery process was logged.)
4117 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4118 lines when AUTH and/or TLS are in use. This is the "received protocol"
4119 field. Exim used to use "asmtp" for authenticated SMTP, without any
4120 indication (in the protocol name) for TLS use. Now it follows the RFC and
4121 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4122 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4123 names appear in log lines as well as in Received: header lines.
4125 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4126 extensions, and comparison tests, and to fix some bugs.
4128 14. Changes to the "personal" filter test:
4130 (1) The test was buggy in that it was just doing the equivalent of
4131 "contains" tests on header lines. For example, if a user's address was
4132 anne@some.where, the "personal" test would incorrectly be true for
4134 To: susanne@some.where
4136 This test is now done by extracting each address from the header in turn,
4137 and checking the entire address. Other tests that are part of "personal"
4138 are now done using regular expressions (for example, to check local parts
4139 of addresses in From: header lines).
4141 (2) The list of non-personal local parts in From: addresses has been
4142 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4143 taken from the Sieve specification recommendations.
4145 (3) If the message contains any header line starting with "List-" it is
4146 treated as non-personal.
4148 (4) The test for "circular" in the Subject: header line has been removed
4149 because it now seems ill-conceived.
4151 15. Minor typos in src/EDITME comments corrected.
4153 16. Installed latest exipick from John Jetmore.
4155 17. If headers_add on a router specified a text string that was too long for
4156 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4157 of string_sprintf() is now avoided.
4159 18. $message_body_size was not set (it was always zero) when running the DATA
4160 ACL and the local_scan() function.
4162 19. For the "mail" command in an Exim filter, no default was being set for
4163 the once_repeat time, causing a random time value to be used if "once" was
4164 specified. (If the value happened to be <= 0, no repeat happened.) The
4165 default is now 0s, meaning "never repeat". The "vacation" command was OK
4166 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4167 (I found it when inspecting the code).
4169 20. There is now an overall timeout for performing a callout verification. It
4170 defaults to 4 times the callout timeout, which applies to individual SMTP
4171 commands during the callout. The overall timeout applies when there is more
4172 than one host that can be tried. The timeout is checked before trying the
4173 next host. This prevents very long delays if there are a large number of
4174 hosts and all are timing out (e.g. when the network connections are timing
4175 out). The value of the overall timeout can be changed by specifying an
4176 additional sub-option for "callout", called "maxwait". For example:
4178 verify = sender/callout=5s,maxwait=20s
4180 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4181 to the end before writing, but this should make it even safer).
4183 22. Exim was forgetting that it had advertised PIPELINING for the second and
4184 subsequent messages on an SMTP connection. It was also not resetting its
4185 memory on STARTTLS and an internal HELO.
4187 23. When Exim logs an SMTP synchronization error within a session, it now
4188 records whether PIPELINING has been advertised or not.
4190 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4191 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4192 rather than long int.
4194 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4196 26. Added the never_mail option to autoreply.
4202 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4203 crash if the getsockname() call failed; this can happen if a connection is
4204 closed very soon after it is established. The problem was simply in the
4205 order in which certain operations were done, causing Exim to try to write
4206 to the SMTP stream before it had set up the file descriptor. The bug has
4207 been fixed by making things happen in the correct order.
4213 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4214 before the connection was closed, thus losing the rejection response.
4216 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4217 some early Solaris releases, but causes trouble in current releases where
4218 socklen_t is defined.
4220 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4223 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4226 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4227 that releasing the top of it at the end releases what was used for sub-
4228 expansions (unless the block got too big). However, discard this block if
4229 the first thing is a variable or header, so that we can use its block when
4230 it is dynamic (useful for very large $message_headers, for example).
4232 6. Lookups now cache *every* query, not just the most recent. A new, separate
4233 store pool is used for this. It can be recovered when all lookup caches are
4234 flushed. Lookups now release memory at the end of their result strings.
4235 This has involved some general refactoring of the lookup sources.
4237 7. Some code has been added to the store_xxx() functions to reduce the amount
4238 of flapping under certain conditions.
4240 8. log_incoming_interface used to affect only the <= reception log lines. Now
4241 it causes the local interface and port to be added to several more SMTP log
4242 lines, for example "SMTP connection from", and rejection lines.
4244 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4246 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4248 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4249 could be overwritten at the end of the current message (or the start of a
4250 new message if it was set in a HELO ACL). The value is now preserved for
4251 the duration of the SMTP connection.
4253 12. If a transport had a headers_rewrite setting, and a matching header line
4254 contained an unqualified address, that address was qualified, even if it
4255 did not match any rewriting rules. The underlying bug was that the values
4256 of the flags that permit the existence of unqualified sender and recipient
4257 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4258 non-local messages, and by -bnq for local messages) were not being
4259 preserved with the message after it was received.
4261 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4262 "next input=" as part of the text comprising the host identity instead of
4263 the correct text. The code was using the same buffer for two different
4264 strings. However, depending on which order the printing function evaluated
4265 its arguments, the bug did not always show up. Under Linux, for example, my
4266 test suite worked just fine.
4268 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4269 doesn't work with some older versions of Perl. It has been changed to "my",
4270 which in any case is probably the better facility to use.
4272 15. A really picky compiler found some instances of statements for creating
4273 error messages that either had too many or two few arguments for the format
4276 16. The size of the buffer for calls to the DNS resolver has been increased
4277 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4278 for addresses that have a lot of PTR records. This alleviates a problem; it
4279 does not fully solve it.
4281 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4282 buffer now truncates the list and logs the incident, which is the same
4283 action as happens when Exim is looking up a host name and its aliases.
4284 Previously in this situation something unpredictable would happen;
4285 sometimes it was "internal error: store_reset failed".
4287 18. If a server dropped the connection unexpectedly when an Exim client was
4288 using GnuTLS and trying to read a response, the client delivery process
4289 crashed while trying to generate an error log message.
4291 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4292 string, for example:
4294 warn message = H1: something\nH2: something
4296 the text was added as a single header line from Exim's point of view
4297 though it ended up OK in the delivered message. However, searching for the
4298 second and subsequent header lines using $h_h2: did not work. This has been
4299 fixed. Similarly, if a system filter added multiple headers in this way,
4300 the routers could not see them.
4302 20. Expanded the error message when iplsearch is called with an invalid key to
4303 suggest using net-iplsearch in a host list.
4305 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4306 ACLs are no longer actually imposed (and a message to that effect is
4309 22. If a "gecos" field in a passwd entry contained escaped characters, in
4310 particular, if it contained a \" sequence, Exim got it wrong when building
4311 a From: or a Sender: header from that name. A second bug also caused
4312 incorrect handling when an unquoted " was present following a character
4313 that needed quoting.
4315 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4316 was not being matched caselessly.
4318 24. Arranged for all hyphens in the exim.8 source to be escaped with
4321 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4322 properly. Recipient callout cache records were still being keyed to include
4323 the sender, even when use_sender was set false. This led to far more
4324 callouts that were necessary. The sender is no longer included in the key
4325 when use_sender is false.
4327 26. Added "control = submission" modifier to ACLs.
4329 27. Added the ${base62d: operator to decode base 62 numbers.
4331 28. dnsdb lookups can now access SRV records.
4333 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4334 the configuration file.
4336 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4337 (-v) mode. This makes the output for a verbose queue run more intelligible.
4339 31. Added a use_postmaster feature to recipient callouts.
4341 32. Added the $body_zerocount variable, containing the number of binary zero
4342 bytes in the message body.
4344 33. The time of last modification of the "new" subdirectory is now used as the
4345 "mailbox time last read" when there is a quota error for a maildir
4348 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4350 35. Added +ignore_unknown as a special item in host lists.
4352 36. Code for decoding IPv6 addresses in host lists is now included, even if
4353 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4354 address was recognized as an IP address, but was then not correctly decoded
4355 into binary, causing unexpected and incorrect effects when compared with
4362 1. Very minor rewording of debugging text in manualroute to say "list of
4363 hosts" instead of "hostlist".
4365 2. If verify=header_syntax was set, and a header line with an unqualified
4366 address (no domain) and a large number of spaces between the end of the
4367 name and the colon was received, the reception process suffered a buffer
4368 overflow, and (when I tested it) crashed. This was caused by some obsolete
4369 code that should have been removed. The fix is to remove it!
4371 3. When running in the test harness, delay a bit after writing a bounce
4372 message to get a bit more predictability in the log output.
4374 4. Added a call to search_tidyup() just before forking a reception process. In
4375 theory, someone could use a lookup in the expansion of smtp_accept_max_
4376 per_host which, without the tidyup, could leave open a database connection.
4378 5. Added the variables $recipient_data and $sender_data which get set from a
4379 lookup success in an ACL "recipients" or "senders" condition, or a router
4380 "senders" option, similar to $domain_data and $local_part_data.
4382 6. Moved the writing of debug_print from before to after the "senders" test
4385 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4386 problems for message scanning, either using a data ACL, or using
4387 local_scan() because the Received: header was not generated till after they
4388 were called (in order to set the time as the time of reception completion).
4389 I have revised the way this works. The header is now generated after the
4390 body is received, but before the ACL or local_scan() are called. After they
4391 are run, the timestamp in the header is updated.
4397 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4398 before starting a queue runner without re-exec. This happened only when
4399 deliver_drop_privilege was set or when the Exim user was set to root. The
4400 effect of the bug was that timeouts during subsequent deliveries caused
4401 crashes instead of being properly handled. The handler is now left at its
4402 default (and expected) setting.
4404 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4405 message, again when deliver_drop_privilege is set or Exim is run as root.
4406 The bug described in (1) was not present in this case, but the tidying up
4407 of the other signals was missing. I have made the two cases consistent.
4409 3. The ignore_target_hosts setting on a manualroute router was being ignored
4410 for hosts that were looked up using the /MX notation.
4412 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4415 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4416 operated on the sender address. After changing the $sender_address to <>
4417 for the sender address verify, Exim was re-instated it as the original
4418 (before rewriting) address, but remembering that it had rewritten it, so it
4419 wasn't rewriting it again. This bug also had the effect of breaking the
4420 sender address verification caching when the sender address was rewritten.
4422 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4423 This has been changed so that if the ip literal address matches
4424 ignore_target_hosts, the router declines.
4426 7. Added expansion conditions match_domain, match_address, and match_local_
4427 part (NOT match_host).
4429 8. The placeholder for the Received: header didn't have a length field set.
4431 9. Added code to Exim itself and to exim_lock to test for a specific race
4432 condition that could lead to file corruption when using MBX delivery. The
4433 issue is with the lockfile that is created in /tmp. If this file is removed
4434 after a process has opened it but before that process has acquired a lock,
4435 there is the potential for a second process to recreate the file and also
4436 acquire a lock. This could lead to two Exim processes writing to the file
4437 at the same time. The added code performs the same test as UW imapd; it
4438 checks after acquiring the lock that its file descriptor still refers to
4439 the same named file.
4441 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4442 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4445 11. Added the smtp_active_hostname option. If used, this will typically be made
4446 to depend on the incoming interface address. Because $interface_address is
4447 not set up until the daemon has forked a reception process, error responses
4448 that can happen earlier (such as "too many connections") no longer contain
4451 12. If an expansion in a condition on a "warn" statement fails because a lookup
4452 defers, the "warn" statement is abandoned, and the next ACL statement is
4453 processed. Previously this caused the whole ACL to be aborted.
4455 13. Added the iplsearch lookup type.
4457 14. Added ident_timeout as a log selector.
4459 15. Added tls_certificate_verified as a log selector.
4461 16. Added a global option tls_require_ciphers (compare the smtp transport
4462 option of the same name). This controls incoming TLS connections.
4464 17. I finally figured out how to make tls_require_ciphers do a similar thing
4465 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4466 before starting the TLS session.
4468 18. Tabs are now shown as \t in -bP output.
4470 19. If the log selector return_path_on_delivery was set, Exim crashed when
4471 bouncing a message because it had too many Received: header lines.
4473 20. If two routers both had headers_remove settings, and the first one included
4474 a superfluous trailing colon, the final name in the first list and the
4475 first name in the second list were incorrectly joined into one item (with a
4476 colon in the middle).
4482 1. Added -C and -D options to the exinext utility, mainly to make it easier
4483 to include in the automated testing, but these could be helpful when
4484 multiple configurations are in use.
4486 2. The exinext utility was not formatting the output nicely when there was
4487 an alternate port involved in the retry record key, nor when there was a
4488 message id as well (for retries that were specific to a specific message
4489 and a specific host). It was also confused by IPv6 addresses, because of
4490 the additional colons they contain. I have fixed the IPv4 problem, and
4491 patched it up to do a reasonable job for IPv6.
4493 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4494 delivery, the log line now contains "pipelined" if PIPELINING was used.
4496 4. An SMTP transport process used to panic and die if the bind() call to set
4497 an explicit outgoing interface failed. This has been changed; it is now
4498 treated in the same way as a connect() failure.
4500 5. A reference to $sender_host_name in the part of a conditional expansion
4501 that was being skipped was still causing a DNS lookup. This no longer
4504 6. The def: expansion condition was not recognizing references to header lines
4505 that used bh_ and bheader_.
4507 7. Added the _cache feature to named lists.
4509 8. The code for checking quota_filecount in the appendfile transport was
4510 allowing one more file than it should have been.
4512 9. For compatibility with Sendmail, the command line option
4520 and sets the incoming protocol and host name (for trusted callers). The
4521 host name and its colon can be omitted when only the protocol is to be set.
4522 Note the Exim already has two private options, -pd and -ps, that refer to
4523 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4524 "s", but I don't think that's a major issue.
4526 10. A number of refactoring changes to the code, none of which should affect
4529 (a) The number of logging options was getting close to filling up the
4530 32-bit word that was used as a bit map. I have split them into two classes:
4531 those that are passed in the argument to log_write(), and those that are
4532 only ever tested independently outside of that function. These are now in
4533 separate 32-bit words, so there is plenty of room for expansion again.
4534 There is no change in the user interface or the logging behaviour.
4536 (b) When building, for example, log lines, the code previously used a
4537 macro that called string_cat() twice, in order to add two strings. This is
4538 not really sufficiently general. Furthermore, there was one instance where
4539 it was actually wrong because one of the argument was used twice, and in
4540 one call a function was used. (As it happened, calling the function twice
4541 did not affect the overall behaviour.) The macro has been replaced by a
4542 function that can join an arbitrary number of extra strings onto a growing
4545 (c) The code for expansion conditions now uses a table and a binary chop
4546 instead of a serial search (which was left over from when there were very
4547 few conditions). Also, it now recognizes conditions like "pam" even when
4548 the relevant support is not compiled in: a suitably worded error message is
4549 given if an attempt is made to use such a condition.
4551 11. Added ${time_interval:xxxxx}.
4553 12. A bug was causing one of the ddress fields not to be passed back correctly
4554 from remote delivery subprocesses. The field in question was not being
4555 subsequently used, so this caused to problems in practice.
4557 13. Added new log selectors queue_time and deliver_time.
4559 14. Might have fixed a bug in maildirsizefile handling that threw up
4560 "unexpected character" debug warnings, and recalculated the data
4561 unnecessarily. In any case, I expanded the warning message to give more
4564 15. Added the message "Restricted characters in address" to the statements in
4565 the default ACL that block characters like @ and % in local parts.
4567 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4568 Three changes have been made:
4570 (a) There was a serious bug; a negative response to MAIL caused the whole
4571 recipient domain to be cached as invalid, thereby blocking all messages
4572 to all local parts at the same domain, from all senders. This bug has
4573 been fixed. The domain is no longer cached after a negative response to
4574 MAIL if the sender used is not empty.
4576 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4579 (c) A new callout option, "use_sender" has been added for people who want
4580 the modified behaviour.
4586 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4589 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4590 indeed breaks things for older releases.
4592 3. Added additional logging to the case where there is a problem reading data
4593 from a filter that is running in a subprocess using a pipe, in order to
4594 try to track down a specific problem.
4596 4. Testing facility fudge: when running in the test harness and attempting
4597 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4598 getting "No route to host". Convert this to a timeout.
4600 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4603 6. Some OS don't have socklen_t but use size_t instead. This affects the
4604 fifth argument of getsockopt() amongst other things. This is now
4605 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4606 can be set for individual OS. I have set it for SunOS5, OSF1, and
4607 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4608 some earlier ones do not.
4610 7. Change 4.30/15 was not doing the test caselessly.
4612 8. The standard form for an IPv6 address literal was being rejected by address
4613 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4614 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4615 this, as well as the form without the "IPv6" on the front (but only when
4616 address literals are enabled, of course).
4618 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4620 10. Exim crashed if a message with an empty sender address specified by -f
4621 encountered a router with an errors_to setting. This could be provoked only
4622 by a command such as
4626 where an empty string was supplied; "<>" did not hit this bug.
4628 11. Installed PCRE release 4.5.
4630 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4631 remained set. It is now erased.
4633 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4634 times from message ids (which are base 36 rather than the normal 62).
4636 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4637 were being counted as actual protocol errors, and logged if the log
4638 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4639 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4640 and DATA following a set of rejected RCPTs do not count as protocol errors.
4641 In other words, Exim assumes they were pipelined, though this may not
4642 actually be the case. Of course, in all cases the client gets an
4643 appropriate error code.
4645 15. If a lookup fails in an ACL condition, a message about the failure may
4646 be available; it is used if testing the ACL cannot continue, because most
4647 such messages specify what the cause of the deferral is. However, some
4648 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4649 that caused an old message to be retained and used if a later statement
4650 caused a defer, replacing the real cause of the deferral.
4652 16. If an IP address had so many PTR records that the DNS lookup buffer
4653 was not large enough to hold them, Exim could crash while trying to process
4654 the truncated data. It now detects and logs this case.
4656 17. Further to 4.21/58, another change has been made: if (and only if) the
4657 first line of a message (the first header line) ends with CRLF, a bare LF
4658 in a subsequent header line has a space inserted after it, so as not to
4659 terminate the header.
4661 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4662 unnecessarily, used atoi() instead of strtol(), and didn't check the
4663 termination when getting file sizes from file names by regex.
4665 19. Completely re-implemented the support for maildirsize files, in the light
4666 of a number of problems with the previous contributed implementation
4667 (4.30/29). In particular:
4669 . If the quota is zero, the maildirsize file is maintained, but no quota is
4672 . If the maildir directory does not exist, it is created before any attempt
4673 to write a maildirsize file.
4675 . The quota value in the file is just a cache; if the quota is changed in
4676 the transport, the new value overrides.
4678 . A regular expression is available for excluding directories from the
4681 20. The autoreply transport checks the characters in options that define the
4682 message's headers; it allows continued headers, but it was checking with
4683 isspace() after an embedded newline instead of explicitly looking for a
4686 21. If all the "regular" hosts to which an address was routed had passed their
4687 expiry times, and had not reached their retry times, the address was
4688 bounced, even if fallback hosts were defined. Now Exim should go on to try
4691 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4692 equivalent code in the SMTP transport. Some hosts send humungous responses
4693 to HELO/EHLO, more than 1024 it seems.
4695 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4696 gives compiler warnings in some environments. I've now done it "properly",
4699 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4700 (because of problems with the built-in one) was declared to return uschar *
4701 instead of char *, causing compiler failure.
4703 25. Fixed a file descriptor leak when processing alias/forward files.
4705 26. Fixed a minor format string issue in dbfn.c.
4707 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4709 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4710 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4713 29. When a sender address is verified, it is cached, to save repeating the test
4714 when there is more than one recipient in a message. However, when the
4715 verification involves a callout, it is possible for different callout
4716 options to be set for different recipients. It is too complicated to keep
4717 track of this in the cache, so now Exim always runs a verification when a
4718 callout is required, relying on the callout cache for the optimization.
4719 The overhead is duplication of the address routing, but this should not be
4722 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4723 to be verified with callout=postmaster, and the main callout worked but the
4724 postmaster check failed, the verification correctly failed. However, if a
4725 subsequent RCPT command asked for sender verification *without* the
4726 postmaster check, incorrect caching caused this verification also to fail,
4729 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4730 it was not caching the DNS options (qualify_single, search_parents) that
4731 were used when the lookup failed. A subsequent lookup with different
4732 options therefore always gave the same answer, though there were cases
4733 where it should not have. (Example: a "domains = !$mx_any" option on a
4734 dnslookup router: the "domains" option is always processed without any
4735 widening, but the router might have qualify_single set.) Now Exim uses the
4736 cached value only when the same options are set.
4738 32. Added John Jetmore's "exipick" utility to the distribution.
4740 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4741 than a timeout (e.g. a certificate is required, and is not provided), an
4742 Exim server now closes the connection immediately. Previously it waited for
4743 the client to close - but if the client is SSL, it seems that they each
4744 wait for each other, leading to a delay before one of them times out.
4746 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4747 maintained 0.8.x compatibility because I don't think many are using it, and
4748 it is clearly obsolete.
4750 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4753 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4754 client certificate was expired. A simple patch fixes this, though I don't
4755 understand the full logic of why the verify callback is called multiple
4758 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4759 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4760 which causes problems with some clients (such as the Certicom SSL Plus
4761 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4762 disables the coutermeasure allowing Eudora to connect."
4764 38. Exim was not checking that a write() to a log file succeeded. This could
4765 lead to Bad Things if a log got too big, in particular if it hit a file
4766 size limit. Exim now panics and dies if it cannot write to a log file, just
4767 as it does if it cannot open a log file.
4769 39. Modified OS/Makefile-Linux so that it now contains
4771 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4773 The two -D definitions ensure that Exim is compiled with large file
4774 support, which makes it possible to handle log files that are bigger than
4777 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4778 instance) a domain was checked against a named list that involved a lookup,
4779 causing $domain_data to be set, then another domain was checked against the
4780 same list, then the first domain was re-checked, the value of $domain_data
4781 after the final check could be wrong. In particular, if the second check
4782 failed, it could be set empty. This bug probably also applied to
4785 41. The strip_trailing_dot option was not being applied to the address given
4786 with the -f command-line option.
4788 42. The code for reading a message's header from the spool was incrementing
4789 $received_count, but never initializing it. This meant that the value was
4790 incorrect (doubled) while delivering a message in the same process in which
4791 it was received. In the most common configuration of Exim, this never
4792 happens - a fresh exec is done - but it can happen when
4793 deliver_drop_privilege is set.
4795 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4796 it now includes up to 150 characters of the unexpected data in the log
4799 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4800 and building data strings. The size of both of these buffers was 10 000
4801 bytes - far larger than anybody would *ever* want, thought I. Needless to
4802 say, somebody hit the limit. I have increased the maximum line length to
4803 20 000 and the maximum data length of concatenated lines to 100 000. I have
4804 also fixed two bugs, because there was no checking on these buffers. Tsk,
4805 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4806 buffer is too small.
4808 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4809 lsearch lookups. Now it does.
4811 46. When parsing a route_list item in a manualroute router, a fixed-length
4812 buffer was used for the list of hosts. I made this 1024 bytes long,
4813 thinking that nobody would ever have a list of hosts that long. Wrong.
4814 Somebody had a whole pile of complicated expansion conditions, and the
4815 string was silently truncated, leading to an expansion error. It turns out
4816 that it is easier to change to an unlimited length (owing to other changes
4817 that have happened since this code was originally written) than to build
4818 structure for giving a limitation error. The length of the item that
4819 expands into the list of hosts is now unlimited.
4821 47. The lsearch lookup could not handle data where the length of text line was
4822 more than 4095 characters. Such lines were truncated, leading to shortened
4823 data being returned. It should now handle lines of any length.
4825 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4826 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4829 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4830 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4831 no longer happens in comment lines. A list of replacements is now placed
4832 at the head of all of the source files, except those whose only change is
4833 to replace PERL_COMMAND in the very first #! line.
4835 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4836 messages on the queue, with a bottom-up merge sort, using code contributed
4837 by Michael Haardt. This should make operations like -bp somewhat faster on
4838 large queues. It won't affect queue runners, except when queue_run_in_order
4841 51. Installed eximstats 1.31 in the distribution.
4843 52. Added support for SRV lookups to the dnslookup router.
4845 53. If an ACL referred to $message_body or $message_body_end, the value was not
4846 reset for any messages that followed in the same SMTP session.
4848 54. The store-handling optimization for building very long strings was not
4849 differentiating between the different store pools. I don't think this
4850 actually made any difference in practice, but I've tidied it.
4852 55. While running the routers to verify a sender address, $sender_address
4853 was still set to the sender address. This is wrong, because when routing to
4854 send a bounce to the sender, it would be empty. Therefore, I have changed
4855 it so that, while verifying a sender address, $sender_address is set to <>.
4856 (There is no change to what happens when verifying a recipient address.)
4858 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4859 target A or AAAA records (if not already returned) without resetting the
4860 qualify_single or search_parents options of the DNS resolver. These are
4861 inappropriate in this case because the targets of MX and SRV records must
4862 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4863 target that, when qualified, matched something in the local domain. These
4864 two options are now turned off when doing these lookups.
4866 57. It seems that at least some releases of Reiserfs (which does not have the
4867 concept of a fixed number of inodes) returns zero and not -1 for the
4868 number of available inodes. This interacted badly with check_spool_inodes,
4869 which assumed that -1 was the "no such thing" setting. What I have done is
4870 to check that the total number of inodes is greater than zero before doing
4871 the test of how many are available.
4873 58. When a "warn" ACL statement has a log_message modifier, the message is
4874 remembered, and not repeated. This is to avoid a lot of repetition when a
4875 message has many recipients that cause the same warning to be written.
4876 Howewer, Exim was preserving the list of already written lines for an
4877 entire SMTP session, which doesn't seem right. The memory is now reset if a
4878 new message is started.
4880 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4881 debugging output unless log_rewrite was also set.
4883 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4884 of (int)(handle) when we know that handle contains (void *)(-1).
4886 61. The Exim daemon panic-logs an error return when it closes the incoming
4887 connection. However "connection reset by peer" seems to be common, and
4888 isn't really an error worthy of noting specially, so that particular error
4891 62. When Exim is trying to find all the local interfaces, it used to panic and
4892 die if the ioctl to get the interface flags failed. However, it seems that
4893 on at least one OS (Solaris 9) it is possible to have an interface that is
4894 included in the list of interfaces, but for which you get a failure error
4895 for this call. This happens when the interface is not "plumbed" into a
4896 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4897 failure of the "get flags" call assumes that the interface is down.
4899 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4900 makes life easier for people who are doing arithmetic on fields extracted
4901 from dates, where you often get leading zeros that should not be
4902 interpreted as octal.
4904 64. Added qualify_domain to the redirect router, to override the global
4907 65. If a pathologically long header line contained very many addresses (the
4908 report of this problem mentioned 10 000) and each of them was rewritten,
4909 Exim could use up a very large amount of memory. (It kept on making new
4910 copies of the header line as it rewrote, and never released the old ones.)
4911 At the expense of a bit more processing, the header rewriting function has
4912 been changed so that it no longer eats memory in this way.
4914 66. The generation of the Received: header has been moved from the time that a
4915 message starts to be received, to the time that it finishes. The timestamp
4916 in the Received: header should now be very close to that of the <= log
4917 line. There are two side-effects of this change:
4919 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4920 logged header lines no longer include the local Received: line, because
4921 it has not yet been created. The same applies to a copy of the message
4922 that is returned to a non-SMTP sender when a message is rejected.
4924 (b) When a filter file is tested using -bf, no additional Received: header
4925 is added to the test message. After some thought, I decided that this
4928 This change does not affect the value of $received_for. It is still set
4929 after address rewriting, but before local_scan() is called.
4931 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4933 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4934 gave an unhelpful panic error message, and a defer error. I have managed to
4935 change this behaviour so that it now rejects any supplied certificate,
4936 which seems right, as the list of acceptable certificates is empty.
4938 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4939 gave an unhelpful defer error. I have not managed to make this reject any
4940 supplied certificates, but the error message it gives is "no certificate
4941 supplied", which is not helpful.
4943 70. exigrep's output now also includes lines that are not associated with any
4944 message, but which match the given pattern. Implemented by a patch from
4945 Martin Sluka, which also tidied up the Perl a bit.
4947 71. Recipient callout verification, like sender verification, was using <> in
4948 the MAIL FROM command. This isn't really the right thing, since the actual
4949 sender may affect whether the remote host accepts the recipient or not. I
4950 have changed it to use the actual sender in the callout; this means that
4951 the cache record is now keyed on a recipient/sender pair, not just the
4952 recipient address. There doesn't seem to be a real danger of callout loops,
4953 since a callout by the remote host to check the sender would use <>.
4954 [SEE ABOVE: changed after hitting problems.]
4956 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4957 temporary errors. However, in the case of such a code being given after
4958 the end of a data transmission (i.e. after ".") Exim was failing to write
4959 a retry record for the message. (Yes, there was some broken host that was
4960 actually sending 8xx at this point.)
4962 73. An unknown lookup type in a host list could cause Exim to panic-die when
4963 the list was checked. (An example that provoked this was putting <; in the
4964 middle of a list instead of at the start.) If this happened during a DATA
4965 ACL check, a -D file could be left lying around. This kind of configuration
4966 error no longer causes Exim to die; instead it causes a defer errror. The
4967 incident is still logged to the main and panic logs.
4969 74. Buglet left over from Exim 3 conversion. The message "too many messages
4970 in one connection" was written to the rejectlog but not the mainlog, except
4971 when address rewriting (yes!) was being logged.
4973 75. Added write_rejectlog option.
4975 76. When a system filter was run not as root (that is, when system_filter_user
4976 was set), the values of the $n variables were not being returned to the
4977 main process; thus, they were not subsequently available in the $sn
4980 77. Added +return_path_on_delivery log selector.
4982 78. A connection timeout was being treated differently from recipients deferred
4983 when testing hosts_max_try with a message that was older than the host's
4984 retry timeout. (The host should not be counted, thus allowing all hosts to
4985 be tried at least once before bouncing.) This may have been the cause of an
4986 occasionally reported bug whereby a message would remain on the queue
4987 longer than the retry timeout, but would be bounced if a delivery was
4988 forced. I say "may" because I never totally pinned down the problem;
4989 setting up timeout/retry tests is difficult. See also the next item.
4991 79. The ultimate address timeout was not being applied to errors that involved
4992 a combination of host plus message (for example, a timeout on a MAIL
4993 command). When an address resolved to a number of possible hosts, and they
4994 were not all tried for each delivery (e.g. because of hosts_max_try), a
4995 message could remain on the queue longer than the retry timeout.
4997 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5000 81. Fixed an obscure SMTP outgoing bug which required at least the following
5001 conditions: (a) there was another message waiting for the same server;
5002 (b) the server returned 5xx to all RCPT commands in the first message so
5003 that the message was not completed; (c) the server dropped the connection
5004 or gave a negative response to the RSET that Exim sends to abort the
5005 transaction. The observed case was a dropped connection after DATA that had
5006 been sent in pipelining mode. That is, the server had advertised PIPELINING
5007 but was not implementing it correctly. The effect of the bug was incorrect
5008 behaviour, such as trying another host, and this could lead to a crash.
5014 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5015 and daemon.c were passed as pointers to ints; they should have been
5016 pointers to socklen_t variables (which are typically unsigned ints).
5018 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5021 3. Fixed a really odd bug that affected only the testing scheme; patching a
5022 certain fixed string in the binary changed the value of another string that
5023 happened to be identical to the end of the original first string.
5025 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5026 name", it returns that address as the IP address. On some operating
5027 systems (e.g. Solaris), it also passes back the IP address string as the
5028 "host name". However, on others (e.g. Linux), it passes back an empty
5029 string. Exim wasn't checking for this, and was changing the host name to an
5030 empty string, assuming it had been canonicized.
5032 5. Although rare, it is permitted to have more than one PTR record for a given
5033 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5034 all the names associated with an address, because they do in Solaris.
5035 However, it seems that they do not in Linux for data that comes from the
5036 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5037 I found this out when I moved to a new Linux workstation and tried to run
5038 the Exim test suite.
5040 To get round this problem I have changed the code so that it now does its
5041 own call to the DNS to look up PTR records when searching for a host name.
5042 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5043 addresses that are only in /etc/hosts are still found.
5045 This behaviour is, however, controlled by an option called host_lookup_
5046 order, which defaults to "bydns:byaddr". If people want to use the other
5047 order, or indeed, just use one or the other means of lookup, they can
5048 specify it in this variable.
5050 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5051 some operating systems, this comes back from gethostbyaddr() as an empty
5052 string, and this is what Exim used to test for. However, it seems that in
5053 other systems, "." is yielded. Exim now tests for this case too.
5055 7. The values of check_spool_space and check_log_space are now held internally
5056 as a number of kilobytes instead of an absolute number of bytes. If a
5057 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5058 kilobyte. This means that much larger values can be stored.
5060 8. Exim monitor: an attempt to get the action menu when not actually pointing
5061 at a message produces an empty menu entitled "No message selected". This
5062 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5063 no entries in it ("Shell widget menu has zero width and/or height"). So I
5064 have added a single, blank menu entry in this case.
5066 9. Added ${quote_local_part.
5068 10. MIME decoding is now applied to the contents of Subject: header lines when
5071 11. Now that a reference to $sender_host_address automatically causes a reverse
5072 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5073 host lookup before query-style lookups in lists that might use this
5074 variable. This has therefore been abolished, and the "net-" prefix is no
5075 longer necessary for query-style lookups.
5077 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5078 have been a typo for LFLAGS, so it has been changed.
5080 13. The install script calls Exim with "-C /dev/null" in order to find the
5081 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5082 to be output. Howeve, since Exim outputs its version number before the
5083 error, it didn't break the script. It just looked ugly. I fixed this by
5084 always allowing "-C /dev/null" if the caller is root.
5086 14. Ignore overlarge ACL variable number when reading spool file - insurance
5087 against a later release with more variables having written the file.
5089 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5090 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5091 this, as well as the form without the "IPv6" on the front.
5093 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5094 OS/Makefile-Darwin file.
5096 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5097 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5099 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5100 variable. However, some OS use ints for inodes. I've added cast to long int
5101 to get rid of the compiler warning.
5103 19. I had forgotten to lock out "/../" in configuration file names when
5104 ALT_CONFIG_PREFIX was set.
5106 20. Routers used for verification do not need to specify transports. However,
5107 if such a router generated a host list, and callout was configured, Exim
5108 crashed, because it could not find a port number from the (non-existent)
5109 transport. It now assumes port 25 in this circumstance.
5111 21. Added the -t option to exigrep.
5113 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5114 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5115 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5116 them is set, LOOKUP_LSEARCH is forced.
5118 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5120 24. Added sender and host information to the "rejected by local_scan()" log
5121 line; previously there was no indication of these.
5123 25. Added .include_if_exists.
5125 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5126 Linux. It turns out that not all file systems support this. Apparently some
5127 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5128 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5129 not supported on the file descriptor, is now ignored when Exim is trying to
5130 sync a directory. This applies only to Linux.
5132 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5134 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5135 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5138 29. Added support for maildirsize files from supplied patch (modified a bit).
5140 30. The use of :fail: followed by an empty string could lead Exim to respond to
5141 sender verification failures with (e.g.):
5143 550 Verification failed for <xxx>
5144 550 Sender verify failed
5146 where the first response line was missing the '-' that indicates it is not
5147 the final line of the response.
5149 31. The loop for finding the name of the user that called Exim had a hardwired
5150 limit of 10; it now uses the value of finduser_retries, which is used for
5151 all other user lookups.
5153 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5156 33. Exim was neglecting to zero errno before one call of strtol() when
5157 expanding a string and expecting an integer value. On some systems this
5158 resulted in spurious "integer overflow" errors. Also, it was casting the
5159 result into an int without checking.
5161 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5162 did not work. The code looks as if it has *never* worked, though it appears
5163 to have been documented since at least releast 1.62. I have made it work.
5165 35. The "timeout_DNS" error in retry rules, also documented since at least
5166 1.62, also never worked. As it isn't clear exactly what this means, and
5167 clearly it isn't a major issue, I have abolished the feature by treating it
5168 as "timeout", and writing a warning to the main and panic logs.
5170 36. The display of retry rules for -brt wasn't always showing the error code
5173 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5174 timeout_connect_A, timeout_connect_MX.
5176 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5177 to the empty sender.
5179 39. The daemon was not analysing the content of -oX till after it had closed
5180 stderr and disconnected from the controlling terminal. This meant that any
5181 syntax errors were only noted on the panic log, and the return code from
5182 the command was 0. By re-arranging the code a little, I've made the
5183 decoding happen first, so such errors now appear on stderr, and the return
5184 code is 1. However, the actual setting up of the sockets still happens in
5185 the disconnected process, so errors there are still only recorded on the
5188 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5189 connections (as happens on some IP stacks) was logged at start up time as
5190 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5191 it from "IPv6 and IPv4", which means that two separate sockets are being
5194 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5195 says whether AF_INET or AF_INET6 was passed as an argument.
5197 42. Exiwhat output was messed up when time zones were included in log
5200 43. Exiwhat now gives more information about the daemon's listening ports,
5201 and whether -tls-on-connect was used.
5203 44. The "port" option of the smtp transport is now expanded.
5205 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5206 silently ignored. Now an error message is written to the main and panic
5209 46. There's a new ACL modifier called "logwrite" which writes to a log file
5210 as soon as it is encountered.
5212 47. Added $local_user_uid and $local_user_gid at routing time.
5214 48. Exim crashed when trying to verify a sender address that was being
5217 49. Exim was recognizing only a space character after ".include". It now also
5218 recognizes a tab character.
5220 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5221 extracting the relevant information from the specification. The man page no
5222 longer contains scrambled data for the -d option, and I've added a section
5223 at the front about calling Exim under different names.
5225 51. Added "extra_headers" argument to the "mail" command in filter files.
5227 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5230 53. Installed eximstats 1.29.
5232 54. Added transport_filter_timeout as a generic transport option.
5234 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5235 Cc: header lines. This was required by RFC 822, but it not required by RFC
5238 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5239 incoming messages that did not have them. Now it does so only if the
5240 message originates locally, that is, if there is no associated remote host
5241 address. When Resent- header lines are present, this applies to the Resent-
5242 lines rather than the non-Resent- lines.
5244 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5245 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5247 58. Messages for configuration errors now include the name of the main
5248 configuration file - useful now that there may be more than one file in a
5249 list (.included file names were always shown).
5251 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5252 for those rare installations that do not start the daemon as root or run it
5253 setuid root. I've cut out the call to initgroups() if the daemon is not
5256 60. The Exim user and group can now be bound into the binary as text strings
5257 that are looked up at the start of Exim's processing.
5259 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5261 62. Added $mailstore_basename variable.
5263 63. Installed patch to sieve.c from Michael Haardt.
5265 64. When Exim failed to open the panic log after failing to open the main log,
5266 the original message it was trying to log was written to stderr and debug
5267 output, but if they were not available (the usual case in production), it
5268 was lost. Now it is written to syslog before the two lines that record the
5269 failures to open the logs.
5271 65. Users' Exim filters run in subprocesses under the user's uid. It is
5272 possible for a "deliver" command or an alias in a "personal" command to
5273 provoke an address rewrite. If logging of address rewriting is configured,
5274 this fails because the process is not running as root or exim. There may be
5275 a better way of dealing with this, but for the moment (because 4.30 needs
5276 to be released), I have disabled address rewrite logging when running a
5277 filter in a non-root, non-exim process.
5283 1. The buildconfig auxiliary program wasn't quoting the value set for
5284 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5285 not defined. This bug was masked in 4.22 by the effect that was fixed in
5288 2. Some messages that were rejected after a message id was allocated were
5289 shown as "incomplete" by exigrep. It no longer does this for messages that
5290 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5292 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5293 have allow_domain_literals set, the ID did not get logged in the <= line.
5294 Domain literals are now always recognized in Message-ID: header lines.
5296 4. The first argument for a ${extract expansion item is the key name or field
5297 number. Leading and trailing spaces in this item were not being ignored,
5298 causing some misleading effects.
5300 5. When deliver_drop_privilege was set, single queue runner processes started
5301 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5302 same command in the process it spins off) were not dropping privilege.
5304 6. When the daemon running as "exim" started a queue runner, it always
5305 re-executed Exim in the spun-off process. This is a waste of effort when
5306 deliver_drop_privilege is set. The new process now just calls the
5307 queue-runner function directly.
5313 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5316 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5317 ignored. Though the use of -oP was forcing the writing of a pid file, it
5318 was always written to the default place.
5320 3. If the message "no IP address found for host xxxx" is generated during
5321 incoming verification, it is now followed by identification of the incoming
5322 connection (so you can more easily find what provoked it).
5324 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5326 5. Added some features to "harden" Exim a bit more against certain attacks:
5328 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5329 be put in Local/Makefile. This is like the never_users runtime option,
5330 but it cannot be overridden. The default setting is "root".
5332 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5333 prefix string with which any file named in a -C command line option
5336 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5337 is retained for -C and -D only if the caller of Exim is root. Without
5338 it, the exim user may also use -C and -D and retain privilege.
5340 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5341 command line option is disabled.
5343 6. Macro names set by the -D option must start with an upper case letter, just
5344 like macro names defined in the configuration file.
5346 7. Added "dereference=" facility to LDAP.
5348 8. Two instances of the typo "uknown" in the source files are fixed.
5350 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5351 the Configure-Makefile script screwed up while processing it.
5353 10. Incorporated PCRE 4.4.
5355 11. The SMTP synchronization check was not operating right at the start of an
5356 SMTP session. For example, it could not catch a HELO sent before the client
5357 waited for the greeting. There is now a check for outstanding input at the
5358 point when the greeting is written. Because of the duplex, asynchronous
5359 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5360 way, but not yet received, when the check is performed.
5362 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5363 on TCP/IP sockets, because this apparently causes some broken clients to
5366 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5367 unchanged) from the Cygwin maintainer.
5369 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5370 was supported instead of testing for mbx. Effectively a typo.
5372 15. The spa authenticator server code was not checking that the input it
5373 received was valid base64.
5375 16. The debug output line for the "set" modifier in ACLs was not showing the
5376 name of the variable that was being set.
5378 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5380 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5381 lookup on its own. Something else was needed to trigger the lookup. For
5382 example, a match in host_lookup or the need for a host name in a host list.
5383 Now, if $sender_host_name is referenced and the host name has not yet been
5384 looked up, a lookup is performed. If the lookup fails, the variable remains
5385 empty, and $host_lookup_failed is set to "1".
5387 19. Added "eqi" as a case-independent comparison operator.
5389 20. The saslauthd authentication condition could segfault if neither service
5390 nor realm was specified.
5392 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5393 error message that was logged was misleading, and incoming SMTP
5394 connections were dropped. The message is now more accurate, and temporary
5395 errors are given to SMTP connections.
5397 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5398 (except RSET) with a 421 error, until QUIT is received. However, it was
5399 failing to send a response to QUIT.
5401 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5402 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5403 ACL is now run after the helo_try_verify_hosts code.
5405 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5406 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5407 case-independent, so other case variants are also recognized). Apparently
5408 some systems use these upper case variants.
5410 25. If more than two messages were waiting for the same host, and a transport
5411 filter was specified for the transport, Exim sent two messages over the
5412 same TCP/IP connection, and then failed with "socket operation on non-
5413 socket" when it tried to send the third.
5415 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5417 27. The extern definition of crypt16() in expand.c was not being excluded when
5418 the OS had its own crypt16() function.
5420 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5421 as a preferred synonym for return_size_limit, both as an option and as an
5424 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5426 30. Changed the default configuration ACL to relax the local part checking rule
5427 for addresses that are not in any local domains. For these addresses,
5428 slashes and pipe symbols are allowed within local parts, but the sequence
5429 /../ is explicitly forbidden.
5431 31. SPA server authentication was not clearing the challenge buffer before
5434 32. log_message in a "warn" ACL statement was writing to the reject log as
5435 well as to the main log, which contradicts the documentation and doesn't
5436 seem right (because no rejection is happening). So I have stopped it.
5438 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5439 However, I am unable to do any testing of this.
5441 34. Fixed an infelicity in the appendfile transport. When checking directories
5442 for a mailbox, to see if any needed to be created, it was accidentally
5443 using path names with one or more superfluous leading slashes; tracing
5444 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5446 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5447 contents are added to the log line that is written for every discarded
5448 recipient. (Previously a log_message setting was ignored.)
5450 36. The ${quote: operator now quotes the string if it is empty.
5452 37. The install script runs exim in order to find its version number. If for
5453 some reason other than non-existence or emptiness, which it checks, it
5454 could not run './exim', it was installing it with an empty version number,
5455 i.e. as "exim-". This error state is now caught, and the installation is
5458 38. An argument was missing from the function that creates an error message
5459 when Exim fails to connect to the socket for saslauthd authentication.
5460 This could cause Exim to crash, or give a corrupted message.
5462 39. Added isip, isip4, and isip6 to ${if conditions.
5464 40. The ACL variables $acl_xx are now saved with the message, and can be
5465 accessed later in routers, transports, and filters.
5467 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5468 strings in the file are not string-expanded.
5470 42. If a MAIL command specified a SIZE value that was too large to fit into an
5471 int variable, the check against message_size_limit failed. Such values are
5472 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5473 day this will have to be increased, but I don't think I want to be around
5474 when emails are that large.
5481 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5482 iconv() is not standard in FreeBSD.
5484 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5485 IPv6 enabled. The observed symptom was a segmentation fault on return from
5486 the function os_common_find_running_interfaces() in src/os.c.
5488 3. In the check_special_case() function in daemon.c I had used "errno" as an
5489 argument name, which causes warnings on some systems. This was basically a
5490 typo, since it was named "eno" in the comments!
5492 4. The code that waits for the clock to tick (at a resolution of some fraction
5493 of a second) so as to ensure message-id uniqueness was always waiting for
5494 at least one whole tick, when it could have waited for less. [This is
5495 almost certainly not relevant at current processor speeds, where it is
5496 unlikely to ever wait at all. But we try to future-proof.]
5498 5. The function that sleeps for a time interval that includes fractions of a
5499 second contained a race. It did not block SIGALRM between setting the
5500 timer, and suspending (a couple of lines later). If the interval was short
5501 and the sigsuspend() was delayed until after it had expired, the suspension
5502 never ended. On busy systems this could lead to processes getting stuck for
5505 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5506 process, before it forks any delivery processes. The open lookup caching
5507 mechanism meant that the open file or database connection was passed into
5508 the delivery process. The problem was that delivery processes always tidy
5509 up cached lookup data. This could cause a problem for the next delivery
5510 process started by the queue runner, because the external queue runner
5511 process does not know about the closure. So the next delivery process
5512 still has data in the lookup cache. In the case of a file lookup, there was
5513 no problem because closing a file descriptor in a subprocess doesn't affect
5514 the parent. However, if the lookup was caching a connection to a database,
5515 the connection was closed, and the second delivery process was likely to
5516 see errors such as "PGSQL: query failed: server closed the connection
5517 unexpectedly". The problem has been fixed by closing all cached lookups
5518 in a queue runner before running a delivery process.
5520 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5521 seem to have the "const" qualifier which it has on other OS. I've
5524 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5525 *of the same type* (client or server) with the same public name that an
5526 error should be diagnosed.
5528 9. When Exim looked up a host name for an IP address, but failed to find the
5529 original IP address when looking up the host name (a safety check), it
5530 output the message "<ip address> does not match any IP for NULL", which was
5531 confusing, to say the least. The bug was that the host name should have
5532 appeared instead of "NULL".
5534 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5535 user that is built into the binary, and the -C or -D options is used, root
5536 privilege is dropped before the configuration file is read. In addition,
5537 logging is switched to stderr instead of the normal log files. If the
5538 configuration then re-defines the Exim user, the unprivileged environment
5539 is probably not what is expected, so Exim logs a panic warning message (but
5542 However, if deliver_drop_privilege is set, the unprivileged state may well
5543 be exactly what is intended, so the warning has been cut out in that case,
5544 and Exim is allowed to try to write to its normal log files.
5550 1. smtp_return_error_details was not giving details for temporary sender
5551 or receiver verification errors.
5553 2. Diagnose a configuration error if two authenticators have the same public
5556 3. Exim used not to create the message log file for a message until the first
5557 delivery attempt. This could be confusing when incoming messages were held
5558 for policy or load reasons. The message log file is now created at the time
5559 the message is received, and an initial "Received" line is written to it.
5561 4. The automatically generated man page for command line options had a minor
5562 bug that caused no ill effects; however, a more serious problem was that
5563 the procedure for building the man page automatically didn't always
5564 operate. Consequently, release 4.20 contains an out-of-date version. This
5565 shouldn't happen again.
5567 5. When building Exim with embedded Perl support, the script that builds the
5568 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5569 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5571 6. The freeze_tell option was not being used for messages that were frozen on
5572 arrival, either by an ACL or by local_scan().
5574 7. Added the smtp_incomplete_transaction log selector.
5576 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5577 was accepting AUTH without a new EHLO.
5579 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5580 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5583 10. Logging of TCP/IP connections (when configured) now happens in the main
5584 daemon process instead of the child process, so that the TCP/IP connection
5585 count is more accurate (but it can never be perfect).
5587 11. The use of "drop" in a nested ACL was not being handled correctly in the
5588 outer ACL. Now, if condition failure induced by the nested "drop" causes
5589 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5590 or "require"), the connection is dropped.
5592 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5593 that yield "discard" can now be used with an "accept" or a "discard" verb,
5594 but an error is generated for any others (because I can't see a useful way
5595 to define what should happen).
5597 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5598 are now processed in the same way as lines in the Exim configuration file.
5599 In particular, continuation lines are supported.
5601 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5603 15. Added -ti meaning -t -i.
5605 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5606 domains, and warn by logging if others are found.
5608 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5609 returned from GIFCONF when Exim is trying to find the list of interfaces on
5610 a host. The code in os.c has been modified to copy each ifreq to an aligned
5611 structure in all cases.
5613 Also, in some cases, the returned ifreq's were being copied to a 'struct
5614 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5615 means the last couple of bytes of an IPv6 address could be chopped if the
5616 ifreq contained only a normal sockaddr (14 bytes storage).
5618 18. Named domain lists were not supported in the hosts_treat_as_local option.
5619 An entry such as +xxxx was not recognized, and was treated as a literal
5622 19. Ensure that header lines added by a DATA ACL are included in the reject log
5623 if the ACL subsequently rejects the message.
5625 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5626 MD5 (which is deprecated).
5628 21. When testing a filter file using -bf, Exim was writing a message when it
5629 took the sender from a "From " line in the message, but it was not doing so
5630 when it took $return_path from a Return-Path: header line. It now does.
5632 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5633 with a valid header line field name (a series of printing characters
5634 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5636 23. Changed "disc" in the source to "disk" to conform to the documentation and
5637 the book and for uniformity.
5639 24. Ignore Sendmail's -Ooption=value command line item.
5641 25. When execve() failed while trying to run a command in a pipe transport,
5642 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5643 could be confused with a return value of 69 from the command itself. This
5644 has been changed to 127, the value the shell returns if it is asked to run
5645 a non-existent command. The wording for the related log line suggests a
5646 non-existent command as the problem.
5648 26. If received_header_text expands to an empty string, do not add a Received:
5649 header line to the message. (Well, it adds a token one on the spool, but
5650 marks it "old" so that it doesn't get used or transmitted.)
5652 27. Installed eximstats 1.28 (addition of -nt option).
5654 28. There was no check for failure on the call to getsockname() in the daemon
5655 code. This can fail if there is a shortage of resources on the system, with
5656 ENOMEM, for example. A temporary error is now given on failure.
5658 29. Contrary to the C standard, it seems that in some environments, the
5659 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5660 program. Exim now does this explicitly; it affects the formatting of
5661 timestamps using strftime().
5663 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5664 complaints. I've now initialized all the variables, to avoid this.
5666 32. Header lines added by a system filter were not being "seen" during
5667 transport-time rewrites.
5669 33. The info_callback() function passed to OpenSSL is set up with type void
5670 (*)(SSL *, int, int), as described somewhere. However, when calling the
5671 function (actually a macro) that sets it up, the type void(*)() is
5672 expected. I've put in a cast to prevent warnings from picky compilers.
5674 34. If a DNS black list lookup found a CNAME record, but there were no A
5675 records associated with the domain it pointed at, Exim crashed.
5677 35. If a DNS black list lookup returned more than one A record, Exim ignored
5678 all but the first. It now scans all returned addresses if a particular IP
5679 value is being sought. In this situation, the contents of the
5680 $dnslist_value variable are a list of all the addresses, separated by a
5683 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5684 to accept a host name and all its aliases if the forward lookup for any of
5685 them yielded the IP address of the incoming connection. Now it accepts only
5686 those names whose forward lookup yields the correct IP address. Any other
5687 names are discarded. This closes a loophole whereby a rogue DNS
5688 administrator could create reverse DNS records to break through a
5689 wildcarded host restriction in an ACL.
5691 37. If a user filter or a system filter that ran in a subprocess used any of
5692 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5693 the wrong values were passed to the pipe command ($thisaddress had the
5694 value of $0, $0 had the value of $1, etc). This bug was introduced by
5695 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5697 38. Improved the line breaking for long SMTP error messages from ACLs.
5698 Previously, if there was no break point between 40 and 75 characters, Exim
5699 left the rest of the message alone. Two changes have been made: (a) I've
5700 reduced the minimum length to 35 characters; (b) if it can't find a break
5701 point between 35 and 75 characters, it looks ahead and uses the first one
5702 that it finds. This may give the occasional overlong line, but at least the
5703 remaining text gets split now.
5705 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5706 file descriptors might be low, and that setting 1000 would always raise it.
5707 It turns out that in some environments, the limit is already over 1000 and
5708 that lowering it causes trouble. So now Exim takes care not to decrease it.
5710 40. When delivering a message, the value of $return_path is set to $sender_
5711 address at the start of routing (routers may change the value). By an
5712 oversight, this default was not being set up when an address was tested by
5713 -bt or -bv, which affected the outcome if any router or filter referred to
5716 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5717 log only when "message" or "log_message" are set. However, if one of the
5718 conditions was an address verification, or a call to a nested ACL, the
5719 messages generated by the underlying test were being passed through. This
5720 no longer happens. The underlying message is available in $acl_verify_
5721 message for both "message" and "log_message" expansions, so it can be
5722 passed through if needed.
5724 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5725 new expansion $bh_ to give the encoded byte string without charset
5726 translation. Translation happens only if iconv() is available; HAVE_ICONV
5727 indicates this at build time. HEADERS_CHARSET gives the charset to
5728 translate to; headers_charset can change it in the configuration, and
5729 "headers charset" can change it in an individual filter file.
5731 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5732 that creates RFC 2047 encoded "words" labels them as that charset instead
5733 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5734 expansion operator; (ii) when Exim creates a From: line for a local
5735 message; (iii) when a header line is rewritten to include a "phrase" part.
5737 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5738 buggy, causing it to skip the first lines of messages whose message ID
5739 ended in 'D'. This would not have bitten before Exim release 4.14, because
5740 message IDs were unlikely to end in 'D' before then. The effect was to have
5741 incorrect size information for certain domains.
5743 45. #include "config.h" was missing at the start of the crypt16.c module. This
5744 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5747 46. If there was a timeout during a "random" callout check, Exim treated it as
5748 a failure of the random address, and carried on sending RSET and the real
5749 address. If the delay was just some slowness somewhere, the response to the
5750 original RCPT would be taken as a response to RSET and so on, causing
5751 mayhem of various kinds.
5753 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5754 when I implemented it. It didn't allow for the fact that some option values
5755 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5756 the right test for positive values.
5758 48. Domain names in DNS records are case-independent. Exim always looks them up
5759 in lower case. Some resolvers return domain names in exactly the case they
5760 appear in the zone file, that is, they may contain uppercase letters. Not
5761 all resolvers do this - some return always lower case. Exim was treating a
5762 change of case by a resolver as a change of domain, similar to a widening
5763 of a domain abbreviation. This triggered its re-routing code and so it was
5764 trying to route what was effectively the same domain again. This normally
5765 caused routing to fail (because the router wouldn't handle the domain
5766 twice). Now Exim checks for this case specially, and just changes the
5767 casing of the domain that it ultimately uses when it transmits the message
5770 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5773 50. If a filter generated a file delivery with a non-absolute name (possible if
5774 no home directory exists for the router), the forbid_file option was not
5777 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5778 the existing equality matching.
5780 52. Exim was using ints instead of ino_t variables in some places where it was
5781 dealing with inode numbers.
5783 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5784 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5785 called TMPDIR, and if it finds it is different, it changes its value.
5787 54. The smtp_printf() function is now made available to local_scan() so
5788 additional output lines can be written before returning. There is also an
5789 smtp_fflush() function to enable the detection of a dropped connection.
5790 The variables smtp_input and smtp_batched_input are exported to
5793 55. Changed the default runtime configuration: the message "Unknown user"
5794 has been removed from the ACL, and instead placed on the localuser router,
5795 using the cannot_route_message feature. This means that any verification
5796 failures that generate their own messages won't get overridden. Similarly,
5797 the "Unrouteable address" message that was in the ACL for unverifiable
5798 relay addresses has also been removed.
5800 56. Added hosts_avoid_esmtp to the smtp transport.
5802 57. The exicyclog script was not checking for the esoteric option
5803 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5804 will work only if exicyclog is run under the appropriate euid.
5806 58. Following a discussion on the list, the rules by which Exim recognises line
5807 endings on incoming messages have been changed. The -dropcr and drop_cr
5808 options are now no-ops, retained only for backwards compatibility. The
5809 following line terminators are recognized: LF CRLF CR. However, special
5810 processing applies to CR:
5812 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5813 nor a local message in the state where . is a terminator.
5815 (ii) If a bare CR is encountered in a header line, an extra space is added
5816 after the line terminator so as not to end the header. The reasoning
5817 behind this is that bare CRs in header lines are most likely either
5818 to be mistakes, or people trying to play silly games.
5820 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5821 was being incorrectly given as 18 bytes larger than it should have been.
5822 This is a VOB (very old bug).
5824 60. This may never have affected anything current, but just in case it has:
5825 When the local host is found other than at the start of a list of hosts,
5826 the local host, those with the same MX, and any that follow, are discarded.
5827 When the list in question was part of a longer list of hosts, the following
5828 hosts (not currently being processed) were also being discarded. This no
5829 longer happens. I'm not sure if this situation could ever has previously
5832 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5835 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5836 header. This is something that is recommended in a new Internet Draft, and
5837 is something that is documented as being done by Sendmail. There are two
5838 possible values. For messages generated by the autoreply transport, Exim
5841 Auto-Submitted: auto-replied
5843 whereas for all other generated messages (e.g. bounces) it adds
5845 Auto-Submitted: auto-generated
5847 63. The "personal" condition in filters now includes a test for the
5848 Auto-Submitted: header. If it contains the string "auto-" the message it
5849 not considered personal.
5851 64. Added rcpt_include_affixes as a generic transport option.
5853 65. Added queue_only_override (default true).
5855 66. Added the syslog_duplication option.
5857 67. If what should have been the first header line of a message consisted of
5858 a space followed by a colon, Exim was mis-interpreting it as a header line.
5859 It isn't of course - it is syntactically invalid and should therefore be
5860 treated as the start of the message body. The misbehaviour could have
5861 caused a number of strange effects, including loss of data in subsequent
5862 header lines, and spool format errors.
5864 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5865 client host had authenticated. This control can now be exercised by an ACL
5866 for more flexibility.
5868 69. By default, callouts do not happen when testing with -bh. There is now a
5869 variant, -bhc, which does actually run the callout code, including
5870 consulting and updating the callout cache.
5872 70. Added support for saslauthd authentication, courtesy of Alexander
5875 71. If statvfs() failed on the spool or log directories while checking their
5876 size for availability, Exim confusingly gave the error "space shortage".
5877 Furthermore, in debugging mode it crashed with a floating point exception.
5878 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5879 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5880 serious problem, Exim now writes to the main and panic logs when this
5881 happens, with details of the failure. It then refuses to accept the
5882 incoming message, giving the message "spool directory problem" or "log
5883 directory problem" with a 421 code for SMTP messages.
5885 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5886 0, 1, and 2 exist, because some OS complain for execs without them (see
5887 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5888 descriptors. However, the code omitted to check that the open succeeded,
5889 causing mysterious errors if for some reason the permissions on /dev/null
5890 got screwed. Now Exim writes a message to the main and panic logs, and
5891 bombs out if it can't open /dev/null.
5893 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5894 interact so that it is all more flexible. It is supposed to remain
5895 backwards compatible. Also added extra_local_interfaces.
5897 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5898 to bomb out with an assertion failure - to the client this appears as a
5899 connection drop. This problem occurs in the part of the code that was taken
5900 from the Samba project. Fortunately, the assertion is in a very simple
5901 function, so I have fixed this by reproducing the function inline in the
5902 one place where it is called, and arranging for authentication to fail
5903 instead of killing the process with assert().
5905 75. The SPA client code was not working when the server requested OEM rather
5906 than Unicode encoding.
5908 76. Added code to make require_files with a specific uid setting more usable in
5909 the case where statting the file as root fails - usually a non-root-mounted
5910 NFS file system. When this happens and the failure is EACCES, Exim now
5911 forks a subprocess and does the per-uid checking as the relevant uid.
5913 77. Added process_log_path.
5915 78. If log_file_path was not explicitly set, a setting of check_log_space or
5916 check_log_inodes was ignored.
5918 79. If a space check for the spool or log partitions fails, the incident is now
5919 logged. Of course, in the latter case the data may get lost...
5921 80. Added the %p formatting code to string_format() so that it can be used to
5922 print addresses in debug_print(). Adjusted all the address printing in the
5923 debugging in store.c to use %p rather than %d.
5925 81. There was a concern that a line of code in smtp_in.c could overflow a
5926 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5927 initially expressed, the concern was not well-founded, because trailing
5928 spaces are removed early. However, if the trailing spaces were followed by
5929 a NULL, they did not get removed, so the overflow was possible. Two fixes
5932 (a) I re-wrote the offending code in a cleaner fashion.
5933 (b) If an incoming SMTP command contains a NULL character, it is rejected
5936 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5937 runs initgroups(), so that if the Exim user is in any additional groups,
5938 they will be used during message reception.
5944 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob