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}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
44 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
46 PP/02 Make -n do something, by making it not do something.
47 When combined with -bP, the name of an option is not output.
49 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
52 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
53 $sender_host_name and config options to manage this, and basic check
56 PP/05 DSCP support for outbound connections and control modifier for inbound.
58 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
59 (Only plugin which currently uses this is kerberos4, which nobody should
60 be using, but we should make it available and other future plugins might
61 conceivably use it, even though it would break NAT; stuff *should* be
62 using channel bindings instead).
64 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
65 name; added for Sendmail compatibility; requires admin caller.
66 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
67 just ignore it); requires trusted caller.
68 Also parse but ignore: -Ac -Am -X<logfile>
71 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
73 TL/02 Add +smtp_confirmation as a default logging option.
75 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
76 Patch by Magnus Holmgren from 2007-02-20.
78 TL/04 Bugzilla 1281 - Spec typo.
79 Bugzilla 1283 - Spec typo.
80 Bugzilla 1290 - Spec grammar fixes.
82 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
84 TL/06 Add Experimental DMARC support using libopendmarc libraries.
86 TL/07 Fix an out of order global option causing a segfault. Reported to dev
87 mailing list by by Dmitry Isaikin.
89 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
91 JH/02 Support "G" suffix to numbers in ${if comparisons.
93 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
95 NM/01 Bugzilla 1197 - Spec typo
96 Bugzilla 1196 - Spec examples corrections
98 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
100 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
101 gnutls_enable_pkcs11, but renamed to more accurately indicate its
104 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
105 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
107 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
108 "acl {{name}{arg}...}", and optional args on acl condition
111 JH/05 Permit multiple router/transport headers_add/remove lines.
113 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
115 JH/07 Avoid using a waiting database for a single-message-only transport.
116 Performance patch from Paul Fisher. Bugzilla 1262.
118 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
121 JH/09 Add $headers_added variable, with content from use of ACL modifier
122 add_header (but not yet added to the message). Bugzilla 199.
124 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
125 Pulled from Bugzilla 817 by Wolfgang Breyha.
127 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
129 (nb: this is the same fix as in Exim 4.80.1)
131 JH/11 Add A= logging on delivery lines, and a client_set_id option on
134 JH/12 Add optional authenticated_sender logging to A= and a log_selector
137 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
139 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
140 advertise SMTP AUTH mechanism to us, instead of a generic
141 protocol violation error. Also, make Exim more robust to bad
142 data from the Dovecot auth socket.
144 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
146 When a queue runner is handling a message, Exim first routes the
147 recipient addresses, during which it prunes them based on the retry
148 hints database. After that it attempts to deliver the message to
149 any remaining recipients. It then updates the hints database using
152 So if a recipient address works intermittently, it can get repeatedly
153 deferred at routing time. The retry hints record remains fresh so the
154 address never reaches the final cutoff time.
156 This is a fairly common occurrence when a user is bumping up against
157 their storage quota. Exim had some logic in its local delivery code
158 to deal with this. However it did not apply to per-recipient defers
159 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
161 This change adds a proper retry rule check during routing so that the
162 final cutoff time is checked against the message's age. We only do
163 this check if there is an address retry record and there is not a
164 domain retry record; this implies that previous attempts to handle
165 the address had the retry_use_local_parts option turned on. We use
166 this as an approximation for the destination being like a local
167 delivery, as in LMTP.
169 I suspect this new check makes the old local delivery cutoff check
170 redundant, but I have not verified this so I left the code in place.
172 TF/02 Correct gecos expansion when From: is a prefix of the username.
174 Test 0254 submits a message to Exim with the header
178 When I ran the test suite under the user fanf2, Exim expanded
179 the header to contain my full name, whereas it should have added
180 a Resent-Sender: header. It erroneously treats any prefix of the
181 username as equal to the username.
183 This change corrects that bug.
185 GF/01 DCC debug and logging tidyup
186 Error conditions log to paniclog rather than rejectlog.
187 Debug lines prefixed by "DCC: " to remove any ambiguity.
189 TF/03 Avoid unnecessary rebuilds of lookup-related code.
191 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
192 Bug spotted by Jeremy Harris; was flawed since initial commit.
193 Would have resulted in OCSP responses post-SNI triggering an Exim
194 NULL dereference and crash.
196 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
198 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
199 Bug detection, analysis and fix by Samuel Thibault.
200 Bugzilla 1331, Debian bug #698092.
202 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
204 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
205 Server implementation by Todd Lyons, client by JH.
206 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
207 config variable "prdr_enable" controls whether the server
208 advertises the facility. If the client requests PRDR a new
209 acl_data_smtp_prdr ACL is called once for each recipient, after
210 the body content is received and before the acl_smtp_data ACL.
211 The client is controlled by bolth of: a hosts_try_prdr option
212 on the smtp transport, and the server advertisement.
213 Default client logging of deliveries and rejections involving
214 PRDR are flagged with the string "PRDR".
216 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
217 fclose(). Diagnosis by Todd Lyons.
219 PP/17 Update configure.default to handle IPv6 localhost better.
220 Patch by Alain Williams (plus minor tweaks).
223 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
224 This is now consistent with GnuTLS, and is now documented: the
225 previous undocumented portable approach to treating the option as
226 unset was to force an expansion failure. That still works, and
227 an empty string is now equivalent.
229 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
230 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
231 not performing validation itself.
233 PP/20 Added force_command boolean option to pipe transport.
234 Patch from Nick Koston, of cPanel Inc.
236 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
239 TF/04 Added udpsend ACL modifer and hexquote expansion operator
241 PP/21 Fix eximon continuous updating with timestamped log-files.
242 Broken in a format-string cleanup in 4.80, missed when I repaired the
243 other false fix of the same issue.
244 Report and fix from Heiko Schlichting.
247 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
248 Report from Prashanth Katuri.
250 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
251 It's SecureTransport, so affects any MacOS clients which use the
252 system-integrated TLS libraries, including email clients.
254 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
255 using a MIME ACL for non-SMTP local injection.
256 Report and assistance in diagnosis by Warren Baker.
258 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
260 JH/16 Fix comparisons for 64b. Bugzilla 1385.
262 TL/09 Add expansion variable $authenticated_fail_id to keep track of
263 last id that failed so it may be referenced in subsequent ACL's.
265 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
268 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
269 ldap library initialization, allowing self-signed CA's to be
270 used. Also properly sets require_cert option later in code by
271 using NULL (global ldap config) instead of ldap handle (per
272 session). Bug diagnosis and testing by alxgomz.
274 TL/12 Enhanced documentation in the ratelimit.pl script provided in
275 the src/util/ subdirectory.
277 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
278 renamed to Transport Post Delivery Action by Jeremy Harris, as
281 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
282 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
283 redis_servers = needs to be configured which will be used by the redis
284 lookup. Patch from Warren Baker, of The Packet Hub.
286 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
288 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
289 hostname or reverse DNS when processing a host list. Used suggestions
290 from multiple comments on this bug.
292 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
294 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
295 Missed a few lines, added it to make the runtest require no keyboard
298 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
299 contains upper case chars. Make router use caseful_local_part.
301 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
302 support when GnuTLS has been built with p11-kit.
308 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
310 This, or similar/improved, will also be change PP/11 of 4.82.
316 PP/01 Handle short writes when writing local log-files.
317 In practice, only affects FreeBSD (8 onwards).
318 Bugzilla 1053, with thanks to Dmitry Isaikin.
320 NM/01 Bugzilla 949 - Documentation tweak
322 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
325 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
327 PP/02 Implemented gsasl authenticator.
329 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
331 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
332 `pkg-config foo` for cflags/libs.
334 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
335 with rest of GSASL and with heimdal_gssapi.
337 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
338 `pkg-config foo` for cflags/libs for the TLS implementation.
340 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
341 properties get this fed in as external SSF. A number of robustness
342 and debugging improvements to the cyrus_sasl authenticator.
344 PP/08 cyrus_sasl server now expands the server_realm option.
346 PP/09 Bugzilla 1214 - Log authentication information in reject log.
347 Patch by Jeremy Harris.
349 PP/10 Added dbmjz lookup type.
351 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
353 PP/12 MAIL args handles TAB as well as SP, for better interop with
354 non-compliant senders.
355 Analysis and variant patch by Todd Lyons.
357 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
358 Bug report from Lars Müller <lars@samba.org> (via SUSE),
359 Patch from Dirk Mueller <dmueller@suse.com>
361 PP/13 tls_peerdn now print-escaped for spool files.
362 Observed some $tls_peerdn in wild which contained \n, which resulted
363 in spool file corruption.
365 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
366 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
367 or write after TLS renegotiation, which otherwise led to messages
370 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
371 as a tracking header (ie: a signed header comes before the signature).
372 Patch from Wolfgang Breyha.
374 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
375 comma-sep list; embedded commas doubled.
377 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
379 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
381 Report and patch from Dmitry Banschikov.
383 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
384 Removed SSL_clear() after SSL_new() which led to protocol negotiation
385 failures. We appear to now support TLS1.1+ with Exim.
387 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
388 lets Exim select keys and certificates based upon TLS SNI from client.
389 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
390 before an outbound SMTP session. New log_selector, +tls_sni.
392 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
393 NULL dereference. Report and patch from Alun Jones.
395 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
396 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
397 Not seeing resolver debug output on NetBSD, but suspect this is a
398 resolver implementation change.
400 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
401 Left warnings. Added "eximon gdb" invocation mode.
403 PP/21 Defaulting "accept_8bitmime" to true, not false.
405 PP/22 Added -bw for inetd wait mode support.
407 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
408 locate the relevant includes and libraries. Made this the default.
410 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
411 Bugzilla 1246, report and most of solution from Tomasz Kusy.
413 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
414 This may cause build issues on older platforms.
416 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
417 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
418 gnutls_require_mac & gnutls_require_protocols (no longer supported).
419 Added SNI support via GnuTLS too.
420 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
422 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
424 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
425 Applied second patch from Janne, implementing suggestion to default
426 multiple-strings-in-record handling to match SPF spec.
428 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
430 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
431 read-only, out of scope).
432 Patch from Wolfgang Breyha, report from Stuart Northfield.
434 PP/29 Fix three issues highlighted by clang analyser static analysis.
435 Only crash-plausible issue would require the Cambridge-specific
436 iplookup router and a misconfiguration.
437 Report from Marcin Mirosław.
439 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
441 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
442 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
443 As part of this, removing so much warning spew let me fix some minor
444 real issues in debug logging.
446 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
447 assignment on my part. Fixed.
449 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
450 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
451 Janne Snabb (who went above and beyond: thank you).
453 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
454 string otherwise requires a connection and a bunch more work and it's
455 relatively easy to get wrong. Should also expose TLS library linkage
458 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
459 64-bit ${eval} (JH/03).
461 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
462 GNU libc to support some of the 64-bit stuff, should not lead to
463 conflicts. Defined before os.h is pulled in, so if a given platform
464 needs to override this, it can.
466 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
467 protection layer was required, which is not implemented.
468 Bugzilla 1254, patch from Wolfgang Breyha.
470 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
471 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
472 tls_dhparam take prime identifiers. Also unbreak combination of
473 OpenSSL+DH_params+TLSSNI.
475 PP/39 Disable SSLv2 by default in OpenSSL support.
481 PP/01 Solaris build fix for Oracle's LDAP libraries.
482 Bugzilla 1109, patch from Stephen Usher.
484 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
486 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
489 TF/02 Fix a couple more cases where we did not log the error message
490 when unlink() failed. See also change 4.74-TF/03.
492 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
493 lock up or crash if it happened to be inside a call to libc when it
494 got a SIGUSR1 from exiwhat.
496 The SIGUSR1 handler appends the current process status to the process
497 log which is later printed by exiwhat. It used to use the general
498 purpose logging code to do this, but several functions it calls are
499 not safe for signals.
501 The new output code in the SIGUSR1 handler is specific to the process
502 log, and simple enough that it's easy to inspect for signal safety.
503 Removing some special cases also simplifies the general logging code.
504 Removing the spurious timestamps from the process log simplifies
507 TF/04 Improved ratelimit ACL condition.
509 The /noupdate option has been deprecated in favour of /readonly which
510 has clearer semantics. The /leaky, /strict, and /readonly update modes
511 are mutually exclusive. The update mode is no longer included in the
512 database key; it just determines when the database is updated. (This
513 means that when you upgrde Exim will forget old rate measurements.)
515 Exim now checks that the per_* options are used with an update mode that
516 makes sense for the current ACL. For example, when Exim is processing a
517 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
518 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
519 must specify per_mail/readonly. If you omit the update mode it defaults to
520 /leaky where that makes sense (as before) or /readonly where required.
522 The /noupdate option is now undocumented but still supported for
523 backwards compatibility. It is equivalent to /readonly except that in
524 ACLs where /readonly is required you may specify /leaky/noupdate or
525 /strict/noupdate which are treated the same as /readonly.
527 A useful new feature is the /count= option. This is a generalization
528 of the per_byte option, so that you can measure the throughput of other
529 aggregate values. For example, the per_byte option is now equivalent
530 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
532 The per_rcpt option has been generalized using the /count= mechanism
533 (though it's more complicated than the per_byte equivalence). When it is
534 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
535 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
536 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
537 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
538 Note that using per_rcpt with a non-readonly update mode in more than
539 one ACL will cause the recipients to be double-counted. (The per_mail
540 and per_byte options don't have this problem.)
542 The handling of very low rates has changed slightly. If the computed rate
543 is less than the event's count (usually one) then this event is the first
544 after a long gap. In this case the rate is set to the same as this event's
545 count, so that the first message of a spam run is counted properly.
547 The major new feature is a mechanism for counting the rate of unique
548 events. The new per_addr option counts the number of different
549 recipients that someone has sent messages to in the last time period. It
550 behaves like per_rcpt if all the recipient addresses are different, but
551 duplicate recipient addresses do not increase the measured rate. Like
552 the /count= option this is a general mechanism, so the per_addr option
553 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
554 example, measure the rate that a client uses different sender addresses
555 with the options per_mail/unique=$sender_address. There are further
556 details in the main documentation.
558 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
560 TF/06 Removed a few PCRE remnants.
562 TF/07 Automatically extract Exim's version number from tags in the git
563 repository when doing development or release builds.
565 PP/02 Raise smtp_cmd_buffer_size to 16kB.
566 Bugzilla 879. Patch from Paul Fisher.
568 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
569 Heavily based on revision 40f9a89a from Simon Arlott's tree.
572 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
574 PP/05 Variable $av_failed, true if the AV scanner deferred.
575 Bugzilla 1078. Patch from John Horne.
577 PP/06 Stop make process more reliably on build failure.
578 Bugzilla 1087. Patch from Heiko Schlittermann.
580 PP/07 Make maildir_use_size_file an _expandable_ boolean.
581 Bugzilla 1089. Patch from Heiko Schlittermann.
583 PP/08 Handle ${run} returning more data than OS pipe buffer size.
584 Bugzilla 1131. Patch from Holger Weiß.
586 PP/09 Handle IPv6 addresses with SPF.
587 Bugzilla 860. Patch from Wolfgang Breyha.
589 PP/10 GnuTLS: support TLS 1.2 & 1.1.
591 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
594 PP/11 match_* no longer expand right-hand-side by default.
595 New compile-time build option, EXPAND_LISTMATCH_RHS.
596 New expansion conditions, "inlist", "inlisti".
598 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
600 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
602 PP/14 fix log_write() format string regression from TF/03.
603 Bugzilla 1152. Patch from Dmitry Isaikin.
609 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
611 PP/02 Harmonised TLS library version reporting; only show if debugging.
612 Layout now matches that introduced for other libraries in 4.74 PP/03.
614 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
616 PP/04 New "dns_use_edns0" global option.
618 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
621 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
622 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
624 TK/01 Updated PolarSSL code to 0.14.2.
625 Bugzilla 1097. Patch from Andreas Metzler.
627 PP/07 Catch divide-by-zero in ${eval:...}.
630 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
633 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
634 format-string attack -- SECURITY: remote arbitrary code execution.
636 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
637 time unintentionally subject to list matching rules, letting the header
638 cause arbitrary Exim lookups (of items which can occur in lists, *not*
639 arbitrary string expansion). This allowed for information disclosure.
641 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
642 INT_MIN/-1 -- value coerced to INT_MAX.
648 NM/01 Workround for PCRE version dependency in version reporting
651 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
652 This fixes portability to compilers other than gcc, notably
653 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
655 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
656 makefiles for portability to HP-UX and POSIX correctness.
658 PP/01 Permit LOOKUP_foo enabling on the make command-line.
659 Also via indented variable definition in the Makefile.
660 (Debugging by Oliver Heesakkers).
662 PP/02 Restore caching of spamd results with expanded spamd_address.
663 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
665 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
666 Improves build reliability. Fix from: Frank Elsner
668 NM/02 Fix wide character breakage in the rfc2047 coding
669 Fixes bug 1064. Patch from Andrey N. Oktyabrski
671 NM/03 Allow underscore in dnslist lookups
672 Fixes bug 1026. Patch from Graeme Fowler
674 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
675 Code patches from Adam Ciarcinski of NetBSD.
677 NM/04 Fixed exiqgrep to cope with mailq missing size issue
680 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
681 is logged, to avoid truncation. Patch from John Horne.
683 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
684 Patch from Jakob Hirsch.
686 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
687 SQL string expansion failure details.
688 Patch from Andrey Oktyabrski.
690 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
691 Patch from Simon Arlott.
693 PP/09 New lookups functionality failed to compile on old gcc which rejects
694 extern declarations in function scope.
695 Patch from Oliver Fleischmann
697 PP/10 Use sig_atomic_t for flags set from signal handlers.
698 Check getgroups() return and improve debugging.
699 Fixed developed for diagnosis in bug 927 (which turned out to be
702 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
703 Patch from Mark Zealey.
705 PP/12 Bugzilla 1056: Improved spamd server selection.
706 Patch from Mark Zealey.
708 PP/13 Bugzilla 1086: Deal with maildir quota file races.
709 Based on patch from Heiko Schlittermann.
711 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
712 Patch from Uwe Doering, sign-off by Michael Haardt.
714 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
715 variable declaration deep within a block. Bug and patch from
718 PP/15 lookups-Makefile IRIX compatibilty coercion.
720 PP/16 Make DISABLE_DKIM build knob functional.
722 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
723 Patch by Simon Arlott
725 TF/03 Fix valgrind.h portability to C89 compilers that do not support
726 variable argument macros. Our copy now differs from upstream.
732 TF/01 Failure to get a lock on a hints database can have serious
733 consequences so log it to the panic log.
735 TF/02 Log LMTP confirmation messages in the same way as SMTP,
736 controlled using the smtp_confirmation log selector.
738 TF/03 Include the error message when we fail to unlink a spool file.
740 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
741 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
742 for maintaining out-of-tree patches for some time.
744 PP/01 Bugzilla 139: Documentation and portability issues.
745 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
746 Handle per-OS dynamic-module compilation flags.
748 PP/02 Let /dev/null have normal permissions.
749 The 4.73 fixes were a little too stringent and complained about the
750 permissions on /dev/null. Exempt it from some checks.
751 Reported by Andreas M. Kirchwitz.
753 PP/03 Report version information for many libraries, including
754 Exim version information for dynamically loaded libraries. Created
755 version.h, now support a version extension string for distributors
756 who patch heavily. Dynamic module ABI change.
758 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
759 privilege escalation vulnerability whereby the Exim run-time user
760 can cause root to append content of the attacker's choosing to
763 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
766 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
767 If dropping privileges for untrusted macros, we disabled normal logging
768 on the basis that it would fail; for the Exim run-time user, this is not
769 the case, and it resulted in successful deliveries going unlogged.
770 Fixed. Reported by Andreas Metzler.
776 PP/01 Date: & Message-Id: revert to normally being appended to a message,
777 only prepend for the Resent-* case. Fixes regression introduced in
778 Exim 4.70 by NM/22 for Bugzilla 607.
780 PP/02 Include check_rfc2047_length in configure.default because we're seeing
781 increasing numbers of administrators be bitten by this.
783 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
785 PP/03 Bugzilla 994: added openssl_options main configuration option.
787 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
789 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
791 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
793 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
794 without a peer certificate, leading to a segfault because of an
795 assumption that peers always have certificates. Be a little more
796 paranoid. Problem reported by Martin Tscholak.
798 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
799 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
800 NB: ClamAV planning to remove STREAM in "middle of 2010".
801 CL also introduces -bmalware, various -d+acl logging additions and
802 more caution in buffer sizes.
804 PP/09 Implemented reverse_ip expansion operator.
806 PP/10 Bugzilla 937: provide a "debug" ACL control.
808 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
810 PP/12 Bugzilla 973: Implement --version.
812 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
814 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
816 PP/15 Bugzilla 816: support multiple condition rules on Routers.
818 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
819 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
820 ignore trailing whitespace.
822 JJ/02 prevent non-panic DKIM error from being sent to paniclog
824 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
827 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
828 Notification from Dr Andrew Aitchison.
830 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
831 ExtendedDetectionInfo response format.
832 Notification from John Horne.
834 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
837 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
838 XSL and documented dependency on system catalogs, with examples of how
841 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
844 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
845 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
846 configuration file which is writeable by the Exim user or group.
848 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
849 of configuration files to cover files specified with the -C option if
850 they are going to be used with root privileges, not just the default
853 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
854 option (effectively making it always true).
856 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
857 files to be used while preserving root privileges.
859 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
860 that rogue child processes cannot use them.
862 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
863 run-time user, instead of root.
865 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
866 Exim run-time user without dropping privileges.
868 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
869 result string, instead of calling string_vformat() twice with the same
872 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
873 for other users. Others should always drop root privileges if they use
874 -C on the command line, even for a whitelisted configure file.
876 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
878 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
884 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
885 $data_path, and $header_path variables; fixed documentation bugs and
888 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
889 exipick to access non-standard spools, including the "frozen" queue
892 NM/01 Bugzilla 965: Support mysql stored procedures.
893 Patch from Alain Williams
895 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
897 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
898 Patch from Andreas Metzler
900 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
901 Patch from Kirill Miazine
903 NM/05 Bugzilla 671: Added umask to procmail example.
905 JJ/03 installed exipick 20100323.0, fixing doc bug
907 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
908 directory. Notification and patch from Dan Rosenberg.
910 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
912 TK/02 Improve log output when DKIM signing operation fails.
914 MH/01 Treat the transport option dkim_domain as a colon separated
915 list, not as a single string, and sign the message with each element,
916 omitting multiple occurences of the same signer.
918 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
919 Bugzilla 985, 986. Patch by Simon Arlott
921 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
922 Patch by Simon Arlott
924 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
925 MBX locking. Notification from Dan Rosenberg.
931 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
933 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
935 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
937 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
939 NM/04 Bugzilla 915: Flag broken perl installation during build.
945 TK/01 Added patch by Johannes Berg that expands the main option
946 "spamd_address" if it starts with a dollar sign.
948 TK/02 Write list of recipients to X-Envelope-Sender header when building
949 the mbox-format spool file for content scanning (suggested by Jakob
952 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
953 (http://www.dcc-servers.net/) support via dccifd. Activated by
954 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
956 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
957 by Mark Daniel Reidel <mr@df.eu>.
959 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
960 When building exim an external PCRE library is now needed -
961 PCRE is a system library on the majority of modern systems.
962 See entry on PCRE_LIBS in EDITME file.
964 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
965 conversation. Added nologin parameter to request.
966 Patch contributed by Kirill Miazine.
968 TF/01 Do not log submission mode rewrites if they do not change the address.
970 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
972 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
973 log files in place. Contributed by Roberto Lima.
975 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
977 TF/03 Bugzilla 615: When checking the local_parts router precondition
978 after a local_part_suffix or local_part_prefix option, Exim now
979 does not use the address's named list lookup cache, since this
980 contains cached lookups for the whole local part.
982 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
983 Robert Millan. Documentation is in experimental-spec.txt.
985 TF/04 Bugzilla 668: Fix parallel build (make -j).
987 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
989 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
990 Patch provided by Jan Srzednicki.
992 TF/05 Leading white space used to be stripped from $spam_report which
993 wrecked the formatting. Now it is preserved.
995 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
996 that they are available at delivery time.
998 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1000 TF/08 TLS error reporting now respects the incoming_interface and
1001 incoming_port log selectors.
1003 TF/09 Produce a more useful error message if an SMTP transport's hosts
1004 setting expands to an empty string.
1006 NM/06 Bugzilla 744: EXPN did not work under TLS.
1007 Patch provided by Phil Pennock.
1009 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1010 Patch provided by Richard Godbee.
1012 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1013 acl_smtp_notquit, added index entry.
1015 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1016 Patch provided by Eugene Bujak.
1018 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1019 accept(). Patch provided by Maxim Dounin.
1021 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1022 Patch provided by Phil Pennock.
1024 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1026 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1027 Patch provided by Brad "anomie" Jorsch.
1029 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1030 Patch provided by Dean Brooks.
1032 TK/05 Add native DKIM support (does not depend on external libraries).
1034 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1035 Patch provided by Graeme Fowler.
1037 NM/16 Bugzilla 851: Documentation example syntax fix.
1039 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1041 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1044 NM/19 Bugzilla 745: TLS version reporting.
1045 Patch provided by Phil Pennock.
1047 NM/20 Bugzilla 167: bool: condition support.
1048 Patch provided by Phil Pennock.
1050 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1051 clients. Patch provided by Phil Pennock.
1053 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1054 Patch provided by Brad "anomie" Jorsch.
1056 NM/23 Bugzilla 687: Fix misparses in eximstats.
1057 Patch provided by Heiko Schlittermann.
1059 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1060 Patch provided by Heiko Schlittermann.
1062 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1063 plus update to original patch.
1065 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1067 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1068 Patch provided by David Brownlee.
1070 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1072 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1074 NM/30 Bugzilla 888: TLS documentation bugfixes.
1076 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1078 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1079 Unlike the original bugzilla I have changed all shell scripts in src tree.
1081 NM/33 Bugzilla 898: Transport filter timeout fix.
1082 Patch by Todd Rinaldo.
1084 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1085 Patch by Serge Demonchaux.
1087 NM/35 Bugzilla 39: Base64 decode bug fixes.
1088 Patch by Jakob Hirsch.
1090 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1092 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1094 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1096 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1102 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1103 ALT-N's libdkim that I have put here:
1104 http://duncanthrax.net/exim-experimental/
1106 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1107 were called 'true' and it seems that C99 defines that as a reserved
1108 keyword to be used with 'bool' variable types. That means you could
1109 not include C99-style headers which use bools without triggering
1110 build errors in sieve.c.
1112 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1113 as mailq or other aliases. Changed the --help handling significantly
1114 to do whats expected. exim_usage() emits usage/help information.
1116 SC/01 Added the -bylocaldomain option to eximstats.
1118 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1120 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1122 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1128 PH/01 Another patch from the Sieve maintainer.
1130 PH/02 When an IPv6 address is converted to a string for single-key lookup
1131 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1132 dots are used instead of colons so that keys in lsearch files need not
1133 contain colons. This was done some time before quoting was made available
1134 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1135 (notated using the quote facility) so as to distinguish them from IPv4
1136 keys. This meant that lookups for IP addresses in host lists did not work
1137 for iplsearch lookups.
1139 This has been fixed by arranging for IPv6 addresses to be expressed with
1140 colons if the lookup type is iplsearch. This is not incompatible, because
1141 previously such lookups could never work.
1143 The situation is now rather anomolous, since one *can* have colons in
1144 ordinary lsearch keys. However, making the change in all cases is
1145 incompatible and would probably break a number of configurations.
1147 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1150 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1151 conversion specification without a maximum field width, thereby enabling
1152 a rogue spamd server to cause a buffer overflow. While nobody in their
1153 right mind would setup Exim to query an untrusted spamd server, an
1154 attacker that gains access to a server running spamd could potentially
1155 exploit this vulnerability to run arbitrary code as the Exim user.
1157 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1158 $primary_hostname instead of what libspf2 thinks the hosts name is.
1160 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1161 a directory entry by the name of the lookup key. Previously, if a
1162 symlink pointed to a non-existing file or a file in a directory that
1163 Exim lacked permissions to read, a lookup for a key matching that
1164 symlink would fail. Now it is enough that a matching directory entry
1165 exists, symlink or not. (Bugzilla 503.)
1167 PH/03 The body_linecount and body_zerocount variables are now exported in the
1170 PH/04 Added the $dnslist_matched variable.
1172 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1173 This means they are set thereafter only if the connection becomes
1176 PH/06 Added the client_condition to authenticators so that some can be skipped
1177 by clients under certain conditions.
1179 PH/07 The error message for a badly-placed control=no_multiline_responses left
1180 "_responses" off the end of the name.
1182 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1184 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1185 (without spaces) instead of just copying the configuration text.
1187 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1189 PH/11 Added $max_received_linelength.
1191 PH/12 Added +ignore_defer and +include_defer to host lists.
1193 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1194 way in which PCRE > 7.0 is built.
1196 PH/14 Implemented queue_only_load_latch.
1198 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1199 MAIL command. The effect was to mangle the value on 64-bit systems.
1201 PH/16 Another patch from the Sieve maintainer.
1203 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1205 PH/18 If a system quota error occurred while trying to create the file for
1206 a maildir delivery, the message "Mailbox is full" was not appended to the
1207 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1208 only to a quota excession during the actual writing of the file.
1210 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1211 characters?) which causes problems in log lines. The DN values are now
1212 passed through string_printing() before being added to log lines.
1214 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1215 and InterBase are left for another time.)
1217 PH/21 Added message_body_newlines option.
1219 PH/22 Guard against possible overflow in moan_check_errorcopy().
1221 PH/23 POSIX allows open() to be a macro; guard against that.
1223 PH/24 If the recipient of an error message contained an @ in the local part
1224 (suitably quoted, of course), incorrect values were put in $domain and
1225 $local_part during the evaluation of errors_copy.
1231 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1232 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1235 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1236 issue a MAIL command.
1238 PH/02 In an ACL statement such as
1240 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1242 if a client was not listed at all, or was listed with a value other than
1243 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1244 the condition was not true (as it should be), so access was not denied.
1245 The bug was that the ! inversion was incorrectly passed on to the second
1246 item. This has been fixed.
1248 PH/03 Added additional dnslists conditions == and =& which are different from
1249 = and & when the dns lookup returns more than one IP address.
1251 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1252 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1254 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1255 FSYNC, which compiles an option called disable_fsync that allows for
1256 bypassing fsync(). The documentation is heavily laced with warnings.
1258 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1260 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1261 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1262 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1263 including adding "make clean"; (3) Added -fPIC when compiling the test
1264 dynamically loaded module, to get rid of a warning.
1266 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1267 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1268 = 0s. The bug is otherwise harmless.
1270 PH/07 There was a bug in the dovecot authenticator such that the value of
1271 $auth1 could be overwritten, and so not correctly preserved, after a
1272 successful authentication. This usually meant that the value preserved by
1273 the server_setid option was incorrect.
1275 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1277 PH/09 Installed PCRE release 7.0.
1279 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1280 run for batched SMTP input. It is now run at the start of every message
1281 in the batch. While fixing this I discovered that the process information
1282 (output by running exiwhat) was not always getting set for -bs and -bS
1283 input. This is fixed, and it now also says "batched" for BSMTP.
1285 PH/11 Added control=no_pipelining.
1287 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1288 patch, slightly modified), and move the expansion of helo_data till after
1289 the connection is made in the smtp transport (so it can use these
1292 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1294 PH/14 Added log_selector = +pid.
1296 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1298 PH/16 Add ${if forany and ${if forall.
1300 PH/17 Added dsn_from option to vary the From: line in DSNs.
1302 PH/18 Flush SMTP output before performing a callout, unless control =
1303 no_callout_flush is set.
1305 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1306 was true (the default) a successful delivery failed to delete the retry
1307 item, thus causing premature timeout of the address. The bug is now
1310 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1312 PH/21 Long custom messages for fakedefer and fakereject are now split up
1313 into multiline reponses in the same way that messages for "deny" and
1314 other ACL rejections are.
1316 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1317 with slight modification.
1319 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1320 draft, changing the syntax and factoring some duplicate code".
1322 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1323 for deliveries of the second and subsequent messages over the same SMTP
1326 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1327 ${reduce, with only minor "tidies".
1329 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1331 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1332 expansion side effects.
1334 PH/27 When a message times out after an over-quota error from an Exim-imposed
1335 quota, the bounce message says "mailbox is full". This message was not
1336 being given when it was a system quota that was exceeded. It now should
1339 MH/03 Made $recipients available in local_scan(). local_scan() already has
1340 better access to the recipient list through recipients_list[], but
1341 $recipients can be useful in postmaster-provided expansion strings.
1343 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1344 in the case of a MAIL command with additional options following the
1345 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1346 were accidentally chopped off.
1348 PH/29 SMTP synchronization checks are implemented when a command is read -
1349 there is a check that no more input is waiting when there shouldn't be
1350 any. However, for some commands, a delay in an ACL can mean that it is
1351 some time before the response is written. In this time, more input might
1352 arrive, invalidly. So now there are extra checks after an ACL has run for
1353 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1354 pipelining has not been advertised.
1356 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1358 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1359 returned in the SMTP response when smtp_return_error_details was false.
1360 This has been fixed.
1362 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1363 instead of the C I/O that was originally supplied, because problems were
1364 reported on Solaris.
1366 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1367 Exim which did not show up earlier: it was assuming that a call to
1368 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1369 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1370 was a macro that became an assignment, so it seemed to work. This has
1371 changed to a proper function call with a void return, hence the compile
1372 error. Exim's code has been fixed.
1374 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1377 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1379 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1381 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1382 because it turns out to be too big or there is a timeout) there is no
1383 "Completed" line in the log. When some messages of this type were
1384 selected by exigrep, they were listed as "not completed". Others were
1385 picked up by some special patterns. I have improved the selection
1386 criteria to be more general.
1388 PH/38 The host_find_failed option in the manualroute router can now be set
1389 to "ignore", to completely ignore a host whose IP address cannot be
1390 found. If all hosts are ignored, the behaviour is controlled by the new
1391 host_all_ignored option.
1393 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1394 homing or because of multiple MX records with /mx) generated more than
1395 one IP address, and the following item turned out to be the local host,
1396 all the secondary addresses of the first item were incorrectly removed
1397 from the list, along with the local host and any following hosts (which
1398 is what is supposed to happen).
1400 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1401 whoever called Exim into the -H file. In the case of the daemon it was
1402 behaving confusingly. When first started, it used values for whoever
1403 started the daemon, but after a SIGHUP it used the Exim user (because it
1404 calls itself on a restart). I have changed the code so that it now always
1407 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1408 message are rejected with the same error (e.g. no authentication or bad
1409 sender address), and a DATA command is nevertheless sent (as can happen
1410 with PIPELINING or a stupid MUA), the error message that was given to the
1411 RCPT commands is included in the rejection of the DATA command. This is
1412 intended to be helpful for MUAs that show only the final error to their
1415 PH/42 Another patch from the Sieve maintainer.
1417 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1418 Eximstats - Fixed some broken HTML links and added missing column headers
1420 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1421 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1423 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1425 PH/43 Yet another patch from the Sieve maintainer.
1427 PH/44 I found a way to check for a TCP/IP connection going away before sending
1428 the response to the final '.' that terminates a message, but only in the
1429 case where the client has not sent further data following the '.'
1430 (unfortunately, this is allowed). However, in many cases there won't be
1431 any further data because there won't be any more messages to send. A call
1432 to select() can be used: if it shows that the input is "ready", there is
1433 either input waiting, or the socket has been closed. An attempt to read
1434 the next input character can distinguish the two cases. Previously, Exim
1435 would have sent an OK response which the client would never have see.
1436 This could lead to message repetition. This fix should cure that, at
1437 least in a lot of common cases.
1439 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1440 advertised in response to EHLO.
1446 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1447 fixed by 4.65/MH/01 (is this a record?) are fixed:
1449 (i) An empty string was always treated as zero by the numeric comparison
1450 operators. This behaviour has been restored.
1452 (ii) It is documented that the numeric comparison operators always treat
1453 their arguments as decimal numbers. This was broken in that numbers
1454 starting with 0 were being interpreted as octal.
1456 While fixing these problems I realized that there was another issue that
1457 hadn't been noticed. Values of message_size_limit (both the global option
1458 and the transport option) were treated as octal if they started with 0.
1459 The documentation was vague. These values are now always treated as
1460 decimal, and I will make that clear in the documentation.
1466 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1467 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1470 MH/01 Don't check that the operands of numeric comparison operators are
1471 integers when their expansion is in "skipping" mode (fixes bug
1472 introduced by 4.64-PH/07).
1474 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1475 child addresses, Exim now panics and dies. Previously, because the count
1476 is held in a short int, deliveries were likely to be lost. As such a
1477 large number of recipients for a single message is ridiculous
1478 (performance will be very, very poor), I have chosen to impose a limit
1479 rather than extend the field.
1485 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1486 leftover -K file (the existence of which was triggered by #402).
1487 While we were at it, introduced process PID as part of the -K
1488 filename. This should rule out race conditions when creating
1491 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1492 processing considerably. Previous code took too long for large mails,
1493 triggering a timeout which in turn triggers #401.
1495 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1496 in the DK code in transports.c. sendfile() is not really portable,
1497 hence the _LINUX specificness.
1499 TF/01 In the add_headers option to the mail command in an Exim filter,
1500 there was a bug that Exim would claim a syntax error in any
1501 header after the first one which had an odd number of characters
1504 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1505 callout verification, Exim cached a "reject" for the entire domain. This
1506 is correct for most verifications, but it is not correct for a recipient
1507 verification with use_sender or use_postmaster set, because in that case
1508 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1509 case of MAIL FROM:<> rejection from other early rejections (e.g.
1510 rejection of HELO). When verifying a recipient using a non-null MAIL
1511 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1512 Whatever the result of the callout, the value of the domain cache is
1513 left unchanged (for any other kind of callout, getting as far as trying
1514 RCPT means that the domain itself is ok).
1516 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1519 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1520 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1521 session, the daemon ignores ECONNECT errors and logs others; it now
1522 ignores EPIPE as well.
1524 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1525 (quoted-printable decoding).
1527 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1528 later the small subsequent patch to fix an introduced bug.
1530 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1532 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1534 PH/08 An error is now given if message_size_limit is specified negative.
1536 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1537 to be given (somewhat) arbitrary names.
1539 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1542 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1543 miscellaneous code fixes
1545 PH/10 Added the log_reject_target ACL modifier to specify where to log
1548 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1549 hostname. This is wrong, because it relates to the incoming message (and
1550 probably the interface on which it is arriving) and not to the outgoing
1551 callout (which could be using a different interface). This has been
1552 changed to use the value of the helo_data option from the smtp transport
1553 instead - this is what is used when a message is actually being sent. If
1554 there is no remote transport (possible with a router that sets up host
1555 addresses), $smtp_active_hostname is used.
1557 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1558 tweaks were necessary in order to get it to work (see also 21 below):
1559 (a) The code assumed that strncpy() returns a negative number on buffer
1560 overflow, which isn't the case. Replaced with Exim's string_format()
1562 (b) There were several signed/unsigned issues. I just did the minimum
1563 hacking in of casts. There is scope for a larger refactoring.
1564 (c) The code used strcasecmp() which is not a standard C function.
1565 Replaced with Exim's strcmpic() function.
1566 (d) The code set only $1; it now sets $auth1 as well.
1567 (e) A simple test gave the error "authentication client didn't specify
1568 service in request". It would seem that Dovecot has changed its
1569 interface. Fortunately there's a specification; I followed it and
1570 changed what the client sends and it appears to be working now.
1572 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1575 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1576 address A is aliased to B and C, where B exists and C does not. Without
1577 -v the output is "A verified" because verification stops after a
1578 successful redirection if more than one address is generated. However,
1579 with -v the child addresses are also verified. Exim was outputting "A
1580 failed to verify" and then showing the successful verification for C,
1581 with its parentage. It now outputs "B failed to verify", showing B's
1582 parentage before showing the successful verification of C.
1584 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1585 look up a TXT record in a specific list after matching in a combined
1588 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1589 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1590 they consult the DNS. I had assumed they would set it the way they
1591 wanted; and indeed my experiments on Linux seem to show that in some
1592 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1593 To be on the safe side, however, I have now made the interface to
1594 host_find_byname() similar to host_find_bydns(), with an argument
1595 containing the DNS resolver options. The host_find_byname() function now
1596 sets these options at its start, just as host_find_bydns() does. The smtp
1597 transport options dns_qualify_single and dns_search_parents are passed to
1598 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1599 of host_find_byname() use the default settings of RES_DEFNAMES
1600 (qualify_single) but not RES_DNSRCH (search_parents).
1602 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1603 spool_read_header() do less string testing, by means of a preliminary
1604 switch on the second character of optional "-foo" lines. (This is
1605 overdue, caused by the large number of possibilities that now exist.
1606 Originally there were few.) While I was there, I also converted the
1607 str(n)cmp tests so they don't re-test the leading "-" and the first
1608 character, in the hope this might squeeze out yet more improvement.
1610 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1611 flag allowing group syntax was set by the header_syntax check but not
1612 turned off, possible causing trouble later; (2) The flag was not being
1613 set at all for the header_verify test, causing "group"-style headers to
1614 be rejected. I have now set it in this case, and also caused header_
1615 verify to ignore an empty address taken from a group. While doing this, I
1616 came across some other cases where the code for allowing group syntax
1617 while scanning a header line wasn't quite right (mostly, not resetting
1618 the flag correctly in the right place). These bugs could have caused
1619 trouble for malformed header lines. I hope it is now all correct.
1621 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1622 with the "reply" argument non-NULL. The code, however (which originally
1623 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1624 but it didn't always do it. This confused somebody who was copying the
1625 code for some other use. I have removed all the tests.
1627 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1628 feature that was used to support insecure browsers during the U.S. crypto
1629 embargo. It requires special client support, and Exim is probably the
1630 only MTA that supported it -- and would never use it because real RSA is
1631 always available. This code has been removed, because it had the bad
1632 effect of slowing Exim down by computing (never used) parameters for the
1633 RSA_EXPORT functionality.
1635 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1636 authenticator to fail if there's a tab character in the incoming data
1637 (there should never be unless someone is messing about, as it's supposed
1638 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1639 if the connection is using TLS or if the remote IP is the same as the
1640 local IP, and the "valid-client-cert option" if a client certificate has
1643 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1644 authenticators. This can be used for authorization after authentication
1645 succeeds. (In the case of plaintext, it servers for both authentication
1648 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1649 if any retry times were supplied.
1651 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1652 connection, where there is no client IP address to check. In this
1653 situation, the verify now always succeeds.
1655 PH/25 Applied John Jetmore's -Mset patch.
1657 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1659 PH/27 In a string expansion for a processed (not raw) header when multiple
1660 headers of the same name were present, leading whitespace was being
1661 removed from all of them, but trailing whitespace was being removed only
1662 from the last one. Now trailing whitespace is removed from each header
1663 before concatenation. Completely empty headers in a concatenation (as
1664 before) are ignored.
1666 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1667 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1669 PH/29 [Removed. This was a change that I later backed out, and forgot to
1670 correct the ChangeLog entry (that I had efficiently created) before
1671 committing the later change.]
1673 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1674 address errors (4xx response to RCPT) over the same connection as other
1675 messages routed to the same hosts. Such deliveries are always "forced",
1676 so retry times are not inspected. This resulted in far too many retries
1677 for the affected addresses. The effect occurred only when there were more
1678 hosts than the hosts_max_try setting in the smtp transport when it had
1679 the 4xx errors. Those hosts that it had tried were not added to the list
1680 of hosts for which the message was waiting, so if all were tried, there
1681 was no problem. Two fixes have been applied:
1683 (i) If there are any address or message errors in an SMTP delivery, none
1684 of the hosts (tried or untried) are now added to the list of hosts
1685 for which the message is waiting, so the message should not be a
1686 candidate for sending over the same connection that was used for a
1687 successful delivery of some other message. This seems entirely
1688 reasonable: after all the message is NOT "waiting for some host".
1689 This is so "obvious" that I'm not sure why it wasn't done
1690 previously. Hope I haven't missed anything, but it can't do any
1691 harm, as the worst effect is to miss an optimization.
1693 (ii) If, despite (i), such a delivery is accidentally attempted, the
1694 routing retry time is respected, so at least it doesn't keep
1695 hammering the server.
1697 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1698 in ${readsocket because some servers need this prod.
1700 PH/32 Added some extra debug output when updating a wait-xxx database.
1702 PH/33 The hint "could be header name not terminated by colon", which has been
1703 given for certain expansion errors for a long time, was not being given
1704 for the ${if def:h_colon_omitted{... case.
1706 PH/34 The spec says: "With one important exception, whenever a domain list is
1707 being scanned, $domain contains the subject domain." There was at least
1708 one case where this was not true.
1710 PH/35 The error "getsockname() failed: connection reset by peer" was being
1711 written to the panic log as well as the main log, but it isn't really
1712 panic-worthy as it just means the connection died rather early on. I have
1713 removed the panic log writing for the ECONNRESET error when getsockname()
1716 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1717 runs only) independently of the message's sender address. This meant
1718 that, if the 4xx error was in fact related to the sender, a different
1719 message to the same recipient with a different sender could confuse
1720 things. In particualar, this can happen when sending to a greylisting
1721 server, but other circumstances could also provoke similar problems.
1722 I have changed the default so that the retry time for these errors is now
1723 based a combination of the sender and recipient addresses. This change
1724 can be overridden by setting address_retry_include_sender=false in the
1727 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1728 remote server are returned as part of bounce messages. This was not
1729 happening for LMTP over a pipe (the lmtp transport), but now it is the
1730 same for both kinds of LMTP.
1732 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1733 in header lines that were in fact CNAMEs. This is no longer the case.
1735 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1736 and queue runs started by the daemon processed all messages. This has
1737 been fixed so that -R and -S can now usefully be given with -q<time>.
1739 PH/40 Import PCRE release 6.7 (fixes some bugs).
1741 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1743 PH/42 Give an error if -q is specified more than once.
1745 PH/43 Renamed the variables $interface_address and $interface_port as
1746 $received_ip_address and $received_port, to make it clear that these
1747 values apply to message reception, and not to the outgoing interface when
1748 a message is delivered. (The old names remain recognized, of course.)
1750 PH/44 There was no timeout on the connect() call when using a Unix domain
1751 socket in the ${readsocket expansion. There now is.
1753 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1754 be meaningful with "accept".
1756 SC/01 Eximstats V1.43
1757 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1759 SC/02 Eximstats V1.44
1760 Use a glob alias rather than an array ref in the generated
1761 parser. This improves both readability and performance.
1763 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1764 Collect SpamAssassin and rejection statistics.
1765 Don't display local sender or destination tables unless
1766 there is data to show.
1767 Added average volumes into the top table text output.
1769 SC/04 Eximstats V1.46
1770 Collect data on the number of addresses (recipients)
1771 as well as the number of messages.
1773 SC/05 Eximstats V1.47
1774 Added 'Message too big' to the list of mail rejection
1775 reasons (thanks to Marco Gaiarin).
1777 SC/06 Eximstats V1.48
1778 Mainlog lines which have GMT offsets and are too short to
1779 have a flag are now skipped.
1781 SC/07 Eximstats V1.49 (Alain Williams)
1782 Added the -emptyok flag.
1784 SC/08 Eximstats V1.50
1785 Fixes for obtaining the IP address from reject messages.
1787 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1788 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1789 whitesspace changes from 4.64-PH/27
1791 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1794 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1795 are found, allow negative numbers in numeric criteria)
1797 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1799 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1800 to match changes made in 4.64-PH/43
1802 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1804 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1805 "domain" is actually an IP address, and force a failure. This locks out
1806 those revolvers/nameservers that support "A-for-A" lookups, in
1807 contravention of the specifications.
1809 PH/48 When a host name was looked up from an IP address, and the subsequent
1810 forward lookup of the name timed out, the host name was left in
1811 $sender_host_name, contrary to the specification.
1813 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1814 restricted to single-key lookups, Exim was not diagnosing an error if
1815 * or *@ was used with a query-style lookup.
1817 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1819 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1820 long ago, but noone interested enough thought of it. Let's just say that
1821 the "1.1" means that there are some new functions that weren't there at
1822 some point in the past.
1824 PH/51 Error processing for expansion failure of helo_data from an smtp
1825 transport during callout processing was broken.
1827 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1828 tested/used via the -bh/-bhc/-bs options.
1830 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1831 bug, fixed in subsequent PCRE releases).
1833 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1834 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1836 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1842 SC/01 Use a glob alias rather than an array ref in eximstats generated
1843 parser. This improves both readability and performance.
1845 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1846 Don't display local sender or destination tables in eximstats unless
1847 there is data to show.
1848 Added average volumes into the eximstats top table text output.
1850 SC/03 Collect data on the number of addresses (recipients) as well
1851 as the number of messages in eximstats.
1853 TF/01 Correct an error in the documentation for the redirect router. Exim
1854 does (usually) call initgroups() when daemonizing.
1856 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1857 with consistent privilege compared to when running as a daemon.
1859 TF/03 Note in the spec that $authenticated_id is not set for local
1860 submissions from trusted users.
1862 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1863 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1865 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1866 by adding some example configuration directives to the default
1867 configuration file. A little bit of work is required to uncomment the
1868 directives and define how usernames and passwords are checked, but
1869 there is now a framework to start from.
1871 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1872 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1873 without this. I don't know how relevant this is to other LDAP libraries.
1875 PH/02 Add the verb name to the "unknown ACL verb" error.
1877 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1879 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1881 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1882 directory not expanded when it should be if an expanded home directory
1883 was set for the address (which is overridden by the transport).
1885 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1888 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1889 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1890 because it is too late at that time, and has no effect.
1892 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1893 security issue with \' (bugzilla #107). I could not use the
1894 PQescapeStringConn() function, because it needs a PGconn value as one of
1897 PH/08 When testing addresses using -bt, indicate those final addresses that
1898 are duplicates that would not cause an additional delivery. At least one
1899 person was confused, thinking that -bt output corresponded to deliveries.
1900 (Suppressing duplicates isn't a good idea as you lose the information
1901 about possibly different redirections that led to the duplicates.)
1903 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1904 systems where poll() doesn't work, in particular OS X.
1906 PH/10 Added more information to debugging output for retry time not reached.
1908 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1909 operations in malware.c.
1911 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1914 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1915 syslog_duplication set false, log lines that would normally be written
1916 both the the main log and to the reject log were not written to syslog at
1919 PH/14 In the default configuration, change the use of "message" in ACL warn
1920 statements to "add_header".
1922 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1923 not followed by a command (e.g. "seen endif").
1925 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1926 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1929 PH/17 Added extra conditions to the default value of delay_warning_condition
1933 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1934 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1935 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1938 The Auto-Submitted: and various List- headers are standardised, whereas I
1939 don't think Precedence: ever was.
1941 PH/18 Refactored debugging code in route_finduser() to show more information,
1942 in particular, the error code if getpwnam() issues one.
1944 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1945 This is apparently needed in addition to the PH/07 change above to avoid
1946 any possible encoding problems.
1948 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1949 but not after initializing Perl.
1951 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1952 output them only if debugging. By default they are written stderr,
1953 apparently, which is not desirable.
1955 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1958 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1961 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1963 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1964 authenticated or an ident call has been made. Suppress the default
1965 values for $authenticated_id and $authenticated_sender (but permit -oMai
1966 and -oMas) when testing with -bh.
1968 PH/24 Re-jigged the order of the tests in the default configuration so that the
1969 tests for valid domains and recipients precede the DNS black list and CSA
1970 tests, on the grounds that those ones are more expensive.
1972 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1973 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1974 This bug exists in every version of Exim that I still have, right back to
1977 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1978 However, an attempt to turn on case-sensitivity in a regex key by
1979 including (?-i) didn't work because the subject string was already
1980 lowercased, and the effects were non-intuitive. It turns out that a
1981 one-line patch can be used to allow (?-i) to work as expected.
1987 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1988 other effects) broke the use of negated acl sub-conditions.
1990 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1993 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1994 "Deny" causes Exim to reject the incoming connection with a 554 error.
1995 Unfortunately, if there is a major crisis, such as a disk failure,
1996 tcp-wrappers gives "deny", whereas what one would like would be some
1997 kind of temporary error. A kludge has been added to help with this.
1998 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1999 554 error is used if errno is still zero or contains ENOENT (which occurs
2000 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2003 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2005 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2006 errors. Otherwise a message that provokes a temporary error (when other
2007 messages do not) can cause a whole host to time out.
2009 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2010 addresses were routed directly to files or pipes from a redirect router.
2011 File deliveries just didn't batch; pipe deliveries might have suffered
2014 PH/06 A failure to get a lock for a hints database would erroneously always say
2015 "Failed to get write lock", even when it was really a read lock.
2017 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2018 of 0600. This has been changed to use the value of the lockfile_mode
2019 option (which defaults to 0600).
2021 PH/08 Applied small patch from the Sieve maintainer.
2023 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2024 folder from quota calculations, a direct delivery into this folder messed
2025 up the contents of the maildirsize file. This was because the regex was
2026 used only to exclude .Trash (or whatever) when the size of the mailbox
2027 was calculated. There was no check that a delivery was happening into an
2028 excluded directory. This bug has been fixed by ignoring all quota
2029 processing for deliveries into excluded directories.
2031 PH/10 Added the maildirfolder_create_regex option to appendfile.
2037 PH/01 The code for finding all the local interface addresses on a FreeBSD
2038 system running IPv6 was broken. This may well have applied to all BSD
2039 systems, as well as to others that have similar system calls. The broken
2040 code found IPv4 interfaces correctly, but gave incorrect values for the
2041 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2042 that it would not match correctly against @[] and not recognize the IPv6
2045 PH/02 The ipliteral router was not recognizing addresses of the form user@
2046 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2048 PH/03 Added disable_ipv6.
2050 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2051 decoded form, because it is most often used to construct To: headers
2052 lines in autoreplies, and the decoded form may well be syntactically
2053 invalid. However, $reply_address has leading white space removed, and all
2054 newlines turned into spaces so that the autoreply transport does not
2057 PH/05 If group was specified without a user on a router, and no group or user
2058 was specified on a transport, the group from the router was ignored.
2060 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2061 for visible compile-time settings that can be used to change these
2062 numbers, for those that want even more. Backwards compatibility with old
2063 spool files has been maintained. However, going back to a previous Exim
2064 release will lost any variables that are in spool files.
2066 PH/07 Two small changes when running in the test harness: increase delay when
2067 passing a TCP/IP connection to a new process, in case the original
2068 process has to generate a bounce, and remove special handling of
2069 127.0.0.2 (sic), which is no longer necessary.
2071 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2072 be the same on different OS.
2074 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2077 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2078 whether --show-vars was specified or not
2080 JJ/02 exipick: Added support for new ACL variable spool format introduced
2083 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2084 syntactically invalid From: or Reply-to: line, and a filter used this to
2085 generate an autoreply, and therefore failed to obtain an address for the
2086 autoreply, Exim could try to deliver to a non-existent relative file
2087 name, causing unrelated and misleading errors. What now happens is that
2088 it logs this as a hard delivery error, but does not attempt to create a
2091 PH/11 The exinext utility has a -C option for testing purposes, but although
2092 the given file was scanned by exinext itself; it wasn't being passed on
2093 when Exim was called.
2095 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2096 an end-of-file indication when reading a command response.
2098 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2099 compiled. In many other places in Exim, IPv6 addresses are always
2100 recognized, so I have changed this. It also means that IPv4 domain
2101 literals of the form [IPV4:n.n.n.n] are now always recognized.
2103 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2104 used if the router is not running as root, for example, when verifying at
2105 ACL time, or when using -bh. The debugging output from this situation was
2106 non-existent - all you got was a failure to exec. I have made two
2109 (a) Failures to set uid/gid, the current directory, or a process leader
2110 in a subprocess such as that created by queryprogram now generate
2111 suitable debugging ouput when -d is set.
2113 (b) The queryprogram router detects when it is not running as root,
2114 outputs suitable debugging information if -d is set, and then runs
2115 the subprocess without attempting to change uid/gid.
2117 PH/15 Minor change to Makefile for building test_host (undocumented testing
2120 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2121 additional section of a DNS packet that returns MX or SRV records.
2122 Instead, it always explicitly searches for A/AAAA records. This avoids
2123 major problems that occur when a DNS server includes only records of one
2124 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2125 fixed another bug: if SRV records were looked up and the corresponding
2126 address records were *not* found in the additional section, the port
2127 values from the SRV records were lost.
2129 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2130 using the correct key (the original address) when searching the retry
2131 rules in order to find which one to use for generating the retry hint.
2133 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2134 adding the default one. Similarly, if it contains a Reply-To: header, the
2135 errors_reply_to option, if set, is not used.
2137 PH/19 When calculating a retry time, Exim used to measure the "time since
2138 failure" by looking at the "first failed" field in the retry record. Now
2139 it does not use this if it is later than than the arrival time of the
2140 message. Instead it uses the arrival time. This makes for better
2141 behaviour in cases where some deliveries succeed, thus re-setting the
2142 "first failed" field. An example is a quota failure for a huge message
2143 when small messages continue to be delivered. Without this change, the
2144 "time since failure" will always be short, possible causing more frequent
2145 delivery attempts for the huge message than are intended.
2146 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2148 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2149 $1, $2, $3) because the numerical variables can be reset during some
2150 expansion items (e.g. "match"), thereby losing the authentication data.
2152 PH/21 Make -bV show the size of off_t variables so that the test suite can
2153 decide whether to run tests for quotas > 2G.
2155 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2156 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2157 filecount value is greater than 2G or if a quota value is greater than 2G
2158 on a system where the size of off_t is not greater than 4, a panic error
2161 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2162 never match. The debug and -bh output now contains an explicit error
2163 message indicating a malformed IPv4 address or mask.
2165 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2166 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2167 PH/23 above applies.
2169 PH/25 Do not write to syslog when running in the test harness. The only
2170 occasion when this arises is a failure to open the main or panic logs
2171 (for which there is an explicit test).
2173 PH/26 Added the /no_tell option to "control=freeze".
2175 PH/27 If a host name lookup failed very early in a connection, for example, if
2176 the IP address matched host_lookup and the reverse lookup yielded a name
2177 that did not have a forward lookup, an error message of the form "no IP
2178 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2179 could be logged. Now it outputs the IP address instead of "NULL".
2181 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2182 allows the sender and the authenticated sender to be set when
2183 submitting a message from within Exim. Since child_open_exim() is
2184 documented for local_scan(), the new function should be too.
2186 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2187 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2188 results in an empty string is now treated as unset.
2190 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2192 PH/31 Added sender_verify_fail logging option.
2194 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2195 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2196 tidied the source and removed it altogether.
2198 PH/33 When a queue run was abandoned because the load average was too high, a
2199 log line was always written; now it is written only if the queue_run log
2200 selector is set. In addition, the log line for abandonment now contains
2201 information about the queue run such as the pid. This is always present
2202 in "start" and "stop" lines but was omitted from the "abandon" line.
2204 PH/34 Omit spaces between a header name and the colon in the error message that
2205 is given when verify = headers_syntax fails (if there are lots of them,
2206 the message gets confusing).
2208 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2209 names, as there are now some PTR records that contain slashes. This check
2210 is only to protect against broken name servers that fall over on strange
2211 characters, so the fact that it applies to all lookups doesn't matter.
2213 PH/36 Now that the new test suite is complete, we can remove some of the
2214 special code in Exim that was needed for the old test suite. For example,
2215 sorting DNS records because real resolvers return them in an arbitrary
2216 order. The new test suite's fake resolver always returns records in the
2219 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2220 bounces) except when queue_only is set, to avoid logging races between
2221 the different processes.
2223 PH/38 Panic-die if .include specifies a non-absolute path.
2225 PH/39 A tweak to the "H" retry rule from its user.
2227 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2228 a label. They prevented compilation on older perls.
2230 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2231 a warning to be raised on newish perls.
2233 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2234 on queue. Changes to match documented behaviour of showing count of
2235 messages matching specified criteria.
2237 PH/40 Changed the default ident timeout from 30s to 5s.
2239 PH/41 Added support for the use of login_cap features, on those BSD systems
2240 that have them, for controlling the resources used by pipe deliveries.
2242 PH/42 The content-scanning code uses fopen() to create files in which to put
2243 message data. Previously it was not paying any attention to the mode of
2244 the files. Exim runs with umask(0) because the rest of the code creates
2245 files with open(), and sets the required mode explicitly. Thus, these
2246 files were ending up world-writeable. This was not a big issue, because,
2247 being within the spool directory, they were not world-accessible. I have
2248 created a function called modefopen, which takes an additional mode
2249 argument. It sets umask(777), creates the file, chmods it to the required
2250 mode, then resets the umask. All the relevant calls to fopen() in the
2251 content scanning code have been changed to use this function.
2253 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2254 to 24 hours. This avoids potential overflow problems when processing G
2255 and H retry rules. I suspect nobody ever tinkers with this value.
2257 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2259 PH/45 When the plaintext authenticator is running as a client, the server's
2260 challenges are checked to ensure they are valid base64 strings. By
2261 default, the authentication attempt is cancelled if an invalid string is
2262 received. Setting client_ignore_invalid_base64 true ignores these errors.
2263 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2264 they are received. Thus, the responses can be made to depend on the
2265 challenges. If an invalid string is ignored, an empty string is placed in
2268 PH/46 Messages that are created by the autoreply transport now contains a
2269 References: header, in accordance with RFCs 2822 and 3834.
2271 PH/47 Added authenticated_sender_force to the smtp transport.
2273 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2275 PH/49 Installed latest patch from the Sieve maintainer.
2277 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2278 was also set, the appendfile transport was unnecessarily scanning a
2279 directory of message files (e.g. for maildir delivery) to find the count
2280 of files (along with the size), even though it did not need this
2281 information. It now does the scan only if it needs to find either the
2282 size of the count of files.
2284 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2286 PH/52 Two bugs concerned with error handling when the smtp transport is
2289 (i) Exim was not creating retry information for temporary errors given
2290 for individual recipients after the DATA command when the smtp transport
2291 was used in LMTP mode. This meant that they could be retried too
2292 frequently, and not timed out correctly.
2294 (ii) Exim was setting the flag that allows error details to be returned
2295 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2296 recipients that were returned after the DATA command.
2298 PH/53 This is related to PH/52, but is more general: for any failing address,
2299 when detailed error information was permitted to be returned to the
2300 sender, but the error was temporary, then after the final timeout, only
2301 "retry timeout exceeded" was returned. Now it returns the full error as
2302 well as "retry timeout exceeded".
2304 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2305 do this, and (what is worse) MTAs that accept it.
2307 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2308 will now be deprecated.
2310 PH/56 New os.c-cygwin from the Cygwin maintainer.
2312 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2313 formats (previously only available in exim formats via -bpr, -bpru,
2314 and -bpra. Now also available in native and exiqgrep formats)
2316 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2317 with very large, slow to parse queues
2319 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2321 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2323 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2324 responses to authentication challenges, though it was showing the
2325 challenges; (ii) I've removed the CR characters from the debug output for
2328 PH/58 Allow for the insertion of a newline as well as a space when a string
2329 is turned into more than one encoded-word during RFC 2047 encoding. The
2330 Sieve code now uses this.
2332 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2333 data_4xx, lost_connection, tls_required.
2335 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2336 message was being sent as an SMTP response.
2338 PH/61 Add -l and -k options to exicyclog.
2340 PH/62 When verifying, if an address was redirected to one new address, so that
2341 verification continued, and the new address failed or deferred after
2342 having set something in $address_data, the value of $address_data was not
2343 passed back to the ACL. This was different to the case when no
2344 redirection occurred. The value is now passed back in both cases.
2346 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2347 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2348 use login_cap.h, so on its own it isn't the distinguishing feature. The
2349 new name refers directly to the setclassresources() function.
2351 PH/65 Added configuration files for NetBSD3.
2353 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2355 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2356 is preferred over IPv4.
2358 PH/68 The bounce_return_message and bounce_return_body options were not being
2359 honoured for bounces generated during the reception of non-SMTP messages.
2360 In particular, this applied to messages rejected by the ACL. This bug has
2361 been fixed. However, if bounce_return_message is true and bounce_return_
2362 body is false, the headers that are returned for a non-SMTP message
2363 include only those that have been read before the error was detected.
2364 (In the case of an ACL rejection, they have all been read.)
2366 PH/69 The HTML version of the specification is now built in a directory called
2367 spec_html instead of spec.html, because the latter looks like a path with
2368 a MIME-type, and this confuses some software.
2370 PH/70 Catch two compiler warnings in sieve.c.
2372 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2373 function verify_get_ident() calls ip_connect() to connect a socket, but
2374 if the "connect()" function timed out, ip_connect() used to close the
2375 socket. However, verify_get_ident() also closes the socket later, and in
2376 between Exim writes to the log, which may get opened at this point. When
2377 the socket was closed in ip_connect(), the log could get the same file
2378 descriptor number as the socket. This naturally causes chaos. The fix is
2379 not to close the socket in ip_connect(); the socket should be closed by
2380 the function that creates it. There was only one place in the code where
2381 this was missing, in the iplookup router, which I don't think anybody now
2382 uses, but I've fixed it anyway.
2384 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2385 well as to direct DNS lookups. Otherwise the handling of names in host
2386 lists is inconsistent and therefore confusing.
2392 PH/01 Two changes to the default runtime configuration:
2394 (1) Move the checks for relay_from_hosts and authenticated clients from
2395 after to before the (commented out) DNS black list checks.
2397 (2) Add control=submission to the relay_from_hosts and authenticated
2398 clients checks, on the grounds that messages accepted by these
2399 statements are most likely to be submissions.
2401 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2403 (1) Generate an error if the third argument for the ${prvs expansion is
2406 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2409 (3) Reset the variables that are obtained from the first argument of
2410 ${prvscheck and used in the second argument before leaving the code,
2411 because their memory is reclaimed, so using them afterwards may do
2414 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2415 one (it's much easier than Tom thought :-).
2417 (5) Because of (4), we can now allow for the use of $prvscheck_result
2418 inside the third argument.
2420 PH/03 For some reason, the default setting of PATH when running a command from
2421 a pipe transport was just "/usr/bin". I have changed it to
2424 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2425 anything to be listed in the output from -bV.
2427 PH/05 When a filter generated an autoreply, the entire To: header line was
2428 quoted in the delivery log line, like this:
2430 => >A.N.Other <ano@some.domain> <original@ddress> ...
2432 This has been changed so that it extracts the operative address. There
2433 may be more than one such address. If so, they are comma-separated, like
2436 => >ano@some.domain,ona@other.domain <original@ddress> ...
2438 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2439 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2440 not being looked up in the rDNS to get a host name, Exim was showing the
2441 IP address twice in Received: lines, even though the IP addresses were
2442 identical. For example:
2444 Received: from [1.2.3.4] (helo=[1.2.3.4])
2446 However, if the real host name was known, it was omitting the HELO data
2447 if it matched the actual IP address. This has been tidied up so that it
2448 doesn't show the same IP address twice.
2450 PH/07 When both +timestamp and +memory debugging was on, the value given by
2451 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2452 called by the debug printing, thereby overwriting the timestamp buffer.
2453 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2455 PH/08 When the original message was included in an autoreply transport, it
2456 always said "this is a copy of the message, including all the headers",
2457 even if body_only or headers_only was set. It now gives an appropriate
2460 PH/09 Applied a patch from the Sieve maintainer which:
2462 o fixes some comments
2463 o adds the (disabled) notify extension core
2464 o adds some debug output for the result of if/elsif tests
2465 o points to the current vacation draft in the documentation
2466 and documents the missing references header update
2470 o fixes a bug in processing the envelope test (when testing
2471 multiple envelope elements, the last element determinted the
2474 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2475 Electronic Mail") by including:
2477 Auto-submitted: auto-generated
2479 in the messages that it generates (bounce messages and others, such as
2480 warnings). In the case of bounce messages for non-SMTP mesages, there was
2481 also a typo: it was using "Auto_submitted" (underscore instead of
2482 hyphen). Since every message generated by Exim is necessarily in response
2483 to another message, thes have all been changed to:
2485 Auto-Submitted: auto-replied
2487 in accordance with these statements in the RFC:
2489 The auto-replied keyword:
2491 - SHOULD be used on messages sent in direct response to another
2492 message by an automatic process,
2494 - MUST NOT be used on manually-generated messages,
2496 - MAY be used on Delivery Status Notifications (DSNs) and Message
2497 Disposition Notifications (MDNs),
2499 - MUST NOT be used on messages generated by automatic or periodic
2500 processes, except for messages which are automatic responses to
2503 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2504 to the default Received: header definition.
2506 PH/12 Added log selector acl_warn_skipped (default on).
2508 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2509 variables because (a) they are in the wrong storage pool and (b) even if
2510 they were copied, it wouldn't work properly because of the caching.
2512 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2513 checking when decoding. Apparently there are clients that generate
2514 overlong encoded strings. Why am I not surprised?
2516 PH/15 If the first argument of "${if match_address" was not empty, but did not
2517 contain an "@" character, Exim crashed. Now it writes a panic log message
2518 and treats the condition as false.
2520 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2522 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2523 extension "envelope-auth". The code is finished and in agreement with
2524 other implementations, but there is no documentation so far and in fact,
2525 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2526 not changing the active code.
2528 Print executed "if" and "elsif" statements when debugging is used. This
2529 helps a great deal to understand what a filter does.
2531 Document more things not specified clearly in RFC3028. I had all this
2532 sorted out, when out of a sudden new issues came to my mind. Oops."
2534 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2537 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2538 canonical form (as documented). However, after a host name lookup from
2539 the IP address, check_host() was doing a simple string comparison with
2540 addresses acquired from the DNS when checking that the found name did
2541 have the original IP as one of its addresses. Since any found IPv6
2542 addresses are likely to be in abbreviated form, the comparison could
2543 fail. Luckily, there already exists a function for doing the comparison
2544 by converting both addresses to binary, so now that is used instead of
2545 the text comparison.
2547 PH/20 There was another similar case to PH/19, when a complete host name was
2548 given in a host list; looking up its IP address could give an abbreviated
2549 form, whereas the current host's name might or might not be abbreviated.
2550 The same fix has been applied.
2556 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2557 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2560 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2561 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2563 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2565 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2566 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2567 Also, for "Auto-Submitted", treat anything other than "no" as
2568 non-personal, in accordance with RFC 3834. (Previously it treated
2569 anything starting "auto-" as non-personal.)
2571 TF/01 The control=submission/name=... option had a problem with syntax
2572 errors if the name included a slash character. The /name= option
2573 now slurps the rest of the string, so it can include any characters
2574 but it must come last in the list of options (after /sender_retain
2577 PH/05 Some modifications to the interface to the fake nameserver for the new
2585 TK/01 Added the "success_on_redirect" address verification option. See
2586 NewStuff for rationale and an example.
2588 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2590 PH/02 Patch to exigrep to allow it to work on syslog lines.
2592 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2593 fread() to skip over the body file's header line, because in Cygwin the
2594 header line is locked and is inaccessible.
2596 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2597 co-exist for some time) to make it clear that it is the Exim ID that is
2598 referenced, not the Message-ID: header line.
2600 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2601 string_format() function, because snprintf() does not exist on all
2604 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2605 ${stat: expansion item.
2607 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2608 protocol synchronization error", to keep the pedants happy.
2610 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2611 well as for IRIX systems, when gcc is being used. See the host.c source
2614 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2616 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2619 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2620 transport and to the smtp transport in LMTP mode.
2622 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2624 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2625 run a filter in a subprocess. This could lead to confusion in subsequent
2626 lookups in the parent process. There should also be a search_tidyup() at
2627 the end of the subprocess.
2629 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2630 only if the host matched helo_try_verify_hosts, which caused the
2631 verification to occur when the EHLO/HELO command was issued. The ACL just
2632 tested the remembered result. Now, if a previous verification attempt has
2633 not happened, "verify = helo" does it there and then.
2635 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2637 TK/03 Fix log output including CR from clamd.
2639 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2640 exist provoked a memory error which could cause a segfault.
2642 PH/15 Installed PCRE 6.2
2644 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2646 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2647 of the problem. Specifically, suggested +O2 rather than +O1 for the
2650 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2652 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2653 record, the port was not correctly logged when the outgoing_port log
2654 selector was set (it logged the transort's default port).
2656 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2657 fallback_hosts, and "hosts" in the smtp transport.
2659 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2660 host errors such as "Connection refused".
2662 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2663 authentication with radiusclient 0.4.9:
2665 - Error returned from rc_read_config was caught wrongly
2666 - Username/password not passed on to radius server due to wrong length.
2668 The presumption is that some radiusclient API changes for 4.51/PH/17
2669 were not taken care of correctly. The code is still untested by me (my
2670 Linux distribution still has 0.3.2 of radiusclient), but it was
2671 contributed by a Radius user.
2673 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2674 expanding the "port" option of the smtp transport.
2676 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2677 while reading a MIME header. Thanks to Tom Hughes for a patch.
2679 PH/24 Include config.h inside local_scan.h so that configuration settings are
2682 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2683 that in an ACL for RCPT (for example), you can examine exactly what was
2686 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2687 commands, but it was not correctly comparing the address with the actual
2688 client host address. Thus, it would show the EHLO address in Received:
2689 header lines when this was not necessary.
2691 PH/27 Added the % operator to ${eval:}.
2693 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2694 it should be ignoring failures (because with -C, for example, it has lost
2695 privilege). It wasn't ignoring creation failures other than "already
2698 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2701 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2702 because an input line was too long, either on its own, or by virtue of
2703 too many continuations, the temporary file was not being removed, and the
2704 return code was incorrect.
2706 PH/31 Missing "BOOL" in function definition in filtertest.c.
2708 PH/32 Applied Sieve patches from the maintainer.
2710 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2712 PH/33 Added "verify = not_blind".
2714 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2715 Local/Makefile (with some defaults set). These are used in built scripts
2716 such as exicyclog, but they have never been used in the exim_install
2717 script (though there are many overriding facilities there). I have
2718 arranged that the exim_install script now takes note of these two
2721 PH/35 Installed configuration files for Dragonfly.
2723 PH/36 When a locally submitted message by a trusted user did not contain a
2724 From: header, and the sender address was obtained from -f or from an SMTP
2725 MAIL command, and the trusted user did not use -F to supply a sender
2726 name, $originator_name was incorrectly used when constructing a From:
2727 header. Furthermore, $originator_name was used for submission mode
2728 messages from external hosts without From: headers in a similar way,
2729 which is clearly wrong.
2731 PH/37 Added control=suppress_local_fixups.
2733 PH/38 When log_selector = +received_sender was set, and the addition of the
2734 sender made the log line's construction buffer exactly full, or one byte
2735 less than full, an overflow happened when the terminating "\n" was
2738 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2739 when the result of a list match is failure because a DNS lookup failed.
2741 PH/40 RM_COMMAND is now used in the building process.
2743 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2744 the "build-* directories that it finds.
2746 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2747 address in a domain literal was a prefix of an interface address.
2749 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2750 when verifying a sender address, unless rewrite_headers is false.
2752 PH/44 Wrote a long comment about why errors_to addresses are verified as
2753 recipients, not senders.
2755 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2756 the ratelimit ACL was added.
2758 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2760 PH/46 Added extra information about PostgreSQL errors to the error string.
2762 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2763 avoiding the need to install special zones in a real server. This is
2764 backwards compatible; if it can't find the fake resolver, it drops back.
2765 Thus, both old and new test suites can be run.
2767 TF/02 Added util/ratelimit.pl
2769 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2770 clock is set back in time.
2772 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2773 Candler <B.Candler@pobox.com>.
2775 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2776 OK for addresses that are the result of redirections.
2778 PH/48 A number of further additions for the benefit of the new test suite,
2779 including a fake gethostbyname() that interfaces to the fake DNS resolver
2782 TF/06 The fix for widen_domains has also been applied to qualify_single and
2783 search_parents which are the other dnslookup options that can cause
2786 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2789 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2791 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2792 service" if it encountered an MX record with an empty target hostname.
2793 The message is now "an MX or SRV record indicated no SMTP service".
2795 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2796 if the DNS of the sending site is misconfigured. This is quite a
2797 common situation. This change restores the behaviour of treating a
2798 helo verification defer as a failure.
2800 PH/51 If self=fail was set on a router, the bounce message did not include the
2801 actual error message.
2807 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2809 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2810 command itself ended in error, the underlying message about the transport
2811 filter timeout was being overwritten with the pipe command error. Now the
2812 underlying error message should be appended to the second error message.
2814 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2816 PH/02 Reduce the amount of output that "make" produces by default. Full output
2817 can still be requested.
2819 PH/03 The warning log line about a condition test deferring for a "warn" verb
2820 was being output only once per connection, rather than after each
2821 occurrence (because it was using the same function as for successful
2822 "warn" verbs). This seems wrong, so I have changed it.
2824 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2825 it should not have, which might have caused a crash in the right
2826 circumstances, but probably never did.
2828 PH/04 Installed a modified version of Tony Finch's patch to make submission
2829 mode fix the return path as well as the Sender: header line, and to
2830 add a /name= option so that you can make the user's friendly name appear
2833 TF/03 Added the control = fakedefer ACL modifier.
2835 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2836 Mark Lowes for thorough testing.
2838 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2840 TK/03 Merged latest SRS patch from Miles Wilton.
2842 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2843 with the definition in sysexits.h (which is #included earlier).
2844 Fortunately, Exim does not actually use EX_OK. The code used to try to
2845 preserve the sysexits.h value, by assumimg that macro definitions were
2846 scanned for macro replacements. I have been disabused of this notion,
2847 so now the code just undefines EX_OK before #including unistd.h.
2849 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2850 in the smtp transport. When a block could not be written in a single
2851 write() function, the timeout was being re-applied to each part-write.
2852 This seems wrong - if the receiver was accepting one byte at a time it
2853 would take for ever. The timeout is now adjusted when this happens. It
2854 doesn't have to be particularly precise.
2856 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2857 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2859 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2860 without a domain if the check to <postmaster@domain> fails.
2862 SC/01 Eximstats: added -xls and the ability to specify output files
2863 (patch written by Frank Heydlauf).
2865 SC/02 Eximstats: use FileHandles for outputing results.
2867 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2869 SC/04 Eximstats: fixed display of large numbers with -nvr option
2871 SC/05 Eximstats: fixed merging of reports with empty tables.
2873 SC/06 Eximstats: added the -include_original_destination flag
2875 SC/07 Eximstats: removed tabs and trailing whitespace.
2877 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2879 TK/06 MBOX spool code: Add real "From " MBOX separator line
2880 so the .eml file is really in mbox format (even though
2881 most programs do not really care). Patch from Alex Miller.
2883 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2884 The latter is generated from $received_to and is only set if the
2885 message has one envelope recipient. SA can use these headers,
2886 obviously out-of-the-box. Patch from Alex Miller.
2888 PH/08 The ${def test on a variable was returning false if the variable's
2889 value was "0", contrary to what the specification has always said!
2890 The result should be true unless the variable is empty.
2892 PH/09 The syntax error of a character other than { following "${if
2893 def:variable_name" (after optional whitespace) was not being diagnosed.
2894 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2895 accidental colon was present, for example, could give incorrect results.
2897 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2898 result is used (not including appendfile, where other changes are about
2901 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2902 This involved changing a lot of size variables from int to off_t. It
2903 should work with maildirs and everything.
2905 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2906 spamd dying while we are connected to it.
2908 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2911 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2912 function for BDB 4.3.
2914 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2916 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2917 the sources for PCRE have been re-arranged and more files are now
2920 PH/15 The code I had for printing potentially long long variables in PH/11
2921 above was not the best (it lost precision). The length of off_t variables
2922 is now inspected at build time, and an appropriate printing format (%ld
2923 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2924 to be "long long int" or "long int". This is needed for the internal
2925 formatting function string_vformat().
2927 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2928 the configuration file to be ":syslog", then the script "guesses" where
2929 the logs files are, rather than using the compiled in default. In our
2930 case the guess is not the same as the compiled default, so the script
2931 suddenly stopped working when I started to use syslog. The patch checks
2932 to see if log_file_path is "". If so, it attempts to read it from exim
2933 with no configuration file to get the compiled in version, before it
2934 falls back to the previous guessing code."
2936 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2937 implementing BATV in an Exim configuration. See NewStuff for the gory
2940 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2941 Makefile that are specific to HP-UX.
2943 PH/18 If the "use_postmaster" option was set for a recipient callout together
2944 with the "random" option, the postmaster address was used as the MAIL
2945 FROM address for the random test, but not for the subsequent recipient
2946 test. It is now used for both.
2948 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2949 patch removes a few documentation additions to RFC 3028, because the
2950 latest draft now contains them. It adds the new en;ascii-case comparator
2951 and a new error check for 8bit text in MIME parts. Comparator and
2952 require names are now matched exactly. I enabled the subaddress
2953 extension, but it is not well tested yet (read: it works for me)."
2955 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2956 rework some of the code of TK/09 above to avoid the hardwired use of
2957 "%lld" and "long long". Replaced the call to snprintf() with a call to
2960 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2961 records point to non-existent hosts", "retry timeout exceeded", and
2962 "retry time not reached for any host after a long failure period".
2964 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2965 experimental DomainKeys support:
2967 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2968 (2) On an error such as an illegally used "control", the wrong name for
2969 the control was given.
2971 These problems did NOT occur unless DomainKeys support was compiled.
2973 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2975 PH/24 Added ${if match_ip condition.
2977 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2978 databases so that it will be absolutely obvious if a crash occurs in the
2979 DB library. This is a regular occurrence (often caused by mis-matched
2982 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2983 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2984 picked up on a user's system that detects such things. There doesn't seem
2985 to be a gcc warning option for this - only an attribute that has to be
2986 put on the function's prototype. It seems that in Fedora Core 4 they have
2987 set this on a number of new functions. No doubt there will be more in due
2990 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2991 specify a transport. However, if an address that was verified by such a
2992 router was the subject of a callout, Exim crashed because it tried to
2993 read the rcpt_include_affixes from the non-existent transport. Now it
2994 just assumes that the setting of that option is false. This bug was
2995 introduced by 4.51/PH/31.
2997 PH/28 Changed -d+all to exclude +memory, because that information is very
2998 rarely of interest, but it makes the output a lot bigger. People tend to
2999 do -d+all out of habit.
3001 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3002 code in os-type was giving problems when libc.so lives in lib64, like on
3005 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3006 aren't the modern standard, and it seems that some systems' include files
3007 don't always have them. Exim was already checking for some of the newer
3008 ones like T_AAAA, and defining it itself. I've added checks for all the
3009 record types that Exim uses.
3011 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3012 not automatically generating a new one, as it is supposed to. This
3013 prevented TLS from working. If the file did exist, but contained invalid
3014 data, a new version was generated, as expected. It was only the case of a
3015 non-existent file that was broken.
3017 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3018 with a change in libdomainkeys > 0.64.
3020 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3021 from DNS. If the selector record carries the flag, it now has
3022 precedence over the domain-wide flag.
3024 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3026 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3027 the use of a port name that isn't defined in /etc/services) occurred, the
3028 message was deferred as in a normal delivery, and thus remained on the
3029 spool, instead of being failed because of the mua_wrapper setting. This
3030 is now fixed, and I tidied up some of the mua_wrapper messages at the
3033 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3034 the messages in a hash of arrays rather than using individual hashes.
3035 This is a bit cleaner and results in dramatic memory savings, albeit
3036 at a slight CPU cost.
3038 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3039 as requested by Marc Sherman.
3041 SC/10 Eximstats: added histograms for user specified patterns as requested
3044 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3046 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3047 fopen() in the content-scanning modules that did not already have it.
3053 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3054 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3056 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3058 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3060 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3061 to test Sieve filters that use "vacation".
3063 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3064 that changes the way the GnuTLS parameters are stored in the cache file.
3065 The new format can be generated externally. For backward compatibility,
3066 if the data in the cache doesn't make sense, Exim assumes it has read an
3067 old-format file, and it generates new data and writes a new file. This
3068 means that you can't go back to an older release without removing the
3071 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3072 work if there are any delivery delays because "one_time" forces the
3073 parent to be marked "delivered", so its unseen clone is never tried
3074 again. For this reason, Exim now forbids the simultaneous setting of
3077 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3078 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3079 entry if you want to know the details. The fix, however, neglected to
3080 consider the case when local delivery batching is involved. The test for
3081 "previously delivered" was not happening when checking to see if an
3082 address could be batched with a previous (undelivered) one; under
3083 certain circumstances this could lead to multiple deliveries to the same
3086 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3087 in its include files, and this causes problems building Exim.
3089 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3090 header_syntax) but Exim was just ignoring anything given after a slash.
3091 In particular, this caused confusion with an attempt to use "verify =
3092 reverse_host_lookup/defer_ok". An error is now given when options are
3093 supplied for verify items that do not have them. (Maybe reverse_host_
3094 lookup should have a defer_ok option, but that's a different point.)
3096 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3097 defined by RFC 821) to 2048, because there were problems with some AUTH
3098 commands, and RFC 1869 says the size should be increased for extended
3099 SMTP commands that take arguments.
3101 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3104 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3105 "unknown" error; now it says that the functionality isn't in the binary.
3107 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3108 an address' error message when a string expansion fails (syntax or
3109 whatever). Otherwise the password may appear in the log. Following change
3110 PH/42 below, there is no longer a chance of it appearing in a bounce
3113 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3115 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3116 of its addresses was ever tried. (Bugzilla bug #2.)
3118 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3119 the result incorrectly in the debug output. (It correctly added a newline
3120 to what was transported.)
3122 TF/01 Added $received_time.
3124 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3125 commented out examples of how to interface to a virus scanner and to
3126 SpamAssassin. Also added commented examples of av_scanner and
3127 spamd_address settings.
3129 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3130 and controls are allowed in which ACLs. There were a couple of minor
3131 errors. Some of the entries in the conditions table (which is a table of
3132 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3133 negation of where the condition IS allowed.
3135 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3137 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3138 header file does not have a version number, so I've had to invent a new
3139 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3140 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3141 radiusclient), but it was contributed by a Radius user.
3143 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3144 files or directories, for OpenSSL.
3146 PH/19 When an Exim process that is running as root has to create an Exim log
3147 file, it does so in a subprocess that runs as exim:exim so as to get the
3148 ownership right at creation (otherwise, other Exim processes might see
3149 the file with the wrong ownership). There was no test for failure of this
3150 fork() call, which would lead to the process getting stuck as it waited
3151 for a non-existent subprocess. Forks do occasionally fail when resources
3152 run out. I reviewed all the other calls to fork(); they all seem to check
3155 PH/20 When checking for unexpected SMTP input at connect time (before writing
3156 the banner), Exim was not dealing correctly with a non-positive return
3157 from the read() function. If the client had disconnected by this time,
3158 the result was a log entry for a synchronization error with an empty
3159 string after "input=" when read() returned zero. If read() returned -1
3160 (an event I could not check), uninitialized data bytes were printed.
3161 There were reports of junk text (parts of files, etc) appearing after
3164 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3166 PH/22 Added support for macro redefinition, and (re)definition in between
3167 driver and ACL definitions.
3169 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3170 forgetting to use the resulting value; it was using the unexpanded value.
3172 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3173 hadn't been configured. The fix is from Juergen Kreileder, who
3174 understands it better than I do:
3176 "Here's what I see happening with three configured cyrus_sasl
3177 authenticators configured (plain, login, cram-md5):
3179 On startup auth_cyrus_sasl_init() gets called for each of these.
3180 This means three calls to sasl_listmech() without a specified mech_list.
3181 => SASL tests which mechs of all available mechs actually work
3182 => three warnings about OTP not working
3183 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3185 With the patch, sasl_listmech() also gets called three times. But now
3186 SASL's mech_list option is set to the server_mech specified in the the
3187 authenticator. Or in other words, the answer from sasl_listmech()
3188 gets limited to just the mech you're testing for (which is different
3190 => the return list contains just 'plain' or 'login', 'cram-md5' or
3191 nothing depending on the value of ob->server_mech.
3193 I've just tested the patch: Authentication still works fine,
3194 unavailable mechs specified in the exim configuration are still
3195 caught, and the auth.log warnings about OTP are gone."
3197 PH/25 When debugging is enabled, the contents of the command line are added
3198 to the debugging output, even when log_selector=+arguments is not
3201 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3202 answer is "GNU", and only if the return is "GNU/something" is the answer
3205 PH/27 $acl_verify_message is now set immediately after the failure of a
3206 verification in an ACL, and so is available in subsequent modifiers. In
3207 particular, the message can be preserved by coding like this:
3209 warn !verify = sender
3210 set acl_m0 = $acl_verify_message
3212 Previously, $acl_verify_message was set only while expanding "message"
3213 and "log_message" when a very denied access.
3215 PH/28 Modified OS/os.c-Linux with
3217 -#ifndef OS_LOAD_AVERAGE
3218 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3220 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3221 nomenclature these days.)
3223 PH/29 Installed patch from the Sieve maintainer that adds the options
3224 sieve_useraddress and sieve_subaddress to the redirect router.
3226 PH/30 In these circumstances:
3227 . Two addresses routed to the same list of hosts;
3228 . First host does not offer TLS;
3229 . First host accepts first address;
3230 . First host gives temporary error to second address;
3231 . Second host offers TLS and a TLS session is established;
3232 . Second host accepts second address.
3233 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3234 and peerdn, if requested) that were in fact used only for the second
3237 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3238 attention to any local part prefix or suffix that was matched by the
3239 router that accepted the address. It now behaves in the same way as it
3240 does for delivery: the affixes are removed from the local part unless
3241 rcpt_include_affixes is set on the transport.
3243 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3244 timeout during the DATA phase of an incoming message.
3246 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3247 applied a patch sanctioned by the Sieve maintainer.
3249 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3250 the uid or gid is negative. A case of a negative gid caused this to be
3251 noticed. The fix allows for either to be negative.
3253 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3254 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3257 PH/36 The change PH/12 above was broken. Fixed it.
3259 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3260 the grounds that routing the same address twice would always produce the
3261 same answer. This might have been true once, but it is certainly no
3262 longer true now. Routing a child address may depend on the previous
3263 routing that produced that child. Some complicated redirection strategies
3264 went wrong when messages had multiple recipients, and made Exim's
3265 behaviour dependent on the order in which the addresses were given.
3267 I have moved the duplicate checking until after the routing is complete.
3268 Exim scans the addresses that are assigned to local and remote
3269 transports, and removes any duplicates. This means that more work will be
3270 done, as duplicates will always all be routed, but duplicates are
3271 presumably rare, so I don't expect this is of any significance.
3273 For deliveries to pipes, files, and autoreplies, the duplicate checking
3274 still happens during the routing process, since they are not going to be
3277 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3278 It corrects a timeout issue with spamd. This is Ian's comment: "The
3279 background is that sometimes spamd either never reads data from a
3280 connection it has accepted, or it never writes response data. The exiscan
3281 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3282 blindly assumes that writes won't block so it may never time out."
3284 PH/39 Allow G after quota size as well as K and M.
3286 PH/40 The value set for $authenticated_id in an authenticator may not contain
3287 binary zeroes or newlines because the value is written to log lines and
3288 to spool files. There was no check on this. Now the value is run through
3289 the string_printing() function so that such characters are converted to
3290 printable escape sequences.
3292 PH/41 $message_linecount is a new variable that contains the total number of
3293 lines in the message. Compare $body_linecount, which is the count for the
3296 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3297 bounce and delay warning messages, except in certain special cases, which
3300 (a) An SMTP error message from a remote host;
3301 (b) A message specified in a :fail: redirection;
3302 (c) A message specified in a "fail" command in a system filter;
3303 (d) A message specified in a FAIL return from the queryprogram router;
3304 (e) A message specified by the cannot_route_message router option.
3306 In these cases only, Exim does include the error details in bounce and
3307 warning messages. There are also a few cases where bland messages such
3308 as "unrouteable address" or "local delivery error" are given.
3310 PH/43 $value is now also set for the "else" part of a ${run expansion.
3312 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3313 being worked on, but at least Exim now implements the latest version to
3316 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3317 process to complete was treated as a delivery failure, a timeout while
3318 writing the message to the pipe was logged, but erroneously treated as a
3319 successful delivery. Such timeouts include transport filter timeouts. For
3320 consistency with the overall process timeout, these timeouts are now
3321 treated as errors, giving rise to delivery failures by default. However,
3322 there is now a new Boolean option for the pipe transport called
3323 timeout_defer, which, if set TRUE, converts the failures into defers for
3324 both kinds of timeout. A transport filter timeout is now identified in
3327 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3328 systems where "make" and "gmake" are different, calling "gmake" at top
3329 level broke things. I've arranged for the value of $(MAKE) to be passed
3330 from the Makefile to this script so that it can call the same version of
3334 A note about Exim versions 4.44 and 4.50
3335 ----------------------------------------
3337 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3338 changes of various kinds. As a consequence, a big documentation update was
3339 needed. This delayed the release for rather longer than seemed good, especially
3340 in the light of a couple of (minor) security issues. Therefore, the changes
3341 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3342 release. So 4.44 and 4.50 are in effect two different branches that both start
3345 I have left the 4.50 change log unchanged; it contains all the changes since
3346 4.43. The change log for 4.44 is below; many of its items are identical to
3347 those for 4.50. This seems to be the most sensible way to preserve the
3348 historical information.
3354 1. Minor wording change to the doc/README.SIEVE file.
3356 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3357 computation of the current number of files was incorrect.
3359 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3360 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3363 4. Give more explanation in the error message when the command for a transport
3364 filter fails to execute.
3366 5. There are several places where Exim runs a non-Exim command in a
3367 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3368 was being done only for the command run by the queryprogram router. It is
3369 now done for all such subprocesses. The other cases are: ${run, transport
3370 filters, and the commands run by the lmtp and pipe transports.
3372 6. Added CONFIGURE_GROUP build-time option.
3374 7. Some older OS have a limit of 256 on the maximum number of file
3375 descriptors. Exim was using setrlimit() to set 1000 as a large value
3376 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3377 systems. I've change it so that if it can't get 1000, it tries for 256.
3379 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3380 was an oversight, and furthermore, ever since the addition of extra
3381 controls (e.g. 4.43/32), the checks on when to allow different forms of
3382 "control" were broken. There should now be diagnostics for all cases when a
3383 control that does not make sense is encountered.
3385 9. Added the /retain_sender option to "control=submission".
3387 10. $recipients is now available in the predata ACL (oversight).
3389 11. Tidy the search cache before the fork to do a delivery from a message
3390 received from the command line. Otherwise the child will trigger a lookup
3391 failure and thereby defer the delivery if it tries to use (for example) a
3392 cached ldap connection that the parent has called unbind on.
3394 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3395 of $address_data from the recipient verification was clobbered by the
3396 sender verification.
3398 13. The value of address_data from a sender verification is now available in
3399 $sender_address_data in subsequent conditions in the ACL statement.
3401 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3403 15. Added a new option "connect=<time>" to callout options, to set a different
3406 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3407 was its contents. (It was OK if the option was not defined at all.)
3409 17. A "Completed" log line is now written for messages that are removed from
3410 the spool by the -Mrm option.
3412 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3413 information about exactly what failed.
3415 19. Added -dd to debug only the daemon process.
3417 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3418 handles timeouts, both on the server side and network timeouts. Renamed the
3419 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3421 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3422 It is now set to "smtps".
3424 22. $host_address is now set to the target address during the checking of
3425 ignore_target_hosts.
3427 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3428 being passed; this would have caused $sender_host_name to have been used if
3429 matching the list had actually called for a host name (not very likely,
3430 since this list is usually IP addresses). A host name is now passed as
3433 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3434 code that specifies a non-restarting handler (typically sigaction() in
3435 modern systems) in an attempt to fix a rare and obscure crash bug.
3437 25. Narrowed the window for a race in the daemon that could cause it to ignore
3438 SIGCHLD signals. This is not a major problem, because they are used only to
3439 wake it up if nothing else does.
3441 26. A malformed maildirsize file could cause Exim to calculate negative values
3442 for the mailbox size or file count. Odd effects could occur as a result.
3443 The maildirsize information is now recalculated if the size or filecount
3446 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3447 support for a long time. Removed HAVE_SYS_VFS_H.
3449 28. Installed the latest version of exipick from John Jetmore.
3451 29. In an address list, if the pattern was not a regular expression, an empty
3452 subject address (from a bounce message) matched only if the pattern was an
3453 empty string. Non-empty patterns were not even tested. This was the wrong
3454 because it is perfectly reasonable to use an empty address as part of a
3455 database query. An empty address is now tested by patterns that are
3456 lookups. However, all the other forms of pattern expect the subject to
3457 contain a local part and a domain, and therefore, for them, an empty
3458 address still always fails if the pattern is not itself empty.
3460 30. Exim went into a mad DNS loop when attempting to do a callout where the
3461 host was specified on an smtp transport, and looking it up yielded more
3462 than one IP address.
3464 31. Re-factored the code for checking spool and log partition space into a
3465 function that finds that data and another that does the check. The former
3466 is then used to implement four new variables: $spool_space, $log_space,
3467 $spool_inodes, and $log_inodes.
3469 32. The RFC2047 encoding function was originally intended for short strings
3470 such as real names; it was not keeping to the 75-character limit for
3471 encoded words that the RFC imposes. It now respects the limit, and
3472 generates multiple encoded words if necessary. To be on the safe side, I
3473 have increased the buffer size for the ${rfc2047: expansion operator from
3476 33. It is now permitted to omit both strings after an "if" condition; if the
3477 condition is true, the result is "true". As before, when the second string
3478 is omitted, a false condition yields an empty string. This makes it less
3479 cumbersome to write custom ACL and router conditions.
3481 34. Failure to deliver a bounce message always caused it to be frozen, even if
3482 there was an errors_to setting on the router. The errors_to setting is now
3485 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3486 canonical form (fully expanded) before being placed in
3487 $sender_host_address.
3489 36. The table in the code that translates DNS record types into text (T_A to
3490 "A" for instance) was missing entries for NS and CNAME. It is just possible
3491 that this could have caused confusion if both these types were looked up
3492 for the same domain, because the text type is used as part of Exim's
3493 per-process caching. But the chance of anyone hitting this buglet seems
3496 37. The dnsdb lookup has been extended in a number of ways.
3498 (1) There is a new type, "zns", which walks up the domain tree until it
3499 finds some nameserver records. It should be used with care.
3501 (2) There is a new type, "mxh", which is like "mx" except that it returns
3502 just the host names, not the priorities.
3504 (3) It is now possible to give a list of domains (or IP addresses) to be
3505 looked up. The behaviour when one of the lookups defers can be
3506 controlled by a keyword.
3508 (4) It is now possible to specify the separator character for use when
3509 multiple records are returned.
3511 38. The dnslists ACL condition has been extended: it is now possible to supply
3512 a list of IP addresses and/or domains to be looked up in a particular DNS
3515 39. Added log_selector=+queue_time_overall.
3517 40. When running the queue in the test harness, wait just a tad after forking a
3518 delivery process, to get repeatability of debugging output.
3520 41. Include certificate and key file names in error message when GnuTLS fails
3521 to set them up, because the GnuTLS error message doesn't include the name
3522 of the failing file when there is a problem reading it.
3524 42. Allow both -bf and -bF in the same test run.
3526 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3528 44. The "Exiscan patch" is now merged into the mainline Exim source.
3530 45. Sometimes the final signoff response after QUIT could fail to get
3531 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3532 < 0 before doing a fflush(). This bug looks as though it goes back to the
3533 introduction of TLS in release 3.20, but "sometimes" must have been rare
3534 because the tests only now provoked it.
3536 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3537 (this can affect the format of dates).
3539 47. exim_tidydb, when checking for the continued existence of a message for
3540 which it has found a message-specific retry record, was not finding
3541 messages that were in split spool directories. Consequently, it was
3542 deleting retry records that should have stayed in existence.
3544 48. Steve fixed some bugs in eximstats.
3546 49. The SPA authentication driver was not abandoning authentication and moving
3547 on to the next authenticator when an expansion was forced to fail,
3548 contradicting the general specification for all authenticators. Instead it
3549 was generating a temporary error. It now behaves as specified.
3551 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3552 (the order specifies the preference for clients). The order is now AES256,
3553 AES128, 3DES, ARCFOUR128.
3555 51. Small patch to Sieve code - explicitly set From: when generating an
3558 52. Exim crashed if a remote delivery caused a very long error message to be
3559 recorded - for instance if somebody sent an entire SpamAssassin report back
3560 as a large number of 550 error lines. This bug was coincidentally fixed by
3561 increasing the size of one of Exim's internal buffers (big_buffer) that
3562 happened as part of the Exiscan merge. However, to be on the safe side, I
3563 have made the code more robust (and fixed the comments that describe what
3566 53. Now that there can be additional text after "Completed" in log lines (if
3567 the queue_time_overall log selector is set), a one-byte patch to exigrep
3568 was needed to allow it to recognize "Completed" as not the last thing in
3571 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3572 patch that reportedly fixes this has been added. I am not expert enough to
3573 create a test for it. This is what the patch creator wrote:
3575 "I found a little strange behaviour of ldap code when working with
3576 Windows 2003 AD Domain, where users was placed in more than one
3577 Organization Units. When I tried to give exim partial DN, the exit code
3578 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3579 But simultaneously result of request was absolutely normal ldap result,
3580 so I produce this patch..."
3582 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3583 REFERENCE, so I have modified the code to exclude the patch when that macro
3586 55. Some experimental protocols are using DNS PTR records for new purposes. The
3587 keys for these records are domain names, not reversed IP addresses. The
3588 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3589 leaves it alone. Component reversal etc. now happens only for IP addresses.
3592 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3594 57. Double the size of the debug message buffer (to 2048) so that more of very
3595 long debug lines gets shown.
3597 58. The exicyclog utility now does better if the number of log files to keep
3598 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3600 59. Two changes related to the smtp_active_hostname option:
3602 (1) $smtp_active_hostname is now available as a variable.
3603 (2) The default for smtp_banner uses $smtp_active_hostname instead
3604 of $primary_hostname.
3606 60. The host_aton() function is supposed to be passed a string that is known
3607 to be a valid IP address. However, in the case of IPv6 addresses, it was
3608 not checking this. This is a hostage to fortune. Exim now panics and dies
3609 if the condition is not met. A case was found where this could be provoked
3610 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3611 components; fortuitously, this particular loophole had already been fixed
3612 by change 4.50/55 above.
3614 If there are any other similar loopholes, the new check in host_aton()
3615 itself should stop them being exploited. The report I received stated that
3616 data on the command line could provoke the exploit when Exim was running as
3617 exim, but did not say which command line option was involved. All I could
3618 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3619 running as the user.
3622 61. There was a buffer overflow vulnerability in the SPA authentication code
3623 (which came originally from the Samba project). I have added a test to the
3624 spa_base64_to_bits() function which I hope fixes it.
3627 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3628 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3630 63. The daemon start-up calls getloadavg() while still root for those OS that
3631 need the first call to be done as root, but it missed one case: when
3632 deliver_queue_load_max is set with deliver_drop_privilege. This is
3633 necessary for the benefit of the queue runner, because there is no re-exec
3634 when deliver_drop_privilege is set.
3636 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3637 This has been fixed.
3639 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3640 was in use, was not putting the data itself into the right store pool;
3641 consequently, it could be overwritten for a subsequent message in the same
3642 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3645 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3647 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3648 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3649 the function were treating the return as a boolean value, which happened to
3650 work because 0=false and not-0=true, but is not correct code.
3652 68. The host_aton() function was not handling scoped IPv6 addresses (those
3653 with, for example, "%eth0" on the end) correctly.
3655 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3656 negated items (that is, ~something) in unsigned ints. Some compilers
3657 apparently mutter when there is no cast.
3659 70. If an address verification called from an ACL failed, and did not produce a
3660 user-specific message (i.e. there was only a "system" message), nothing was
3661 put in $acl_verify_message. In this situation, it now puts the system
3664 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3665 session; change 4.31/43 added the unwanted input to the log line - except
3666 that it did not do this in the start of session case. It now does.
3668 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3669 This is wrong and can cause the other end to generate a synchronization
3670 error if it is another Exim or anything else that does the synchronization
3671 check. A QUIT command is no longer sent after a timeout.
3673 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3674 during host lookups.
3676 74. The defer_ok option of callout verification was not working if it was used
3677 when verifying addresses in header lines, that is, for this case:
3679 verify = header_sender/callout=defer_ok
3681 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3682 those file descriptors could be used for SMTP connections. If anything
3683 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3684 could be sent to the SMTP client, causing chaos. The daemon now opens
3685 stdin, stdout, and stderr to /dev/null when it puts itself into the
3688 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3689 log by default. The user can override this with suitable Perl magic.
3691 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3692 the log message when discard triggers, was not working for the DATA ACL or
3693 for the non-SMTP ACL.
3695 78. Error message wording change in sieve.c.
3697 79. If smtp_accept_max_per_host was set, the number of connections could be
3698 restricted to fewer than expected, because the daemon was trying to set up
3699 a new connection before checking whether the processes handling previous
3700 connections had finished. The check for completed processes is now done
3701 earlier. On busy systems, this bug wouldn't be noticed because something
3702 else would have woken the daemon, and it would have reaped the completed
3705 80. If a message was submitted locally by a user whose login name contained one
3706 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3707 It caused a spool format error. I have fixed the spool reading code. A
3708 related problem was that the "from" clause in the Received: line became
3709 illegal because of the space(s). It is now covered by ${quote_local_part.
3711 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3714 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3716 83. Patch from Sieve maintainer to fix unterminated string problem in
3717 "vacation" handling.
3719 84. Some minor changes to the Linux configuration files to help with other
3720 OS variants using glibc.
3722 85. One more patch for Sieve to update vacation handling to latest spec.
3725 ----------------------------------------------------
3726 See the note above about the 4.44 and 4.50 releases.
3727 ----------------------------------------------------
3733 1. Change 4.43/35 introduced a bug that caused file counts to be
3734 incorrectly computed when quota_filecount was set in an appendfile
3737 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3738 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3741 3. Give more explanation in the error message when the command for a transport
3742 filter fails to execute.
3744 4. There are several places where Exim runs a non-Exim command in a
3745 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3746 was being done only for the command run by the queryprogram router. It is
3747 now done for all such subprocesses. The other cases are: ${run, transport
3748 filters, and the commands run by the lmtp and pipe transports.
3750 5. Some older OS have a limit of 256 on the maximum number of file
3751 descriptors. Exim was using setrlimit() to set 1000 as a large value
3752 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3753 systems. I've change it so that if it can't get 1000, it tries for 256.
3755 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3756 was an oversight, and furthermore, ever since the addition of extra
3757 controls (e.g. 4.43/32), the checks on when to allow different forms of
3758 "control" were broken. There should now be diagnostics for all cases when a
3759 control that does not make sense is encountered.
3761 7. $recipients is now available in the predata ACL (oversight).
3763 8. Tidy the search cache before the fork to do a delivery from a message
3764 received from the command line. Otherwise the child will trigger a lookup
3765 failure and thereby defer the delivery if it tries to use (for example) a
3766 cached ldap connection that the parent has called unbind on.
3768 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3769 of $address_data from the recipient verification was clobbered by the
3770 sender verification.
3772 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3773 was its contents. (It was OK if the option was not defined at all.)
3775 11. A "Completed" log line is now written for messages that are removed from
3776 the spool by the -Mrm option.
3778 12. $host_address is now set to the target address during the checking of
3779 ignore_target_hosts.
3781 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3782 being passed; this would have caused $sender_host_name to have been used if
3783 matching the list had actually called for a host name (not very likely,
3784 since this list is usually IP addresses). A host name is now passed as
3787 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3788 code that specifies a non-restarting handler (typically sigaction() in
3789 modern systems) in an attempt to fix a rare and obscure crash bug.
3791 15. Narrowed the window for a race in the daemon that could cause it to ignore
3792 SIGCHLD signals. This is not a major problem, because they are used only to
3793 wake it up if nothing else does.
3795 16. A malformed maildirsize file could cause Exim to calculate negative values
3796 for the mailbox size or file count. Odd effects could occur as a result.
3797 The maildirsize information is now recalculated if the size or filecount
3800 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3801 support for a long time. Removed HAVE_SYS_VFS_H.
3803 18. Updated exipick to current release from John Jetmore.
3805 19. Allow an empty sender to be matched against a lookup in an address list.
3806 Previously the only cases considered were a regular expression, or an
3809 20. Exim went into a mad DNS lookup loop when doing a callout where the
3810 host was specified on the transport, if the DNS lookup yielded more than
3813 21. The RFC2047 encoding function was originally intended for short strings
3814 such as real names; it was not keeping to the 75-character limit for
3815 encoded words that the RFC imposes. It now respects the limit, and
3816 generates multiple encoded words if necessary. To be on the safe side, I
3817 have increased the buffer size for the ${rfc2047: expansion operator from
3820 22. Failure to deliver a bounce message always caused it to be frozen, even if
3821 there was an errors_to setting on the router. The errors_to setting is now
3824 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3825 canonical form (fully expanded) before being placed in
3826 $sender_host_address.
3828 24. Updated eximstats to version 1.33
3830 25. Include certificate and key file names in error message when GnuTLS fails
3831 to set them up, because the GnuTLS error message doesn't include the name
3832 of the failing file when there is a problem reading it.
3834 26. Expand error message when OpenSSL has problems setting up cert/key files.
3837 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3838 (this can affect the format of dates).
3840 28. exim_tidydb, when checking for the continued existence of a message for
3841 which it has found a message-specific retry record, was not finding
3842 messages that were in split spool directories. Consequently, it was
3843 deleting retry records that should have stayed in existence.
3845 29. eximstats updated to version 1.35
3846 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3847 1.35 - bugfix such that pie charts by volume are generated correctly
3849 30. The SPA authentication driver was not abandoning authentication and moving
3850 on to the next authenticator when an expansion was forced to fail,
3851 contradicting the general specification for all authenticators. Instead it
3852 was generating a temporary error. It now behaves as specified.
3854 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3855 (the order specifies the preference for clients). The order is now AES256,
3856 AES128, 3DES, ARCFOUR128.
3858 31. Small patch to Sieve code - explicitly set From: when generating an
3861 32. Exim crashed if a remote delivery caused a very long error message to be
3862 recorded - for instance if somebody sent an entire SpamAssassin report back
3863 as a large number of 550 error lines. This bug was coincidentally fixed by
3864 increasing the size of one of Exim's internal buffers (big_buffer) that
3865 happened as part of the Exiscan merge. However, to be on the safe side, I
3866 have made the code more robust (and fixed the comments that describe what
3869 33. Some experimental protocols are using DNS PTR records for new purposes. The
3870 keys for these records are domain names, not reversed IP addresses. The
3871 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3872 leaves it alone. Component reversal etc. now happens only for IP addresses.
3875 34. The host_aton() function is supposed to be passed a string that is known
3876 to be a valid IP address. However, in the case of IPv6 addresses, it was
3877 not checking this. This is a hostage to fortune. Exim now panics and dies
3878 if the condition is not met. A case was found where this could be provoked
3879 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3880 components; fortuitously, this particular loophole had already been fixed
3881 by change 4.50/55 or 4.44/33 above.
3883 If there are any other similar loopholes, the new check in host_aton()
3884 itself should stop them being exploited. The report I received stated that
3885 data on the command line could provoke the exploit when Exim was running as
3886 exim, but did not say which command line option was involved. All I could
3887 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3888 running as the user.
3891 35. There was a buffer overflow vulnerability in the SPA authentication code
3892 (which came originally from the Samba project). I have added a test to the
3893 spa_base64_to_bits() function which I hope fixes it.
3896 36. The daemon start-up calls getloadavg() while still root for those OS that
3897 need the first call to be done as root, but it missed one case: when
3898 deliver_queue_load_max is set with deliver_drop_privilege. This is
3899 necessary for the benefit of the queue runner, because there is no re-exec
3900 when deliver_drop_privilege is set.
3902 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3903 was in use, was not putting the data itself into the right store pool;
3904 consequently, it could be overwritten for a subsequent message in the same
3905 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3908 38. Sometimes the final signoff response after QUIT could fail to get
3909 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3910 < 0 before doing a fflush(). This bug looks as though it goes back to the
3911 introduction of TLS in release 3.20, but "sometimes" must have been rare
3912 because the tests only now provoked it.
3918 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3919 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3920 Now it returns an integer. A number of calls were still expecting a T/F
3921 return. Fortuitously, in all cases, the tests worked in OK situations,
3922 which is the norm. However, things would have gone wrong on any write
3923 failures on the smtp file descriptor. This function is used when sending
3924 messages over SMTP and also when doing verify callouts.
3926 2. When Exim is called to do synchronous delivery of a locally submitted
3927 message (the -odf or -odi options), it no longer closes stderr before doing
3930 3. Implemented the mua_wrapper option.
3932 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3934 5. Implemented the functions header_remove(), header_testname(),
3935 header_add_at_position(), and receive_remove_recipient(), and exported them
3938 6. If an ACL "warn" statement specified the addition of headers, Exim already
3939 inserted X-ACL-Warn: at the start if there was no header name. However, it
3940 was not making this test for the second and subsequent header lines if
3941 there were newlines in the string. This meant that an invalid header could
3942 be inserted if Exim was badly configured.
3944 7. Allow an ACL "warn" statement to add header lines at the start or after all
3945 the Received: headers, as well as at the end.
3947 8. Added the rcpt_4xx retry error code.
3949 9. Added postmaster_mailfrom=xxx to callout verification option.
3951 10. Added mailfrom=xxxx to the callout verification option, for verify=
3954 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3955 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3957 12. Inserted some casts to stop certain compilers warning when using pointer
3958 differences as field lengths or precisions in printf-type calls (mostly
3959 affecting debugging statements).
3961 13. Added optional readline() support for -be (dynamically loaded).
3963 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3964 same clock tick as a message's arrival, so that its received time was the
3965 same as the "first fail" time on the retry record, and that message
3966 remained on the queue past the ultimate address timeout, every queue runner
3967 would try a delivery (because it was past the ultimate address timeout) but
3968 after another failure, the ultimate address timeout, which should have then
3969 bounced the address, did not kick in. This was a "< instead of <=" error;
3970 in most cases the first failure would have been in the next clock tick
3971 after the received time, and all would be well.
3973 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3974 being recognized when the domain list was tested by the match_domain
3975 condition in an expansion string.
3977 16. Added the ${str2b64: operator.
3979 17. Exim was always calling setrlimit() to set a large limit for the number of
3980 processes, without checking whether the existing limit was already
3981 adequate. (It did check for the limit on file descriptors.) Furthermore,
3982 errors from getrlimit() and setrlimit() were being ignored. Now they are
3983 logged to the main and panic logs, but Exim does carry on, to try to do its
3984 job under whatever limits there are.
3986 18. Imported PCRE 5.0.
3988 19. Trivial typo in log message " temporarily refused connection" (the leading
3991 20. If the log selector return_path_on_delivery was set and an address was
3992 redirected to /dev/null, the delivery process crashed because it assumed
3993 that a return path would always be set for a "successful" delivery. In this
3994 case, the whole delivery is bypassed as an optimization, and therefore no
3997 21. Internal re-arrangement: the function for sending a challenge and reading
3998 a response while authentication was assuming a zero-terminated challenge
3999 string. It's now changed to take a pointer and a length, to allow for
4000 binary data in such strings.
4002 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4004 23. Exim was not respecting finduser_retries when seeking the login of the
4005 uid under which it was called; it was always trying 10 times. (The default
4006 setting of finduser_retries is zero.) Also, it was sleeping after the final
4007 failure, which is pointless.
4009 24. Implemented tls_on_connect_ports.
4011 25. Implemented acl_smtp_predata.
4013 26. If the domain in control=submission is set empty, Exim assumes that the
4014 authenticated id is a complete email address when it generates From: or
4015 Sender: header lines.
4017 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4018 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4019 chown and chgrp in /bin and hostname in /usr/bin.
4021 28. Exim was keeping the "process log" file open after each use, just as it
4022 does for the main log. This opens the possibility of it remaining open for
4023 long periods when the USR1 signal hits a daemon. Occasional processlog
4024 errors were reported, that could have been caused by this. Anyway, it seems
4025 much more sensible not to leave this file open at all, so that is what now
4028 29. The long-running daemon process does not normally write to the log once it
4029 has entered its main loop, and it closes the log before doing so. This is
4030 so that log files can straightforwardly be renamed and moved. However,
4031 there are a couple of unusual error situations where the daemon does write
4032 log entries, and I had neglected to close the log afterwards.
4034 30. The text of an SMTP error response that was received during a remote
4035 delivery was being truncated at 512 bytes. This is too short for some of
4036 the long messages that one sometimes sees. I've increased the limit to
4039 31. It is now possible to make retry rules that apply only when a message has a
4040 specific sender, in particular, an empty sender.
4042 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4043 it possible to be selective about when SMTP synchronization is enforced.
4045 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4047 32. Implemented hosts_connection_nolog.
4049 33. Added an ACL for QUIT.
4051 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4054 35. Added mailbox_size and mailbox_filecount to appendfile.
4056 36. Added control = no_multiline_responses to ACLs.
4058 37. There was a bug in the logic of the code that waits for the clock to tick
4059 in the case where the clock went backwards by a substantial amount such
4060 that the microsecond fraction of "now" was more than the microsecond
4061 fraction of "then" (but the whole seconds number was less).
4063 38. Added support for the libradius Radius client library this is found on
4064 FreeBSD (previously only the radiusclient library was supported).
4070 1. When certain lookups returned multiple values in the form name=value, the
4071 quoting of the values was not always being done properly. Specifically:
4072 (a) If the value started with a double quote, but contained no whitespace,
4074 (b) If the value contained whitespace other than a space character (i.e.
4075 tabs or newlines or carriage returns) it was not quoted.
4076 This fix has been applied to the mysql and pgsql lookups by writing a
4077 separate quoting function and calling it from the lookup code. The fix
4078 should probably also be applied to nisplus, ibase and oracle lookups, but
4079 since I cannot test any of those, I have not disturbed their existing code.
4081 2. A hit in the callout cache for a specific address caused a log line with no
4082 reason for rejecting RCPT. Now it says "Previous (cached) callout
4083 verification failure".
4085 3. There was an off-by-one bug in the queryprogram router. An over-long
4086 return line was truncated at 256 instead of 255 characters, thereby
4087 overflowing its buffer with the terminating zero. As well as fixing this, I
4088 have increased the buffer size to 1024 (and made a note to document this).
4090 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4091 when Exim is waiting for an SMTP response from a remote server, Exim
4092 restarts its select() call on the socket, thereby resetting its timeout.
4093 This is not a problem when such interrupts are rare. Somebody set up a cron
4094 job to run exiwhat every 2 minutes, which is less than the normal select()
4095 timeout (5 or 10 minutes). This meant that the select() timeout never
4096 kicked in because it was always reset. I have fixed this by comparing the
4097 time when an interrupt arrives with the time at the start of the first call
4098 to select(). If more time than the timeout has elapsed, the interrupt is
4099 treated as a timeout.
4101 5. Some internal re-factoring in preparation for the addition of Sieve
4102 extensions (by MH). In particular, the "personal" test is moved to a
4103 separate function, and given an option for scanning Cc: and Bcc: (which is
4104 not set for Exim filters).
4106 6. When Exim created an email address using the login of the caller as the
4107 local part (e.g. when creating a From: or Sender: header line), it was not
4108 quoting the local part when it contained special characters such as @.
4110 7. Installed new OpenBSD configuration files.
4112 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4113 try to make them clearer.
4115 9. Callout options, other than the timeout value, were being ignored when
4116 verifying sender addresses in header lines. For example, when using
4118 verify = header_sender/callout=no_cache
4120 the cache was (incorrectly) being used.
4122 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4123 only the Cygwin environment.
4125 11. When return_path_on_delivery was set as a log selector, if different remote
4126 addresses in the same message used different return paths and parallel
4127 remote delivery occurred, the wrong values would sometimes be logged.
4128 (Whenever a remote delivery process finished, the return path value from
4129 the most recently started remote delivery process was logged.)
4131 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4132 lines when AUTH and/or TLS are in use. This is the "received protocol"
4133 field. Exim used to use "asmtp" for authenticated SMTP, without any
4134 indication (in the protocol name) for TLS use. Now it follows the RFC and
4135 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4136 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4137 names appear in log lines as well as in Received: header lines.
4139 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4140 extensions, and comparison tests, and to fix some bugs.
4142 14. Changes to the "personal" filter test:
4144 (1) The test was buggy in that it was just doing the equivalent of
4145 "contains" tests on header lines. For example, if a user's address was
4146 anne@some.where, the "personal" test would incorrectly be true for
4148 To: susanne@some.where
4150 This test is now done by extracting each address from the header in turn,
4151 and checking the entire address. Other tests that are part of "personal"
4152 are now done using regular expressions (for example, to check local parts
4153 of addresses in From: header lines).
4155 (2) The list of non-personal local parts in From: addresses has been
4156 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4157 taken from the Sieve specification recommendations.
4159 (3) If the message contains any header line starting with "List-" it is
4160 treated as non-personal.
4162 (4) The test for "circular" in the Subject: header line has been removed
4163 because it now seems ill-conceived.
4165 15. Minor typos in src/EDITME comments corrected.
4167 16. Installed latest exipick from John Jetmore.
4169 17. If headers_add on a router specified a text string that was too long for
4170 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4171 of string_sprintf() is now avoided.
4173 18. $message_body_size was not set (it was always zero) when running the DATA
4174 ACL and the local_scan() function.
4176 19. For the "mail" command in an Exim filter, no default was being set for
4177 the once_repeat time, causing a random time value to be used if "once" was
4178 specified. (If the value happened to be <= 0, no repeat happened.) The
4179 default is now 0s, meaning "never repeat". The "vacation" command was OK
4180 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4181 (I found it when inspecting the code).
4183 20. There is now an overall timeout for performing a callout verification. It
4184 defaults to 4 times the callout timeout, which applies to individual SMTP
4185 commands during the callout. The overall timeout applies when there is more
4186 than one host that can be tried. The timeout is checked before trying the
4187 next host. This prevents very long delays if there are a large number of
4188 hosts and all are timing out (e.g. when the network connections are timing
4189 out). The value of the overall timeout can be changed by specifying an
4190 additional sub-option for "callout", called "maxwait". For example:
4192 verify = sender/callout=5s,maxwait=20s
4194 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4195 to the end before writing, but this should make it even safer).
4197 22. Exim was forgetting that it had advertised PIPELINING for the second and
4198 subsequent messages on an SMTP connection. It was also not resetting its
4199 memory on STARTTLS and an internal HELO.
4201 23. When Exim logs an SMTP synchronization error within a session, it now
4202 records whether PIPELINING has been advertised or not.
4204 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4205 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4206 rather than long int.
4208 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4210 26. Added the never_mail option to autoreply.
4216 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4217 crash if the getsockname() call failed; this can happen if a connection is
4218 closed very soon after it is established. The problem was simply in the
4219 order in which certain operations were done, causing Exim to try to write
4220 to the SMTP stream before it had set up the file descriptor. The bug has
4221 been fixed by making things happen in the correct order.
4227 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4228 before the connection was closed, thus losing the rejection response.
4230 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4231 some early Solaris releases, but causes trouble in current releases where
4232 socklen_t is defined.
4234 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4237 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4240 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4241 that releasing the top of it at the end releases what was used for sub-
4242 expansions (unless the block got too big). However, discard this block if
4243 the first thing is a variable or header, so that we can use its block when
4244 it is dynamic (useful for very large $message_headers, for example).
4246 6. Lookups now cache *every* query, not just the most recent. A new, separate
4247 store pool is used for this. It can be recovered when all lookup caches are
4248 flushed. Lookups now release memory at the end of their result strings.
4249 This has involved some general refactoring of the lookup sources.
4251 7. Some code has been added to the store_xxx() functions to reduce the amount
4252 of flapping under certain conditions.
4254 8. log_incoming_interface used to affect only the <= reception log lines. Now
4255 it causes the local interface and port to be added to several more SMTP log
4256 lines, for example "SMTP connection from", and rejection lines.
4258 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4260 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4262 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4263 could be overwritten at the end of the current message (or the start of a
4264 new message if it was set in a HELO ACL). The value is now preserved for
4265 the duration of the SMTP connection.
4267 12. If a transport had a headers_rewrite setting, and a matching header line
4268 contained an unqualified address, that address was qualified, even if it
4269 did not match any rewriting rules. The underlying bug was that the values
4270 of the flags that permit the existence of unqualified sender and recipient
4271 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4272 non-local messages, and by -bnq for local messages) were not being
4273 preserved with the message after it was received.
4275 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4276 "next input=" as part of the text comprising the host identity instead of
4277 the correct text. The code was using the same buffer for two different
4278 strings. However, depending on which order the printing function evaluated
4279 its arguments, the bug did not always show up. Under Linux, for example, my
4280 test suite worked just fine.
4282 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4283 doesn't work with some older versions of Perl. It has been changed to "my",
4284 which in any case is probably the better facility to use.
4286 15. A really picky compiler found some instances of statements for creating
4287 error messages that either had too many or two few arguments for the format
4290 16. The size of the buffer for calls to the DNS resolver has been increased
4291 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4292 for addresses that have a lot of PTR records. This alleviates a problem; it
4293 does not fully solve it.
4295 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4296 buffer now truncates the list and logs the incident, which is the same
4297 action as happens when Exim is looking up a host name and its aliases.
4298 Previously in this situation something unpredictable would happen;
4299 sometimes it was "internal error: store_reset failed".
4301 18. If a server dropped the connection unexpectedly when an Exim client was
4302 using GnuTLS and trying to read a response, the client delivery process
4303 crashed while trying to generate an error log message.
4305 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4306 string, for example:
4308 warn message = H1: something\nH2: something
4310 the text was added as a single header line from Exim's point of view
4311 though it ended up OK in the delivered message. However, searching for the
4312 second and subsequent header lines using $h_h2: did not work. This has been
4313 fixed. Similarly, if a system filter added multiple headers in this way,
4314 the routers could not see them.
4316 20. Expanded the error message when iplsearch is called with an invalid key to
4317 suggest using net-iplsearch in a host list.
4319 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4320 ACLs are no longer actually imposed (and a message to that effect is
4323 22. If a "gecos" field in a passwd entry contained escaped characters, in
4324 particular, if it contained a \" sequence, Exim got it wrong when building
4325 a From: or a Sender: header from that name. A second bug also caused
4326 incorrect handling when an unquoted " was present following a character
4327 that needed quoting.
4329 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4330 was not being matched caselessly.
4332 24. Arranged for all hyphens in the exim.8 source to be escaped with
4335 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4336 properly. Recipient callout cache records were still being keyed to include
4337 the sender, even when use_sender was set false. This led to far more
4338 callouts that were necessary. The sender is no longer included in the key
4339 when use_sender is false.
4341 26. Added "control = submission" modifier to ACLs.
4343 27. Added the ${base62d: operator to decode base 62 numbers.
4345 28. dnsdb lookups can now access SRV records.
4347 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4348 the configuration file.
4350 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4351 (-v) mode. This makes the output for a verbose queue run more intelligible.
4353 31. Added a use_postmaster feature to recipient callouts.
4355 32. Added the $body_zerocount variable, containing the number of binary zero
4356 bytes in the message body.
4358 33. The time of last modification of the "new" subdirectory is now used as the
4359 "mailbox time last read" when there is a quota error for a maildir
4362 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4364 35. Added +ignore_unknown as a special item in host lists.
4366 36. Code for decoding IPv6 addresses in host lists is now included, even if
4367 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4368 address was recognized as an IP address, but was then not correctly decoded
4369 into binary, causing unexpected and incorrect effects when compared with
4376 1. Very minor rewording of debugging text in manualroute to say "list of
4377 hosts" instead of "hostlist".
4379 2. If verify=header_syntax was set, and a header line with an unqualified
4380 address (no domain) and a large number of spaces between the end of the
4381 name and the colon was received, the reception process suffered a buffer
4382 overflow, and (when I tested it) crashed. This was caused by some obsolete
4383 code that should have been removed. The fix is to remove it!
4385 3. When running in the test harness, delay a bit after writing a bounce
4386 message to get a bit more predictability in the log output.
4388 4. Added a call to search_tidyup() just before forking a reception process. In
4389 theory, someone could use a lookup in the expansion of smtp_accept_max_
4390 per_host which, without the tidyup, could leave open a database connection.
4392 5. Added the variables $recipient_data and $sender_data which get set from a
4393 lookup success in an ACL "recipients" or "senders" condition, or a router
4394 "senders" option, similar to $domain_data and $local_part_data.
4396 6. Moved the writing of debug_print from before to after the "senders" test
4399 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4400 problems for message scanning, either using a data ACL, or using
4401 local_scan() because the Received: header was not generated till after they
4402 were called (in order to set the time as the time of reception completion).
4403 I have revised the way this works. The header is now generated after the
4404 body is received, but before the ACL or local_scan() are called. After they
4405 are run, the timestamp in the header is updated.
4411 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4412 before starting a queue runner without re-exec. This happened only when
4413 deliver_drop_privilege was set or when the Exim user was set to root. The
4414 effect of the bug was that timeouts during subsequent deliveries caused
4415 crashes instead of being properly handled. The handler is now left at its
4416 default (and expected) setting.
4418 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4419 message, again when deliver_drop_privilege is set or Exim is run as root.
4420 The bug described in (1) was not present in this case, but the tidying up
4421 of the other signals was missing. I have made the two cases consistent.
4423 3. The ignore_target_hosts setting on a manualroute router was being ignored
4424 for hosts that were looked up using the /MX notation.
4426 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4429 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4430 operated on the sender address. After changing the $sender_address to <>
4431 for the sender address verify, Exim was re-instated it as the original
4432 (before rewriting) address, but remembering that it had rewritten it, so it
4433 wasn't rewriting it again. This bug also had the effect of breaking the
4434 sender address verification caching when the sender address was rewritten.
4436 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4437 This has been changed so that if the ip literal address matches
4438 ignore_target_hosts, the router declines.
4440 7. Added expansion conditions match_domain, match_address, and match_local_
4441 part (NOT match_host).
4443 8. The placeholder for the Received: header didn't have a length field set.
4445 9. Added code to Exim itself and to exim_lock to test for a specific race
4446 condition that could lead to file corruption when using MBX delivery. The
4447 issue is with the lockfile that is created in /tmp. If this file is removed
4448 after a process has opened it but before that process has acquired a lock,
4449 there is the potential for a second process to recreate the file and also
4450 acquire a lock. This could lead to two Exim processes writing to the file
4451 at the same time. The added code performs the same test as UW imapd; it
4452 checks after acquiring the lock that its file descriptor still refers to
4453 the same named file.
4455 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4456 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4459 11. Added the smtp_active_hostname option. If used, this will typically be made
4460 to depend on the incoming interface address. Because $interface_address is
4461 not set up until the daemon has forked a reception process, error responses
4462 that can happen earlier (such as "too many connections") no longer contain
4465 12. If an expansion in a condition on a "warn" statement fails because a lookup
4466 defers, the "warn" statement is abandoned, and the next ACL statement is
4467 processed. Previously this caused the whole ACL to be aborted.
4469 13. Added the iplsearch lookup type.
4471 14. Added ident_timeout as a log selector.
4473 15. Added tls_certificate_verified as a log selector.
4475 16. Added a global option tls_require_ciphers (compare the smtp transport
4476 option of the same name). This controls incoming TLS connections.
4478 17. I finally figured out how to make tls_require_ciphers do a similar thing
4479 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4480 before starting the TLS session.
4482 18. Tabs are now shown as \t in -bP output.
4484 19. If the log selector return_path_on_delivery was set, Exim crashed when
4485 bouncing a message because it had too many Received: header lines.
4487 20. If two routers both had headers_remove settings, and the first one included
4488 a superfluous trailing colon, the final name in the first list and the
4489 first name in the second list were incorrectly joined into one item (with a
4490 colon in the middle).
4496 1. Added -C and -D options to the exinext utility, mainly to make it easier
4497 to include in the automated testing, but these could be helpful when
4498 multiple configurations are in use.
4500 2. The exinext utility was not formatting the output nicely when there was
4501 an alternate port involved in the retry record key, nor when there was a
4502 message id as well (for retries that were specific to a specific message
4503 and a specific host). It was also confused by IPv6 addresses, because of
4504 the additional colons they contain. I have fixed the IPv4 problem, and
4505 patched it up to do a reasonable job for IPv6.
4507 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4508 delivery, the log line now contains "pipelined" if PIPELINING was used.
4510 4. An SMTP transport process used to panic and die if the bind() call to set
4511 an explicit outgoing interface failed. This has been changed; it is now
4512 treated in the same way as a connect() failure.
4514 5. A reference to $sender_host_name in the part of a conditional expansion
4515 that was being skipped was still causing a DNS lookup. This no longer
4518 6. The def: expansion condition was not recognizing references to header lines
4519 that used bh_ and bheader_.
4521 7. Added the _cache feature to named lists.
4523 8. The code for checking quota_filecount in the appendfile transport was
4524 allowing one more file than it should have been.
4526 9. For compatibility with Sendmail, the command line option
4534 and sets the incoming protocol and host name (for trusted callers). The
4535 host name and its colon can be omitted when only the protocol is to be set.
4536 Note the Exim already has two private options, -pd and -ps, that refer to
4537 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4538 "s", but I don't think that's a major issue.
4540 10. A number of refactoring changes to the code, none of which should affect
4543 (a) The number of logging options was getting close to filling up the
4544 32-bit word that was used as a bit map. I have split them into two classes:
4545 those that are passed in the argument to log_write(), and those that are
4546 only ever tested independently outside of that function. These are now in
4547 separate 32-bit words, so there is plenty of room for expansion again.
4548 There is no change in the user interface or the logging behaviour.
4550 (b) When building, for example, log lines, the code previously used a
4551 macro that called string_cat() twice, in order to add two strings. This is
4552 not really sufficiently general. Furthermore, there was one instance where
4553 it was actually wrong because one of the argument was used twice, and in
4554 one call a function was used. (As it happened, calling the function twice
4555 did not affect the overall behaviour.) The macro has been replaced by a
4556 function that can join an arbitrary number of extra strings onto a growing
4559 (c) The code for expansion conditions now uses a table and a binary chop
4560 instead of a serial search (which was left over from when there were very
4561 few conditions). Also, it now recognizes conditions like "pam" even when
4562 the relevant support is not compiled in: a suitably worded error message is
4563 given if an attempt is made to use such a condition.
4565 11. Added ${time_interval:xxxxx}.
4567 12. A bug was causing one of the ddress fields not to be passed back correctly
4568 from remote delivery subprocesses. The field in question was not being
4569 subsequently used, so this caused to problems in practice.
4571 13. Added new log selectors queue_time and deliver_time.
4573 14. Might have fixed a bug in maildirsizefile handling that threw up
4574 "unexpected character" debug warnings, and recalculated the data
4575 unnecessarily. In any case, I expanded the warning message to give more
4578 15. Added the message "Restricted characters in address" to the statements in
4579 the default ACL that block characters like @ and % in local parts.
4581 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4582 Three changes have been made:
4584 (a) There was a serious bug; a negative response to MAIL caused the whole
4585 recipient domain to be cached as invalid, thereby blocking all messages
4586 to all local parts at the same domain, from all senders. This bug has
4587 been fixed. The domain is no longer cached after a negative response to
4588 MAIL if the sender used is not empty.
4590 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4593 (c) A new callout option, "use_sender" has been added for people who want
4594 the modified behaviour.
4600 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4603 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4604 indeed breaks things for older releases.
4606 3. Added additional logging to the case where there is a problem reading data
4607 from a filter that is running in a subprocess using a pipe, in order to
4608 try to track down a specific problem.
4610 4. Testing facility fudge: when running in the test harness and attempting
4611 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4612 getting "No route to host". Convert this to a timeout.
4614 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4617 6. Some OS don't have socklen_t but use size_t instead. This affects the
4618 fifth argument of getsockopt() amongst other things. This is now
4619 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4620 can be set for individual OS. I have set it for SunOS5, OSF1, and
4621 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4622 some earlier ones do not.
4624 7. Change 4.30/15 was not doing the test caselessly.
4626 8. The standard form for an IPv6 address literal was being rejected by address
4627 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4628 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4629 this, as well as the form without the "IPv6" on the front (but only when
4630 address literals are enabled, of course).
4632 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4634 10. Exim crashed if a message with an empty sender address specified by -f
4635 encountered a router with an errors_to setting. This could be provoked only
4636 by a command such as
4640 where an empty string was supplied; "<>" did not hit this bug.
4642 11. Installed PCRE release 4.5.
4644 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4645 remained set. It is now erased.
4647 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4648 times from message ids (which are base 36 rather than the normal 62).
4650 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4651 were being counted as actual protocol errors, and logged if the log
4652 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4653 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4654 and DATA following a set of rejected RCPTs do not count as protocol errors.
4655 In other words, Exim assumes they were pipelined, though this may not
4656 actually be the case. Of course, in all cases the client gets an
4657 appropriate error code.
4659 15. If a lookup fails in an ACL condition, a message about the failure may
4660 be available; it is used if testing the ACL cannot continue, because most
4661 such messages specify what the cause of the deferral is. However, some
4662 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4663 that caused an old message to be retained and used if a later statement
4664 caused a defer, replacing the real cause of the deferral.
4666 16. If an IP address had so many PTR records that the DNS lookup buffer
4667 was not large enough to hold them, Exim could crash while trying to process
4668 the truncated data. It now detects and logs this case.
4670 17. Further to 4.21/58, another change has been made: if (and only if) the
4671 first line of a message (the first header line) ends with CRLF, a bare LF
4672 in a subsequent header line has a space inserted after it, so as not to
4673 terminate the header.
4675 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4676 unnecessarily, used atoi() instead of strtol(), and didn't check the
4677 termination when getting file sizes from file names by regex.
4679 19. Completely re-implemented the support for maildirsize files, in the light
4680 of a number of problems with the previous contributed implementation
4681 (4.30/29). In particular:
4683 . If the quota is zero, the maildirsize file is maintained, but no quota is
4686 . If the maildir directory does not exist, it is created before any attempt
4687 to write a maildirsize file.
4689 . The quota value in the file is just a cache; if the quota is changed in
4690 the transport, the new value overrides.
4692 . A regular expression is available for excluding directories from the
4695 20. The autoreply transport checks the characters in options that define the
4696 message's headers; it allows continued headers, but it was checking with
4697 isspace() after an embedded newline instead of explicitly looking for a
4700 21. If all the "regular" hosts to which an address was routed had passed their
4701 expiry times, and had not reached their retry times, the address was
4702 bounced, even if fallback hosts were defined. Now Exim should go on to try
4705 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4706 equivalent code in the SMTP transport. Some hosts send humungous responses
4707 to HELO/EHLO, more than 1024 it seems.
4709 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4710 gives compiler warnings in some environments. I've now done it "properly",
4713 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4714 (because of problems with the built-in one) was declared to return uschar *
4715 instead of char *, causing compiler failure.
4717 25. Fixed a file descriptor leak when processing alias/forward files.
4719 26. Fixed a minor format string issue in dbfn.c.
4721 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4723 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4724 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4727 29. When a sender address is verified, it is cached, to save repeating the test
4728 when there is more than one recipient in a message. However, when the
4729 verification involves a callout, it is possible for different callout
4730 options to be set for different recipients. It is too complicated to keep
4731 track of this in the cache, so now Exim always runs a verification when a
4732 callout is required, relying on the callout cache for the optimization.
4733 The overhead is duplication of the address routing, but this should not be
4736 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4737 to be verified with callout=postmaster, and the main callout worked but the
4738 postmaster check failed, the verification correctly failed. However, if a
4739 subsequent RCPT command asked for sender verification *without* the
4740 postmaster check, incorrect caching caused this verification also to fail,
4743 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4744 it was not caching the DNS options (qualify_single, search_parents) that
4745 were used when the lookup failed. A subsequent lookup with different
4746 options therefore always gave the same answer, though there were cases
4747 where it should not have. (Example: a "domains = !$mx_any" option on a
4748 dnslookup router: the "domains" option is always processed without any
4749 widening, but the router might have qualify_single set.) Now Exim uses the
4750 cached value only when the same options are set.
4752 32. Added John Jetmore's "exipick" utility to the distribution.
4754 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4755 than a timeout (e.g. a certificate is required, and is not provided), an
4756 Exim server now closes the connection immediately. Previously it waited for
4757 the client to close - but if the client is SSL, it seems that they each
4758 wait for each other, leading to a delay before one of them times out.
4760 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4761 maintained 0.8.x compatibility because I don't think many are using it, and
4762 it is clearly obsolete.
4764 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4767 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4768 client certificate was expired. A simple patch fixes this, though I don't
4769 understand the full logic of why the verify callback is called multiple
4772 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4773 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4774 which causes problems with some clients (such as the Certicom SSL Plus
4775 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4776 disables the coutermeasure allowing Eudora to connect."
4778 38. Exim was not checking that a write() to a log file succeeded. This could
4779 lead to Bad Things if a log got too big, in particular if it hit a file
4780 size limit. Exim now panics and dies if it cannot write to a log file, just
4781 as it does if it cannot open a log file.
4783 39. Modified OS/Makefile-Linux so that it now contains
4785 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4787 The two -D definitions ensure that Exim is compiled with large file
4788 support, which makes it possible to handle log files that are bigger than
4791 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4792 instance) a domain was checked against a named list that involved a lookup,
4793 causing $domain_data to be set, then another domain was checked against the
4794 same list, then the first domain was re-checked, the value of $domain_data
4795 after the final check could be wrong. In particular, if the second check
4796 failed, it could be set empty. This bug probably also applied to
4799 41. The strip_trailing_dot option was not being applied to the address given
4800 with the -f command-line option.
4802 42. The code for reading a message's header from the spool was incrementing
4803 $received_count, but never initializing it. This meant that the value was
4804 incorrect (doubled) while delivering a message in the same process in which
4805 it was received. In the most common configuration of Exim, this never
4806 happens - a fresh exec is done - but it can happen when
4807 deliver_drop_privilege is set.
4809 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4810 it now includes up to 150 characters of the unexpected data in the log
4813 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4814 and building data strings. The size of both of these buffers was 10 000
4815 bytes - far larger than anybody would *ever* want, thought I. Needless to
4816 say, somebody hit the limit. I have increased the maximum line length to
4817 20 000 and the maximum data length of concatenated lines to 100 000. I have
4818 also fixed two bugs, because there was no checking on these buffers. Tsk,
4819 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4820 buffer is too small.
4822 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4823 lsearch lookups. Now it does.
4825 46. When parsing a route_list item in a manualroute router, a fixed-length
4826 buffer was used for the list of hosts. I made this 1024 bytes long,
4827 thinking that nobody would ever have a list of hosts that long. Wrong.
4828 Somebody had a whole pile of complicated expansion conditions, and the
4829 string was silently truncated, leading to an expansion error. It turns out
4830 that it is easier to change to an unlimited length (owing to other changes
4831 that have happened since this code was originally written) than to build
4832 structure for giving a limitation error. The length of the item that
4833 expands into the list of hosts is now unlimited.
4835 47. The lsearch lookup could not handle data where the length of text line was
4836 more than 4095 characters. Such lines were truncated, leading to shortened
4837 data being returned. It should now handle lines of any length.
4839 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4840 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4843 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4844 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4845 no longer happens in comment lines. A list of replacements is now placed
4846 at the head of all of the source files, except those whose only change is
4847 to replace PERL_COMMAND in the very first #! line.
4849 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4850 messages on the queue, with a bottom-up merge sort, using code contributed
4851 by Michael Haardt. This should make operations like -bp somewhat faster on
4852 large queues. It won't affect queue runners, except when queue_run_in_order
4855 51. Installed eximstats 1.31 in the distribution.
4857 52. Added support for SRV lookups to the dnslookup router.
4859 53. If an ACL referred to $message_body or $message_body_end, the value was not
4860 reset for any messages that followed in the same SMTP session.
4862 54. The store-handling optimization for building very long strings was not
4863 differentiating between the different store pools. I don't think this
4864 actually made any difference in practice, but I've tidied it.
4866 55. While running the routers to verify a sender address, $sender_address
4867 was still set to the sender address. This is wrong, because when routing to
4868 send a bounce to the sender, it would be empty. Therefore, I have changed
4869 it so that, while verifying a sender address, $sender_address is set to <>.
4870 (There is no change to what happens when verifying a recipient address.)
4872 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4873 target A or AAAA records (if not already returned) without resetting the
4874 qualify_single or search_parents options of the DNS resolver. These are
4875 inappropriate in this case because the targets of MX and SRV records must
4876 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4877 target that, when qualified, matched something in the local domain. These
4878 two options are now turned off when doing these lookups.
4880 57. It seems that at least some releases of Reiserfs (which does not have the
4881 concept of a fixed number of inodes) returns zero and not -1 for the
4882 number of available inodes. This interacted badly with check_spool_inodes,
4883 which assumed that -1 was the "no such thing" setting. What I have done is
4884 to check that the total number of inodes is greater than zero before doing
4885 the test of how many are available.
4887 58. When a "warn" ACL statement has a log_message modifier, the message is
4888 remembered, and not repeated. This is to avoid a lot of repetition when a
4889 message has many recipients that cause the same warning to be written.
4890 Howewer, Exim was preserving the list of already written lines for an
4891 entire SMTP session, which doesn't seem right. The memory is now reset if a
4892 new message is started.
4894 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4895 debugging output unless log_rewrite was also set.
4897 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4898 of (int)(handle) when we know that handle contains (void *)(-1).
4900 61. The Exim daemon panic-logs an error return when it closes the incoming
4901 connection. However "connection reset by peer" seems to be common, and
4902 isn't really an error worthy of noting specially, so that particular error
4905 62. When Exim is trying to find all the local interfaces, it used to panic and
4906 die if the ioctl to get the interface flags failed. However, it seems that
4907 on at least one OS (Solaris 9) it is possible to have an interface that is
4908 included in the list of interfaces, but for which you get a failure error
4909 for this call. This happens when the interface is not "plumbed" into a
4910 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4911 failure of the "get flags" call assumes that the interface is down.
4913 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4914 makes life easier for people who are doing arithmetic on fields extracted
4915 from dates, where you often get leading zeros that should not be
4916 interpreted as octal.
4918 64. Added qualify_domain to the redirect router, to override the global
4921 65. If a pathologically long header line contained very many addresses (the
4922 report of this problem mentioned 10 000) and each of them was rewritten,
4923 Exim could use up a very large amount of memory. (It kept on making new
4924 copies of the header line as it rewrote, and never released the old ones.)
4925 At the expense of a bit more processing, the header rewriting function has
4926 been changed so that it no longer eats memory in this way.
4928 66. The generation of the Received: header has been moved from the time that a
4929 message starts to be received, to the time that it finishes. The timestamp
4930 in the Received: header should now be very close to that of the <= log
4931 line. There are two side-effects of this change:
4933 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4934 logged header lines no longer include the local Received: line, because
4935 it has not yet been created. The same applies to a copy of the message
4936 that is returned to a non-SMTP sender when a message is rejected.
4938 (b) When a filter file is tested using -bf, no additional Received: header
4939 is added to the test message. After some thought, I decided that this
4942 This change does not affect the value of $received_for. It is still set
4943 after address rewriting, but before local_scan() is called.
4945 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4947 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4948 gave an unhelpful panic error message, and a defer error. I have managed to
4949 change this behaviour so that it now rejects any supplied certificate,
4950 which seems right, as the list of acceptable certificates is empty.
4952 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4953 gave an unhelpful defer error. I have not managed to make this reject any
4954 supplied certificates, but the error message it gives is "no certificate
4955 supplied", which is not helpful.
4957 70. exigrep's output now also includes lines that are not associated with any
4958 message, but which match the given pattern. Implemented by a patch from
4959 Martin Sluka, which also tidied up the Perl a bit.
4961 71. Recipient callout verification, like sender verification, was using <> in
4962 the MAIL FROM command. This isn't really the right thing, since the actual
4963 sender may affect whether the remote host accepts the recipient or not. I
4964 have changed it to use the actual sender in the callout; this means that
4965 the cache record is now keyed on a recipient/sender pair, not just the
4966 recipient address. There doesn't seem to be a real danger of callout loops,
4967 since a callout by the remote host to check the sender would use <>.
4968 [SEE ABOVE: changed after hitting problems.]
4970 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4971 temporary errors. However, in the case of such a code being given after
4972 the end of a data transmission (i.e. after ".") Exim was failing to write
4973 a retry record for the message. (Yes, there was some broken host that was
4974 actually sending 8xx at this point.)
4976 73. An unknown lookup type in a host list could cause Exim to panic-die when
4977 the list was checked. (An example that provoked this was putting <; in the
4978 middle of a list instead of at the start.) If this happened during a DATA
4979 ACL check, a -D file could be left lying around. This kind of configuration
4980 error no longer causes Exim to die; instead it causes a defer errror. The
4981 incident is still logged to the main and panic logs.
4983 74. Buglet left over from Exim 3 conversion. The message "too many messages
4984 in one connection" was written to the rejectlog but not the mainlog, except
4985 when address rewriting (yes!) was being logged.
4987 75. Added write_rejectlog option.
4989 76. When a system filter was run not as root (that is, when system_filter_user
4990 was set), the values of the $n variables were not being returned to the
4991 main process; thus, they were not subsequently available in the $sn
4994 77. Added +return_path_on_delivery log selector.
4996 78. A connection timeout was being treated differently from recipients deferred
4997 when testing hosts_max_try with a message that was older than the host's
4998 retry timeout. (The host should not be counted, thus allowing all hosts to
4999 be tried at least once before bouncing.) This may have been the cause of an
5000 occasionally reported bug whereby a message would remain on the queue
5001 longer than the retry timeout, but would be bounced if a delivery was
5002 forced. I say "may" because I never totally pinned down the problem;
5003 setting up timeout/retry tests is difficult. See also the next item.
5005 79. The ultimate address timeout was not being applied to errors that involved
5006 a combination of host plus message (for example, a timeout on a MAIL
5007 command). When an address resolved to a number of possible hosts, and they
5008 were not all tried for each delivery (e.g. because of hosts_max_try), a
5009 message could remain on the queue longer than the retry timeout.
5011 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5014 81. Fixed an obscure SMTP outgoing bug which required at least the following
5015 conditions: (a) there was another message waiting for the same server;
5016 (b) the server returned 5xx to all RCPT commands in the first message so
5017 that the message was not completed; (c) the server dropped the connection
5018 or gave a negative response to the RSET that Exim sends to abort the
5019 transaction. The observed case was a dropped connection after DATA that had
5020 been sent in pipelining mode. That is, the server had advertised PIPELINING
5021 but was not implementing it correctly. The effect of the bug was incorrect
5022 behaviour, such as trying another host, and this could lead to a crash.
5028 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5029 and daemon.c were passed as pointers to ints; they should have been
5030 pointers to socklen_t variables (which are typically unsigned ints).
5032 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5035 3. Fixed a really odd bug that affected only the testing scheme; patching a
5036 certain fixed string in the binary changed the value of another string that
5037 happened to be identical to the end of the original first string.
5039 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5040 name", it returns that address as the IP address. On some operating
5041 systems (e.g. Solaris), it also passes back the IP address string as the
5042 "host name". However, on others (e.g. Linux), it passes back an empty
5043 string. Exim wasn't checking for this, and was changing the host name to an
5044 empty string, assuming it had been canonicized.
5046 5. Although rare, it is permitted to have more than one PTR record for a given
5047 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5048 all the names associated with an address, because they do in Solaris.
5049 However, it seems that they do not in Linux for data that comes from the
5050 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5051 I found this out when I moved to a new Linux workstation and tried to run
5052 the Exim test suite.
5054 To get round this problem I have changed the code so that it now does its
5055 own call to the DNS to look up PTR records when searching for a host name.
5056 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5057 addresses that are only in /etc/hosts are still found.
5059 This behaviour is, however, controlled by an option called host_lookup_
5060 order, which defaults to "bydns:byaddr". If people want to use the other
5061 order, or indeed, just use one or the other means of lookup, they can
5062 specify it in this variable.
5064 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5065 some operating systems, this comes back from gethostbyaddr() as an empty
5066 string, and this is what Exim used to test for. However, it seems that in
5067 other systems, "." is yielded. Exim now tests for this case too.
5069 7. The values of check_spool_space and check_log_space are now held internally
5070 as a number of kilobytes instead of an absolute number of bytes. If a
5071 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5072 kilobyte. This means that much larger values can be stored.
5074 8. Exim monitor: an attempt to get the action menu when not actually pointing
5075 at a message produces an empty menu entitled "No message selected". This
5076 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5077 no entries in it ("Shell widget menu has zero width and/or height"). So I
5078 have added a single, blank menu entry in this case.
5080 9. Added ${quote_local_part.
5082 10. MIME decoding is now applied to the contents of Subject: header lines when
5085 11. Now that a reference to $sender_host_address automatically causes a reverse
5086 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5087 host lookup before query-style lookups in lists that might use this
5088 variable. This has therefore been abolished, and the "net-" prefix is no
5089 longer necessary for query-style lookups.
5091 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5092 have been a typo for LFLAGS, so it has been changed.
5094 13. The install script calls Exim with "-C /dev/null" in order to find the
5095 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5096 to be output. Howeve, since Exim outputs its version number before the
5097 error, it didn't break the script. It just looked ugly. I fixed this by
5098 always allowing "-C /dev/null" if the caller is root.
5100 14. Ignore overlarge ACL variable number when reading spool file - insurance
5101 against a later release with more variables having written the file.
5103 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5104 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5105 this, as well as the form without the "IPv6" on the front.
5107 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5108 OS/Makefile-Darwin file.
5110 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5111 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5113 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5114 variable. However, some OS use ints for inodes. I've added cast to long int
5115 to get rid of the compiler warning.
5117 19. I had forgotten to lock out "/../" in configuration file names when
5118 ALT_CONFIG_PREFIX was set.
5120 20. Routers used for verification do not need to specify transports. However,
5121 if such a router generated a host list, and callout was configured, Exim
5122 crashed, because it could not find a port number from the (non-existent)
5123 transport. It now assumes port 25 in this circumstance.
5125 21. Added the -t option to exigrep.
5127 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5128 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5129 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5130 them is set, LOOKUP_LSEARCH is forced.
5132 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5134 24. Added sender and host information to the "rejected by local_scan()" log
5135 line; previously there was no indication of these.
5137 25. Added .include_if_exists.
5139 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5140 Linux. It turns out that not all file systems support this. Apparently some
5141 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5142 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5143 not supported on the file descriptor, is now ignored when Exim is trying to
5144 sync a directory. This applies only to Linux.
5146 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5148 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5149 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5152 29. Added support for maildirsize files from supplied patch (modified a bit).
5154 30. The use of :fail: followed by an empty string could lead Exim to respond to
5155 sender verification failures with (e.g.):
5157 550 Verification failed for <xxx>
5158 550 Sender verify failed
5160 where the first response line was missing the '-' that indicates it is not
5161 the final line of the response.
5163 31. The loop for finding the name of the user that called Exim had a hardwired
5164 limit of 10; it now uses the value of finduser_retries, which is used for
5165 all other user lookups.
5167 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5170 33. Exim was neglecting to zero errno before one call of strtol() when
5171 expanding a string and expecting an integer value. On some systems this
5172 resulted in spurious "integer overflow" errors. Also, it was casting the
5173 result into an int without checking.
5175 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5176 did not work. The code looks as if it has *never* worked, though it appears
5177 to have been documented since at least releast 1.62. I have made it work.
5179 35. The "timeout_DNS" error in retry rules, also documented since at least
5180 1.62, also never worked. As it isn't clear exactly what this means, and
5181 clearly it isn't a major issue, I have abolished the feature by treating it
5182 as "timeout", and writing a warning to the main and panic logs.
5184 36. The display of retry rules for -brt wasn't always showing the error code
5187 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5188 timeout_connect_A, timeout_connect_MX.
5190 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5191 to the empty sender.
5193 39. The daemon was not analysing the content of -oX till after it had closed
5194 stderr and disconnected from the controlling terminal. This meant that any
5195 syntax errors were only noted on the panic log, and the return code from
5196 the command was 0. By re-arranging the code a little, I've made the
5197 decoding happen first, so such errors now appear on stderr, and the return
5198 code is 1. However, the actual setting up of the sockets still happens in
5199 the disconnected process, so errors there are still only recorded on the
5202 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5203 connections (as happens on some IP stacks) was logged at start up time as
5204 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5205 it from "IPv6 and IPv4", which means that two separate sockets are being
5208 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5209 says whether AF_INET or AF_INET6 was passed as an argument.
5211 42. Exiwhat output was messed up when time zones were included in log
5214 43. Exiwhat now gives more information about the daemon's listening ports,
5215 and whether -tls-on-connect was used.
5217 44. The "port" option of the smtp transport is now expanded.
5219 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5220 silently ignored. Now an error message is written to the main and panic
5223 46. There's a new ACL modifier called "logwrite" which writes to a log file
5224 as soon as it is encountered.
5226 47. Added $local_user_uid and $local_user_gid at routing time.
5228 48. Exim crashed when trying to verify a sender address that was being
5231 49. Exim was recognizing only a space character after ".include". It now also
5232 recognizes a tab character.
5234 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5235 extracting the relevant information from the specification. The man page no
5236 longer contains scrambled data for the -d option, and I've added a section
5237 at the front about calling Exim under different names.
5239 51. Added "extra_headers" argument to the "mail" command in filter files.
5241 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5244 53. Installed eximstats 1.29.
5246 54. Added transport_filter_timeout as a generic transport option.
5248 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5249 Cc: header lines. This was required by RFC 822, but it not required by RFC
5252 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5253 incoming messages that did not have them. Now it does so only if the
5254 message originates locally, that is, if there is no associated remote host
5255 address. When Resent- header lines are present, this applies to the Resent-
5256 lines rather than the non-Resent- lines.
5258 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5259 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5261 58. Messages for configuration errors now include the name of the main
5262 configuration file - useful now that there may be more than one file in a
5263 list (.included file names were always shown).
5265 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5266 for those rare installations that do not start the daemon as root or run it
5267 setuid root. I've cut out the call to initgroups() if the daemon is not
5270 60. The Exim user and group can now be bound into the binary as text strings
5271 that are looked up at the start of Exim's processing.
5273 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5275 62. Added $mailstore_basename variable.
5277 63. Installed patch to sieve.c from Michael Haardt.
5279 64. When Exim failed to open the panic log after failing to open the main log,
5280 the original message it was trying to log was written to stderr and debug
5281 output, but if they were not available (the usual case in production), it
5282 was lost. Now it is written to syslog before the two lines that record the
5283 failures to open the logs.
5285 65. Users' Exim filters run in subprocesses under the user's uid. It is
5286 possible for a "deliver" command or an alias in a "personal" command to
5287 provoke an address rewrite. If logging of address rewriting is configured,
5288 this fails because the process is not running as root or exim. There may be
5289 a better way of dealing with this, but for the moment (because 4.30 needs
5290 to be released), I have disabled address rewrite logging when running a
5291 filter in a non-root, non-exim process.
5297 1. The buildconfig auxiliary program wasn't quoting the value set for
5298 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5299 not defined. This bug was masked in 4.22 by the effect that was fixed in
5302 2. Some messages that were rejected after a message id was allocated were
5303 shown as "incomplete" by exigrep. It no longer does this for messages that
5304 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5306 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5307 have allow_domain_literals set, the ID did not get logged in the <= line.
5308 Domain literals are now always recognized in Message-ID: header lines.
5310 4. The first argument for a ${extract expansion item is the key name or field
5311 number. Leading and trailing spaces in this item were not being ignored,
5312 causing some misleading effects.
5314 5. When deliver_drop_privilege was set, single queue runner processes started
5315 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5316 same command in the process it spins off) were not dropping privilege.
5318 6. When the daemon running as "exim" started a queue runner, it always
5319 re-executed Exim in the spun-off process. This is a waste of effort when
5320 deliver_drop_privilege is set. The new process now just calls the
5321 queue-runner function directly.
5327 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5330 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5331 ignored. Though the use of -oP was forcing the writing of a pid file, it
5332 was always written to the default place.
5334 3. If the message "no IP address found for host xxxx" is generated during
5335 incoming verification, it is now followed by identification of the incoming
5336 connection (so you can more easily find what provoked it).
5338 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5340 5. Added some features to "harden" Exim a bit more against certain attacks:
5342 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5343 be put in Local/Makefile. This is like the never_users runtime option,
5344 but it cannot be overridden. The default setting is "root".
5346 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5347 prefix string with which any file named in a -C command line option
5350 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5351 is retained for -C and -D only if the caller of Exim is root. Without
5352 it, the exim user may also use -C and -D and retain privilege.
5354 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5355 command line option is disabled.
5357 6. Macro names set by the -D option must start with an upper case letter, just
5358 like macro names defined in the configuration file.
5360 7. Added "dereference=" facility to LDAP.
5362 8. Two instances of the typo "uknown" in the source files are fixed.
5364 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5365 the Configure-Makefile script screwed up while processing it.
5367 10. Incorporated PCRE 4.4.
5369 11. The SMTP synchronization check was not operating right at the start of an
5370 SMTP session. For example, it could not catch a HELO sent before the client
5371 waited for the greeting. There is now a check for outstanding input at the
5372 point when the greeting is written. Because of the duplex, asynchronous
5373 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5374 way, but not yet received, when the check is performed.
5376 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5377 on TCP/IP sockets, because this apparently causes some broken clients to
5380 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5381 unchanged) from the Cygwin maintainer.
5383 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5384 was supported instead of testing for mbx. Effectively a typo.
5386 15. The spa authenticator server code was not checking that the input it
5387 received was valid base64.
5389 16. The debug output line for the "set" modifier in ACLs was not showing the
5390 name of the variable that was being set.
5392 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5394 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5395 lookup on its own. Something else was needed to trigger the lookup. For
5396 example, a match in host_lookup or the need for a host name in a host list.
5397 Now, if $sender_host_name is referenced and the host name has not yet been
5398 looked up, a lookup is performed. If the lookup fails, the variable remains
5399 empty, and $host_lookup_failed is set to "1".
5401 19. Added "eqi" as a case-independent comparison operator.
5403 20. The saslauthd authentication condition could segfault if neither service
5404 nor realm was specified.
5406 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5407 error message that was logged was misleading, and incoming SMTP
5408 connections were dropped. The message is now more accurate, and temporary
5409 errors are given to SMTP connections.
5411 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5412 (except RSET) with a 421 error, until QUIT is received. However, it was
5413 failing to send a response to QUIT.
5415 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5416 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5417 ACL is now run after the helo_try_verify_hosts code.
5419 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5420 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5421 case-independent, so other case variants are also recognized). Apparently
5422 some systems use these upper case variants.
5424 25. If more than two messages were waiting for the same host, and a transport
5425 filter was specified for the transport, Exim sent two messages over the
5426 same TCP/IP connection, and then failed with "socket operation on non-
5427 socket" when it tried to send the third.
5429 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5431 27. The extern definition of crypt16() in expand.c was not being excluded when
5432 the OS had its own crypt16() function.
5434 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5435 as a preferred synonym for return_size_limit, both as an option and as an
5438 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5440 30. Changed the default configuration ACL to relax the local part checking rule
5441 for addresses that are not in any local domains. For these addresses,
5442 slashes and pipe symbols are allowed within local parts, but the sequence
5443 /../ is explicitly forbidden.
5445 31. SPA server authentication was not clearing the challenge buffer before
5448 32. log_message in a "warn" ACL statement was writing to the reject log as
5449 well as to the main log, which contradicts the documentation and doesn't
5450 seem right (because no rejection is happening). So I have stopped it.
5452 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5453 However, I am unable to do any testing of this.
5455 34. Fixed an infelicity in the appendfile transport. When checking directories
5456 for a mailbox, to see if any needed to be created, it was accidentally
5457 using path names with one or more superfluous leading slashes; tracing
5458 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5460 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5461 contents are added to the log line that is written for every discarded
5462 recipient. (Previously a log_message setting was ignored.)
5464 36. The ${quote: operator now quotes the string if it is empty.
5466 37. The install script runs exim in order to find its version number. If for
5467 some reason other than non-existence or emptiness, which it checks, it
5468 could not run './exim', it was installing it with an empty version number,
5469 i.e. as "exim-". This error state is now caught, and the installation is
5472 38. An argument was missing from the function that creates an error message
5473 when Exim fails to connect to the socket for saslauthd authentication.
5474 This could cause Exim to crash, or give a corrupted message.
5476 39. Added isip, isip4, and isip6 to ${if conditions.
5478 40. The ACL variables $acl_xx are now saved with the message, and can be
5479 accessed later in routers, transports, and filters.
5481 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5482 strings in the file are not string-expanded.
5484 42. If a MAIL command specified a SIZE value that was too large to fit into an
5485 int variable, the check against message_size_limit failed. Such values are
5486 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5487 day this will have to be increased, but I don't think I want to be around
5488 when emails are that large.
5495 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5496 iconv() is not standard in FreeBSD.
5498 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5499 IPv6 enabled. The observed symptom was a segmentation fault on return from
5500 the function os_common_find_running_interfaces() in src/os.c.
5502 3. In the check_special_case() function in daemon.c I had used "errno" as an
5503 argument name, which causes warnings on some systems. This was basically a
5504 typo, since it was named "eno" in the comments!
5506 4. The code that waits for the clock to tick (at a resolution of some fraction
5507 of a second) so as to ensure message-id uniqueness was always waiting for
5508 at least one whole tick, when it could have waited for less. [This is
5509 almost certainly not relevant at current processor speeds, where it is
5510 unlikely to ever wait at all. But we try to future-proof.]
5512 5. The function that sleeps for a time interval that includes fractions of a
5513 second contained a race. It did not block SIGALRM between setting the
5514 timer, and suspending (a couple of lines later). If the interval was short
5515 and the sigsuspend() was delayed until after it had expired, the suspension
5516 never ended. On busy systems this could lead to processes getting stuck for
5519 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5520 process, before it forks any delivery processes. The open lookup caching
5521 mechanism meant that the open file or database connection was passed into
5522 the delivery process. The problem was that delivery processes always tidy
5523 up cached lookup data. This could cause a problem for the next delivery
5524 process started by the queue runner, because the external queue runner
5525 process does not know about the closure. So the next delivery process
5526 still has data in the lookup cache. In the case of a file lookup, there was
5527 no problem because closing a file descriptor in a subprocess doesn't affect
5528 the parent. However, if the lookup was caching a connection to a database,
5529 the connection was closed, and the second delivery process was likely to
5530 see errors such as "PGSQL: query failed: server closed the connection
5531 unexpectedly". The problem has been fixed by closing all cached lookups
5532 in a queue runner before running a delivery process.
5534 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5535 seem to have the "const" qualifier which it has on other OS. I've
5538 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5539 *of the same type* (client or server) with the same public name that an
5540 error should be diagnosed.
5542 9. When Exim looked up a host name for an IP address, but failed to find the
5543 original IP address when looking up the host name (a safety check), it
5544 output the message "<ip address> does not match any IP for NULL", which was
5545 confusing, to say the least. The bug was that the host name should have
5546 appeared instead of "NULL".
5548 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5549 user that is built into the binary, and the -C or -D options is used, root
5550 privilege is dropped before the configuration file is read. In addition,
5551 logging is switched to stderr instead of the normal log files. If the
5552 configuration then re-defines the Exim user, the unprivileged environment
5553 is probably not what is expected, so Exim logs a panic warning message (but
5556 However, if deliver_drop_privilege is set, the unprivileged state may well
5557 be exactly what is intended, so the warning has been cut out in that case,
5558 and Exim is allowed to try to write to its normal log files.
5564 1. smtp_return_error_details was not giving details for temporary sender
5565 or receiver verification errors.
5567 2. Diagnose a configuration error if two authenticators have the same public
5570 3. Exim used not to create the message log file for a message until the first
5571 delivery attempt. This could be confusing when incoming messages were held
5572 for policy or load reasons. The message log file is now created at the time
5573 the message is received, and an initial "Received" line is written to it.
5575 4. The automatically generated man page for command line options had a minor
5576 bug that caused no ill effects; however, a more serious problem was that
5577 the procedure for building the man page automatically didn't always
5578 operate. Consequently, release 4.20 contains an out-of-date version. This
5579 shouldn't happen again.
5581 5. When building Exim with embedded Perl support, the script that builds the
5582 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5583 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5585 6. The freeze_tell option was not being used for messages that were frozen on
5586 arrival, either by an ACL or by local_scan().
5588 7. Added the smtp_incomplete_transaction log selector.
5590 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5591 was accepting AUTH without a new EHLO.
5593 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5594 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5597 10. Logging of TCP/IP connections (when configured) now happens in the main
5598 daemon process instead of the child process, so that the TCP/IP connection
5599 count is more accurate (but it can never be perfect).
5601 11. The use of "drop" in a nested ACL was not being handled correctly in the
5602 outer ACL. Now, if condition failure induced by the nested "drop" causes
5603 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5604 or "require"), the connection is dropped.
5606 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5607 that yield "discard" can now be used with an "accept" or a "discard" verb,
5608 but an error is generated for any others (because I can't see a useful way
5609 to define what should happen).
5611 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5612 are now processed in the same way as lines in the Exim configuration file.
5613 In particular, continuation lines are supported.
5615 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5617 15. Added -ti meaning -t -i.
5619 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5620 domains, and warn by logging if others are found.
5622 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5623 returned from GIFCONF when Exim is trying to find the list of interfaces on
5624 a host. The code in os.c has been modified to copy each ifreq to an aligned
5625 structure in all cases.
5627 Also, in some cases, the returned ifreq's were being copied to a 'struct
5628 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5629 means the last couple of bytes of an IPv6 address could be chopped if the
5630 ifreq contained only a normal sockaddr (14 bytes storage).
5632 18. Named domain lists were not supported in the hosts_treat_as_local option.
5633 An entry such as +xxxx was not recognized, and was treated as a literal
5636 19. Ensure that header lines added by a DATA ACL are included in the reject log
5637 if the ACL subsequently rejects the message.
5639 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5640 MD5 (which is deprecated).
5642 21. When testing a filter file using -bf, Exim was writing a message when it
5643 took the sender from a "From " line in the message, but it was not doing so
5644 when it took $return_path from a Return-Path: header line. It now does.
5646 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5647 with a valid header line field name (a series of printing characters
5648 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5650 23. Changed "disc" in the source to "disk" to conform to the documentation and
5651 the book and for uniformity.
5653 24. Ignore Sendmail's -Ooption=value command line item.
5655 25. When execve() failed while trying to run a command in a pipe transport,
5656 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5657 could be confused with a return value of 69 from the command itself. This
5658 has been changed to 127, the value the shell returns if it is asked to run
5659 a non-existent command. The wording for the related log line suggests a
5660 non-existent command as the problem.
5662 26. If received_header_text expands to an empty string, do not add a Received:
5663 header line to the message. (Well, it adds a token one on the spool, but
5664 marks it "old" so that it doesn't get used or transmitted.)
5666 27. Installed eximstats 1.28 (addition of -nt option).
5668 28. There was no check for failure on the call to getsockname() in the daemon
5669 code. This can fail if there is a shortage of resources on the system, with
5670 ENOMEM, for example. A temporary error is now given on failure.
5672 29. Contrary to the C standard, it seems that in some environments, the
5673 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5674 program. Exim now does this explicitly; it affects the formatting of
5675 timestamps using strftime().
5677 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5678 complaints. I've now initialized all the variables, to avoid this.
5680 32. Header lines added by a system filter were not being "seen" during
5681 transport-time rewrites.
5683 33. The info_callback() function passed to OpenSSL is set up with type void
5684 (*)(SSL *, int, int), as described somewhere. However, when calling the
5685 function (actually a macro) that sets it up, the type void(*)() is
5686 expected. I've put in a cast to prevent warnings from picky compilers.
5688 34. If a DNS black list lookup found a CNAME record, but there were no A
5689 records associated with the domain it pointed at, Exim crashed.
5691 35. If a DNS black list lookup returned more than one A record, Exim ignored
5692 all but the first. It now scans all returned addresses if a particular IP
5693 value is being sought. In this situation, the contents of the
5694 $dnslist_value variable are a list of all the addresses, separated by a
5697 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5698 to accept a host name and all its aliases if the forward lookup for any of
5699 them yielded the IP address of the incoming connection. Now it accepts only
5700 those names whose forward lookup yields the correct IP address. Any other
5701 names are discarded. This closes a loophole whereby a rogue DNS
5702 administrator could create reverse DNS records to break through a
5703 wildcarded host restriction in an ACL.
5705 37. If a user filter or a system filter that ran in a subprocess used any of
5706 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5707 the wrong values were passed to the pipe command ($thisaddress had the
5708 value of $0, $0 had the value of $1, etc). This bug was introduced by
5709 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5711 38. Improved the line breaking for long SMTP error messages from ACLs.
5712 Previously, if there was no break point between 40 and 75 characters, Exim
5713 left the rest of the message alone. Two changes have been made: (a) I've
5714 reduced the minimum length to 35 characters; (b) if it can't find a break
5715 point between 35 and 75 characters, it looks ahead and uses the first one
5716 that it finds. This may give the occasional overlong line, but at least the
5717 remaining text gets split now.
5719 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5720 file descriptors might be low, and that setting 1000 would always raise it.
5721 It turns out that in some environments, the limit is already over 1000 and
5722 that lowering it causes trouble. So now Exim takes care not to decrease it.
5724 40. When delivering a message, the value of $return_path is set to $sender_
5725 address at the start of routing (routers may change the value). By an
5726 oversight, this default was not being set up when an address was tested by
5727 -bt or -bv, which affected the outcome if any router or filter referred to
5730 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5731 log only when "message" or "log_message" are set. However, if one of the
5732 conditions was an address verification, or a call to a nested ACL, the
5733 messages generated by the underlying test were being passed through. This
5734 no longer happens. The underlying message is available in $acl_verify_
5735 message for both "message" and "log_message" expansions, so it can be
5736 passed through if needed.
5738 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5739 new expansion $bh_ to give the encoded byte string without charset
5740 translation. Translation happens only if iconv() is available; HAVE_ICONV
5741 indicates this at build time. HEADERS_CHARSET gives the charset to
5742 translate to; headers_charset can change it in the configuration, and
5743 "headers charset" can change it in an individual filter file.
5745 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5746 that creates RFC 2047 encoded "words" labels them as that charset instead
5747 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5748 expansion operator; (ii) when Exim creates a From: line for a local
5749 message; (iii) when a header line is rewritten to include a "phrase" part.
5751 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5752 buggy, causing it to skip the first lines of messages whose message ID
5753 ended in 'D'. This would not have bitten before Exim release 4.14, because
5754 message IDs were unlikely to end in 'D' before then. The effect was to have
5755 incorrect size information for certain domains.
5757 45. #include "config.h" was missing at the start of the crypt16.c module. This
5758 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5761 46. If there was a timeout during a "random" callout check, Exim treated it as
5762 a failure of the random address, and carried on sending RSET and the real
5763 address. If the delay was just some slowness somewhere, the response to the
5764 original RCPT would be taken as a response to RSET and so on, causing
5765 mayhem of various kinds.
5767 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5768 when I implemented it. It didn't allow for the fact that some option values
5769 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5770 the right test for positive values.
5772 48. Domain names in DNS records are case-independent. Exim always looks them up
5773 in lower case. Some resolvers return domain names in exactly the case they
5774 appear in the zone file, that is, they may contain uppercase letters. Not
5775 all resolvers do this - some return always lower case. Exim was treating a
5776 change of case by a resolver as a change of domain, similar to a widening
5777 of a domain abbreviation. This triggered its re-routing code and so it was
5778 trying to route what was effectively the same domain again. This normally
5779 caused routing to fail (because the router wouldn't handle the domain
5780 twice). Now Exim checks for this case specially, and just changes the
5781 casing of the domain that it ultimately uses when it transmits the message
5784 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5787 50. If a filter generated a file delivery with a non-absolute name (possible if
5788 no home directory exists for the router), the forbid_file option was not
5791 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5792 the existing equality matching.
5794 52. Exim was using ints instead of ino_t variables in some places where it was
5795 dealing with inode numbers.
5797 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5798 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5799 called TMPDIR, and if it finds it is different, it changes its value.
5801 54. The smtp_printf() function is now made available to local_scan() so
5802 additional output lines can be written before returning. There is also an
5803 smtp_fflush() function to enable the detection of a dropped connection.
5804 The variables smtp_input and smtp_batched_input are exported to
5807 55. Changed the default runtime configuration: the message "Unknown user"
5808 has been removed from the ACL, and instead placed on the localuser router,
5809 using the cannot_route_message feature. This means that any verification
5810 failures that generate their own messages won't get overridden. Similarly,
5811 the "Unrouteable address" message that was in the ACL for unverifiable
5812 relay addresses has also been removed.
5814 56. Added hosts_avoid_esmtp to the smtp transport.
5816 57. The exicyclog script was not checking for the esoteric option
5817 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5818 will work only if exicyclog is run under the appropriate euid.
5820 58. Following a discussion on the list, the rules by which Exim recognises line
5821 endings on incoming messages have been changed. The -dropcr and drop_cr
5822 options are now no-ops, retained only for backwards compatibility. The
5823 following line terminators are recognized: LF CRLF CR. However, special
5824 processing applies to CR:
5826 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5827 nor a local message in the state where . is a terminator.
5829 (ii) If a bare CR is encountered in a header line, an extra space is added
5830 after the line terminator so as not to end the header. The reasoning
5831 behind this is that bare CRs in header lines are most likely either
5832 to be mistakes, or people trying to play silly games.
5834 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5835 was being incorrectly given as 18 bytes larger than it should have been.
5836 This is a VOB (very old bug).
5838 60. This may never have affected anything current, but just in case it has:
5839 When the local host is found other than at the start of a list of hosts,
5840 the local host, those with the same MX, and any that follow, are discarded.
5841 When the list in question was part of a longer list of hosts, the following
5842 hosts (not currently being processed) were also being discarded. This no
5843 longer happens. I'm not sure if this situation could ever has previously
5846 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5849 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5850 header. This is something that is recommended in a new Internet Draft, and
5851 is something that is documented as being done by Sendmail. There are two
5852 possible values. For messages generated by the autoreply transport, Exim
5855 Auto-Submitted: auto-replied
5857 whereas for all other generated messages (e.g. bounces) it adds
5859 Auto-Submitted: auto-generated
5861 63. The "personal" condition in filters now includes a test for the
5862 Auto-Submitted: header. If it contains the string "auto-" the message it
5863 not considered personal.
5865 64. Added rcpt_include_affixes as a generic transport option.
5867 65. Added queue_only_override (default true).
5869 66. Added the syslog_duplication option.
5871 67. If what should have been the first header line of a message consisted of
5872 a space followed by a colon, Exim was mis-interpreting it as a header line.
5873 It isn't of course - it is syntactically invalid and should therefore be
5874 treated as the start of the message body. The misbehaviour could have
5875 caused a number of strange effects, including loss of data in subsequent
5876 header lines, and spool format errors.
5878 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5879 client host had authenticated. This control can now be exercised by an ACL
5880 for more flexibility.
5882 69. By default, callouts do not happen when testing with -bh. There is now a
5883 variant, -bhc, which does actually run the callout code, including
5884 consulting and updating the callout cache.
5886 70. Added support for saslauthd authentication, courtesy of Alexander
5889 71. If statvfs() failed on the spool or log directories while checking their
5890 size for availability, Exim confusingly gave the error "space shortage".
5891 Furthermore, in debugging mode it crashed with a floating point exception.
5892 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5893 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5894 serious problem, Exim now writes to the main and panic logs when this
5895 happens, with details of the failure. It then refuses to accept the
5896 incoming message, giving the message "spool directory problem" or "log
5897 directory problem" with a 421 code for SMTP messages.
5899 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5900 0, 1, and 2 exist, because some OS complain for execs without them (see
5901 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5902 descriptors. However, the code omitted to check that the open succeeded,
5903 causing mysterious errors if for some reason the permissions on /dev/null
5904 got screwed. Now Exim writes a message to the main and panic logs, and
5905 bombs out if it can't open /dev/null.
5907 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5908 interact so that it is all more flexible. It is supposed to remain
5909 backwards compatible. Also added extra_local_interfaces.
5911 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5912 to bomb out with an assertion failure - to the client this appears as a
5913 connection drop. This problem occurs in the part of the code that was taken
5914 from the Samba project. Fortunately, the assertion is in a very simple
5915 function, so I have fixed this by reproducing the function inline in the
5916 one place where it is called, and arranging for authentication to fail
5917 instead of killing the process with assert().
5919 75. The SPA client code was not working when the server requested OEM rather
5920 than Unicode encoding.
5922 76. Added code to make require_files with a specific uid setting more usable in
5923 the case where statting the file as root fails - usually a non-root-mounted
5924 NFS file system. When this happens and the failure is EACCES, Exim now
5925 forks a subprocess and does the per-uid checking as the relevant uid.
5927 77. Added process_log_path.
5929 78. If log_file_path was not explicitly set, a setting of check_log_space or
5930 check_log_inodes was ignored.
5932 79. If a space check for the spool or log partitions fails, the incident is now
5933 logged. Of course, in the latter case the data may get lost...
5935 80. Added the %p formatting code to string_format() so that it can be used to
5936 print addresses in debug_print(). Adjusted all the address printing in the
5937 debugging in store.c to use %p rather than %d.
5939 81. There was a concern that a line of code in smtp_in.c could overflow a
5940 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5941 initially expressed, the concern was not well-founded, because trailing
5942 spaces are removed early. However, if the trailing spaces were followed by
5943 a NULL, they did not get removed, so the overflow was possible. Two fixes
5946 (a) I re-wrote the offending code in a cleaner fashion.
5947 (b) If an incoming SMTP command contains a NULL character, it is rejected
5950 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5951 runs initgroups(), so that if the Exim user is in any additional groups,
5952 they will be used during message reception.
5958 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob