1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
36 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
38 PP/02 Make -n do something, by making it not do something.
39 When combined with -bP, the name of an option is not output.
41 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
44 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
45 $sender_host_name and config options to manage this, and basic check
48 PP/05 DSCP support for outbound connections and control modifier for inbound.
50 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
51 (Only plugin which currently uses this is kerberos4, which nobody should
52 be using, but we should make it available and other future plugins might
53 conceivably use it, even though it would break NAT; stuff *should* be
54 using channel bindings instead).
56 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
57 name; added for Sendmail compatibility; requires admin caller.
58 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
59 just ignore it); requires trusted caller.
60 Also parse but ignore: -Ac -Am -X<logfile>
63 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
65 TL/02 Add +smtp_confirmation as a default logging option.
67 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
68 Patch by Magnus Holmgren from 2007-02-20.
70 TL/04 Bugzilla 1281 - Spec typo.
71 Bugzilla 1283 - Spec typo.
72 Bugzilla 1290 - Spec grammar fixes.
74 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
76 TL/06 Add Experimental DMARC support using libopendmarc libraries.
78 TL/07 Fix an out of order global option causing a segfault. Reported to dev
79 mailing list by by Dmitry Isaikin.
81 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
83 JH/02 Support "G" suffix to numbers in ${if comparisons.
85 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
87 NM/01 Bugzilla 1197 - Spec typo
88 Bugzilla 1196 - Spec examples corrections
90 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
92 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
93 gnutls_enable_pkcs11, but renamed to more accurately indicate its
96 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
97 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
99 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
100 "acl {{name}{arg}...}", and optional args on acl condition
103 JH/05 Permit multiple router/transport headers_add/remove lines.
105 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
107 JH/07 Avoid using a waiting database for a single-message-only transport.
108 Performance patch from Paul Fisher. Bugzilla 1262.
110 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
113 JH/09 Add $headers_added variable, with content from use of ACL modifier
114 add_header (but not yet added to the message). Bugzilla 199.
116 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
117 Pulled from Bugzilla 817 by Wolfgang Breyha.
119 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
121 (nb: this is the same fix as in Exim 4.80.1)
123 JH/11 Add A= logging on delivery lines, and a client_set_id option on
126 JH/12 Add optional authenticated_sender logging to A= and a log_selector
129 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
131 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
132 advertise SMTP AUTH mechanism to us, instead of a generic
133 protocol violation error. Also, make Exim more robust to bad
134 data from the Dovecot auth socket.
136 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
138 When a queue runner is handling a message, Exim first routes the
139 recipient addresses, during which it prunes them based on the retry
140 hints database. After that it attempts to deliver the message to
141 any remaining recipients. It then updates the hints database using
144 So if a recipient address works intermittently, it can get repeatedly
145 deferred at routing time. The retry hints record remains fresh so the
146 address never reaches the final cutoff time.
148 This is a fairly common occurrence when a user is bumping up against
149 their storage quota. Exim had some logic in its local delivery code
150 to deal with this. However it did not apply to per-recipient defers
151 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
153 This change adds a proper retry rule check during routing so that the
154 final cutoff time is checked against the message's age. We only do
155 this check if there is an address retry record and there is not a
156 domain retry record; this implies that previous attempts to handle
157 the address had the retry_use_local_parts option turned on. We use
158 this as an approximation for the destination being like a local
159 delivery, as in LMTP.
161 I suspect this new check makes the old local delivery cutoff check
162 redundant, but I have not verified this so I left the code in place.
164 TF/02 Correct gecos expansion when From: is a prefix of the username.
166 Test 0254 submits a message to Exim with the header
170 When I ran the test suite under the user fanf2, Exim expanded
171 the header to contain my full name, whereas it should have added
172 a Resent-Sender: header. It erroneously treats any prefix of the
173 username as equal to the username.
175 This change corrects that bug.
177 GF/01 DCC debug and logging tidyup
178 Error conditions log to paniclog rather than rejectlog.
179 Debug lines prefixed by "DCC: " to remove any ambiguity.
181 TF/03 Avoid unnecessary rebuilds of lookup-related code.
183 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
184 Bug spotted by Jeremy Harris; was flawed since initial commit.
185 Would have resulted in OCSP responses post-SNI triggering an Exim
186 NULL dereference and crash.
188 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
190 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
191 Bug detection, analysis and fix by Samuel Thibault.
192 Bugzilla 1331, Debian bug #698092.
194 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
196 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
197 Server implementation by Todd Lyons, client by JH.
198 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
199 config variable "prdr_enable" controls whether the server
200 advertises the facility. If the client requests PRDR a new
201 acl_data_smtp_prdr ACL is called once for each recipient, after
202 the body content is received and before the acl_smtp_data ACL.
203 The client is controlled by bolth of: a hosts_try_prdr option
204 on the smtp transport, and the server advertisement.
205 Default client logging of deliveries and rejections involving
206 PRDR are flagged with the string "PRDR".
208 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
209 fclose(). Diagnosis by Todd Lyons.
211 PP/17 Update configure.default to handle IPv6 localhost better.
212 Patch by Alain Williams (plus minor tweaks).
215 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
216 This is now consistent with GnuTLS, and is now documented: the
217 previous undocumented portable approach to treating the option as
218 unset was to force an expansion failure. That still works, and
219 an empty string is now equivalent.
221 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
222 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
223 not performing validation itself.
225 PP/20 Added force_command boolean option to pipe transport.
226 Patch from Nick Koston, of cPanel Inc.
228 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
231 TF/04 Added udpsend ACL modifer and hexquote expansion operator
233 PP/21 Fix eximon continuous updating with timestamped log-files.
234 Broken in a format-string cleanup in 4.80, missed when I repaired the
235 other false fix of the same issue.
236 Report and fix from Heiko Schlichting.
239 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
240 Report from Prashanth Katuri.
242 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
243 It's SecureTransport, so affects any MacOS clients which use the
244 system-integrated TLS libraries, including email clients.
246 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
247 using a MIME ACL for non-SMTP local injection.
248 Report and assistance in diagnosis by Warren Baker.
250 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
252 JH/16 Fix comparisons for 64b. Bugzilla 1385.
254 TL/09 Add expansion variable $authenticated_fail_id to keep track of
255 last id that failed so it may be referenced in subsequent ACL's.
257 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
260 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
261 ldap library initialization, allowing self-signed CA's to be
262 used. Also properly sets require_cert option later in code by
263 using NULL (global ldap config) instead of ldap handle (per
264 session). Bug diagnosis and testing by alxgomz.
266 TL/12 Enhanced documentation in the ratelimit.pl script provided in
267 the src/util/ subdirectory.
269 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
270 renamed to Transport Post Delivery Action by Jeremy Harris, as
273 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
274 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
275 redis_servers = needs to be configured which will be used by the redis
276 lookup. Patch from Warren Baker, of The Packet Hub.
278 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
280 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
281 hostname or reverse DNS when processing a host list. Used suggestions
282 from multiple comments on this bug.
284 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
286 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
287 Missed a few lines, added it to make the runtest require no keyboard
290 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
291 contains upper case chars. Make router use caseful_local_part.
293 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
294 support when GnuTLS has been built with p11-kit.
300 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
302 This, or similar/improved, will also be change PP/11 of 4.82.
308 PP/01 Handle short writes when writing local log-files.
309 In practice, only affects FreeBSD (8 onwards).
310 Bugzilla 1053, with thanks to Dmitry Isaikin.
312 NM/01 Bugzilla 949 - Documentation tweak
314 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
317 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
319 PP/02 Implemented gsasl authenticator.
321 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
323 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
324 `pkg-config foo` for cflags/libs.
326 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
327 with rest of GSASL and with heimdal_gssapi.
329 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
330 `pkg-config foo` for cflags/libs for the TLS implementation.
332 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
333 properties get this fed in as external SSF. A number of robustness
334 and debugging improvements to the cyrus_sasl authenticator.
336 PP/08 cyrus_sasl server now expands the server_realm option.
338 PP/09 Bugzilla 1214 - Log authentication information in reject log.
339 Patch by Jeremy Harris.
341 PP/10 Added dbmjz lookup type.
343 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
345 PP/12 MAIL args handles TAB as well as SP, for better interop with
346 non-compliant senders.
347 Analysis and variant patch by Todd Lyons.
349 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
350 Bug report from Lars Müller <lars@samba.org> (via SUSE),
351 Patch from Dirk Mueller <dmueller@suse.com>
353 PP/13 tls_peerdn now print-escaped for spool files.
354 Observed some $tls_peerdn in wild which contained \n, which resulted
355 in spool file corruption.
357 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
358 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
359 or write after TLS renegotiation, which otherwise led to messages
362 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
363 as a tracking header (ie: a signed header comes before the signature).
364 Patch from Wolfgang Breyha.
366 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
367 comma-sep list; embedded commas doubled.
369 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
371 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
373 Report and patch from Dmitry Banschikov.
375 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
376 Removed SSL_clear() after SSL_new() which led to protocol negotiation
377 failures. We appear to now support TLS1.1+ with Exim.
379 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
380 lets Exim select keys and certificates based upon TLS SNI from client.
381 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
382 before an outbound SMTP session. New log_selector, +tls_sni.
384 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
385 NULL dereference. Report and patch from Alun Jones.
387 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
388 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
389 Not seeing resolver debug output on NetBSD, but suspect this is a
390 resolver implementation change.
392 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
393 Left warnings. Added "eximon gdb" invocation mode.
395 PP/21 Defaulting "accept_8bitmime" to true, not false.
397 PP/22 Added -bw for inetd wait mode support.
399 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
400 locate the relevant includes and libraries. Made this the default.
402 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
403 Bugzilla 1246, report and most of solution from Tomasz Kusy.
405 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
406 This may cause build issues on older platforms.
408 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
409 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
410 gnutls_require_mac & gnutls_require_protocols (no longer supported).
411 Added SNI support via GnuTLS too.
412 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
414 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
416 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
417 Applied second patch from Janne, implementing suggestion to default
418 multiple-strings-in-record handling to match SPF spec.
420 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
422 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
423 read-only, out of scope).
424 Patch from Wolfgang Breyha, report from Stuart Northfield.
426 PP/29 Fix three issues highlighted by clang analyser static analysis.
427 Only crash-plausible issue would require the Cambridge-specific
428 iplookup router and a misconfiguration.
429 Report from Marcin Mirosław.
431 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
433 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
434 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
435 As part of this, removing so much warning spew let me fix some minor
436 real issues in debug logging.
438 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
439 assignment on my part. Fixed.
441 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
442 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
443 Janne Snabb (who went above and beyond: thank you).
445 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
446 string otherwise requires a connection and a bunch more work and it's
447 relatively easy to get wrong. Should also expose TLS library linkage
450 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
451 64-bit ${eval} (JH/03).
453 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
454 GNU libc to support some of the 64-bit stuff, should not lead to
455 conflicts. Defined before os.h is pulled in, so if a given platform
456 needs to override this, it can.
458 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
459 protection layer was required, which is not implemented.
460 Bugzilla 1254, patch from Wolfgang Breyha.
462 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
463 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
464 tls_dhparam take prime identifiers. Also unbreak combination of
465 OpenSSL+DH_params+TLSSNI.
467 PP/39 Disable SSLv2 by default in OpenSSL support.
473 PP/01 Solaris build fix for Oracle's LDAP libraries.
474 Bugzilla 1109, patch from Stephen Usher.
476 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
478 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
481 TF/02 Fix a couple more cases where we did not log the error message
482 when unlink() failed. See also change 4.74-TF/03.
484 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
485 lock up or crash if it happened to be inside a call to libc when it
486 got a SIGUSR1 from exiwhat.
488 The SIGUSR1 handler appends the current process status to the process
489 log which is later printed by exiwhat. It used to use the general
490 purpose logging code to do this, but several functions it calls are
491 not safe for signals.
493 The new output code in the SIGUSR1 handler is specific to the process
494 log, and simple enough that it's easy to inspect for signal safety.
495 Removing some special cases also simplifies the general logging code.
496 Removing the spurious timestamps from the process log simplifies
499 TF/04 Improved ratelimit ACL condition.
501 The /noupdate option has been deprecated in favour of /readonly which
502 has clearer semantics. The /leaky, /strict, and /readonly update modes
503 are mutually exclusive. The update mode is no longer included in the
504 database key; it just determines when the database is updated. (This
505 means that when you upgrde Exim will forget old rate measurements.)
507 Exim now checks that the per_* options are used with an update mode that
508 makes sense for the current ACL. For example, when Exim is processing a
509 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
510 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
511 must specify per_mail/readonly. If you omit the update mode it defaults to
512 /leaky where that makes sense (as before) or /readonly where required.
514 The /noupdate option is now undocumented but still supported for
515 backwards compatibility. It is equivalent to /readonly except that in
516 ACLs where /readonly is required you may specify /leaky/noupdate or
517 /strict/noupdate which are treated the same as /readonly.
519 A useful new feature is the /count= option. This is a generalization
520 of the per_byte option, so that you can measure the throughput of other
521 aggregate values. For example, the per_byte option is now equivalent
522 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
524 The per_rcpt option has been generalized using the /count= mechanism
525 (though it's more complicated than the per_byte equivalence). When it is
526 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
527 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
528 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
529 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
530 Note that using per_rcpt with a non-readonly update mode in more than
531 one ACL will cause the recipients to be double-counted. (The per_mail
532 and per_byte options don't have this problem.)
534 The handling of very low rates has changed slightly. If the computed rate
535 is less than the event's count (usually one) then this event is the first
536 after a long gap. In this case the rate is set to the same as this event's
537 count, so that the first message of a spam run is counted properly.
539 The major new feature is a mechanism for counting the rate of unique
540 events. The new per_addr option counts the number of different
541 recipients that someone has sent messages to in the last time period. It
542 behaves like per_rcpt if all the recipient addresses are different, but
543 duplicate recipient addresses do not increase the measured rate. Like
544 the /count= option this is a general mechanism, so the per_addr option
545 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
546 example, measure the rate that a client uses different sender addresses
547 with the options per_mail/unique=$sender_address. There are further
548 details in the main documentation.
550 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
552 TF/06 Removed a few PCRE remnants.
554 TF/07 Automatically extract Exim's version number from tags in the git
555 repository when doing development or release builds.
557 PP/02 Raise smtp_cmd_buffer_size to 16kB.
558 Bugzilla 879. Patch from Paul Fisher.
560 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
561 Heavily based on revision 40f9a89a from Simon Arlott's tree.
564 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
566 PP/05 Variable $av_failed, true if the AV scanner deferred.
567 Bugzilla 1078. Patch from John Horne.
569 PP/06 Stop make process more reliably on build failure.
570 Bugzilla 1087. Patch from Heiko Schlittermann.
572 PP/07 Make maildir_use_size_file an _expandable_ boolean.
573 Bugzilla 1089. Patch from Heiko Schlittermann.
575 PP/08 Handle ${run} returning more data than OS pipe buffer size.
576 Bugzilla 1131. Patch from Holger Weiß.
578 PP/09 Handle IPv6 addresses with SPF.
579 Bugzilla 860. Patch from Wolfgang Breyha.
581 PP/10 GnuTLS: support TLS 1.2 & 1.1.
583 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
586 PP/11 match_* no longer expand right-hand-side by default.
587 New compile-time build option, EXPAND_LISTMATCH_RHS.
588 New expansion conditions, "inlist", "inlisti".
590 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
592 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
594 PP/14 fix log_write() format string regression from TF/03.
595 Bugzilla 1152. Patch from Dmitry Isaikin.
601 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
603 PP/02 Harmonised TLS library version reporting; only show if debugging.
604 Layout now matches that introduced for other libraries in 4.74 PP/03.
606 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
608 PP/04 New "dns_use_edns0" global option.
610 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
613 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
614 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
616 TK/01 Updated PolarSSL code to 0.14.2.
617 Bugzilla 1097. Patch from Andreas Metzler.
619 PP/07 Catch divide-by-zero in ${eval:...}.
622 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
625 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
626 format-string attack -- SECURITY: remote arbitrary code execution.
628 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
629 time unintentionally subject to list matching rules, letting the header
630 cause arbitrary Exim lookups (of items which can occur in lists, *not*
631 arbitrary string expansion). This allowed for information disclosure.
633 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
634 INT_MIN/-1 -- value coerced to INT_MAX.
640 NM/01 Workround for PCRE version dependency in version reporting
643 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
644 This fixes portability to compilers other than gcc, notably
645 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
647 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
648 makefiles for portability to HP-UX and POSIX correctness.
650 PP/01 Permit LOOKUP_foo enabling on the make command-line.
651 Also via indented variable definition in the Makefile.
652 (Debugging by Oliver Heesakkers).
654 PP/02 Restore caching of spamd results with expanded spamd_address.
655 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
657 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
658 Improves build reliability. Fix from: Frank Elsner
660 NM/02 Fix wide character breakage in the rfc2047 coding
661 Fixes bug 1064. Patch from Andrey N. Oktyabrski
663 NM/03 Allow underscore in dnslist lookups
664 Fixes bug 1026. Patch from Graeme Fowler
666 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
667 Code patches from Adam Ciarcinski of NetBSD.
669 NM/04 Fixed exiqgrep to cope with mailq missing size issue
672 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
673 is logged, to avoid truncation. Patch from John Horne.
675 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
676 Patch from Jakob Hirsch.
678 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
679 SQL string expansion failure details.
680 Patch from Andrey Oktyabrski.
682 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
683 Patch from Simon Arlott.
685 PP/09 New lookups functionality failed to compile on old gcc which rejects
686 extern declarations in function scope.
687 Patch from Oliver Fleischmann
689 PP/10 Use sig_atomic_t for flags set from signal handlers.
690 Check getgroups() return and improve debugging.
691 Fixed developed for diagnosis in bug 927 (which turned out to be
694 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
695 Patch from Mark Zealey.
697 PP/12 Bugzilla 1056: Improved spamd server selection.
698 Patch from Mark Zealey.
700 PP/13 Bugzilla 1086: Deal with maildir quota file races.
701 Based on patch from Heiko Schlittermann.
703 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
704 Patch from Uwe Doering, sign-off by Michael Haardt.
706 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
707 variable declaration deep within a block. Bug and patch from
710 PP/15 lookups-Makefile IRIX compatibilty coercion.
712 PP/16 Make DISABLE_DKIM build knob functional.
714 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
715 Patch by Simon Arlott
717 TF/03 Fix valgrind.h portability to C89 compilers that do not support
718 variable argument macros. Our copy now differs from upstream.
724 TF/01 Failure to get a lock on a hints database can have serious
725 consequences so log it to the panic log.
727 TF/02 Log LMTP confirmation messages in the same way as SMTP,
728 controlled using the smtp_confirmation log selector.
730 TF/03 Include the error message when we fail to unlink a spool file.
732 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
733 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
734 for maintaining out-of-tree patches for some time.
736 PP/01 Bugzilla 139: Documentation and portability issues.
737 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
738 Handle per-OS dynamic-module compilation flags.
740 PP/02 Let /dev/null have normal permissions.
741 The 4.73 fixes were a little too stringent and complained about the
742 permissions on /dev/null. Exempt it from some checks.
743 Reported by Andreas M. Kirchwitz.
745 PP/03 Report version information for many libraries, including
746 Exim version information for dynamically loaded libraries. Created
747 version.h, now support a version extension string for distributors
748 who patch heavily. Dynamic module ABI change.
750 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
751 privilege escalation vulnerability whereby the Exim run-time user
752 can cause root to append content of the attacker's choosing to
755 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
758 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
759 If dropping privileges for untrusted macros, we disabled normal logging
760 on the basis that it would fail; for the Exim run-time user, this is not
761 the case, and it resulted in successful deliveries going unlogged.
762 Fixed. Reported by Andreas Metzler.
768 PP/01 Date: & Message-Id: revert to normally being appended to a message,
769 only prepend for the Resent-* case. Fixes regression introduced in
770 Exim 4.70 by NM/22 for Bugzilla 607.
772 PP/02 Include check_rfc2047_length in configure.default because we're seeing
773 increasing numbers of administrators be bitten by this.
775 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
777 PP/03 Bugzilla 994: added openssl_options main configuration option.
779 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
781 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
783 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
785 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
786 without a peer certificate, leading to a segfault because of an
787 assumption that peers always have certificates. Be a little more
788 paranoid. Problem reported by Martin Tscholak.
790 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
791 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
792 NB: ClamAV planning to remove STREAM in "middle of 2010".
793 CL also introduces -bmalware, various -d+acl logging additions and
794 more caution in buffer sizes.
796 PP/09 Implemented reverse_ip expansion operator.
798 PP/10 Bugzilla 937: provide a "debug" ACL control.
800 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
802 PP/12 Bugzilla 973: Implement --version.
804 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
806 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
808 PP/15 Bugzilla 816: support multiple condition rules on Routers.
810 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
811 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
812 ignore trailing whitespace.
814 JJ/02 prevent non-panic DKIM error from being sent to paniclog
816 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
819 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
820 Notification from Dr Andrew Aitchison.
822 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
823 ExtendedDetectionInfo response format.
824 Notification from John Horne.
826 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
829 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
830 XSL and documented dependency on system catalogs, with examples of how
833 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
836 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
837 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
838 configuration file which is writeable by the Exim user or group.
840 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
841 of configuration files to cover files specified with the -C option if
842 they are going to be used with root privileges, not just the default
845 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
846 option (effectively making it always true).
848 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
849 files to be used while preserving root privileges.
851 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
852 that rogue child processes cannot use them.
854 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
855 run-time user, instead of root.
857 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
858 Exim run-time user without dropping privileges.
860 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
861 result string, instead of calling string_vformat() twice with the same
864 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
865 for other users. Others should always drop root privileges if they use
866 -C on the command line, even for a whitelisted configure file.
868 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
870 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
876 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
877 $data_path, and $header_path variables; fixed documentation bugs and
880 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
881 exipick to access non-standard spools, including the "frozen" queue
884 NM/01 Bugzilla 965: Support mysql stored procedures.
885 Patch from Alain Williams
887 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
889 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
890 Patch from Andreas Metzler
892 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
893 Patch from Kirill Miazine
895 NM/05 Bugzilla 671: Added umask to procmail example.
897 JJ/03 installed exipick 20100323.0, fixing doc bug
899 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
900 directory. Notification and patch from Dan Rosenberg.
902 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
904 TK/02 Improve log output when DKIM signing operation fails.
906 MH/01 Treat the transport option dkim_domain as a colon separated
907 list, not as a single string, and sign the message with each element,
908 omitting multiple occurences of the same signer.
910 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
911 Bugzilla 985, 986. Patch by Simon Arlott
913 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
914 Patch by Simon Arlott
916 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
917 MBX locking. Notification from Dan Rosenberg.
923 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
925 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
927 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
929 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
931 NM/04 Bugzilla 915: Flag broken perl installation during build.
937 TK/01 Added patch by Johannes Berg that expands the main option
938 "spamd_address" if it starts with a dollar sign.
940 TK/02 Write list of recipients to X-Envelope-Sender header when building
941 the mbox-format spool file for content scanning (suggested by Jakob
944 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
945 (http://www.dcc-servers.net/) support via dccifd. Activated by
946 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
948 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
949 by Mark Daniel Reidel <mr@df.eu>.
951 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
952 When building exim an external PCRE library is now needed -
953 PCRE is a system library on the majority of modern systems.
954 See entry on PCRE_LIBS in EDITME file.
956 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
957 conversation. Added nologin parameter to request.
958 Patch contributed by Kirill Miazine.
960 TF/01 Do not log submission mode rewrites if they do not change the address.
962 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
964 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
965 log files in place. Contributed by Roberto Lima.
967 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
969 TF/03 Bugzilla 615: When checking the local_parts router precondition
970 after a local_part_suffix or local_part_prefix option, Exim now
971 does not use the address's named list lookup cache, since this
972 contains cached lookups for the whole local part.
974 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
975 Robert Millan. Documentation is in experimental-spec.txt.
977 TF/04 Bugzilla 668: Fix parallel build (make -j).
979 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
981 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
982 Patch provided by Jan Srzednicki.
984 TF/05 Leading white space used to be stripped from $spam_report which
985 wrecked the formatting. Now it is preserved.
987 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
988 that they are available at delivery time.
990 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
992 TF/08 TLS error reporting now respects the incoming_interface and
993 incoming_port log selectors.
995 TF/09 Produce a more useful error message if an SMTP transport's hosts
996 setting expands to an empty string.
998 NM/06 Bugzilla 744: EXPN did not work under TLS.
999 Patch provided by Phil Pennock.
1001 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1002 Patch provided by Richard Godbee.
1004 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1005 acl_smtp_notquit, added index entry.
1007 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1008 Patch provided by Eugene Bujak.
1010 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1011 accept(). Patch provided by Maxim Dounin.
1013 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1014 Patch provided by Phil Pennock.
1016 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1018 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1019 Patch provided by Brad "anomie" Jorsch.
1021 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1022 Patch provided by Dean Brooks.
1024 TK/05 Add native DKIM support (does not depend on external libraries).
1026 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1027 Patch provided by Graeme Fowler.
1029 NM/16 Bugzilla 851: Documentation example syntax fix.
1031 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1033 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1036 NM/19 Bugzilla 745: TLS version reporting.
1037 Patch provided by Phil Pennock.
1039 NM/20 Bugzilla 167: bool: condition support.
1040 Patch provided by Phil Pennock.
1042 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1043 clients. Patch provided by Phil Pennock.
1045 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1046 Patch provided by Brad "anomie" Jorsch.
1048 NM/23 Bugzilla 687: Fix misparses in eximstats.
1049 Patch provided by Heiko Schlittermann.
1051 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1052 Patch provided by Heiko Schlittermann.
1054 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1055 plus update to original patch.
1057 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1059 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1060 Patch provided by David Brownlee.
1062 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1064 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1066 NM/30 Bugzilla 888: TLS documentation bugfixes.
1068 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1070 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1071 Unlike the original bugzilla I have changed all shell scripts in src tree.
1073 NM/33 Bugzilla 898: Transport filter timeout fix.
1074 Patch by Todd Rinaldo.
1076 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1077 Patch by Serge Demonchaux.
1079 NM/35 Bugzilla 39: Base64 decode bug fixes.
1080 Patch by Jakob Hirsch.
1082 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1084 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1086 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1088 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1094 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1095 ALT-N's libdkim that I have put here:
1096 http://duncanthrax.net/exim-experimental/
1098 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1099 were called 'true' and it seems that C99 defines that as a reserved
1100 keyword to be used with 'bool' variable types. That means you could
1101 not include C99-style headers which use bools without triggering
1102 build errors in sieve.c.
1104 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1105 as mailq or other aliases. Changed the --help handling significantly
1106 to do whats expected. exim_usage() emits usage/help information.
1108 SC/01 Added the -bylocaldomain option to eximstats.
1110 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1112 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1114 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1120 PH/01 Another patch from the Sieve maintainer.
1122 PH/02 When an IPv6 address is converted to a string for single-key lookup
1123 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1124 dots are used instead of colons so that keys in lsearch files need not
1125 contain colons. This was done some time before quoting was made available
1126 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1127 (notated using the quote facility) so as to distinguish them from IPv4
1128 keys. This meant that lookups for IP addresses in host lists did not work
1129 for iplsearch lookups.
1131 This has been fixed by arranging for IPv6 addresses to be expressed with
1132 colons if the lookup type is iplsearch. This is not incompatible, because
1133 previously such lookups could never work.
1135 The situation is now rather anomolous, since one *can* have colons in
1136 ordinary lsearch keys. However, making the change in all cases is
1137 incompatible and would probably break a number of configurations.
1139 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1142 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1143 conversion specification without a maximum field width, thereby enabling
1144 a rogue spamd server to cause a buffer overflow. While nobody in their
1145 right mind would setup Exim to query an untrusted spamd server, an
1146 attacker that gains access to a server running spamd could potentially
1147 exploit this vulnerability to run arbitrary code as the Exim user.
1149 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1150 $primary_hostname instead of what libspf2 thinks the hosts name is.
1152 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1153 a directory entry by the name of the lookup key. Previously, if a
1154 symlink pointed to a non-existing file or a file in a directory that
1155 Exim lacked permissions to read, a lookup for a key matching that
1156 symlink would fail. Now it is enough that a matching directory entry
1157 exists, symlink or not. (Bugzilla 503.)
1159 PH/03 The body_linecount and body_zerocount variables are now exported in the
1162 PH/04 Added the $dnslist_matched variable.
1164 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1165 This means they are set thereafter only if the connection becomes
1168 PH/06 Added the client_condition to authenticators so that some can be skipped
1169 by clients under certain conditions.
1171 PH/07 The error message for a badly-placed control=no_multiline_responses left
1172 "_responses" off the end of the name.
1174 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1176 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1177 (without spaces) instead of just copying the configuration text.
1179 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1181 PH/11 Added $max_received_linelength.
1183 PH/12 Added +ignore_defer and +include_defer to host lists.
1185 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1186 way in which PCRE > 7.0 is built.
1188 PH/14 Implemented queue_only_load_latch.
1190 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1191 MAIL command. The effect was to mangle the value on 64-bit systems.
1193 PH/16 Another patch from the Sieve maintainer.
1195 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1197 PH/18 If a system quota error occurred while trying to create the file for
1198 a maildir delivery, the message "Mailbox is full" was not appended to the
1199 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1200 only to a quota excession during the actual writing of the file.
1202 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1203 characters?) which causes problems in log lines. The DN values are now
1204 passed through string_printing() before being added to log lines.
1206 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1207 and InterBase are left for another time.)
1209 PH/21 Added message_body_newlines option.
1211 PH/22 Guard against possible overflow in moan_check_errorcopy().
1213 PH/23 POSIX allows open() to be a macro; guard against that.
1215 PH/24 If the recipient of an error message contained an @ in the local part
1216 (suitably quoted, of course), incorrect values were put in $domain and
1217 $local_part during the evaluation of errors_copy.
1223 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1224 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1227 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1228 issue a MAIL command.
1230 PH/02 In an ACL statement such as
1232 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1234 if a client was not listed at all, or was listed with a value other than
1235 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1236 the condition was not true (as it should be), so access was not denied.
1237 The bug was that the ! inversion was incorrectly passed on to the second
1238 item. This has been fixed.
1240 PH/03 Added additional dnslists conditions == and =& which are different from
1241 = and & when the dns lookup returns more than one IP address.
1243 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1244 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1246 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1247 FSYNC, which compiles an option called disable_fsync that allows for
1248 bypassing fsync(). The documentation is heavily laced with warnings.
1250 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1252 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1253 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1254 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1255 including adding "make clean"; (3) Added -fPIC when compiling the test
1256 dynamically loaded module, to get rid of a warning.
1258 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1259 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1260 = 0s. The bug is otherwise harmless.
1262 PH/07 There was a bug in the dovecot authenticator such that the value of
1263 $auth1 could be overwritten, and so not correctly preserved, after a
1264 successful authentication. This usually meant that the value preserved by
1265 the server_setid option was incorrect.
1267 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1269 PH/09 Installed PCRE release 7.0.
1271 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1272 run for batched SMTP input. It is now run at the start of every message
1273 in the batch. While fixing this I discovered that the process information
1274 (output by running exiwhat) was not always getting set for -bs and -bS
1275 input. This is fixed, and it now also says "batched" for BSMTP.
1277 PH/11 Added control=no_pipelining.
1279 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1280 patch, slightly modified), and move the expansion of helo_data till after
1281 the connection is made in the smtp transport (so it can use these
1284 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1286 PH/14 Added log_selector = +pid.
1288 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1290 PH/16 Add ${if forany and ${if forall.
1292 PH/17 Added dsn_from option to vary the From: line in DSNs.
1294 PH/18 Flush SMTP output before performing a callout, unless control =
1295 no_callout_flush is set.
1297 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1298 was true (the default) a successful delivery failed to delete the retry
1299 item, thus causing premature timeout of the address. The bug is now
1302 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1304 PH/21 Long custom messages for fakedefer and fakereject are now split up
1305 into multiline reponses in the same way that messages for "deny" and
1306 other ACL rejections are.
1308 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1309 with slight modification.
1311 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1312 draft, changing the syntax and factoring some duplicate code".
1314 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1315 for deliveries of the second and subsequent messages over the same SMTP
1318 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1319 ${reduce, with only minor "tidies".
1321 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1323 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1324 expansion side effects.
1326 PH/27 When a message times out after an over-quota error from an Exim-imposed
1327 quota, the bounce message says "mailbox is full". This message was not
1328 being given when it was a system quota that was exceeded. It now should
1331 MH/03 Made $recipients available in local_scan(). local_scan() already has
1332 better access to the recipient list through recipients_list[], but
1333 $recipients can be useful in postmaster-provided expansion strings.
1335 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1336 in the case of a MAIL command with additional options following the
1337 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1338 were accidentally chopped off.
1340 PH/29 SMTP synchronization checks are implemented when a command is read -
1341 there is a check that no more input is waiting when there shouldn't be
1342 any. However, for some commands, a delay in an ACL can mean that it is
1343 some time before the response is written. In this time, more input might
1344 arrive, invalidly. So now there are extra checks after an ACL has run for
1345 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1346 pipelining has not been advertised.
1348 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1350 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1351 returned in the SMTP response when smtp_return_error_details was false.
1352 This has been fixed.
1354 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1355 instead of the C I/O that was originally supplied, because problems were
1356 reported on Solaris.
1358 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1359 Exim which did not show up earlier: it was assuming that a call to
1360 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1361 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1362 was a macro that became an assignment, so it seemed to work. This has
1363 changed to a proper function call with a void return, hence the compile
1364 error. Exim's code has been fixed.
1366 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1369 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1371 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1373 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1374 because it turns out to be too big or there is a timeout) there is no
1375 "Completed" line in the log. When some messages of this type were
1376 selected by exigrep, they were listed as "not completed". Others were
1377 picked up by some special patterns. I have improved the selection
1378 criteria to be more general.
1380 PH/38 The host_find_failed option in the manualroute router can now be set
1381 to "ignore", to completely ignore a host whose IP address cannot be
1382 found. If all hosts are ignored, the behaviour is controlled by the new
1383 host_all_ignored option.
1385 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1386 homing or because of multiple MX records with /mx) generated more than
1387 one IP address, and the following item turned out to be the local host,
1388 all the secondary addresses of the first item were incorrectly removed
1389 from the list, along with the local host and any following hosts (which
1390 is what is supposed to happen).
1392 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1393 whoever called Exim into the -H file. In the case of the daemon it was
1394 behaving confusingly. When first started, it used values for whoever
1395 started the daemon, but after a SIGHUP it used the Exim user (because it
1396 calls itself on a restart). I have changed the code so that it now always
1399 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1400 message are rejected with the same error (e.g. no authentication or bad
1401 sender address), and a DATA command is nevertheless sent (as can happen
1402 with PIPELINING or a stupid MUA), the error message that was given to the
1403 RCPT commands is included in the rejection of the DATA command. This is
1404 intended to be helpful for MUAs that show only the final error to their
1407 PH/42 Another patch from the Sieve maintainer.
1409 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1410 Eximstats - Fixed some broken HTML links and added missing column headers
1412 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1413 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1415 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1417 PH/43 Yet another patch from the Sieve maintainer.
1419 PH/44 I found a way to check for a TCP/IP connection going away before sending
1420 the response to the final '.' that terminates a message, but only in the
1421 case where the client has not sent further data following the '.'
1422 (unfortunately, this is allowed). However, in many cases there won't be
1423 any further data because there won't be any more messages to send. A call
1424 to select() can be used: if it shows that the input is "ready", there is
1425 either input waiting, or the socket has been closed. An attempt to read
1426 the next input character can distinguish the two cases. Previously, Exim
1427 would have sent an OK response which the client would never have see.
1428 This could lead to message repetition. This fix should cure that, at
1429 least in a lot of common cases.
1431 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1432 advertised in response to EHLO.
1438 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1439 fixed by 4.65/MH/01 (is this a record?) are fixed:
1441 (i) An empty string was always treated as zero by the numeric comparison
1442 operators. This behaviour has been restored.
1444 (ii) It is documented that the numeric comparison operators always treat
1445 their arguments as decimal numbers. This was broken in that numbers
1446 starting with 0 were being interpreted as octal.
1448 While fixing these problems I realized that there was another issue that
1449 hadn't been noticed. Values of message_size_limit (both the global option
1450 and the transport option) were treated as octal if they started with 0.
1451 The documentation was vague. These values are now always treated as
1452 decimal, and I will make that clear in the documentation.
1458 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1459 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1462 MH/01 Don't check that the operands of numeric comparison operators are
1463 integers when their expansion is in "skipping" mode (fixes bug
1464 introduced by 4.64-PH/07).
1466 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1467 child addresses, Exim now panics and dies. Previously, because the count
1468 is held in a short int, deliveries were likely to be lost. As such a
1469 large number of recipients for a single message is ridiculous
1470 (performance will be very, very poor), I have chosen to impose a limit
1471 rather than extend the field.
1477 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1478 leftover -K file (the existence of which was triggered by #402).
1479 While we were at it, introduced process PID as part of the -K
1480 filename. This should rule out race conditions when creating
1483 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1484 processing considerably. Previous code took too long for large mails,
1485 triggering a timeout which in turn triggers #401.
1487 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1488 in the DK code in transports.c. sendfile() is not really portable,
1489 hence the _LINUX specificness.
1491 TF/01 In the add_headers option to the mail command in an Exim filter,
1492 there was a bug that Exim would claim a syntax error in any
1493 header after the first one which had an odd number of characters
1496 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1497 callout verification, Exim cached a "reject" for the entire domain. This
1498 is correct for most verifications, but it is not correct for a recipient
1499 verification with use_sender or use_postmaster set, because in that case
1500 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1501 case of MAIL FROM:<> rejection from other early rejections (e.g.
1502 rejection of HELO). When verifying a recipient using a non-null MAIL
1503 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1504 Whatever the result of the callout, the value of the domain cache is
1505 left unchanged (for any other kind of callout, getting as far as trying
1506 RCPT means that the domain itself is ok).
1508 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1511 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1512 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1513 session, the daemon ignores ECONNECT errors and logs others; it now
1514 ignores EPIPE as well.
1516 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1517 (quoted-printable decoding).
1519 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1520 later the small subsequent patch to fix an introduced bug.
1522 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1524 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1526 PH/08 An error is now given if message_size_limit is specified negative.
1528 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1529 to be given (somewhat) arbitrary names.
1531 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1534 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1535 miscellaneous code fixes
1537 PH/10 Added the log_reject_target ACL modifier to specify where to log
1540 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1541 hostname. This is wrong, because it relates to the incoming message (and
1542 probably the interface on which it is arriving) and not to the outgoing
1543 callout (which could be using a different interface). This has been
1544 changed to use the value of the helo_data option from the smtp transport
1545 instead - this is what is used when a message is actually being sent. If
1546 there is no remote transport (possible with a router that sets up host
1547 addresses), $smtp_active_hostname is used.
1549 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1550 tweaks were necessary in order to get it to work (see also 21 below):
1551 (a) The code assumed that strncpy() returns a negative number on buffer
1552 overflow, which isn't the case. Replaced with Exim's string_format()
1554 (b) There were several signed/unsigned issues. I just did the minimum
1555 hacking in of casts. There is scope for a larger refactoring.
1556 (c) The code used strcasecmp() which is not a standard C function.
1557 Replaced with Exim's strcmpic() function.
1558 (d) The code set only $1; it now sets $auth1 as well.
1559 (e) A simple test gave the error "authentication client didn't specify
1560 service in request". It would seem that Dovecot has changed its
1561 interface. Fortunately there's a specification; I followed it and
1562 changed what the client sends and it appears to be working now.
1564 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1567 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1568 address A is aliased to B and C, where B exists and C does not. Without
1569 -v the output is "A verified" because verification stops after a
1570 successful redirection if more than one address is generated. However,
1571 with -v the child addresses are also verified. Exim was outputting "A
1572 failed to verify" and then showing the successful verification for C,
1573 with its parentage. It now outputs "B failed to verify", showing B's
1574 parentage before showing the successful verification of C.
1576 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1577 look up a TXT record in a specific list after matching in a combined
1580 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1581 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1582 they consult the DNS. I had assumed they would set it the way they
1583 wanted; and indeed my experiments on Linux seem to show that in some
1584 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1585 To be on the safe side, however, I have now made the interface to
1586 host_find_byname() similar to host_find_bydns(), with an argument
1587 containing the DNS resolver options. The host_find_byname() function now
1588 sets these options at its start, just as host_find_bydns() does. The smtp
1589 transport options dns_qualify_single and dns_search_parents are passed to
1590 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1591 of host_find_byname() use the default settings of RES_DEFNAMES
1592 (qualify_single) but not RES_DNSRCH (search_parents).
1594 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1595 spool_read_header() do less string testing, by means of a preliminary
1596 switch on the second character of optional "-foo" lines. (This is
1597 overdue, caused by the large number of possibilities that now exist.
1598 Originally there were few.) While I was there, I also converted the
1599 str(n)cmp tests so they don't re-test the leading "-" and the first
1600 character, in the hope this might squeeze out yet more improvement.
1602 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1603 flag allowing group syntax was set by the header_syntax check but not
1604 turned off, possible causing trouble later; (2) The flag was not being
1605 set at all for the header_verify test, causing "group"-style headers to
1606 be rejected. I have now set it in this case, and also caused header_
1607 verify to ignore an empty address taken from a group. While doing this, I
1608 came across some other cases where the code for allowing group syntax
1609 while scanning a header line wasn't quite right (mostly, not resetting
1610 the flag correctly in the right place). These bugs could have caused
1611 trouble for malformed header lines. I hope it is now all correct.
1613 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1614 with the "reply" argument non-NULL. The code, however (which originally
1615 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1616 but it didn't always do it. This confused somebody who was copying the
1617 code for some other use. I have removed all the tests.
1619 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1620 feature that was used to support insecure browsers during the U.S. crypto
1621 embargo. It requires special client support, and Exim is probably the
1622 only MTA that supported it -- and would never use it because real RSA is
1623 always available. This code has been removed, because it had the bad
1624 effect of slowing Exim down by computing (never used) parameters for the
1625 RSA_EXPORT functionality.
1627 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1628 authenticator to fail if there's a tab character in the incoming data
1629 (there should never be unless someone is messing about, as it's supposed
1630 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1631 if the connection is using TLS or if the remote IP is the same as the
1632 local IP, and the "valid-client-cert option" if a client certificate has
1635 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1636 authenticators. This can be used for authorization after authentication
1637 succeeds. (In the case of plaintext, it servers for both authentication
1640 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1641 if any retry times were supplied.
1643 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1644 connection, where there is no client IP address to check. In this
1645 situation, the verify now always succeeds.
1647 PH/25 Applied John Jetmore's -Mset patch.
1649 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1651 PH/27 In a string expansion for a processed (not raw) header when multiple
1652 headers of the same name were present, leading whitespace was being
1653 removed from all of them, but trailing whitespace was being removed only
1654 from the last one. Now trailing whitespace is removed from each header
1655 before concatenation. Completely empty headers in a concatenation (as
1656 before) are ignored.
1658 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1659 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1661 PH/29 [Removed. This was a change that I later backed out, and forgot to
1662 correct the ChangeLog entry (that I had efficiently created) before
1663 committing the later change.]
1665 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1666 address errors (4xx response to RCPT) over the same connection as other
1667 messages routed to the same hosts. Such deliveries are always "forced",
1668 so retry times are not inspected. This resulted in far too many retries
1669 for the affected addresses. The effect occurred only when there were more
1670 hosts than the hosts_max_try setting in the smtp transport when it had
1671 the 4xx errors. Those hosts that it had tried were not added to the list
1672 of hosts for which the message was waiting, so if all were tried, there
1673 was no problem. Two fixes have been applied:
1675 (i) If there are any address or message errors in an SMTP delivery, none
1676 of the hosts (tried or untried) are now added to the list of hosts
1677 for which the message is waiting, so the message should not be a
1678 candidate for sending over the same connection that was used for a
1679 successful delivery of some other message. This seems entirely
1680 reasonable: after all the message is NOT "waiting for some host".
1681 This is so "obvious" that I'm not sure why it wasn't done
1682 previously. Hope I haven't missed anything, but it can't do any
1683 harm, as the worst effect is to miss an optimization.
1685 (ii) If, despite (i), such a delivery is accidentally attempted, the
1686 routing retry time is respected, so at least it doesn't keep
1687 hammering the server.
1689 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1690 in ${readsocket because some servers need this prod.
1692 PH/32 Added some extra debug output when updating a wait-xxx database.
1694 PH/33 The hint "could be header name not terminated by colon", which has been
1695 given for certain expansion errors for a long time, was not being given
1696 for the ${if def:h_colon_omitted{... case.
1698 PH/34 The spec says: "With one important exception, whenever a domain list is
1699 being scanned, $domain contains the subject domain." There was at least
1700 one case where this was not true.
1702 PH/35 The error "getsockname() failed: connection reset by peer" was being
1703 written to the panic log as well as the main log, but it isn't really
1704 panic-worthy as it just means the connection died rather early on. I have
1705 removed the panic log writing for the ECONNRESET error when getsockname()
1708 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1709 runs only) independently of the message's sender address. This meant
1710 that, if the 4xx error was in fact related to the sender, a different
1711 message to the same recipient with a different sender could confuse
1712 things. In particualar, this can happen when sending to a greylisting
1713 server, but other circumstances could also provoke similar problems.
1714 I have changed the default so that the retry time for these errors is now
1715 based a combination of the sender and recipient addresses. This change
1716 can be overridden by setting address_retry_include_sender=false in the
1719 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1720 remote server are returned as part of bounce messages. This was not
1721 happening for LMTP over a pipe (the lmtp transport), but now it is the
1722 same for both kinds of LMTP.
1724 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1725 in header lines that were in fact CNAMEs. This is no longer the case.
1727 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1728 and queue runs started by the daemon processed all messages. This has
1729 been fixed so that -R and -S can now usefully be given with -q<time>.
1731 PH/40 Import PCRE release 6.7 (fixes some bugs).
1733 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1735 PH/42 Give an error if -q is specified more than once.
1737 PH/43 Renamed the variables $interface_address and $interface_port as
1738 $received_ip_address and $received_port, to make it clear that these
1739 values apply to message reception, and not to the outgoing interface when
1740 a message is delivered. (The old names remain recognized, of course.)
1742 PH/44 There was no timeout on the connect() call when using a Unix domain
1743 socket in the ${readsocket expansion. There now is.
1745 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1746 be meaningful with "accept".
1748 SC/01 Eximstats V1.43
1749 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1751 SC/02 Eximstats V1.44
1752 Use a glob alias rather than an array ref in the generated
1753 parser. This improves both readability and performance.
1755 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1756 Collect SpamAssassin and rejection statistics.
1757 Don't display local sender or destination tables unless
1758 there is data to show.
1759 Added average volumes into the top table text output.
1761 SC/04 Eximstats V1.46
1762 Collect data on the number of addresses (recipients)
1763 as well as the number of messages.
1765 SC/05 Eximstats V1.47
1766 Added 'Message too big' to the list of mail rejection
1767 reasons (thanks to Marco Gaiarin).
1769 SC/06 Eximstats V1.48
1770 Mainlog lines which have GMT offsets and are too short to
1771 have a flag are now skipped.
1773 SC/07 Eximstats V1.49 (Alain Williams)
1774 Added the -emptyok flag.
1776 SC/08 Eximstats V1.50
1777 Fixes for obtaining the IP address from reject messages.
1779 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1780 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1781 whitesspace changes from 4.64-PH/27
1783 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1786 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1787 are found, allow negative numbers in numeric criteria)
1789 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1791 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1792 to match changes made in 4.64-PH/43
1794 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1796 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1797 "domain" is actually an IP address, and force a failure. This locks out
1798 those revolvers/nameservers that support "A-for-A" lookups, in
1799 contravention of the specifications.
1801 PH/48 When a host name was looked up from an IP address, and the subsequent
1802 forward lookup of the name timed out, the host name was left in
1803 $sender_host_name, contrary to the specification.
1805 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1806 restricted to single-key lookups, Exim was not diagnosing an error if
1807 * or *@ was used with a query-style lookup.
1809 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1811 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1812 long ago, but noone interested enough thought of it. Let's just say that
1813 the "1.1" means that there are some new functions that weren't there at
1814 some point in the past.
1816 PH/51 Error processing for expansion failure of helo_data from an smtp
1817 transport during callout processing was broken.
1819 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1820 tested/used via the -bh/-bhc/-bs options.
1822 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1823 bug, fixed in subsequent PCRE releases).
1825 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1826 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1828 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1834 SC/01 Use a glob alias rather than an array ref in eximstats generated
1835 parser. This improves both readability and performance.
1837 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1838 Don't display local sender or destination tables in eximstats unless
1839 there is data to show.
1840 Added average volumes into the eximstats top table text output.
1842 SC/03 Collect data on the number of addresses (recipients) as well
1843 as the number of messages in eximstats.
1845 TF/01 Correct an error in the documentation for the redirect router. Exim
1846 does (usually) call initgroups() when daemonizing.
1848 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1849 with consistent privilege compared to when running as a daemon.
1851 TF/03 Note in the spec that $authenticated_id is not set for local
1852 submissions from trusted users.
1854 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1855 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1857 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1858 by adding some example configuration directives to the default
1859 configuration file. A little bit of work is required to uncomment the
1860 directives and define how usernames and passwords are checked, but
1861 there is now a framework to start from.
1863 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1864 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1865 without this. I don't know how relevant this is to other LDAP libraries.
1867 PH/02 Add the verb name to the "unknown ACL verb" error.
1869 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1871 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1873 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1874 directory not expanded when it should be if an expanded home directory
1875 was set for the address (which is overridden by the transport).
1877 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1880 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1881 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1882 because it is too late at that time, and has no effect.
1884 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1885 security issue with \' (bugzilla #107). I could not use the
1886 PQescapeStringConn() function, because it needs a PGconn value as one of
1889 PH/08 When testing addresses using -bt, indicate those final addresses that
1890 are duplicates that would not cause an additional delivery. At least one
1891 person was confused, thinking that -bt output corresponded to deliveries.
1892 (Suppressing duplicates isn't a good idea as you lose the information
1893 about possibly different redirections that led to the duplicates.)
1895 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1896 systems where poll() doesn't work, in particular OS X.
1898 PH/10 Added more information to debugging output for retry time not reached.
1900 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1901 operations in malware.c.
1903 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1906 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1907 syslog_duplication set false, log lines that would normally be written
1908 both the the main log and to the reject log were not written to syslog at
1911 PH/14 In the default configuration, change the use of "message" in ACL warn
1912 statements to "add_header".
1914 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1915 not followed by a command (e.g. "seen endif").
1917 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1918 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1921 PH/17 Added extra conditions to the default value of delay_warning_condition
1925 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1926 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1927 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1930 The Auto-Submitted: and various List- headers are standardised, whereas I
1931 don't think Precedence: ever was.
1933 PH/18 Refactored debugging code in route_finduser() to show more information,
1934 in particular, the error code if getpwnam() issues one.
1936 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1937 This is apparently needed in addition to the PH/07 change above to avoid
1938 any possible encoding problems.
1940 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1941 but not after initializing Perl.
1943 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1944 output them only if debugging. By default they are written stderr,
1945 apparently, which is not desirable.
1947 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1950 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1953 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1955 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1956 authenticated or an ident call has been made. Suppress the default
1957 values for $authenticated_id and $authenticated_sender (but permit -oMai
1958 and -oMas) when testing with -bh.
1960 PH/24 Re-jigged the order of the tests in the default configuration so that the
1961 tests for valid domains and recipients precede the DNS black list and CSA
1962 tests, on the grounds that those ones are more expensive.
1964 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1965 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1966 This bug exists in every version of Exim that I still have, right back to
1969 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1970 However, an attempt to turn on case-sensitivity in a regex key by
1971 including (?-i) didn't work because the subject string was already
1972 lowercased, and the effects were non-intuitive. It turns out that a
1973 one-line patch can be used to allow (?-i) to work as expected.
1979 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1980 other effects) broke the use of negated acl sub-conditions.
1982 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1985 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1986 "Deny" causes Exim to reject the incoming connection with a 554 error.
1987 Unfortunately, if there is a major crisis, such as a disk failure,
1988 tcp-wrappers gives "deny", whereas what one would like would be some
1989 kind of temporary error. A kludge has been added to help with this.
1990 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1991 554 error is used if errno is still zero or contains ENOENT (which occurs
1992 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1995 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1997 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1998 errors. Otherwise a message that provokes a temporary error (when other
1999 messages do not) can cause a whole host to time out.
2001 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2002 addresses were routed directly to files or pipes from a redirect router.
2003 File deliveries just didn't batch; pipe deliveries might have suffered
2006 PH/06 A failure to get a lock for a hints database would erroneously always say
2007 "Failed to get write lock", even when it was really a read lock.
2009 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2010 of 0600. This has been changed to use the value of the lockfile_mode
2011 option (which defaults to 0600).
2013 PH/08 Applied small patch from the Sieve maintainer.
2015 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2016 folder from quota calculations, a direct delivery into this folder messed
2017 up the contents of the maildirsize file. This was because the regex was
2018 used only to exclude .Trash (or whatever) when the size of the mailbox
2019 was calculated. There was no check that a delivery was happening into an
2020 excluded directory. This bug has been fixed by ignoring all quota
2021 processing for deliveries into excluded directories.
2023 PH/10 Added the maildirfolder_create_regex option to appendfile.
2029 PH/01 The code for finding all the local interface addresses on a FreeBSD
2030 system running IPv6 was broken. This may well have applied to all BSD
2031 systems, as well as to others that have similar system calls. The broken
2032 code found IPv4 interfaces correctly, but gave incorrect values for the
2033 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2034 that it would not match correctly against @[] and not recognize the IPv6
2037 PH/02 The ipliteral router was not recognizing addresses of the form user@
2038 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2040 PH/03 Added disable_ipv6.
2042 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2043 decoded form, because it is most often used to construct To: headers
2044 lines in autoreplies, and the decoded form may well be syntactically
2045 invalid. However, $reply_address has leading white space removed, and all
2046 newlines turned into spaces so that the autoreply transport does not
2049 PH/05 If group was specified without a user on a router, and no group or user
2050 was specified on a transport, the group from the router was ignored.
2052 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2053 for visible compile-time settings that can be used to change these
2054 numbers, for those that want even more. Backwards compatibility with old
2055 spool files has been maintained. However, going back to a previous Exim
2056 release will lost any variables that are in spool files.
2058 PH/07 Two small changes when running in the test harness: increase delay when
2059 passing a TCP/IP connection to a new process, in case the original
2060 process has to generate a bounce, and remove special handling of
2061 127.0.0.2 (sic), which is no longer necessary.
2063 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2064 be the same on different OS.
2066 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2069 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2070 whether --show-vars was specified or not
2072 JJ/02 exipick: Added support for new ACL variable spool format introduced
2075 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2076 syntactically invalid From: or Reply-to: line, and a filter used this to
2077 generate an autoreply, and therefore failed to obtain an address for the
2078 autoreply, Exim could try to deliver to a non-existent relative file
2079 name, causing unrelated and misleading errors. What now happens is that
2080 it logs this as a hard delivery error, but does not attempt to create a
2083 PH/11 The exinext utility has a -C option for testing purposes, but although
2084 the given file was scanned by exinext itself; it wasn't being passed on
2085 when Exim was called.
2087 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2088 an end-of-file indication when reading a command response.
2090 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2091 compiled. In many other places in Exim, IPv6 addresses are always
2092 recognized, so I have changed this. It also means that IPv4 domain
2093 literals of the form [IPV4:n.n.n.n] are now always recognized.
2095 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2096 used if the router is not running as root, for example, when verifying at
2097 ACL time, or when using -bh. The debugging output from this situation was
2098 non-existent - all you got was a failure to exec. I have made two
2101 (a) Failures to set uid/gid, the current directory, or a process leader
2102 in a subprocess such as that created by queryprogram now generate
2103 suitable debugging ouput when -d is set.
2105 (b) The queryprogram router detects when it is not running as root,
2106 outputs suitable debugging information if -d is set, and then runs
2107 the subprocess without attempting to change uid/gid.
2109 PH/15 Minor change to Makefile for building test_host (undocumented testing
2112 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2113 additional section of a DNS packet that returns MX or SRV records.
2114 Instead, it always explicitly searches for A/AAAA records. This avoids
2115 major problems that occur when a DNS server includes only records of one
2116 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2117 fixed another bug: if SRV records were looked up and the corresponding
2118 address records were *not* found in the additional section, the port
2119 values from the SRV records were lost.
2121 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2122 using the correct key (the original address) when searching the retry
2123 rules in order to find which one to use for generating the retry hint.
2125 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2126 adding the default one. Similarly, if it contains a Reply-To: header, the
2127 errors_reply_to option, if set, is not used.
2129 PH/19 When calculating a retry time, Exim used to measure the "time since
2130 failure" by looking at the "first failed" field in the retry record. Now
2131 it does not use this if it is later than than the arrival time of the
2132 message. Instead it uses the arrival time. This makes for better
2133 behaviour in cases where some deliveries succeed, thus re-setting the
2134 "first failed" field. An example is a quota failure for a huge message
2135 when small messages continue to be delivered. Without this change, the
2136 "time since failure" will always be short, possible causing more frequent
2137 delivery attempts for the huge message than are intended.
2138 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2140 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2141 $1, $2, $3) because the numerical variables can be reset during some
2142 expansion items (e.g. "match"), thereby losing the authentication data.
2144 PH/21 Make -bV show the size of off_t variables so that the test suite can
2145 decide whether to run tests for quotas > 2G.
2147 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2148 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2149 filecount value is greater than 2G or if a quota value is greater than 2G
2150 on a system where the size of off_t is not greater than 4, a panic error
2153 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2154 never match. The debug and -bh output now contains an explicit error
2155 message indicating a malformed IPv4 address or mask.
2157 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2158 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2159 PH/23 above applies.
2161 PH/25 Do not write to syslog when running in the test harness. The only
2162 occasion when this arises is a failure to open the main or panic logs
2163 (for which there is an explicit test).
2165 PH/26 Added the /no_tell option to "control=freeze".
2167 PH/27 If a host name lookup failed very early in a connection, for example, if
2168 the IP address matched host_lookup and the reverse lookup yielded a name
2169 that did not have a forward lookup, an error message of the form "no IP
2170 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2171 could be logged. Now it outputs the IP address instead of "NULL".
2173 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2174 allows the sender and the authenticated sender to be set when
2175 submitting a message from within Exim. Since child_open_exim() is
2176 documented for local_scan(), the new function should be too.
2178 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2179 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2180 results in an empty string is now treated as unset.
2182 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2184 PH/31 Added sender_verify_fail logging option.
2186 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2187 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2188 tidied the source and removed it altogether.
2190 PH/33 When a queue run was abandoned because the load average was too high, a
2191 log line was always written; now it is written only if the queue_run log
2192 selector is set. In addition, the log line for abandonment now contains
2193 information about the queue run such as the pid. This is always present
2194 in "start" and "stop" lines but was omitted from the "abandon" line.
2196 PH/34 Omit spaces between a header name and the colon in the error message that
2197 is given when verify = headers_syntax fails (if there are lots of them,
2198 the message gets confusing).
2200 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2201 names, as there are now some PTR records that contain slashes. This check
2202 is only to protect against broken name servers that fall over on strange
2203 characters, so the fact that it applies to all lookups doesn't matter.
2205 PH/36 Now that the new test suite is complete, we can remove some of the
2206 special code in Exim that was needed for the old test suite. For example,
2207 sorting DNS records because real resolvers return them in an arbitrary
2208 order. The new test suite's fake resolver always returns records in the
2211 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2212 bounces) except when queue_only is set, to avoid logging races between
2213 the different processes.
2215 PH/38 Panic-die if .include specifies a non-absolute path.
2217 PH/39 A tweak to the "H" retry rule from its user.
2219 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2220 a label. They prevented compilation on older perls.
2222 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2223 a warning to be raised on newish perls.
2225 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2226 on queue. Changes to match documented behaviour of showing count of
2227 messages matching specified criteria.
2229 PH/40 Changed the default ident timeout from 30s to 5s.
2231 PH/41 Added support for the use of login_cap features, on those BSD systems
2232 that have them, for controlling the resources used by pipe deliveries.
2234 PH/42 The content-scanning code uses fopen() to create files in which to put
2235 message data. Previously it was not paying any attention to the mode of
2236 the files. Exim runs with umask(0) because the rest of the code creates
2237 files with open(), and sets the required mode explicitly. Thus, these
2238 files were ending up world-writeable. This was not a big issue, because,
2239 being within the spool directory, they were not world-accessible. I have
2240 created a function called modefopen, which takes an additional mode
2241 argument. It sets umask(777), creates the file, chmods it to the required
2242 mode, then resets the umask. All the relevant calls to fopen() in the
2243 content scanning code have been changed to use this function.
2245 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2246 to 24 hours. This avoids potential overflow problems when processing G
2247 and H retry rules. I suspect nobody ever tinkers with this value.
2249 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2251 PH/45 When the plaintext authenticator is running as a client, the server's
2252 challenges are checked to ensure they are valid base64 strings. By
2253 default, the authentication attempt is cancelled if an invalid string is
2254 received. Setting client_ignore_invalid_base64 true ignores these errors.
2255 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2256 they are received. Thus, the responses can be made to depend on the
2257 challenges. If an invalid string is ignored, an empty string is placed in
2260 PH/46 Messages that are created by the autoreply transport now contains a
2261 References: header, in accordance with RFCs 2822 and 3834.
2263 PH/47 Added authenticated_sender_force to the smtp transport.
2265 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2267 PH/49 Installed latest patch from the Sieve maintainer.
2269 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2270 was also set, the appendfile transport was unnecessarily scanning a
2271 directory of message files (e.g. for maildir delivery) to find the count
2272 of files (along with the size), even though it did not need this
2273 information. It now does the scan only if it needs to find either the
2274 size of the count of files.
2276 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2278 PH/52 Two bugs concerned with error handling when the smtp transport is
2281 (i) Exim was not creating retry information for temporary errors given
2282 for individual recipients after the DATA command when the smtp transport
2283 was used in LMTP mode. This meant that they could be retried too
2284 frequently, and not timed out correctly.
2286 (ii) Exim was setting the flag that allows error details to be returned
2287 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2288 recipients that were returned after the DATA command.
2290 PH/53 This is related to PH/52, but is more general: for any failing address,
2291 when detailed error information was permitted to be returned to the
2292 sender, but the error was temporary, then after the final timeout, only
2293 "retry timeout exceeded" was returned. Now it returns the full error as
2294 well as "retry timeout exceeded".
2296 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2297 do this, and (what is worse) MTAs that accept it.
2299 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2300 will now be deprecated.
2302 PH/56 New os.c-cygwin from the Cygwin maintainer.
2304 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2305 formats (previously only available in exim formats via -bpr, -bpru,
2306 and -bpra. Now also available in native and exiqgrep formats)
2308 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2309 with very large, slow to parse queues
2311 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2313 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2315 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2316 responses to authentication challenges, though it was showing the
2317 challenges; (ii) I've removed the CR characters from the debug output for
2320 PH/58 Allow for the insertion of a newline as well as a space when a string
2321 is turned into more than one encoded-word during RFC 2047 encoding. The
2322 Sieve code now uses this.
2324 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2325 data_4xx, lost_connection, tls_required.
2327 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2328 message was being sent as an SMTP response.
2330 PH/61 Add -l and -k options to exicyclog.
2332 PH/62 When verifying, if an address was redirected to one new address, so that
2333 verification continued, and the new address failed or deferred after
2334 having set something in $address_data, the value of $address_data was not
2335 passed back to the ACL. This was different to the case when no
2336 redirection occurred. The value is now passed back in both cases.
2338 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2339 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2340 use login_cap.h, so on its own it isn't the distinguishing feature. The
2341 new name refers directly to the setclassresources() function.
2343 PH/65 Added configuration files for NetBSD3.
2345 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2347 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2348 is preferred over IPv4.
2350 PH/68 The bounce_return_message and bounce_return_body options were not being
2351 honoured for bounces generated during the reception of non-SMTP messages.
2352 In particular, this applied to messages rejected by the ACL. This bug has
2353 been fixed. However, if bounce_return_message is true and bounce_return_
2354 body is false, the headers that are returned for a non-SMTP message
2355 include only those that have been read before the error was detected.
2356 (In the case of an ACL rejection, they have all been read.)
2358 PH/69 The HTML version of the specification is now built in a directory called
2359 spec_html instead of spec.html, because the latter looks like a path with
2360 a MIME-type, and this confuses some software.
2362 PH/70 Catch two compiler warnings in sieve.c.
2364 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2365 function verify_get_ident() calls ip_connect() to connect a socket, but
2366 if the "connect()" function timed out, ip_connect() used to close the
2367 socket. However, verify_get_ident() also closes the socket later, and in
2368 between Exim writes to the log, which may get opened at this point. When
2369 the socket was closed in ip_connect(), the log could get the same file
2370 descriptor number as the socket. This naturally causes chaos. The fix is
2371 not to close the socket in ip_connect(); the socket should be closed by
2372 the function that creates it. There was only one place in the code where
2373 this was missing, in the iplookup router, which I don't think anybody now
2374 uses, but I've fixed it anyway.
2376 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2377 well as to direct DNS lookups. Otherwise the handling of names in host
2378 lists is inconsistent and therefore confusing.
2384 PH/01 Two changes to the default runtime configuration:
2386 (1) Move the checks for relay_from_hosts and authenticated clients from
2387 after to before the (commented out) DNS black list checks.
2389 (2) Add control=submission to the relay_from_hosts and authenticated
2390 clients checks, on the grounds that messages accepted by these
2391 statements are most likely to be submissions.
2393 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2395 (1) Generate an error if the third argument for the ${prvs expansion is
2398 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2401 (3) Reset the variables that are obtained from the first argument of
2402 ${prvscheck and used in the second argument before leaving the code,
2403 because their memory is reclaimed, so using them afterwards may do
2406 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2407 one (it's much easier than Tom thought :-).
2409 (5) Because of (4), we can now allow for the use of $prvscheck_result
2410 inside the third argument.
2412 PH/03 For some reason, the default setting of PATH when running a command from
2413 a pipe transport was just "/usr/bin". I have changed it to
2416 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2417 anything to be listed in the output from -bV.
2419 PH/05 When a filter generated an autoreply, the entire To: header line was
2420 quoted in the delivery log line, like this:
2422 => >A.N.Other <ano@some.domain> <original@ddress> ...
2424 This has been changed so that it extracts the operative address. There
2425 may be more than one such address. If so, they are comma-separated, like
2428 => >ano@some.domain,ona@other.domain <original@ddress> ...
2430 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2431 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2432 not being looked up in the rDNS to get a host name, Exim was showing the
2433 IP address twice in Received: lines, even though the IP addresses were
2434 identical. For example:
2436 Received: from [1.2.3.4] (helo=[1.2.3.4])
2438 However, if the real host name was known, it was omitting the HELO data
2439 if it matched the actual IP address. This has been tidied up so that it
2440 doesn't show the same IP address twice.
2442 PH/07 When both +timestamp and +memory debugging was on, the value given by
2443 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2444 called by the debug printing, thereby overwriting the timestamp buffer.
2445 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2447 PH/08 When the original message was included in an autoreply transport, it
2448 always said "this is a copy of the message, including all the headers",
2449 even if body_only or headers_only was set. It now gives an appropriate
2452 PH/09 Applied a patch from the Sieve maintainer which:
2454 o fixes some comments
2455 o adds the (disabled) notify extension core
2456 o adds some debug output for the result of if/elsif tests
2457 o points to the current vacation draft in the documentation
2458 and documents the missing references header update
2462 o fixes a bug in processing the envelope test (when testing
2463 multiple envelope elements, the last element determinted the
2466 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2467 Electronic Mail") by including:
2469 Auto-submitted: auto-generated
2471 in the messages that it generates (bounce messages and others, such as
2472 warnings). In the case of bounce messages for non-SMTP mesages, there was
2473 also a typo: it was using "Auto_submitted" (underscore instead of
2474 hyphen). Since every message generated by Exim is necessarily in response
2475 to another message, thes have all been changed to:
2477 Auto-Submitted: auto-replied
2479 in accordance with these statements in the RFC:
2481 The auto-replied keyword:
2483 - SHOULD be used on messages sent in direct response to another
2484 message by an automatic process,
2486 - MUST NOT be used on manually-generated messages,
2488 - MAY be used on Delivery Status Notifications (DSNs) and Message
2489 Disposition Notifications (MDNs),
2491 - MUST NOT be used on messages generated by automatic or periodic
2492 processes, except for messages which are automatic responses to
2495 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2496 to the default Received: header definition.
2498 PH/12 Added log selector acl_warn_skipped (default on).
2500 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2501 variables because (a) they are in the wrong storage pool and (b) even if
2502 they were copied, it wouldn't work properly because of the caching.
2504 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2505 checking when decoding. Apparently there are clients that generate
2506 overlong encoded strings. Why am I not surprised?
2508 PH/15 If the first argument of "${if match_address" was not empty, but did not
2509 contain an "@" character, Exim crashed. Now it writes a panic log message
2510 and treats the condition as false.
2512 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2514 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2515 extension "envelope-auth". The code is finished and in agreement with
2516 other implementations, but there is no documentation so far and in fact,
2517 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2518 not changing the active code.
2520 Print executed "if" and "elsif" statements when debugging is used. This
2521 helps a great deal to understand what a filter does.
2523 Document more things not specified clearly in RFC3028. I had all this
2524 sorted out, when out of a sudden new issues came to my mind. Oops."
2526 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2529 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2530 canonical form (as documented). However, after a host name lookup from
2531 the IP address, check_host() was doing a simple string comparison with
2532 addresses acquired from the DNS when checking that the found name did
2533 have the original IP as one of its addresses. Since any found IPv6
2534 addresses are likely to be in abbreviated form, the comparison could
2535 fail. Luckily, there already exists a function for doing the comparison
2536 by converting both addresses to binary, so now that is used instead of
2537 the text comparison.
2539 PH/20 There was another similar case to PH/19, when a complete host name was
2540 given in a host list; looking up its IP address could give an abbreviated
2541 form, whereas the current host's name might or might not be abbreviated.
2542 The same fix has been applied.
2548 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2549 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2552 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2553 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2555 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2557 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2558 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2559 Also, for "Auto-Submitted", treat anything other than "no" as
2560 non-personal, in accordance with RFC 3834. (Previously it treated
2561 anything starting "auto-" as non-personal.)
2563 TF/01 The control=submission/name=... option had a problem with syntax
2564 errors if the name included a slash character. The /name= option
2565 now slurps the rest of the string, so it can include any characters
2566 but it must come last in the list of options (after /sender_retain
2569 PH/05 Some modifications to the interface to the fake nameserver for the new
2577 TK/01 Added the "success_on_redirect" address verification option. See
2578 NewStuff for rationale and an example.
2580 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2582 PH/02 Patch to exigrep to allow it to work on syslog lines.
2584 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2585 fread() to skip over the body file's header line, because in Cygwin the
2586 header line is locked and is inaccessible.
2588 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2589 co-exist for some time) to make it clear that it is the Exim ID that is
2590 referenced, not the Message-ID: header line.
2592 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2593 string_format() function, because snprintf() does not exist on all
2596 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2597 ${stat: expansion item.
2599 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2600 protocol synchronization error", to keep the pedants happy.
2602 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2603 well as for IRIX systems, when gcc is being used. See the host.c source
2606 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2608 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2611 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2612 transport and to the smtp transport in LMTP mode.
2614 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2616 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2617 run a filter in a subprocess. This could lead to confusion in subsequent
2618 lookups in the parent process. There should also be a search_tidyup() at
2619 the end of the subprocess.
2621 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2622 only if the host matched helo_try_verify_hosts, which caused the
2623 verification to occur when the EHLO/HELO command was issued. The ACL just
2624 tested the remembered result. Now, if a previous verification attempt has
2625 not happened, "verify = helo" does it there and then.
2627 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2629 TK/03 Fix log output including CR from clamd.
2631 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2632 exist provoked a memory error which could cause a segfault.
2634 PH/15 Installed PCRE 6.2
2636 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2638 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2639 of the problem. Specifically, suggested +O2 rather than +O1 for the
2642 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2644 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2645 record, the port was not correctly logged when the outgoing_port log
2646 selector was set (it logged the transort's default port).
2648 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2649 fallback_hosts, and "hosts" in the smtp transport.
2651 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2652 host errors such as "Connection refused".
2654 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2655 authentication with radiusclient 0.4.9:
2657 - Error returned from rc_read_config was caught wrongly
2658 - Username/password not passed on to radius server due to wrong length.
2660 The presumption is that some radiusclient API changes for 4.51/PH/17
2661 were not taken care of correctly. The code is still untested by me (my
2662 Linux distribution still has 0.3.2 of radiusclient), but it was
2663 contributed by a Radius user.
2665 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2666 expanding the "port" option of the smtp transport.
2668 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2669 while reading a MIME header. Thanks to Tom Hughes for a patch.
2671 PH/24 Include config.h inside local_scan.h so that configuration settings are
2674 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2675 that in an ACL for RCPT (for example), you can examine exactly what was
2678 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2679 commands, but it was not correctly comparing the address with the actual
2680 client host address. Thus, it would show the EHLO address in Received:
2681 header lines when this was not necessary.
2683 PH/27 Added the % operator to ${eval:}.
2685 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2686 it should be ignoring failures (because with -C, for example, it has lost
2687 privilege). It wasn't ignoring creation failures other than "already
2690 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2693 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2694 because an input line was too long, either on its own, or by virtue of
2695 too many continuations, the temporary file was not being removed, and the
2696 return code was incorrect.
2698 PH/31 Missing "BOOL" in function definition in filtertest.c.
2700 PH/32 Applied Sieve patches from the maintainer.
2702 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2704 PH/33 Added "verify = not_blind".
2706 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2707 Local/Makefile (with some defaults set). These are used in built scripts
2708 such as exicyclog, but they have never been used in the exim_install
2709 script (though there are many overriding facilities there). I have
2710 arranged that the exim_install script now takes note of these two
2713 PH/35 Installed configuration files for Dragonfly.
2715 PH/36 When a locally submitted message by a trusted user did not contain a
2716 From: header, and the sender address was obtained from -f or from an SMTP
2717 MAIL command, and the trusted user did not use -F to supply a sender
2718 name, $originator_name was incorrectly used when constructing a From:
2719 header. Furthermore, $originator_name was used for submission mode
2720 messages from external hosts without From: headers in a similar way,
2721 which is clearly wrong.
2723 PH/37 Added control=suppress_local_fixups.
2725 PH/38 When log_selector = +received_sender was set, and the addition of the
2726 sender made the log line's construction buffer exactly full, or one byte
2727 less than full, an overflow happened when the terminating "\n" was
2730 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2731 when the result of a list match is failure because a DNS lookup failed.
2733 PH/40 RM_COMMAND is now used in the building process.
2735 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2736 the "build-* directories that it finds.
2738 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2739 address in a domain literal was a prefix of an interface address.
2741 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2742 when verifying a sender address, unless rewrite_headers is false.
2744 PH/44 Wrote a long comment about why errors_to addresses are verified as
2745 recipients, not senders.
2747 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2748 the ratelimit ACL was added.
2750 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2752 PH/46 Added extra information about PostgreSQL errors to the error string.
2754 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2755 avoiding the need to install special zones in a real server. This is
2756 backwards compatible; if it can't find the fake resolver, it drops back.
2757 Thus, both old and new test suites can be run.
2759 TF/02 Added util/ratelimit.pl
2761 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2762 clock is set back in time.
2764 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2765 Candler <B.Candler@pobox.com>.
2767 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2768 OK for addresses that are the result of redirections.
2770 PH/48 A number of further additions for the benefit of the new test suite,
2771 including a fake gethostbyname() that interfaces to the fake DNS resolver
2774 TF/06 The fix for widen_domains has also been applied to qualify_single and
2775 search_parents which are the other dnslookup options that can cause
2778 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2781 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2783 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2784 service" if it encountered an MX record with an empty target hostname.
2785 The message is now "an MX or SRV record indicated no SMTP service".
2787 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2788 if the DNS of the sending site is misconfigured. This is quite a
2789 common situation. This change restores the behaviour of treating a
2790 helo verification defer as a failure.
2792 PH/51 If self=fail was set on a router, the bounce message did not include the
2793 actual error message.
2799 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2801 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2802 command itself ended in error, the underlying message about the transport
2803 filter timeout was being overwritten with the pipe command error. Now the
2804 underlying error message should be appended to the second error message.
2806 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2808 PH/02 Reduce the amount of output that "make" produces by default. Full output
2809 can still be requested.
2811 PH/03 The warning log line about a condition test deferring for a "warn" verb
2812 was being output only once per connection, rather than after each
2813 occurrence (because it was using the same function as for successful
2814 "warn" verbs). This seems wrong, so I have changed it.
2816 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2817 it should not have, which might have caused a crash in the right
2818 circumstances, but probably never did.
2820 PH/04 Installed a modified version of Tony Finch's patch to make submission
2821 mode fix the return path as well as the Sender: header line, and to
2822 add a /name= option so that you can make the user's friendly name appear
2825 TF/03 Added the control = fakedefer ACL modifier.
2827 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2828 Mark Lowes for thorough testing.
2830 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2832 TK/03 Merged latest SRS patch from Miles Wilton.
2834 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2835 with the definition in sysexits.h (which is #included earlier).
2836 Fortunately, Exim does not actually use EX_OK. The code used to try to
2837 preserve the sysexits.h value, by assumimg that macro definitions were
2838 scanned for macro replacements. I have been disabused of this notion,
2839 so now the code just undefines EX_OK before #including unistd.h.
2841 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2842 in the smtp transport. When a block could not be written in a single
2843 write() function, the timeout was being re-applied to each part-write.
2844 This seems wrong - if the receiver was accepting one byte at a time it
2845 would take for ever. The timeout is now adjusted when this happens. It
2846 doesn't have to be particularly precise.
2848 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2849 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2851 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2852 without a domain if the check to <postmaster@domain> fails.
2854 SC/01 Eximstats: added -xls and the ability to specify output files
2855 (patch written by Frank Heydlauf).
2857 SC/02 Eximstats: use FileHandles for outputing results.
2859 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2861 SC/04 Eximstats: fixed display of large numbers with -nvr option
2863 SC/05 Eximstats: fixed merging of reports with empty tables.
2865 SC/06 Eximstats: added the -include_original_destination flag
2867 SC/07 Eximstats: removed tabs and trailing whitespace.
2869 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2871 TK/06 MBOX spool code: Add real "From " MBOX separator line
2872 so the .eml file is really in mbox format (even though
2873 most programs do not really care). Patch from Alex Miller.
2875 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2876 The latter is generated from $received_to and is only set if the
2877 message has one envelope recipient. SA can use these headers,
2878 obviously out-of-the-box. Patch from Alex Miller.
2880 PH/08 The ${def test on a variable was returning false if the variable's
2881 value was "0", contrary to what the specification has always said!
2882 The result should be true unless the variable is empty.
2884 PH/09 The syntax error of a character other than { following "${if
2885 def:variable_name" (after optional whitespace) was not being diagnosed.
2886 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2887 accidental colon was present, for example, could give incorrect results.
2889 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2890 result is used (not including appendfile, where other changes are about
2893 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2894 This involved changing a lot of size variables from int to off_t. It
2895 should work with maildirs and everything.
2897 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2898 spamd dying while we are connected to it.
2900 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2903 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2904 function for BDB 4.3.
2906 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2908 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2909 the sources for PCRE have been re-arranged and more files are now
2912 PH/15 The code I had for printing potentially long long variables in PH/11
2913 above was not the best (it lost precision). The length of off_t variables
2914 is now inspected at build time, and an appropriate printing format (%ld
2915 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2916 to be "long long int" or "long int". This is needed for the internal
2917 formatting function string_vformat().
2919 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2920 the configuration file to be ":syslog", then the script "guesses" where
2921 the logs files are, rather than using the compiled in default. In our
2922 case the guess is not the same as the compiled default, so the script
2923 suddenly stopped working when I started to use syslog. The patch checks
2924 to see if log_file_path is "". If so, it attempts to read it from exim
2925 with no configuration file to get the compiled in version, before it
2926 falls back to the previous guessing code."
2928 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2929 implementing BATV in an Exim configuration. See NewStuff for the gory
2932 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2933 Makefile that are specific to HP-UX.
2935 PH/18 If the "use_postmaster" option was set for a recipient callout together
2936 with the "random" option, the postmaster address was used as the MAIL
2937 FROM address for the random test, but not for the subsequent recipient
2938 test. It is now used for both.
2940 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2941 patch removes a few documentation additions to RFC 3028, because the
2942 latest draft now contains them. It adds the new en;ascii-case comparator
2943 and a new error check for 8bit text in MIME parts. Comparator and
2944 require names are now matched exactly. I enabled the subaddress
2945 extension, but it is not well tested yet (read: it works for me)."
2947 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2948 rework some of the code of TK/09 above to avoid the hardwired use of
2949 "%lld" and "long long". Replaced the call to snprintf() with a call to
2952 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2953 records point to non-existent hosts", "retry timeout exceeded", and
2954 "retry time not reached for any host after a long failure period".
2956 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2957 experimental DomainKeys support:
2959 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2960 (2) On an error such as an illegally used "control", the wrong name for
2961 the control was given.
2963 These problems did NOT occur unless DomainKeys support was compiled.
2965 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2967 PH/24 Added ${if match_ip condition.
2969 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2970 databases so that it will be absolutely obvious if a crash occurs in the
2971 DB library. This is a regular occurrence (often caused by mis-matched
2974 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2975 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2976 picked up on a user's system that detects such things. There doesn't seem
2977 to be a gcc warning option for this - only an attribute that has to be
2978 put on the function's prototype. It seems that in Fedora Core 4 they have
2979 set this on a number of new functions. No doubt there will be more in due
2982 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2983 specify a transport. However, if an address that was verified by such a
2984 router was the subject of a callout, Exim crashed because it tried to
2985 read the rcpt_include_affixes from the non-existent transport. Now it
2986 just assumes that the setting of that option is false. This bug was
2987 introduced by 4.51/PH/31.
2989 PH/28 Changed -d+all to exclude +memory, because that information is very
2990 rarely of interest, but it makes the output a lot bigger. People tend to
2991 do -d+all out of habit.
2993 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2994 code in os-type was giving problems when libc.so lives in lib64, like on
2997 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2998 aren't the modern standard, and it seems that some systems' include files
2999 don't always have them. Exim was already checking for some of the newer
3000 ones like T_AAAA, and defining it itself. I've added checks for all the
3001 record types that Exim uses.
3003 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3004 not automatically generating a new one, as it is supposed to. This
3005 prevented TLS from working. If the file did exist, but contained invalid
3006 data, a new version was generated, as expected. It was only the case of a
3007 non-existent file that was broken.
3009 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3010 with a change in libdomainkeys > 0.64.
3012 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3013 from DNS. If the selector record carries the flag, it now has
3014 precedence over the domain-wide flag.
3016 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3018 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3019 the use of a port name that isn't defined in /etc/services) occurred, the
3020 message was deferred as in a normal delivery, and thus remained on the
3021 spool, instead of being failed because of the mua_wrapper setting. This
3022 is now fixed, and I tidied up some of the mua_wrapper messages at the
3025 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3026 the messages in a hash of arrays rather than using individual hashes.
3027 This is a bit cleaner and results in dramatic memory savings, albeit
3028 at a slight CPU cost.
3030 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3031 as requested by Marc Sherman.
3033 SC/10 Eximstats: added histograms for user specified patterns as requested
3036 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3038 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3039 fopen() in the content-scanning modules that did not already have it.
3045 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3046 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3048 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3050 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3052 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3053 to test Sieve filters that use "vacation".
3055 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3056 that changes the way the GnuTLS parameters are stored in the cache file.
3057 The new format can be generated externally. For backward compatibility,
3058 if the data in the cache doesn't make sense, Exim assumes it has read an
3059 old-format file, and it generates new data and writes a new file. This
3060 means that you can't go back to an older release without removing the
3063 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3064 work if there are any delivery delays because "one_time" forces the
3065 parent to be marked "delivered", so its unseen clone is never tried
3066 again. For this reason, Exim now forbids the simultaneous setting of
3069 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3070 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3071 entry if you want to know the details. The fix, however, neglected to
3072 consider the case when local delivery batching is involved. The test for
3073 "previously delivered" was not happening when checking to see if an
3074 address could be batched with a previous (undelivered) one; under
3075 certain circumstances this could lead to multiple deliveries to the same
3078 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3079 in its include files, and this causes problems building Exim.
3081 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3082 header_syntax) but Exim was just ignoring anything given after a slash.
3083 In particular, this caused confusion with an attempt to use "verify =
3084 reverse_host_lookup/defer_ok". An error is now given when options are
3085 supplied for verify items that do not have them. (Maybe reverse_host_
3086 lookup should have a defer_ok option, but that's a different point.)
3088 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3089 defined by RFC 821) to 2048, because there were problems with some AUTH
3090 commands, and RFC 1869 says the size should be increased for extended
3091 SMTP commands that take arguments.
3093 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3096 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3097 "unknown" error; now it says that the functionality isn't in the binary.
3099 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3100 an address' error message when a string expansion fails (syntax or
3101 whatever). Otherwise the password may appear in the log. Following change
3102 PH/42 below, there is no longer a chance of it appearing in a bounce
3105 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3107 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3108 of its addresses was ever tried. (Bugzilla bug #2.)
3110 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3111 the result incorrectly in the debug output. (It correctly added a newline
3112 to what was transported.)
3114 TF/01 Added $received_time.
3116 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3117 commented out examples of how to interface to a virus scanner and to
3118 SpamAssassin. Also added commented examples of av_scanner and
3119 spamd_address settings.
3121 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3122 and controls are allowed in which ACLs. There were a couple of minor
3123 errors. Some of the entries in the conditions table (which is a table of
3124 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3125 negation of where the condition IS allowed.
3127 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3129 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3130 header file does not have a version number, so I've had to invent a new
3131 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3132 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3133 radiusclient), but it was contributed by a Radius user.
3135 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3136 files or directories, for OpenSSL.
3138 PH/19 When an Exim process that is running as root has to create an Exim log
3139 file, it does so in a subprocess that runs as exim:exim so as to get the
3140 ownership right at creation (otherwise, other Exim processes might see
3141 the file with the wrong ownership). There was no test for failure of this
3142 fork() call, which would lead to the process getting stuck as it waited
3143 for a non-existent subprocess. Forks do occasionally fail when resources
3144 run out. I reviewed all the other calls to fork(); they all seem to check
3147 PH/20 When checking for unexpected SMTP input at connect time (before writing
3148 the banner), Exim was not dealing correctly with a non-positive return
3149 from the read() function. If the client had disconnected by this time,
3150 the result was a log entry for a synchronization error with an empty
3151 string after "input=" when read() returned zero. If read() returned -1
3152 (an event I could not check), uninitialized data bytes were printed.
3153 There were reports of junk text (parts of files, etc) appearing after
3156 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3158 PH/22 Added support for macro redefinition, and (re)definition in between
3159 driver and ACL definitions.
3161 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3162 forgetting to use the resulting value; it was using the unexpanded value.
3164 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3165 hadn't been configured. The fix is from Juergen Kreileder, who
3166 understands it better than I do:
3168 "Here's what I see happening with three configured cyrus_sasl
3169 authenticators configured (plain, login, cram-md5):
3171 On startup auth_cyrus_sasl_init() gets called for each of these.
3172 This means three calls to sasl_listmech() without a specified mech_list.
3173 => SASL tests which mechs of all available mechs actually work
3174 => three warnings about OTP not working
3175 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3177 With the patch, sasl_listmech() also gets called three times. But now
3178 SASL's mech_list option is set to the server_mech specified in the the
3179 authenticator. Or in other words, the answer from sasl_listmech()
3180 gets limited to just the mech you're testing for (which is different
3182 => the return list contains just 'plain' or 'login', 'cram-md5' or
3183 nothing depending on the value of ob->server_mech.
3185 I've just tested the patch: Authentication still works fine,
3186 unavailable mechs specified in the exim configuration are still
3187 caught, and the auth.log warnings about OTP are gone."
3189 PH/25 When debugging is enabled, the contents of the command line are added
3190 to the debugging output, even when log_selector=+arguments is not
3193 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3194 answer is "GNU", and only if the return is "GNU/something" is the answer
3197 PH/27 $acl_verify_message is now set immediately after the failure of a
3198 verification in an ACL, and so is available in subsequent modifiers. In
3199 particular, the message can be preserved by coding like this:
3201 warn !verify = sender
3202 set acl_m0 = $acl_verify_message
3204 Previously, $acl_verify_message was set only while expanding "message"
3205 and "log_message" when a very denied access.
3207 PH/28 Modified OS/os.c-Linux with
3209 -#ifndef OS_LOAD_AVERAGE
3210 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3212 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3213 nomenclature these days.)
3215 PH/29 Installed patch from the Sieve maintainer that adds the options
3216 sieve_useraddress and sieve_subaddress to the redirect router.
3218 PH/30 In these circumstances:
3219 . Two addresses routed to the same list of hosts;
3220 . First host does not offer TLS;
3221 . First host accepts first address;
3222 . First host gives temporary error to second address;
3223 . Second host offers TLS and a TLS session is established;
3224 . Second host accepts second address.
3225 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3226 and peerdn, if requested) that were in fact used only for the second
3229 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3230 attention to any local part prefix or suffix that was matched by the
3231 router that accepted the address. It now behaves in the same way as it
3232 does for delivery: the affixes are removed from the local part unless
3233 rcpt_include_affixes is set on the transport.
3235 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3236 timeout during the DATA phase of an incoming message.
3238 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3239 applied a patch sanctioned by the Sieve maintainer.
3241 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3242 the uid or gid is negative. A case of a negative gid caused this to be
3243 noticed. The fix allows for either to be negative.
3245 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3246 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3249 PH/36 The change PH/12 above was broken. Fixed it.
3251 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3252 the grounds that routing the same address twice would always produce the
3253 same answer. This might have been true once, but it is certainly no
3254 longer true now. Routing a child address may depend on the previous
3255 routing that produced that child. Some complicated redirection strategies
3256 went wrong when messages had multiple recipients, and made Exim's
3257 behaviour dependent on the order in which the addresses were given.
3259 I have moved the duplicate checking until after the routing is complete.
3260 Exim scans the addresses that are assigned to local and remote
3261 transports, and removes any duplicates. This means that more work will be
3262 done, as duplicates will always all be routed, but duplicates are
3263 presumably rare, so I don't expect this is of any significance.
3265 For deliveries to pipes, files, and autoreplies, the duplicate checking
3266 still happens during the routing process, since they are not going to be
3269 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3270 It corrects a timeout issue with spamd. This is Ian's comment: "The
3271 background is that sometimes spamd either never reads data from a
3272 connection it has accepted, or it never writes response data. The exiscan
3273 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3274 blindly assumes that writes won't block so it may never time out."
3276 PH/39 Allow G after quota size as well as K and M.
3278 PH/40 The value set for $authenticated_id in an authenticator may not contain
3279 binary zeroes or newlines because the value is written to log lines and
3280 to spool files. There was no check on this. Now the value is run through
3281 the string_printing() function so that such characters are converted to
3282 printable escape sequences.
3284 PH/41 $message_linecount is a new variable that contains the total number of
3285 lines in the message. Compare $body_linecount, which is the count for the
3288 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3289 bounce and delay warning messages, except in certain special cases, which
3292 (a) An SMTP error message from a remote host;
3293 (b) A message specified in a :fail: redirection;
3294 (c) A message specified in a "fail" command in a system filter;
3295 (d) A message specified in a FAIL return from the queryprogram router;
3296 (e) A message specified by the cannot_route_message router option.
3298 In these cases only, Exim does include the error details in bounce and
3299 warning messages. There are also a few cases where bland messages such
3300 as "unrouteable address" or "local delivery error" are given.
3302 PH/43 $value is now also set for the "else" part of a ${run expansion.
3304 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3305 being worked on, but at least Exim now implements the latest version to
3308 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3309 process to complete was treated as a delivery failure, a timeout while
3310 writing the message to the pipe was logged, but erroneously treated as a
3311 successful delivery. Such timeouts include transport filter timeouts. For
3312 consistency with the overall process timeout, these timeouts are now
3313 treated as errors, giving rise to delivery failures by default. However,
3314 there is now a new Boolean option for the pipe transport called
3315 timeout_defer, which, if set TRUE, converts the failures into defers for
3316 both kinds of timeout. A transport filter timeout is now identified in
3319 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3320 systems where "make" and "gmake" are different, calling "gmake" at top
3321 level broke things. I've arranged for the value of $(MAKE) to be passed
3322 from the Makefile to this script so that it can call the same version of
3326 A note about Exim versions 4.44 and 4.50
3327 ----------------------------------------
3329 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3330 changes of various kinds. As a consequence, a big documentation update was
3331 needed. This delayed the release for rather longer than seemed good, especially
3332 in the light of a couple of (minor) security issues. Therefore, the changes
3333 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3334 release. So 4.44 and 4.50 are in effect two different branches that both start
3337 I have left the 4.50 change log unchanged; it contains all the changes since
3338 4.43. The change log for 4.44 is below; many of its items are identical to
3339 those for 4.50. This seems to be the most sensible way to preserve the
3340 historical information.
3346 1. Minor wording change to the doc/README.SIEVE file.
3348 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3349 computation of the current number of files was incorrect.
3351 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3352 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3355 4. Give more explanation in the error message when the command for a transport
3356 filter fails to execute.
3358 5. There are several places where Exim runs a non-Exim command in a
3359 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3360 was being done only for the command run by the queryprogram router. It is
3361 now done for all such subprocesses. The other cases are: ${run, transport
3362 filters, and the commands run by the lmtp and pipe transports.
3364 6. Added CONFIGURE_GROUP build-time option.
3366 7. Some older OS have a limit of 256 on the maximum number of file
3367 descriptors. Exim was using setrlimit() to set 1000 as a large value
3368 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3369 systems. I've change it so that if it can't get 1000, it tries for 256.
3371 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3372 was an oversight, and furthermore, ever since the addition of extra
3373 controls (e.g. 4.43/32), the checks on when to allow different forms of
3374 "control" were broken. There should now be diagnostics for all cases when a
3375 control that does not make sense is encountered.
3377 9. Added the /retain_sender option to "control=submission".
3379 10. $recipients is now available in the predata ACL (oversight).
3381 11. Tidy the search cache before the fork to do a delivery from a message
3382 received from the command line. Otherwise the child will trigger a lookup
3383 failure and thereby defer the delivery if it tries to use (for example) a
3384 cached ldap connection that the parent has called unbind on.
3386 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3387 of $address_data from the recipient verification was clobbered by the
3388 sender verification.
3390 13. The value of address_data from a sender verification is now available in
3391 $sender_address_data in subsequent conditions in the ACL statement.
3393 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3395 15. Added a new option "connect=<time>" to callout options, to set a different
3398 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3399 was its contents. (It was OK if the option was not defined at all.)
3401 17. A "Completed" log line is now written for messages that are removed from
3402 the spool by the -Mrm option.
3404 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3405 information about exactly what failed.
3407 19. Added -dd to debug only the daemon process.
3409 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3410 handles timeouts, both on the server side and network timeouts. Renamed the
3411 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3413 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3414 It is now set to "smtps".
3416 22. $host_address is now set to the target address during the checking of
3417 ignore_target_hosts.
3419 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3420 being passed; this would have caused $sender_host_name to have been used if
3421 matching the list had actually called for a host name (not very likely,
3422 since this list is usually IP addresses). A host name is now passed as
3425 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3426 code that specifies a non-restarting handler (typically sigaction() in
3427 modern systems) in an attempt to fix a rare and obscure crash bug.
3429 25. Narrowed the window for a race in the daemon that could cause it to ignore
3430 SIGCHLD signals. This is not a major problem, because they are used only to
3431 wake it up if nothing else does.
3433 26. A malformed maildirsize file could cause Exim to calculate negative values
3434 for the mailbox size or file count. Odd effects could occur as a result.
3435 The maildirsize information is now recalculated if the size or filecount
3438 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3439 support for a long time. Removed HAVE_SYS_VFS_H.
3441 28. Installed the latest version of exipick from John Jetmore.
3443 29. In an address list, if the pattern was not a regular expression, an empty
3444 subject address (from a bounce message) matched only if the pattern was an
3445 empty string. Non-empty patterns were not even tested. This was the wrong
3446 because it is perfectly reasonable to use an empty address as part of a
3447 database query. An empty address is now tested by patterns that are
3448 lookups. However, all the other forms of pattern expect the subject to
3449 contain a local part and a domain, and therefore, for them, an empty
3450 address still always fails if the pattern is not itself empty.
3452 30. Exim went into a mad DNS loop when attempting to do a callout where the
3453 host was specified on an smtp transport, and looking it up yielded more
3454 than one IP address.
3456 31. Re-factored the code for checking spool and log partition space into a
3457 function that finds that data and another that does the check. The former
3458 is then used to implement four new variables: $spool_space, $log_space,
3459 $spool_inodes, and $log_inodes.
3461 32. The RFC2047 encoding function was originally intended for short strings
3462 such as real names; it was not keeping to the 75-character limit for
3463 encoded words that the RFC imposes. It now respects the limit, and
3464 generates multiple encoded words if necessary. To be on the safe side, I
3465 have increased the buffer size for the ${rfc2047: expansion operator from
3468 33. It is now permitted to omit both strings after an "if" condition; if the
3469 condition is true, the result is "true". As before, when the second string
3470 is omitted, a false condition yields an empty string. This makes it less
3471 cumbersome to write custom ACL and router conditions.
3473 34. Failure to deliver a bounce message always caused it to be frozen, even if
3474 there was an errors_to setting on the router. The errors_to setting is now
3477 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3478 canonical form (fully expanded) before being placed in
3479 $sender_host_address.
3481 36. The table in the code that translates DNS record types into text (T_A to
3482 "A" for instance) was missing entries for NS and CNAME. It is just possible
3483 that this could have caused confusion if both these types were looked up
3484 for the same domain, because the text type is used as part of Exim's
3485 per-process caching. But the chance of anyone hitting this buglet seems
3488 37. The dnsdb lookup has been extended in a number of ways.
3490 (1) There is a new type, "zns", which walks up the domain tree until it
3491 finds some nameserver records. It should be used with care.
3493 (2) There is a new type, "mxh", which is like "mx" except that it returns
3494 just the host names, not the priorities.
3496 (3) It is now possible to give a list of domains (or IP addresses) to be
3497 looked up. The behaviour when one of the lookups defers can be
3498 controlled by a keyword.
3500 (4) It is now possible to specify the separator character for use when
3501 multiple records are returned.
3503 38. The dnslists ACL condition has been extended: it is now possible to supply
3504 a list of IP addresses and/or domains to be looked up in a particular DNS
3507 39. Added log_selector=+queue_time_overall.
3509 40. When running the queue in the test harness, wait just a tad after forking a
3510 delivery process, to get repeatability of debugging output.
3512 41. Include certificate and key file names in error message when GnuTLS fails
3513 to set them up, because the GnuTLS error message doesn't include the name
3514 of the failing file when there is a problem reading it.
3516 42. Allow both -bf and -bF in the same test run.
3518 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3520 44. The "Exiscan patch" is now merged into the mainline Exim source.
3522 45. Sometimes the final signoff response after QUIT could fail to get
3523 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3524 < 0 before doing a fflush(). This bug looks as though it goes back to the
3525 introduction of TLS in release 3.20, but "sometimes" must have been rare
3526 because the tests only now provoked it.
3528 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3529 (this can affect the format of dates).
3531 47. exim_tidydb, when checking for the continued existence of a message for
3532 which it has found a message-specific retry record, was not finding
3533 messages that were in split spool directories. Consequently, it was
3534 deleting retry records that should have stayed in existence.
3536 48. Steve fixed some bugs in eximstats.
3538 49. The SPA authentication driver was not abandoning authentication and moving
3539 on to the next authenticator when an expansion was forced to fail,
3540 contradicting the general specification for all authenticators. Instead it
3541 was generating a temporary error. It now behaves as specified.
3543 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3544 (the order specifies the preference for clients). The order is now AES256,
3545 AES128, 3DES, ARCFOUR128.
3547 51. Small patch to Sieve code - explicitly set From: when generating an
3550 52. Exim crashed if a remote delivery caused a very long error message to be
3551 recorded - for instance if somebody sent an entire SpamAssassin report back
3552 as a large number of 550 error lines. This bug was coincidentally fixed by
3553 increasing the size of one of Exim's internal buffers (big_buffer) that
3554 happened as part of the Exiscan merge. However, to be on the safe side, I
3555 have made the code more robust (and fixed the comments that describe what
3558 53. Now that there can be additional text after "Completed" in log lines (if
3559 the queue_time_overall log selector is set), a one-byte patch to exigrep
3560 was needed to allow it to recognize "Completed" as not the last thing in
3563 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3564 patch that reportedly fixes this has been added. I am not expert enough to
3565 create a test for it. This is what the patch creator wrote:
3567 "I found a little strange behaviour of ldap code when working with
3568 Windows 2003 AD Domain, where users was placed in more than one
3569 Organization Units. When I tried to give exim partial DN, the exit code
3570 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3571 But simultaneously result of request was absolutely normal ldap result,
3572 so I produce this patch..."
3574 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3575 REFERENCE, so I have modified the code to exclude the patch when that macro
3578 55. Some experimental protocols are using DNS PTR records for new purposes. The
3579 keys for these records are domain names, not reversed IP addresses. The
3580 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3581 leaves it alone. Component reversal etc. now happens only for IP addresses.
3584 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3586 57. Double the size of the debug message buffer (to 2048) so that more of very
3587 long debug lines gets shown.
3589 58. The exicyclog utility now does better if the number of log files to keep
3590 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3592 59. Two changes related to the smtp_active_hostname option:
3594 (1) $smtp_active_hostname is now available as a variable.
3595 (2) The default for smtp_banner uses $smtp_active_hostname instead
3596 of $primary_hostname.
3598 60. The host_aton() function is supposed to be passed a string that is known
3599 to be a valid IP address. However, in the case of IPv6 addresses, it was
3600 not checking this. This is a hostage to fortune. Exim now panics and dies
3601 if the condition is not met. A case was found where this could be provoked
3602 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3603 components; fortuitously, this particular loophole had already been fixed
3604 by change 4.50/55 above.
3606 If there are any other similar loopholes, the new check in host_aton()
3607 itself should stop them being exploited. The report I received stated that
3608 data on the command line could provoke the exploit when Exim was running as
3609 exim, but did not say which command line option was involved. All I could
3610 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3611 running as the user.
3614 61. There was a buffer overflow vulnerability in the SPA authentication code
3615 (which came originally from the Samba project). I have added a test to the
3616 spa_base64_to_bits() function which I hope fixes it.
3619 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3620 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3622 63. The daemon start-up calls getloadavg() while still root for those OS that
3623 need the first call to be done as root, but it missed one case: when
3624 deliver_queue_load_max is set with deliver_drop_privilege. This is
3625 necessary for the benefit of the queue runner, because there is no re-exec
3626 when deliver_drop_privilege is set.
3628 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3629 This has been fixed.
3631 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3632 was in use, was not putting the data itself into the right store pool;
3633 consequently, it could be overwritten for a subsequent message in the same
3634 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3637 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3639 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3640 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3641 the function were treating the return as a boolean value, which happened to
3642 work because 0=false and not-0=true, but is not correct code.
3644 68. The host_aton() function was not handling scoped IPv6 addresses (those
3645 with, for example, "%eth0" on the end) correctly.
3647 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3648 negated items (that is, ~something) in unsigned ints. Some compilers
3649 apparently mutter when there is no cast.
3651 70. If an address verification called from an ACL failed, and did not produce a
3652 user-specific message (i.e. there was only a "system" message), nothing was
3653 put in $acl_verify_message. In this situation, it now puts the system
3656 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3657 session; change 4.31/43 added the unwanted input to the log line - except
3658 that it did not do this in the start of session case. It now does.
3660 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3661 This is wrong and can cause the other end to generate a synchronization
3662 error if it is another Exim or anything else that does the synchronization
3663 check. A QUIT command is no longer sent after a timeout.
3665 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3666 during host lookups.
3668 74. The defer_ok option of callout verification was not working if it was used
3669 when verifying addresses in header lines, that is, for this case:
3671 verify = header_sender/callout=defer_ok
3673 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3674 those file descriptors could be used for SMTP connections. If anything
3675 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3676 could be sent to the SMTP client, causing chaos. The daemon now opens
3677 stdin, stdout, and stderr to /dev/null when it puts itself into the
3680 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3681 log by default. The user can override this with suitable Perl magic.
3683 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3684 the log message when discard triggers, was not working for the DATA ACL or
3685 for the non-SMTP ACL.
3687 78. Error message wording change in sieve.c.
3689 79. If smtp_accept_max_per_host was set, the number of connections could be
3690 restricted to fewer than expected, because the daemon was trying to set up
3691 a new connection before checking whether the processes handling previous
3692 connections had finished. The check for completed processes is now done
3693 earlier. On busy systems, this bug wouldn't be noticed because something
3694 else would have woken the daemon, and it would have reaped the completed
3697 80. If a message was submitted locally by a user whose login name contained one
3698 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3699 It caused a spool format error. I have fixed the spool reading code. A
3700 related problem was that the "from" clause in the Received: line became
3701 illegal because of the space(s). It is now covered by ${quote_local_part.
3703 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3706 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3708 83. Patch from Sieve maintainer to fix unterminated string problem in
3709 "vacation" handling.
3711 84. Some minor changes to the Linux configuration files to help with other
3712 OS variants using glibc.
3714 85. One more patch for Sieve to update vacation handling to latest spec.
3717 ----------------------------------------------------
3718 See the note above about the 4.44 and 4.50 releases.
3719 ----------------------------------------------------
3725 1. Change 4.43/35 introduced a bug that caused file counts to be
3726 incorrectly computed when quota_filecount was set in an appendfile
3729 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3730 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3733 3. Give more explanation in the error message when the command for a transport
3734 filter fails to execute.
3736 4. There are several places where Exim runs a non-Exim command in a
3737 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3738 was being done only for the command run by the queryprogram router. It is
3739 now done for all such subprocesses. The other cases are: ${run, transport
3740 filters, and the commands run by the lmtp and pipe transports.
3742 5. Some older OS have a limit of 256 on the maximum number of file
3743 descriptors. Exim was using setrlimit() to set 1000 as a large value
3744 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3745 systems. I've change it so that if it can't get 1000, it tries for 256.
3747 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3748 was an oversight, and furthermore, ever since the addition of extra
3749 controls (e.g. 4.43/32), the checks on when to allow different forms of
3750 "control" were broken. There should now be diagnostics for all cases when a
3751 control that does not make sense is encountered.
3753 7. $recipients is now available in the predata ACL (oversight).
3755 8. Tidy the search cache before the fork to do a delivery from a message
3756 received from the command line. Otherwise the child will trigger a lookup
3757 failure and thereby defer the delivery if it tries to use (for example) a
3758 cached ldap connection that the parent has called unbind on.
3760 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3761 of $address_data from the recipient verification was clobbered by the
3762 sender verification.
3764 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3765 was its contents. (It was OK if the option was not defined at all.)
3767 11. A "Completed" log line is now written for messages that are removed from
3768 the spool by the -Mrm option.
3770 12. $host_address is now set to the target address during the checking of
3771 ignore_target_hosts.
3773 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3774 being passed; this would have caused $sender_host_name to have been used if
3775 matching the list had actually called for a host name (not very likely,
3776 since this list is usually IP addresses). A host name is now passed as
3779 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3780 code that specifies a non-restarting handler (typically sigaction() in
3781 modern systems) in an attempt to fix a rare and obscure crash bug.
3783 15. Narrowed the window for a race in the daemon that could cause it to ignore
3784 SIGCHLD signals. This is not a major problem, because they are used only to
3785 wake it up if nothing else does.
3787 16. A malformed maildirsize file could cause Exim to calculate negative values
3788 for the mailbox size or file count. Odd effects could occur as a result.
3789 The maildirsize information is now recalculated if the size or filecount
3792 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3793 support for a long time. Removed HAVE_SYS_VFS_H.
3795 18. Updated exipick to current release from John Jetmore.
3797 19. Allow an empty sender to be matched against a lookup in an address list.
3798 Previously the only cases considered were a regular expression, or an
3801 20. Exim went into a mad DNS lookup loop when doing a callout where the
3802 host was specified on the transport, if the DNS lookup yielded more than
3805 21. The RFC2047 encoding function was originally intended for short strings
3806 such as real names; it was not keeping to the 75-character limit for
3807 encoded words that the RFC imposes. It now respects the limit, and
3808 generates multiple encoded words if necessary. To be on the safe side, I
3809 have increased the buffer size for the ${rfc2047: expansion operator from
3812 22. Failure to deliver a bounce message always caused it to be frozen, even if
3813 there was an errors_to setting on the router. The errors_to setting is now
3816 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3817 canonical form (fully expanded) before being placed in
3818 $sender_host_address.
3820 24. Updated eximstats to version 1.33
3822 25. Include certificate and key file names in error message when GnuTLS fails
3823 to set them up, because the GnuTLS error message doesn't include the name
3824 of the failing file when there is a problem reading it.
3826 26. Expand error message when OpenSSL has problems setting up cert/key files.
3829 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3830 (this can affect the format of dates).
3832 28. exim_tidydb, when checking for the continued existence of a message for
3833 which it has found a message-specific retry record, was not finding
3834 messages that were in split spool directories. Consequently, it was
3835 deleting retry records that should have stayed in existence.
3837 29. eximstats updated to version 1.35
3838 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3839 1.35 - bugfix such that pie charts by volume are generated correctly
3841 30. The SPA authentication driver was not abandoning authentication and moving
3842 on to the next authenticator when an expansion was forced to fail,
3843 contradicting the general specification for all authenticators. Instead it
3844 was generating a temporary error. It now behaves as specified.
3846 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3847 (the order specifies the preference for clients). The order is now AES256,
3848 AES128, 3DES, ARCFOUR128.
3850 31. Small patch to Sieve code - explicitly set From: when generating an
3853 32. Exim crashed if a remote delivery caused a very long error message to be
3854 recorded - for instance if somebody sent an entire SpamAssassin report back
3855 as a large number of 550 error lines. This bug was coincidentally fixed by
3856 increasing the size of one of Exim's internal buffers (big_buffer) that
3857 happened as part of the Exiscan merge. However, to be on the safe side, I
3858 have made the code more robust (and fixed the comments that describe what
3861 33. Some experimental protocols are using DNS PTR records for new purposes. The
3862 keys for these records are domain names, not reversed IP addresses. The
3863 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3864 leaves it alone. Component reversal etc. now happens only for IP addresses.
3867 34. The host_aton() function is supposed to be passed a string that is known
3868 to be a valid IP address. However, in the case of IPv6 addresses, it was
3869 not checking this. This is a hostage to fortune. Exim now panics and dies
3870 if the condition is not met. A case was found where this could be provoked
3871 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3872 components; fortuitously, this particular loophole had already been fixed
3873 by change 4.50/55 or 4.44/33 above.
3875 If there are any other similar loopholes, the new check in host_aton()
3876 itself should stop them being exploited. The report I received stated that
3877 data on the command line could provoke the exploit when Exim was running as
3878 exim, but did not say which command line option was involved. All I could
3879 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3880 running as the user.
3883 35. There was a buffer overflow vulnerability in the SPA authentication code
3884 (which came originally from the Samba project). I have added a test to the
3885 spa_base64_to_bits() function which I hope fixes it.
3888 36. The daemon start-up calls getloadavg() while still root for those OS that
3889 need the first call to be done as root, but it missed one case: when
3890 deliver_queue_load_max is set with deliver_drop_privilege. This is
3891 necessary for the benefit of the queue runner, because there is no re-exec
3892 when deliver_drop_privilege is set.
3894 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3895 was in use, was not putting the data itself into the right store pool;
3896 consequently, it could be overwritten for a subsequent message in the same
3897 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3900 38. Sometimes the final signoff response after QUIT could fail to get
3901 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3902 < 0 before doing a fflush(). This bug looks as though it goes back to the
3903 introduction of TLS in release 3.20, but "sometimes" must have been rare
3904 because the tests only now provoked it.
3910 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3911 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3912 Now it returns an integer. A number of calls were still expecting a T/F
3913 return. Fortuitously, in all cases, the tests worked in OK situations,
3914 which is the norm. However, things would have gone wrong on any write
3915 failures on the smtp file descriptor. This function is used when sending
3916 messages over SMTP and also when doing verify callouts.
3918 2. When Exim is called to do synchronous delivery of a locally submitted
3919 message (the -odf or -odi options), it no longer closes stderr before doing
3922 3. Implemented the mua_wrapper option.
3924 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3926 5. Implemented the functions header_remove(), header_testname(),
3927 header_add_at_position(), and receive_remove_recipient(), and exported them
3930 6. If an ACL "warn" statement specified the addition of headers, Exim already
3931 inserted X-ACL-Warn: at the start if there was no header name. However, it
3932 was not making this test for the second and subsequent header lines if
3933 there were newlines in the string. This meant that an invalid header could
3934 be inserted if Exim was badly configured.
3936 7. Allow an ACL "warn" statement to add header lines at the start or after all
3937 the Received: headers, as well as at the end.
3939 8. Added the rcpt_4xx retry error code.
3941 9. Added postmaster_mailfrom=xxx to callout verification option.
3943 10. Added mailfrom=xxxx to the callout verification option, for verify=
3946 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3947 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3949 12. Inserted some casts to stop certain compilers warning when using pointer
3950 differences as field lengths or precisions in printf-type calls (mostly
3951 affecting debugging statements).
3953 13. Added optional readline() support for -be (dynamically loaded).
3955 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3956 same clock tick as a message's arrival, so that its received time was the
3957 same as the "first fail" time on the retry record, and that message
3958 remained on the queue past the ultimate address timeout, every queue runner
3959 would try a delivery (because it was past the ultimate address timeout) but
3960 after another failure, the ultimate address timeout, which should have then
3961 bounced the address, did not kick in. This was a "< instead of <=" error;
3962 in most cases the first failure would have been in the next clock tick
3963 after the received time, and all would be well.
3965 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3966 being recognized when the domain list was tested by the match_domain
3967 condition in an expansion string.
3969 16. Added the ${str2b64: operator.
3971 17. Exim was always calling setrlimit() to set a large limit for the number of
3972 processes, without checking whether the existing limit was already
3973 adequate. (It did check for the limit on file descriptors.) Furthermore,
3974 errors from getrlimit() and setrlimit() were being ignored. Now they are
3975 logged to the main and panic logs, but Exim does carry on, to try to do its
3976 job under whatever limits there are.
3978 18. Imported PCRE 5.0.
3980 19. Trivial typo in log message " temporarily refused connection" (the leading
3983 20. If the log selector return_path_on_delivery was set and an address was
3984 redirected to /dev/null, the delivery process crashed because it assumed
3985 that a return path would always be set for a "successful" delivery. In this
3986 case, the whole delivery is bypassed as an optimization, and therefore no
3989 21. Internal re-arrangement: the function for sending a challenge and reading
3990 a response while authentication was assuming a zero-terminated challenge
3991 string. It's now changed to take a pointer and a length, to allow for
3992 binary data in such strings.
3994 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3996 23. Exim was not respecting finduser_retries when seeking the login of the
3997 uid under which it was called; it was always trying 10 times. (The default
3998 setting of finduser_retries is zero.) Also, it was sleeping after the final
3999 failure, which is pointless.
4001 24. Implemented tls_on_connect_ports.
4003 25. Implemented acl_smtp_predata.
4005 26. If the domain in control=submission is set empty, Exim assumes that the
4006 authenticated id is a complete email address when it generates From: or
4007 Sender: header lines.
4009 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4010 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4011 chown and chgrp in /bin and hostname in /usr/bin.
4013 28. Exim was keeping the "process log" file open after each use, just as it
4014 does for the main log. This opens the possibility of it remaining open for
4015 long periods when the USR1 signal hits a daemon. Occasional processlog
4016 errors were reported, that could have been caused by this. Anyway, it seems
4017 much more sensible not to leave this file open at all, so that is what now
4020 29. The long-running daemon process does not normally write to the log once it
4021 has entered its main loop, and it closes the log before doing so. This is
4022 so that log files can straightforwardly be renamed and moved. However,
4023 there are a couple of unusual error situations where the daemon does write
4024 log entries, and I had neglected to close the log afterwards.
4026 30. The text of an SMTP error response that was received during a remote
4027 delivery was being truncated at 512 bytes. This is too short for some of
4028 the long messages that one sometimes sees. I've increased the limit to
4031 31. It is now possible to make retry rules that apply only when a message has a
4032 specific sender, in particular, an empty sender.
4034 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4035 it possible to be selective about when SMTP synchronization is enforced.
4037 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4039 32. Implemented hosts_connection_nolog.
4041 33. Added an ACL for QUIT.
4043 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4046 35. Added mailbox_size and mailbox_filecount to appendfile.
4048 36. Added control = no_multiline_responses to ACLs.
4050 37. There was a bug in the logic of the code that waits for the clock to tick
4051 in the case where the clock went backwards by a substantial amount such
4052 that the microsecond fraction of "now" was more than the microsecond
4053 fraction of "then" (but the whole seconds number was less).
4055 38. Added support for the libradius Radius client library this is found on
4056 FreeBSD (previously only the radiusclient library was supported).
4062 1. When certain lookups returned multiple values in the form name=value, the
4063 quoting of the values was not always being done properly. Specifically:
4064 (a) If the value started with a double quote, but contained no whitespace,
4066 (b) If the value contained whitespace other than a space character (i.e.
4067 tabs or newlines or carriage returns) it was not quoted.
4068 This fix has been applied to the mysql and pgsql lookups by writing a
4069 separate quoting function and calling it from the lookup code. The fix
4070 should probably also be applied to nisplus, ibase and oracle lookups, but
4071 since I cannot test any of those, I have not disturbed their existing code.
4073 2. A hit in the callout cache for a specific address caused a log line with no
4074 reason for rejecting RCPT. Now it says "Previous (cached) callout
4075 verification failure".
4077 3. There was an off-by-one bug in the queryprogram router. An over-long
4078 return line was truncated at 256 instead of 255 characters, thereby
4079 overflowing its buffer with the terminating zero. As well as fixing this, I
4080 have increased the buffer size to 1024 (and made a note to document this).
4082 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4083 when Exim is waiting for an SMTP response from a remote server, Exim
4084 restarts its select() call on the socket, thereby resetting its timeout.
4085 This is not a problem when such interrupts are rare. Somebody set up a cron
4086 job to run exiwhat every 2 minutes, which is less than the normal select()
4087 timeout (5 or 10 minutes). This meant that the select() timeout never
4088 kicked in because it was always reset. I have fixed this by comparing the
4089 time when an interrupt arrives with the time at the start of the first call
4090 to select(). If more time than the timeout has elapsed, the interrupt is
4091 treated as a timeout.
4093 5. Some internal re-factoring in preparation for the addition of Sieve
4094 extensions (by MH). In particular, the "personal" test is moved to a
4095 separate function, and given an option for scanning Cc: and Bcc: (which is
4096 not set for Exim filters).
4098 6. When Exim created an email address using the login of the caller as the
4099 local part (e.g. when creating a From: or Sender: header line), it was not
4100 quoting the local part when it contained special characters such as @.
4102 7. Installed new OpenBSD configuration files.
4104 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4105 try to make them clearer.
4107 9. Callout options, other than the timeout value, were being ignored when
4108 verifying sender addresses in header lines. For example, when using
4110 verify = header_sender/callout=no_cache
4112 the cache was (incorrectly) being used.
4114 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4115 only the Cygwin environment.
4117 11. When return_path_on_delivery was set as a log selector, if different remote
4118 addresses in the same message used different return paths and parallel
4119 remote delivery occurred, the wrong values would sometimes be logged.
4120 (Whenever a remote delivery process finished, the return path value from
4121 the most recently started remote delivery process was logged.)
4123 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4124 lines when AUTH and/or TLS are in use. This is the "received protocol"
4125 field. Exim used to use "asmtp" for authenticated SMTP, without any
4126 indication (in the protocol name) for TLS use. Now it follows the RFC and
4127 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4128 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4129 names appear in log lines as well as in Received: header lines.
4131 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4132 extensions, and comparison tests, and to fix some bugs.
4134 14. Changes to the "personal" filter test:
4136 (1) The test was buggy in that it was just doing the equivalent of
4137 "contains" tests on header lines. For example, if a user's address was
4138 anne@some.where, the "personal" test would incorrectly be true for
4140 To: susanne@some.where
4142 This test is now done by extracting each address from the header in turn,
4143 and checking the entire address. Other tests that are part of "personal"
4144 are now done using regular expressions (for example, to check local parts
4145 of addresses in From: header lines).
4147 (2) The list of non-personal local parts in From: addresses has been
4148 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4149 taken from the Sieve specification recommendations.
4151 (3) If the message contains any header line starting with "List-" it is
4152 treated as non-personal.
4154 (4) The test for "circular" in the Subject: header line has been removed
4155 because it now seems ill-conceived.
4157 15. Minor typos in src/EDITME comments corrected.
4159 16. Installed latest exipick from John Jetmore.
4161 17. If headers_add on a router specified a text string that was too long for
4162 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4163 of string_sprintf() is now avoided.
4165 18. $message_body_size was not set (it was always zero) when running the DATA
4166 ACL and the local_scan() function.
4168 19. For the "mail" command in an Exim filter, no default was being set for
4169 the once_repeat time, causing a random time value to be used if "once" was
4170 specified. (If the value happened to be <= 0, no repeat happened.) The
4171 default is now 0s, meaning "never repeat". The "vacation" command was OK
4172 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4173 (I found it when inspecting the code).
4175 20. There is now an overall timeout for performing a callout verification. It
4176 defaults to 4 times the callout timeout, which applies to individual SMTP
4177 commands during the callout. The overall timeout applies when there is more
4178 than one host that can be tried. The timeout is checked before trying the
4179 next host. This prevents very long delays if there are a large number of
4180 hosts and all are timing out (e.g. when the network connections are timing
4181 out). The value of the overall timeout can be changed by specifying an
4182 additional sub-option for "callout", called "maxwait". For example:
4184 verify = sender/callout=5s,maxwait=20s
4186 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4187 to the end before writing, but this should make it even safer).
4189 22. Exim was forgetting that it had advertised PIPELINING for the second and
4190 subsequent messages on an SMTP connection. It was also not resetting its
4191 memory on STARTTLS and an internal HELO.
4193 23. When Exim logs an SMTP synchronization error within a session, it now
4194 records whether PIPELINING has been advertised or not.
4196 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4197 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4198 rather than long int.
4200 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4202 26. Added the never_mail option to autoreply.
4208 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4209 crash if the getsockname() call failed; this can happen if a connection is
4210 closed very soon after it is established. The problem was simply in the
4211 order in which certain operations were done, causing Exim to try to write
4212 to the SMTP stream before it had set up the file descriptor. The bug has
4213 been fixed by making things happen in the correct order.
4219 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4220 before the connection was closed, thus losing the rejection response.
4222 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4223 some early Solaris releases, but causes trouble in current releases where
4224 socklen_t is defined.
4226 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4229 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4232 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4233 that releasing the top of it at the end releases what was used for sub-
4234 expansions (unless the block got too big). However, discard this block if
4235 the first thing is a variable or header, so that we can use its block when
4236 it is dynamic (useful for very large $message_headers, for example).
4238 6. Lookups now cache *every* query, not just the most recent. A new, separate
4239 store pool is used for this. It can be recovered when all lookup caches are
4240 flushed. Lookups now release memory at the end of their result strings.
4241 This has involved some general refactoring of the lookup sources.
4243 7. Some code has been added to the store_xxx() functions to reduce the amount
4244 of flapping under certain conditions.
4246 8. log_incoming_interface used to affect only the <= reception log lines. Now
4247 it causes the local interface and port to be added to several more SMTP log
4248 lines, for example "SMTP connection from", and rejection lines.
4250 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4252 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4254 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4255 could be overwritten at the end of the current message (or the start of a
4256 new message if it was set in a HELO ACL). The value is now preserved for
4257 the duration of the SMTP connection.
4259 12. If a transport had a headers_rewrite setting, and a matching header line
4260 contained an unqualified address, that address was qualified, even if it
4261 did not match any rewriting rules. The underlying bug was that the values
4262 of the flags that permit the existence of unqualified sender and recipient
4263 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4264 non-local messages, and by -bnq for local messages) were not being
4265 preserved with the message after it was received.
4267 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4268 "next input=" as part of the text comprising the host identity instead of
4269 the correct text. The code was using the same buffer for two different
4270 strings. However, depending on which order the printing function evaluated
4271 its arguments, the bug did not always show up. Under Linux, for example, my
4272 test suite worked just fine.
4274 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4275 doesn't work with some older versions of Perl. It has been changed to "my",
4276 which in any case is probably the better facility to use.
4278 15. A really picky compiler found some instances of statements for creating
4279 error messages that either had too many or two few arguments for the format
4282 16. The size of the buffer for calls to the DNS resolver has been increased
4283 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4284 for addresses that have a lot of PTR records. This alleviates a problem; it
4285 does not fully solve it.
4287 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4288 buffer now truncates the list and logs the incident, which is the same
4289 action as happens when Exim is looking up a host name and its aliases.
4290 Previously in this situation something unpredictable would happen;
4291 sometimes it was "internal error: store_reset failed".
4293 18. If a server dropped the connection unexpectedly when an Exim client was
4294 using GnuTLS and trying to read a response, the client delivery process
4295 crashed while trying to generate an error log message.
4297 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4298 string, for example:
4300 warn message = H1: something\nH2: something
4302 the text was added as a single header line from Exim's point of view
4303 though it ended up OK in the delivered message. However, searching for the
4304 second and subsequent header lines using $h_h2: did not work. This has been
4305 fixed. Similarly, if a system filter added multiple headers in this way,
4306 the routers could not see them.
4308 20. Expanded the error message when iplsearch is called with an invalid key to
4309 suggest using net-iplsearch in a host list.
4311 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4312 ACLs are no longer actually imposed (and a message to that effect is
4315 22. If a "gecos" field in a passwd entry contained escaped characters, in
4316 particular, if it contained a \" sequence, Exim got it wrong when building
4317 a From: or a Sender: header from that name. A second bug also caused
4318 incorrect handling when an unquoted " was present following a character
4319 that needed quoting.
4321 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4322 was not being matched caselessly.
4324 24. Arranged for all hyphens in the exim.8 source to be escaped with
4327 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4328 properly. Recipient callout cache records were still being keyed to include
4329 the sender, even when use_sender was set false. This led to far more
4330 callouts that were necessary. The sender is no longer included in the key
4331 when use_sender is false.
4333 26. Added "control = submission" modifier to ACLs.
4335 27. Added the ${base62d: operator to decode base 62 numbers.
4337 28. dnsdb lookups can now access SRV records.
4339 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4340 the configuration file.
4342 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4343 (-v) mode. This makes the output for a verbose queue run more intelligible.
4345 31. Added a use_postmaster feature to recipient callouts.
4347 32. Added the $body_zerocount variable, containing the number of binary zero
4348 bytes in the message body.
4350 33. The time of last modification of the "new" subdirectory is now used as the
4351 "mailbox time last read" when there is a quota error for a maildir
4354 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4356 35. Added +ignore_unknown as a special item in host lists.
4358 36. Code for decoding IPv6 addresses in host lists is now included, even if
4359 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4360 address was recognized as an IP address, but was then not correctly decoded
4361 into binary, causing unexpected and incorrect effects when compared with
4368 1. Very minor rewording of debugging text in manualroute to say "list of
4369 hosts" instead of "hostlist".
4371 2. If verify=header_syntax was set, and a header line with an unqualified
4372 address (no domain) and a large number of spaces between the end of the
4373 name and the colon was received, the reception process suffered a buffer
4374 overflow, and (when I tested it) crashed. This was caused by some obsolete
4375 code that should have been removed. The fix is to remove it!
4377 3. When running in the test harness, delay a bit after writing a bounce
4378 message to get a bit more predictability in the log output.
4380 4. Added a call to search_tidyup() just before forking a reception process. In
4381 theory, someone could use a lookup in the expansion of smtp_accept_max_
4382 per_host which, without the tidyup, could leave open a database connection.
4384 5. Added the variables $recipient_data and $sender_data which get set from a
4385 lookup success in an ACL "recipients" or "senders" condition, or a router
4386 "senders" option, similar to $domain_data and $local_part_data.
4388 6. Moved the writing of debug_print from before to after the "senders" test
4391 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4392 problems for message scanning, either using a data ACL, or using
4393 local_scan() because the Received: header was not generated till after they
4394 were called (in order to set the time as the time of reception completion).
4395 I have revised the way this works. The header is now generated after the
4396 body is received, but before the ACL or local_scan() are called. After they
4397 are run, the timestamp in the header is updated.
4403 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4404 before starting a queue runner without re-exec. This happened only when
4405 deliver_drop_privilege was set or when the Exim user was set to root. The
4406 effect of the bug was that timeouts during subsequent deliveries caused
4407 crashes instead of being properly handled. The handler is now left at its
4408 default (and expected) setting.
4410 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4411 message, again when deliver_drop_privilege is set or Exim is run as root.
4412 The bug described in (1) was not present in this case, but the tidying up
4413 of the other signals was missing. I have made the two cases consistent.
4415 3. The ignore_target_hosts setting on a manualroute router was being ignored
4416 for hosts that were looked up using the /MX notation.
4418 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4421 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4422 operated on the sender address. After changing the $sender_address to <>
4423 for the sender address verify, Exim was re-instated it as the original
4424 (before rewriting) address, but remembering that it had rewritten it, so it
4425 wasn't rewriting it again. This bug also had the effect of breaking the
4426 sender address verification caching when the sender address was rewritten.
4428 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4429 This has been changed so that if the ip literal address matches
4430 ignore_target_hosts, the router declines.
4432 7. Added expansion conditions match_domain, match_address, and match_local_
4433 part (NOT match_host).
4435 8. The placeholder for the Received: header didn't have a length field set.
4437 9. Added code to Exim itself and to exim_lock to test for a specific race
4438 condition that could lead to file corruption when using MBX delivery. The
4439 issue is with the lockfile that is created in /tmp. If this file is removed
4440 after a process has opened it but before that process has acquired a lock,
4441 there is the potential for a second process to recreate the file and also
4442 acquire a lock. This could lead to two Exim processes writing to the file
4443 at the same time. The added code performs the same test as UW imapd; it
4444 checks after acquiring the lock that its file descriptor still refers to
4445 the same named file.
4447 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4448 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4451 11. Added the smtp_active_hostname option. If used, this will typically be made
4452 to depend on the incoming interface address. Because $interface_address is
4453 not set up until the daemon has forked a reception process, error responses
4454 that can happen earlier (such as "too many connections") no longer contain
4457 12. If an expansion in a condition on a "warn" statement fails because a lookup
4458 defers, the "warn" statement is abandoned, and the next ACL statement is
4459 processed. Previously this caused the whole ACL to be aborted.
4461 13. Added the iplsearch lookup type.
4463 14. Added ident_timeout as a log selector.
4465 15. Added tls_certificate_verified as a log selector.
4467 16. Added a global option tls_require_ciphers (compare the smtp transport
4468 option of the same name). This controls incoming TLS connections.
4470 17. I finally figured out how to make tls_require_ciphers do a similar thing
4471 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4472 before starting the TLS session.
4474 18. Tabs are now shown as \t in -bP output.
4476 19. If the log selector return_path_on_delivery was set, Exim crashed when
4477 bouncing a message because it had too many Received: header lines.
4479 20. If two routers both had headers_remove settings, and the first one included
4480 a superfluous trailing colon, the final name in the first list and the
4481 first name in the second list were incorrectly joined into one item (with a
4482 colon in the middle).
4488 1. Added -C and -D options to the exinext utility, mainly to make it easier
4489 to include in the automated testing, but these could be helpful when
4490 multiple configurations are in use.
4492 2. The exinext utility was not formatting the output nicely when there was
4493 an alternate port involved in the retry record key, nor when there was a
4494 message id as well (for retries that were specific to a specific message
4495 and a specific host). It was also confused by IPv6 addresses, because of
4496 the additional colons they contain. I have fixed the IPv4 problem, and
4497 patched it up to do a reasonable job for IPv6.
4499 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4500 delivery, the log line now contains "pipelined" if PIPELINING was used.
4502 4. An SMTP transport process used to panic and die if the bind() call to set
4503 an explicit outgoing interface failed. This has been changed; it is now
4504 treated in the same way as a connect() failure.
4506 5. A reference to $sender_host_name in the part of a conditional expansion
4507 that was being skipped was still causing a DNS lookup. This no longer
4510 6. The def: expansion condition was not recognizing references to header lines
4511 that used bh_ and bheader_.
4513 7. Added the _cache feature to named lists.
4515 8. The code for checking quota_filecount in the appendfile transport was
4516 allowing one more file than it should have been.
4518 9. For compatibility with Sendmail, the command line option
4526 and sets the incoming protocol and host name (for trusted callers). The
4527 host name and its colon can be omitted when only the protocol is to be set.
4528 Note the Exim already has two private options, -pd and -ps, that refer to
4529 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4530 "s", but I don't think that's a major issue.
4532 10. A number of refactoring changes to the code, none of which should affect
4535 (a) The number of logging options was getting close to filling up the
4536 32-bit word that was used as a bit map. I have split them into two classes:
4537 those that are passed in the argument to log_write(), and those that are
4538 only ever tested independently outside of that function. These are now in
4539 separate 32-bit words, so there is plenty of room for expansion again.
4540 There is no change in the user interface or the logging behaviour.
4542 (b) When building, for example, log lines, the code previously used a
4543 macro that called string_cat() twice, in order to add two strings. This is
4544 not really sufficiently general. Furthermore, there was one instance where
4545 it was actually wrong because one of the argument was used twice, and in
4546 one call a function was used. (As it happened, calling the function twice
4547 did not affect the overall behaviour.) The macro has been replaced by a
4548 function that can join an arbitrary number of extra strings onto a growing
4551 (c) The code for expansion conditions now uses a table and a binary chop
4552 instead of a serial search (which was left over from when there were very
4553 few conditions). Also, it now recognizes conditions like "pam" even when
4554 the relevant support is not compiled in: a suitably worded error message is
4555 given if an attempt is made to use such a condition.
4557 11. Added ${time_interval:xxxxx}.
4559 12. A bug was causing one of the ddress fields not to be passed back correctly
4560 from remote delivery subprocesses. The field in question was not being
4561 subsequently used, so this caused to problems in practice.
4563 13. Added new log selectors queue_time and deliver_time.
4565 14. Might have fixed a bug in maildirsizefile handling that threw up
4566 "unexpected character" debug warnings, and recalculated the data
4567 unnecessarily. In any case, I expanded the warning message to give more
4570 15. Added the message "Restricted characters in address" to the statements in
4571 the default ACL that block characters like @ and % in local parts.
4573 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4574 Three changes have been made:
4576 (a) There was a serious bug; a negative response to MAIL caused the whole
4577 recipient domain to be cached as invalid, thereby blocking all messages
4578 to all local parts at the same domain, from all senders. This bug has
4579 been fixed. The domain is no longer cached after a negative response to
4580 MAIL if the sender used is not empty.
4582 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4585 (c) A new callout option, "use_sender" has been added for people who want
4586 the modified behaviour.
4592 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4595 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4596 indeed breaks things for older releases.
4598 3. Added additional logging to the case where there is a problem reading data
4599 from a filter that is running in a subprocess using a pipe, in order to
4600 try to track down a specific problem.
4602 4. Testing facility fudge: when running in the test harness and attempting
4603 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4604 getting "No route to host". Convert this to a timeout.
4606 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4609 6. Some OS don't have socklen_t but use size_t instead. This affects the
4610 fifth argument of getsockopt() amongst other things. This is now
4611 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4612 can be set for individual OS. I have set it for SunOS5, OSF1, and
4613 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4614 some earlier ones do not.
4616 7. Change 4.30/15 was not doing the test caselessly.
4618 8. The standard form for an IPv6 address literal was being rejected by address
4619 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4620 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4621 this, as well as the form without the "IPv6" on the front (but only when
4622 address literals are enabled, of course).
4624 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4626 10. Exim crashed if a message with an empty sender address specified by -f
4627 encountered a router with an errors_to setting. This could be provoked only
4628 by a command such as
4632 where an empty string was supplied; "<>" did not hit this bug.
4634 11. Installed PCRE release 4.5.
4636 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4637 remained set. It is now erased.
4639 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4640 times from message ids (which are base 36 rather than the normal 62).
4642 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4643 were being counted as actual protocol errors, and logged if the log
4644 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4645 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4646 and DATA following a set of rejected RCPTs do not count as protocol errors.
4647 In other words, Exim assumes they were pipelined, though this may not
4648 actually be the case. Of course, in all cases the client gets an
4649 appropriate error code.
4651 15. If a lookup fails in an ACL condition, a message about the failure may
4652 be available; it is used if testing the ACL cannot continue, because most
4653 such messages specify what the cause of the deferral is. However, some
4654 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4655 that caused an old message to be retained and used if a later statement
4656 caused a defer, replacing the real cause of the deferral.
4658 16. If an IP address had so many PTR records that the DNS lookup buffer
4659 was not large enough to hold them, Exim could crash while trying to process
4660 the truncated data. It now detects and logs this case.
4662 17. Further to 4.21/58, another change has been made: if (and only if) the
4663 first line of a message (the first header line) ends with CRLF, a bare LF
4664 in a subsequent header line has a space inserted after it, so as not to
4665 terminate the header.
4667 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4668 unnecessarily, used atoi() instead of strtol(), and didn't check the
4669 termination when getting file sizes from file names by regex.
4671 19. Completely re-implemented the support for maildirsize files, in the light
4672 of a number of problems with the previous contributed implementation
4673 (4.30/29). In particular:
4675 . If the quota is zero, the maildirsize file is maintained, but no quota is
4678 . If the maildir directory does not exist, it is created before any attempt
4679 to write a maildirsize file.
4681 . The quota value in the file is just a cache; if the quota is changed in
4682 the transport, the new value overrides.
4684 . A regular expression is available for excluding directories from the
4687 20. The autoreply transport checks the characters in options that define the
4688 message's headers; it allows continued headers, but it was checking with
4689 isspace() after an embedded newline instead of explicitly looking for a
4692 21. If all the "regular" hosts to which an address was routed had passed their
4693 expiry times, and had not reached their retry times, the address was
4694 bounced, even if fallback hosts were defined. Now Exim should go on to try
4697 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4698 equivalent code in the SMTP transport. Some hosts send humungous responses
4699 to HELO/EHLO, more than 1024 it seems.
4701 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4702 gives compiler warnings in some environments. I've now done it "properly",
4705 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4706 (because of problems with the built-in one) was declared to return uschar *
4707 instead of char *, causing compiler failure.
4709 25. Fixed a file descriptor leak when processing alias/forward files.
4711 26. Fixed a minor format string issue in dbfn.c.
4713 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4715 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4716 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4719 29. When a sender address is verified, it is cached, to save repeating the test
4720 when there is more than one recipient in a message. However, when the
4721 verification involves a callout, it is possible for different callout
4722 options to be set for different recipients. It is too complicated to keep
4723 track of this in the cache, so now Exim always runs a verification when a
4724 callout is required, relying on the callout cache for the optimization.
4725 The overhead is duplication of the address routing, but this should not be
4728 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4729 to be verified with callout=postmaster, and the main callout worked but the
4730 postmaster check failed, the verification correctly failed. However, if a
4731 subsequent RCPT command asked for sender verification *without* the
4732 postmaster check, incorrect caching caused this verification also to fail,
4735 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4736 it was not caching the DNS options (qualify_single, search_parents) that
4737 were used when the lookup failed. A subsequent lookup with different
4738 options therefore always gave the same answer, though there were cases
4739 where it should not have. (Example: a "domains = !$mx_any" option on a
4740 dnslookup router: the "domains" option is always processed without any
4741 widening, but the router might have qualify_single set.) Now Exim uses the
4742 cached value only when the same options are set.
4744 32. Added John Jetmore's "exipick" utility to the distribution.
4746 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4747 than a timeout (e.g. a certificate is required, and is not provided), an
4748 Exim server now closes the connection immediately. Previously it waited for
4749 the client to close - but if the client is SSL, it seems that they each
4750 wait for each other, leading to a delay before one of them times out.
4752 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4753 maintained 0.8.x compatibility because I don't think many are using it, and
4754 it is clearly obsolete.
4756 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4759 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4760 client certificate was expired. A simple patch fixes this, though I don't
4761 understand the full logic of why the verify callback is called multiple
4764 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4765 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4766 which causes problems with some clients (such as the Certicom SSL Plus
4767 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4768 disables the coutermeasure allowing Eudora to connect."
4770 38. Exim was not checking that a write() to a log file succeeded. This could
4771 lead to Bad Things if a log got too big, in particular if it hit a file
4772 size limit. Exim now panics and dies if it cannot write to a log file, just
4773 as it does if it cannot open a log file.
4775 39. Modified OS/Makefile-Linux so that it now contains
4777 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4779 The two -D definitions ensure that Exim is compiled with large file
4780 support, which makes it possible to handle log files that are bigger than
4783 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4784 instance) a domain was checked against a named list that involved a lookup,
4785 causing $domain_data to be set, then another domain was checked against the
4786 same list, then the first domain was re-checked, the value of $domain_data
4787 after the final check could be wrong. In particular, if the second check
4788 failed, it could be set empty. This bug probably also applied to
4791 41. The strip_trailing_dot option was not being applied to the address given
4792 with the -f command-line option.
4794 42. The code for reading a message's header from the spool was incrementing
4795 $received_count, but never initializing it. This meant that the value was
4796 incorrect (doubled) while delivering a message in the same process in which
4797 it was received. In the most common configuration of Exim, this never
4798 happens - a fresh exec is done - but it can happen when
4799 deliver_drop_privilege is set.
4801 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4802 it now includes up to 150 characters of the unexpected data in the log
4805 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4806 and building data strings. The size of both of these buffers was 10 000
4807 bytes - far larger than anybody would *ever* want, thought I. Needless to
4808 say, somebody hit the limit. I have increased the maximum line length to
4809 20 000 and the maximum data length of concatenated lines to 100 000. I have
4810 also fixed two bugs, because there was no checking on these buffers. Tsk,
4811 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4812 buffer is too small.
4814 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4815 lsearch lookups. Now it does.
4817 46. When parsing a route_list item in a manualroute router, a fixed-length
4818 buffer was used for the list of hosts. I made this 1024 bytes long,
4819 thinking that nobody would ever have a list of hosts that long. Wrong.
4820 Somebody had a whole pile of complicated expansion conditions, and the
4821 string was silently truncated, leading to an expansion error. It turns out
4822 that it is easier to change to an unlimited length (owing to other changes
4823 that have happened since this code was originally written) than to build
4824 structure for giving a limitation error. The length of the item that
4825 expands into the list of hosts is now unlimited.
4827 47. The lsearch lookup could not handle data where the length of text line was
4828 more than 4095 characters. Such lines were truncated, leading to shortened
4829 data being returned. It should now handle lines of any length.
4831 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4832 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4835 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4836 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4837 no longer happens in comment lines. A list of replacements is now placed
4838 at the head of all of the source files, except those whose only change is
4839 to replace PERL_COMMAND in the very first #! line.
4841 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4842 messages on the queue, with a bottom-up merge sort, using code contributed
4843 by Michael Haardt. This should make operations like -bp somewhat faster on
4844 large queues. It won't affect queue runners, except when queue_run_in_order
4847 51. Installed eximstats 1.31 in the distribution.
4849 52. Added support for SRV lookups to the dnslookup router.
4851 53. If an ACL referred to $message_body or $message_body_end, the value was not
4852 reset for any messages that followed in the same SMTP session.
4854 54. The store-handling optimization for building very long strings was not
4855 differentiating between the different store pools. I don't think this
4856 actually made any difference in practice, but I've tidied it.
4858 55. While running the routers to verify a sender address, $sender_address
4859 was still set to the sender address. This is wrong, because when routing to
4860 send a bounce to the sender, it would be empty. Therefore, I have changed
4861 it so that, while verifying a sender address, $sender_address is set to <>.
4862 (There is no change to what happens when verifying a recipient address.)
4864 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4865 target A or AAAA records (if not already returned) without resetting the
4866 qualify_single or search_parents options of the DNS resolver. These are
4867 inappropriate in this case because the targets of MX and SRV records must
4868 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4869 target that, when qualified, matched something in the local domain. These
4870 two options are now turned off when doing these lookups.
4872 57. It seems that at least some releases of Reiserfs (which does not have the
4873 concept of a fixed number of inodes) returns zero and not -1 for the
4874 number of available inodes. This interacted badly with check_spool_inodes,
4875 which assumed that -1 was the "no such thing" setting. What I have done is
4876 to check that the total number of inodes is greater than zero before doing
4877 the test of how many are available.
4879 58. When a "warn" ACL statement has a log_message modifier, the message is
4880 remembered, and not repeated. This is to avoid a lot of repetition when a
4881 message has many recipients that cause the same warning to be written.
4882 Howewer, Exim was preserving the list of already written lines for an
4883 entire SMTP session, which doesn't seem right. The memory is now reset if a
4884 new message is started.
4886 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4887 debugging output unless log_rewrite was also set.
4889 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4890 of (int)(handle) when we know that handle contains (void *)(-1).
4892 61. The Exim daemon panic-logs an error return when it closes the incoming
4893 connection. However "connection reset by peer" seems to be common, and
4894 isn't really an error worthy of noting specially, so that particular error
4897 62. When Exim is trying to find all the local interfaces, it used to panic and
4898 die if the ioctl to get the interface flags failed. However, it seems that
4899 on at least one OS (Solaris 9) it is possible to have an interface that is
4900 included in the list of interfaces, but for which you get a failure error
4901 for this call. This happens when the interface is not "plumbed" into a
4902 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4903 failure of the "get flags" call assumes that the interface is down.
4905 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4906 makes life easier for people who are doing arithmetic on fields extracted
4907 from dates, where you often get leading zeros that should not be
4908 interpreted as octal.
4910 64. Added qualify_domain to the redirect router, to override the global
4913 65. If a pathologically long header line contained very many addresses (the
4914 report of this problem mentioned 10 000) and each of them was rewritten,
4915 Exim could use up a very large amount of memory. (It kept on making new
4916 copies of the header line as it rewrote, and never released the old ones.)
4917 At the expense of a bit more processing, the header rewriting function has
4918 been changed so that it no longer eats memory in this way.
4920 66. The generation of the Received: header has been moved from the time that a
4921 message starts to be received, to the time that it finishes. The timestamp
4922 in the Received: header should now be very close to that of the <= log
4923 line. There are two side-effects of this change:
4925 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4926 logged header lines no longer include the local Received: line, because
4927 it has not yet been created. The same applies to a copy of the message
4928 that is returned to a non-SMTP sender when a message is rejected.
4930 (b) When a filter file is tested using -bf, no additional Received: header
4931 is added to the test message. After some thought, I decided that this
4934 This change does not affect the value of $received_for. It is still set
4935 after address rewriting, but before local_scan() is called.
4937 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4939 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4940 gave an unhelpful panic error message, and a defer error. I have managed to
4941 change this behaviour so that it now rejects any supplied certificate,
4942 which seems right, as the list of acceptable certificates is empty.
4944 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4945 gave an unhelpful defer error. I have not managed to make this reject any
4946 supplied certificates, but the error message it gives is "no certificate
4947 supplied", which is not helpful.
4949 70. exigrep's output now also includes lines that are not associated with any
4950 message, but which match the given pattern. Implemented by a patch from
4951 Martin Sluka, which also tidied up the Perl a bit.
4953 71. Recipient callout verification, like sender verification, was using <> in
4954 the MAIL FROM command. This isn't really the right thing, since the actual
4955 sender may affect whether the remote host accepts the recipient or not. I
4956 have changed it to use the actual sender in the callout; this means that
4957 the cache record is now keyed on a recipient/sender pair, not just the
4958 recipient address. There doesn't seem to be a real danger of callout loops,
4959 since a callout by the remote host to check the sender would use <>.
4960 [SEE ABOVE: changed after hitting problems.]
4962 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4963 temporary errors. However, in the case of such a code being given after
4964 the end of a data transmission (i.e. after ".") Exim was failing to write
4965 a retry record for the message. (Yes, there was some broken host that was
4966 actually sending 8xx at this point.)
4968 73. An unknown lookup type in a host list could cause Exim to panic-die when
4969 the list was checked. (An example that provoked this was putting <; in the
4970 middle of a list instead of at the start.) If this happened during a DATA
4971 ACL check, a -D file could be left lying around. This kind of configuration
4972 error no longer causes Exim to die; instead it causes a defer errror. The
4973 incident is still logged to the main and panic logs.
4975 74. Buglet left over from Exim 3 conversion. The message "too many messages
4976 in one connection" was written to the rejectlog but not the mainlog, except
4977 when address rewriting (yes!) was being logged.
4979 75. Added write_rejectlog option.
4981 76. When a system filter was run not as root (that is, when system_filter_user
4982 was set), the values of the $n variables were not being returned to the
4983 main process; thus, they were not subsequently available in the $sn
4986 77. Added +return_path_on_delivery log selector.
4988 78. A connection timeout was being treated differently from recipients deferred
4989 when testing hosts_max_try with a message that was older than the host's
4990 retry timeout. (The host should not be counted, thus allowing all hosts to
4991 be tried at least once before bouncing.) This may have been the cause of an
4992 occasionally reported bug whereby a message would remain on the queue
4993 longer than the retry timeout, but would be bounced if a delivery was
4994 forced. I say "may" because I never totally pinned down the problem;
4995 setting up timeout/retry tests is difficult. See also the next item.
4997 79. The ultimate address timeout was not being applied to errors that involved
4998 a combination of host plus message (for example, a timeout on a MAIL
4999 command). When an address resolved to a number of possible hosts, and they
5000 were not all tried for each delivery (e.g. because of hosts_max_try), a
5001 message could remain on the queue longer than the retry timeout.
5003 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5006 81. Fixed an obscure SMTP outgoing bug which required at least the following
5007 conditions: (a) there was another message waiting for the same server;
5008 (b) the server returned 5xx to all RCPT commands in the first message so
5009 that the message was not completed; (c) the server dropped the connection
5010 or gave a negative response to the RSET that Exim sends to abort the
5011 transaction. The observed case was a dropped connection after DATA that had
5012 been sent in pipelining mode. That is, the server had advertised PIPELINING
5013 but was not implementing it correctly. The effect of the bug was incorrect
5014 behaviour, such as trying another host, and this could lead to a crash.
5020 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5021 and daemon.c were passed as pointers to ints; they should have been
5022 pointers to socklen_t variables (which are typically unsigned ints).
5024 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5027 3. Fixed a really odd bug that affected only the testing scheme; patching a
5028 certain fixed string in the binary changed the value of another string that
5029 happened to be identical to the end of the original first string.
5031 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5032 name", it returns that address as the IP address. On some operating
5033 systems (e.g. Solaris), it also passes back the IP address string as the
5034 "host name". However, on others (e.g. Linux), it passes back an empty
5035 string. Exim wasn't checking for this, and was changing the host name to an
5036 empty string, assuming it had been canonicized.
5038 5. Although rare, it is permitted to have more than one PTR record for a given
5039 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5040 all the names associated with an address, because they do in Solaris.
5041 However, it seems that they do not in Linux for data that comes from the
5042 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5043 I found this out when I moved to a new Linux workstation and tried to run
5044 the Exim test suite.
5046 To get round this problem I have changed the code so that it now does its
5047 own call to the DNS to look up PTR records when searching for a host name.
5048 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5049 addresses that are only in /etc/hosts are still found.
5051 This behaviour is, however, controlled by an option called host_lookup_
5052 order, which defaults to "bydns:byaddr". If people want to use the other
5053 order, or indeed, just use one or the other means of lookup, they can
5054 specify it in this variable.
5056 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5057 some operating systems, this comes back from gethostbyaddr() as an empty
5058 string, and this is what Exim used to test for. However, it seems that in
5059 other systems, "." is yielded. Exim now tests for this case too.
5061 7. The values of check_spool_space and check_log_space are now held internally
5062 as a number of kilobytes instead of an absolute number of bytes. If a
5063 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5064 kilobyte. This means that much larger values can be stored.
5066 8. Exim monitor: an attempt to get the action menu when not actually pointing
5067 at a message produces an empty menu entitled "No message selected". This
5068 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5069 no entries in it ("Shell widget menu has zero width and/or height"). So I
5070 have added a single, blank menu entry in this case.
5072 9. Added ${quote_local_part.
5074 10. MIME decoding is now applied to the contents of Subject: header lines when
5077 11. Now that a reference to $sender_host_address automatically causes a reverse
5078 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5079 host lookup before query-style lookups in lists that might use this
5080 variable. This has therefore been abolished, and the "net-" prefix is no
5081 longer necessary for query-style lookups.
5083 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5084 have been a typo for LFLAGS, so it has been changed.
5086 13. The install script calls Exim with "-C /dev/null" in order to find the
5087 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5088 to be output. Howeve, since Exim outputs its version number before the
5089 error, it didn't break the script. It just looked ugly. I fixed this by
5090 always allowing "-C /dev/null" if the caller is root.
5092 14. Ignore overlarge ACL variable number when reading spool file - insurance
5093 against a later release with more variables having written the file.
5095 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5096 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5097 this, as well as the form without the "IPv6" on the front.
5099 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5100 OS/Makefile-Darwin file.
5102 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5103 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5105 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5106 variable. However, some OS use ints for inodes. I've added cast to long int
5107 to get rid of the compiler warning.
5109 19. I had forgotten to lock out "/../" in configuration file names when
5110 ALT_CONFIG_PREFIX was set.
5112 20. Routers used for verification do not need to specify transports. However,
5113 if such a router generated a host list, and callout was configured, Exim
5114 crashed, because it could not find a port number from the (non-existent)
5115 transport. It now assumes port 25 in this circumstance.
5117 21. Added the -t option to exigrep.
5119 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5120 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5121 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5122 them is set, LOOKUP_LSEARCH is forced.
5124 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5126 24. Added sender and host information to the "rejected by local_scan()" log
5127 line; previously there was no indication of these.
5129 25. Added .include_if_exists.
5131 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5132 Linux. It turns out that not all file systems support this. Apparently some
5133 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5134 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5135 not supported on the file descriptor, is now ignored when Exim is trying to
5136 sync a directory. This applies only to Linux.
5138 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5140 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5141 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5144 29. Added support for maildirsize files from supplied patch (modified a bit).
5146 30. The use of :fail: followed by an empty string could lead Exim to respond to
5147 sender verification failures with (e.g.):
5149 550 Verification failed for <xxx>
5150 550 Sender verify failed
5152 where the first response line was missing the '-' that indicates it is not
5153 the final line of the response.
5155 31. The loop for finding the name of the user that called Exim had a hardwired
5156 limit of 10; it now uses the value of finduser_retries, which is used for
5157 all other user lookups.
5159 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5162 33. Exim was neglecting to zero errno before one call of strtol() when
5163 expanding a string and expecting an integer value. On some systems this
5164 resulted in spurious "integer overflow" errors. Also, it was casting the
5165 result into an int without checking.
5167 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5168 did not work. The code looks as if it has *never* worked, though it appears
5169 to have been documented since at least releast 1.62. I have made it work.
5171 35. The "timeout_DNS" error in retry rules, also documented since at least
5172 1.62, also never worked. As it isn't clear exactly what this means, and
5173 clearly it isn't a major issue, I have abolished the feature by treating it
5174 as "timeout", and writing a warning to the main and panic logs.
5176 36. The display of retry rules for -brt wasn't always showing the error code
5179 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5180 timeout_connect_A, timeout_connect_MX.
5182 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5183 to the empty sender.
5185 39. The daemon was not analysing the content of -oX till after it had closed
5186 stderr and disconnected from the controlling terminal. This meant that any
5187 syntax errors were only noted on the panic log, and the return code from
5188 the command was 0. By re-arranging the code a little, I've made the
5189 decoding happen first, so such errors now appear on stderr, and the return
5190 code is 1. However, the actual setting up of the sockets still happens in
5191 the disconnected process, so errors there are still only recorded on the
5194 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5195 connections (as happens on some IP stacks) was logged at start up time as
5196 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5197 it from "IPv6 and IPv4", which means that two separate sockets are being
5200 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5201 says whether AF_INET or AF_INET6 was passed as an argument.
5203 42. Exiwhat output was messed up when time zones were included in log
5206 43. Exiwhat now gives more information about the daemon's listening ports,
5207 and whether -tls-on-connect was used.
5209 44. The "port" option of the smtp transport is now expanded.
5211 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5212 silently ignored. Now an error message is written to the main and panic
5215 46. There's a new ACL modifier called "logwrite" which writes to a log file
5216 as soon as it is encountered.
5218 47. Added $local_user_uid and $local_user_gid at routing time.
5220 48. Exim crashed when trying to verify a sender address that was being
5223 49. Exim was recognizing only a space character after ".include". It now also
5224 recognizes a tab character.
5226 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5227 extracting the relevant information from the specification. The man page no
5228 longer contains scrambled data for the -d option, and I've added a section
5229 at the front about calling Exim under different names.
5231 51. Added "extra_headers" argument to the "mail" command in filter files.
5233 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5236 53. Installed eximstats 1.29.
5238 54. Added transport_filter_timeout as a generic transport option.
5240 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5241 Cc: header lines. This was required by RFC 822, but it not required by RFC
5244 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5245 incoming messages that did not have them. Now it does so only if the
5246 message originates locally, that is, if there is no associated remote host
5247 address. When Resent- header lines are present, this applies to the Resent-
5248 lines rather than the non-Resent- lines.
5250 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5251 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5253 58. Messages for configuration errors now include the name of the main
5254 configuration file - useful now that there may be more than one file in a
5255 list (.included file names were always shown).
5257 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5258 for those rare installations that do not start the daemon as root or run it
5259 setuid root. I've cut out the call to initgroups() if the daemon is not
5262 60. The Exim user and group can now be bound into the binary as text strings
5263 that are looked up at the start of Exim's processing.
5265 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5267 62. Added $mailstore_basename variable.
5269 63. Installed patch to sieve.c from Michael Haardt.
5271 64. When Exim failed to open the panic log after failing to open the main log,
5272 the original message it was trying to log was written to stderr and debug
5273 output, but if they were not available (the usual case in production), it
5274 was lost. Now it is written to syslog before the two lines that record the
5275 failures to open the logs.
5277 65. Users' Exim filters run in subprocesses under the user's uid. It is
5278 possible for a "deliver" command or an alias in a "personal" command to
5279 provoke an address rewrite. If logging of address rewriting is configured,
5280 this fails because the process is not running as root or exim. There may be
5281 a better way of dealing with this, but for the moment (because 4.30 needs
5282 to be released), I have disabled address rewrite logging when running a
5283 filter in a non-root, non-exim process.
5289 1. The buildconfig auxiliary program wasn't quoting the value set for
5290 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5291 not defined. This bug was masked in 4.22 by the effect that was fixed in
5294 2. Some messages that were rejected after a message id was allocated were
5295 shown as "incomplete" by exigrep. It no longer does this for messages that
5296 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5298 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5299 have allow_domain_literals set, the ID did not get logged in the <= line.
5300 Domain literals are now always recognized in Message-ID: header lines.
5302 4. The first argument for a ${extract expansion item is the key name or field
5303 number. Leading and trailing spaces in this item were not being ignored,
5304 causing some misleading effects.
5306 5. When deliver_drop_privilege was set, single queue runner processes started
5307 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5308 same command in the process it spins off) were not dropping privilege.
5310 6. When the daemon running as "exim" started a queue runner, it always
5311 re-executed Exim in the spun-off process. This is a waste of effort when
5312 deliver_drop_privilege is set. The new process now just calls the
5313 queue-runner function directly.
5319 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5322 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5323 ignored. Though the use of -oP was forcing the writing of a pid file, it
5324 was always written to the default place.
5326 3. If the message "no IP address found for host xxxx" is generated during
5327 incoming verification, it is now followed by identification of the incoming
5328 connection (so you can more easily find what provoked it).
5330 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5332 5. Added some features to "harden" Exim a bit more against certain attacks:
5334 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5335 be put in Local/Makefile. This is like the never_users runtime option,
5336 but it cannot be overridden. The default setting is "root".
5338 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5339 prefix string with which any file named in a -C command line option
5342 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5343 is retained for -C and -D only if the caller of Exim is root. Without
5344 it, the exim user may also use -C and -D and retain privilege.
5346 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5347 command line option is disabled.
5349 6. Macro names set by the -D option must start with an upper case letter, just
5350 like macro names defined in the configuration file.
5352 7. Added "dereference=" facility to LDAP.
5354 8. Two instances of the typo "uknown" in the source files are fixed.
5356 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5357 the Configure-Makefile script screwed up while processing it.
5359 10. Incorporated PCRE 4.4.
5361 11. The SMTP synchronization check was not operating right at the start of an
5362 SMTP session. For example, it could not catch a HELO sent before the client
5363 waited for the greeting. There is now a check for outstanding input at the
5364 point when the greeting is written. Because of the duplex, asynchronous
5365 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5366 way, but not yet received, when the check is performed.
5368 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5369 on TCP/IP sockets, because this apparently causes some broken clients to
5372 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5373 unchanged) from the Cygwin maintainer.
5375 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5376 was supported instead of testing for mbx. Effectively a typo.
5378 15. The spa authenticator server code was not checking that the input it
5379 received was valid base64.
5381 16. The debug output line for the "set" modifier in ACLs was not showing the
5382 name of the variable that was being set.
5384 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5386 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5387 lookup on its own. Something else was needed to trigger the lookup. For
5388 example, a match in host_lookup or the need for a host name in a host list.
5389 Now, if $sender_host_name is referenced and the host name has not yet been
5390 looked up, a lookup is performed. If the lookup fails, the variable remains
5391 empty, and $host_lookup_failed is set to "1".
5393 19. Added "eqi" as a case-independent comparison operator.
5395 20. The saslauthd authentication condition could segfault if neither service
5396 nor realm was specified.
5398 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5399 error message that was logged was misleading, and incoming SMTP
5400 connections were dropped. The message is now more accurate, and temporary
5401 errors are given to SMTP connections.
5403 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5404 (except RSET) with a 421 error, until QUIT is received. However, it was
5405 failing to send a response to QUIT.
5407 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5408 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5409 ACL is now run after the helo_try_verify_hosts code.
5411 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5412 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5413 case-independent, so other case variants are also recognized). Apparently
5414 some systems use these upper case variants.
5416 25. If more than two messages were waiting for the same host, and a transport
5417 filter was specified for the transport, Exim sent two messages over the
5418 same TCP/IP connection, and then failed with "socket operation on non-
5419 socket" when it tried to send the third.
5421 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5423 27. The extern definition of crypt16() in expand.c was not being excluded when
5424 the OS had its own crypt16() function.
5426 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5427 as a preferred synonym for return_size_limit, both as an option and as an
5430 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5432 30. Changed the default configuration ACL to relax the local part checking rule
5433 for addresses that are not in any local domains. For these addresses,
5434 slashes and pipe symbols are allowed within local parts, but the sequence
5435 /../ is explicitly forbidden.
5437 31. SPA server authentication was not clearing the challenge buffer before
5440 32. log_message in a "warn" ACL statement was writing to the reject log as
5441 well as to the main log, which contradicts the documentation and doesn't
5442 seem right (because no rejection is happening). So I have stopped it.
5444 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5445 However, I am unable to do any testing of this.
5447 34. Fixed an infelicity in the appendfile transport. When checking directories
5448 for a mailbox, to see if any needed to be created, it was accidentally
5449 using path names with one or more superfluous leading slashes; tracing
5450 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5452 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5453 contents are added to the log line that is written for every discarded
5454 recipient. (Previously a log_message setting was ignored.)
5456 36. The ${quote: operator now quotes the string if it is empty.
5458 37. The install script runs exim in order to find its version number. If for
5459 some reason other than non-existence or emptiness, which it checks, it
5460 could not run './exim', it was installing it with an empty version number,
5461 i.e. as "exim-". This error state is now caught, and the installation is
5464 38. An argument was missing from the function that creates an error message
5465 when Exim fails to connect to the socket for saslauthd authentication.
5466 This could cause Exim to crash, or give a corrupted message.
5468 39. Added isip, isip4, and isip6 to ${if conditions.
5470 40. The ACL variables $acl_xx are now saved with the message, and can be
5471 accessed later in routers, transports, and filters.
5473 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5474 strings in the file are not string-expanded.
5476 42. If a MAIL command specified a SIZE value that was too large to fit into an
5477 int variable, the check against message_size_limit failed. Such values are
5478 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5479 day this will have to be increased, but I don't think I want to be around
5480 when emails are that large.
5487 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5488 iconv() is not standard in FreeBSD.
5490 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5491 IPv6 enabled. The observed symptom was a segmentation fault on return from
5492 the function os_common_find_running_interfaces() in src/os.c.
5494 3. In the check_special_case() function in daemon.c I had used "errno" as an
5495 argument name, which causes warnings on some systems. This was basically a
5496 typo, since it was named "eno" in the comments!
5498 4. The code that waits for the clock to tick (at a resolution of some fraction
5499 of a second) so as to ensure message-id uniqueness was always waiting for
5500 at least one whole tick, when it could have waited for less. [This is
5501 almost certainly not relevant at current processor speeds, where it is
5502 unlikely to ever wait at all. But we try to future-proof.]
5504 5. The function that sleeps for a time interval that includes fractions of a
5505 second contained a race. It did not block SIGALRM between setting the
5506 timer, and suspending (a couple of lines later). If the interval was short
5507 and the sigsuspend() was delayed until after it had expired, the suspension
5508 never ended. On busy systems this could lead to processes getting stuck for
5511 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5512 process, before it forks any delivery processes. The open lookup caching
5513 mechanism meant that the open file or database connection was passed into
5514 the delivery process. The problem was that delivery processes always tidy
5515 up cached lookup data. This could cause a problem for the next delivery
5516 process started by the queue runner, because the external queue runner
5517 process does not know about the closure. So the next delivery process
5518 still has data in the lookup cache. In the case of a file lookup, there was
5519 no problem because closing a file descriptor in a subprocess doesn't affect
5520 the parent. However, if the lookup was caching a connection to a database,
5521 the connection was closed, and the second delivery process was likely to
5522 see errors such as "PGSQL: query failed: server closed the connection
5523 unexpectedly". The problem has been fixed by closing all cached lookups
5524 in a queue runner before running a delivery process.
5526 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5527 seem to have the "const" qualifier which it has on other OS. I've
5530 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5531 *of the same type* (client or server) with the same public name that an
5532 error should be diagnosed.
5534 9. When Exim looked up a host name for an IP address, but failed to find the
5535 original IP address when looking up the host name (a safety check), it
5536 output the message "<ip address> does not match any IP for NULL", which was
5537 confusing, to say the least. The bug was that the host name should have
5538 appeared instead of "NULL".
5540 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5541 user that is built into the binary, and the -C or -D options is used, root
5542 privilege is dropped before the configuration file is read. In addition,
5543 logging is switched to stderr instead of the normal log files. If the
5544 configuration then re-defines the Exim user, the unprivileged environment
5545 is probably not what is expected, so Exim logs a panic warning message (but
5548 However, if deliver_drop_privilege is set, the unprivileged state may well
5549 be exactly what is intended, so the warning has been cut out in that case,
5550 and Exim is allowed to try to write to its normal log files.
5556 1. smtp_return_error_details was not giving details for temporary sender
5557 or receiver verification errors.
5559 2. Diagnose a configuration error if two authenticators have the same public
5562 3. Exim used not to create the message log file for a message until the first
5563 delivery attempt. This could be confusing when incoming messages were held
5564 for policy or load reasons. The message log file is now created at the time
5565 the message is received, and an initial "Received" line is written to it.
5567 4. The automatically generated man page for command line options had a minor
5568 bug that caused no ill effects; however, a more serious problem was that
5569 the procedure for building the man page automatically didn't always
5570 operate. Consequently, release 4.20 contains an out-of-date version. This
5571 shouldn't happen again.
5573 5. When building Exim with embedded Perl support, the script that builds the
5574 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5575 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5577 6. The freeze_tell option was not being used for messages that were frozen on
5578 arrival, either by an ACL or by local_scan().
5580 7. Added the smtp_incomplete_transaction log selector.
5582 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5583 was accepting AUTH without a new EHLO.
5585 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5586 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5589 10. Logging of TCP/IP connections (when configured) now happens in the main
5590 daemon process instead of the child process, so that the TCP/IP connection
5591 count is more accurate (but it can never be perfect).
5593 11. The use of "drop" in a nested ACL was not being handled correctly in the
5594 outer ACL. Now, if condition failure induced by the nested "drop" causes
5595 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5596 or "require"), the connection is dropped.
5598 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5599 that yield "discard" can now be used with an "accept" or a "discard" verb,
5600 but an error is generated for any others (because I can't see a useful way
5601 to define what should happen).
5603 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5604 are now processed in the same way as lines in the Exim configuration file.
5605 In particular, continuation lines are supported.
5607 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5609 15. Added -ti meaning -t -i.
5611 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5612 domains, and warn by logging if others are found.
5614 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5615 returned from GIFCONF when Exim is trying to find the list of interfaces on
5616 a host. The code in os.c has been modified to copy each ifreq to an aligned
5617 structure in all cases.
5619 Also, in some cases, the returned ifreq's were being copied to a 'struct
5620 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5621 means the last couple of bytes of an IPv6 address could be chopped if the
5622 ifreq contained only a normal sockaddr (14 bytes storage).
5624 18. Named domain lists were not supported in the hosts_treat_as_local option.
5625 An entry such as +xxxx was not recognized, and was treated as a literal
5628 19. Ensure that header lines added by a DATA ACL are included in the reject log
5629 if the ACL subsequently rejects the message.
5631 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5632 MD5 (which is deprecated).
5634 21. When testing a filter file using -bf, Exim was writing a message when it
5635 took the sender from a "From " line in the message, but it was not doing so
5636 when it took $return_path from a Return-Path: header line. It now does.
5638 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5639 with a valid header line field name (a series of printing characters
5640 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5642 23. Changed "disc" in the source to "disk" to conform to the documentation and
5643 the book and for uniformity.
5645 24. Ignore Sendmail's -Ooption=value command line item.
5647 25. When execve() failed while trying to run a command in a pipe transport,
5648 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5649 could be confused with a return value of 69 from the command itself. This
5650 has been changed to 127, the value the shell returns if it is asked to run
5651 a non-existent command. The wording for the related log line suggests a
5652 non-existent command as the problem.
5654 26. If received_header_text expands to an empty string, do not add a Received:
5655 header line to the message. (Well, it adds a token one on the spool, but
5656 marks it "old" so that it doesn't get used or transmitted.)
5658 27. Installed eximstats 1.28 (addition of -nt option).
5660 28. There was no check for failure on the call to getsockname() in the daemon
5661 code. This can fail if there is a shortage of resources on the system, with
5662 ENOMEM, for example. A temporary error is now given on failure.
5664 29. Contrary to the C standard, it seems that in some environments, the
5665 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5666 program. Exim now does this explicitly; it affects the formatting of
5667 timestamps using strftime().
5669 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5670 complaints. I've now initialized all the variables, to avoid this.
5672 32. Header lines added by a system filter were not being "seen" during
5673 transport-time rewrites.
5675 33. The info_callback() function passed to OpenSSL is set up with type void
5676 (*)(SSL *, int, int), as described somewhere. However, when calling the
5677 function (actually a macro) that sets it up, the type void(*)() is
5678 expected. I've put in a cast to prevent warnings from picky compilers.
5680 34. If a DNS black list lookup found a CNAME record, but there were no A
5681 records associated with the domain it pointed at, Exim crashed.
5683 35. If a DNS black list lookup returned more than one A record, Exim ignored
5684 all but the first. It now scans all returned addresses if a particular IP
5685 value is being sought. In this situation, the contents of the
5686 $dnslist_value variable are a list of all the addresses, separated by a
5689 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5690 to accept a host name and all its aliases if the forward lookup for any of
5691 them yielded the IP address of the incoming connection. Now it accepts only
5692 those names whose forward lookup yields the correct IP address. Any other
5693 names are discarded. This closes a loophole whereby a rogue DNS
5694 administrator could create reverse DNS records to break through a
5695 wildcarded host restriction in an ACL.
5697 37. If a user filter or a system filter that ran in a subprocess used any of
5698 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5699 the wrong values were passed to the pipe command ($thisaddress had the
5700 value of $0, $0 had the value of $1, etc). This bug was introduced by
5701 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5703 38. Improved the line breaking for long SMTP error messages from ACLs.
5704 Previously, if there was no break point between 40 and 75 characters, Exim
5705 left the rest of the message alone. Two changes have been made: (a) I've
5706 reduced the minimum length to 35 characters; (b) if it can't find a break
5707 point between 35 and 75 characters, it looks ahead and uses the first one
5708 that it finds. This may give the occasional overlong line, but at least the
5709 remaining text gets split now.
5711 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5712 file descriptors might be low, and that setting 1000 would always raise it.
5713 It turns out that in some environments, the limit is already over 1000 and
5714 that lowering it causes trouble. So now Exim takes care not to decrease it.
5716 40. When delivering a message, the value of $return_path is set to $sender_
5717 address at the start of routing (routers may change the value). By an
5718 oversight, this default was not being set up when an address was tested by
5719 -bt or -bv, which affected the outcome if any router or filter referred to
5722 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5723 log only when "message" or "log_message" are set. However, if one of the
5724 conditions was an address verification, or a call to a nested ACL, the
5725 messages generated by the underlying test were being passed through. This
5726 no longer happens. The underlying message is available in $acl_verify_
5727 message for both "message" and "log_message" expansions, so it can be
5728 passed through if needed.
5730 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5731 new expansion $bh_ to give the encoded byte string without charset
5732 translation. Translation happens only if iconv() is available; HAVE_ICONV
5733 indicates this at build time. HEADERS_CHARSET gives the charset to
5734 translate to; headers_charset can change it in the configuration, and
5735 "headers charset" can change it in an individual filter file.
5737 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5738 that creates RFC 2047 encoded "words" labels them as that charset instead
5739 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5740 expansion operator; (ii) when Exim creates a From: line for a local
5741 message; (iii) when a header line is rewritten to include a "phrase" part.
5743 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5744 buggy, causing it to skip the first lines of messages whose message ID
5745 ended in 'D'. This would not have bitten before Exim release 4.14, because
5746 message IDs were unlikely to end in 'D' before then. The effect was to have
5747 incorrect size information for certain domains.
5749 45. #include "config.h" was missing at the start of the crypt16.c module. This
5750 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5753 46. If there was a timeout during a "random" callout check, Exim treated it as
5754 a failure of the random address, and carried on sending RSET and the real
5755 address. If the delay was just some slowness somewhere, the response to the
5756 original RCPT would be taken as a response to RSET and so on, causing
5757 mayhem of various kinds.
5759 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5760 when I implemented it. It didn't allow for the fact that some option values
5761 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5762 the right test for positive values.
5764 48. Domain names in DNS records are case-independent. Exim always looks them up
5765 in lower case. Some resolvers return domain names in exactly the case they
5766 appear in the zone file, that is, they may contain uppercase letters. Not
5767 all resolvers do this - some return always lower case. Exim was treating a
5768 change of case by a resolver as a change of domain, similar to a widening
5769 of a domain abbreviation. This triggered its re-routing code and so it was
5770 trying to route what was effectively the same domain again. This normally
5771 caused routing to fail (because the router wouldn't handle the domain
5772 twice). Now Exim checks for this case specially, and just changes the
5773 casing of the domain that it ultimately uses when it transmits the message
5776 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5779 50. If a filter generated a file delivery with a non-absolute name (possible if
5780 no home directory exists for the router), the forbid_file option was not
5783 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5784 the existing equality matching.
5786 52. Exim was using ints instead of ino_t variables in some places where it was
5787 dealing with inode numbers.
5789 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5790 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5791 called TMPDIR, and if it finds it is different, it changes its value.
5793 54. The smtp_printf() function is now made available to local_scan() so
5794 additional output lines can be written before returning. There is also an
5795 smtp_fflush() function to enable the detection of a dropped connection.
5796 The variables smtp_input and smtp_batched_input are exported to
5799 55. Changed the default runtime configuration: the message "Unknown user"
5800 has been removed from the ACL, and instead placed on the localuser router,
5801 using the cannot_route_message feature. This means that any verification
5802 failures that generate their own messages won't get overridden. Similarly,
5803 the "Unrouteable address" message that was in the ACL for unverifiable
5804 relay addresses has also been removed.
5806 56. Added hosts_avoid_esmtp to the smtp transport.
5808 57. The exicyclog script was not checking for the esoteric option
5809 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5810 will work only if exicyclog is run under the appropriate euid.
5812 58. Following a discussion on the list, the rules by which Exim recognises line
5813 endings on incoming messages have been changed. The -dropcr and drop_cr
5814 options are now no-ops, retained only for backwards compatibility. The
5815 following line terminators are recognized: LF CRLF CR. However, special
5816 processing applies to CR:
5818 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5819 nor a local message in the state where . is a terminator.
5821 (ii) If a bare CR is encountered in a header line, an extra space is added
5822 after the line terminator so as not to end the header. The reasoning
5823 behind this is that bare CRs in header lines are most likely either
5824 to be mistakes, or people trying to play silly games.
5826 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5827 was being incorrectly given as 18 bytes larger than it should have been.
5828 This is a VOB (very old bug).
5830 60. This may never have affected anything current, but just in case it has:
5831 When the local host is found other than at the start of a list of hosts,
5832 the local host, those with the same MX, and any that follow, are discarded.
5833 When the list in question was part of a longer list of hosts, the following
5834 hosts (not currently being processed) were also being discarded. This no
5835 longer happens. I'm not sure if this situation could ever has previously
5838 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5841 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5842 header. This is something that is recommended in a new Internet Draft, and
5843 is something that is documented as being done by Sendmail. There are two
5844 possible values. For messages generated by the autoreply transport, Exim
5847 Auto-Submitted: auto-replied
5849 whereas for all other generated messages (e.g. bounces) it adds
5851 Auto-Submitted: auto-generated
5853 63. The "personal" condition in filters now includes a test for the
5854 Auto-Submitted: header. If it contains the string "auto-" the message it
5855 not considered personal.
5857 64. Added rcpt_include_affixes as a generic transport option.
5859 65. Added queue_only_override (default true).
5861 66. Added the syslog_duplication option.
5863 67. If what should have been the first header line of a message consisted of
5864 a space followed by a colon, Exim was mis-interpreting it as a header line.
5865 It isn't of course - it is syntactically invalid and should therefore be
5866 treated as the start of the message body. The misbehaviour could have
5867 caused a number of strange effects, including loss of data in subsequent
5868 header lines, and spool format errors.
5870 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5871 client host had authenticated. This control can now be exercised by an ACL
5872 for more flexibility.
5874 69. By default, callouts do not happen when testing with -bh. There is now a
5875 variant, -bhc, which does actually run the callout code, including
5876 consulting and updating the callout cache.
5878 70. Added support for saslauthd authentication, courtesy of Alexander
5881 71. If statvfs() failed on the spool or log directories while checking their
5882 size for availability, Exim confusingly gave the error "space shortage".
5883 Furthermore, in debugging mode it crashed with a floating point exception.
5884 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5885 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5886 serious problem, Exim now writes to the main and panic logs when this
5887 happens, with details of the failure. It then refuses to accept the
5888 incoming message, giving the message "spool directory problem" or "log
5889 directory problem" with a 421 code for SMTP messages.
5891 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5892 0, 1, and 2 exist, because some OS complain for execs without them (see
5893 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5894 descriptors. However, the code omitted to check that the open succeeded,
5895 causing mysterious errors if for some reason the permissions on /dev/null
5896 got screwed. Now Exim writes a message to the main and panic logs, and
5897 bombs out if it can't open /dev/null.
5899 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5900 interact so that it is all more flexible. It is supposed to remain
5901 backwards compatible. Also added extra_local_interfaces.
5903 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5904 to bomb out with an assertion failure - to the client this appears as a
5905 connection drop. This problem occurs in the part of the code that was taken
5906 from the Samba project. Fortunately, the assertion is in a very simple
5907 function, so I have fixed this by reproducing the function inline in the
5908 one place where it is called, and arranging for authentication to fail
5909 instead of killing the process with assert().
5911 75. The SPA client code was not working when the server requested OEM rather
5912 than Unicode encoding.
5914 76. Added code to make require_files with a specific uid setting more usable in
5915 the case where statting the file as root fails - usually a non-root-mounted
5916 NFS file system. When this happens and the failure is EACCES, Exim now
5917 forks a subprocess and does the per-uid checking as the relevant uid.
5919 77. Added process_log_path.
5921 78. If log_file_path was not explicitly set, a setting of check_log_space or
5922 check_log_inodes was ignored.
5924 79. If a space check for the spool or log partitions fails, the incident is now
5925 logged. Of course, in the latter case the data may get lost...
5927 80. Added the %p formatting code to string_format() so that it can be used to
5928 print addresses in debug_print(). Adjusted all the address printing in the
5929 debugging in store.c to use %p rather than %d.
5931 81. There was a concern that a line of code in smtp_in.c could overflow a
5932 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5933 initially expressed, the concern was not well-founded, because trailing
5934 spaces are removed early. However, if the trailing spaces were followed by
5935 a NULL, they did not get removed, so the overflow was possible. Two fixes
5938 (a) I re-wrote the offending code in a cleaner fashion.
5939 (b) If an incoming SMTP command contains a NULL character, it is rejected
5942 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5943 runs initgroups(), so that if the Exim user is in any additional groups,
5944 they will be used during message reception.
5950 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob