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.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
49 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
56 PP/02 Make -n do something, by making it not do something.
57 When combined with -bP, the name of an option is not output.
59 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
62 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
63 $sender_host_name and config options to manage this, and basic check
66 PP/05 DSCP support for outbound connections and control modifier for inbound.
68 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
69 (Only plugin which currently uses this is kerberos4, which nobody should
70 be using, but we should make it available and other future plugins might
71 conceivably use it, even though it would break NAT; stuff *should* be
72 using channel bindings instead).
74 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
75 name; added for Sendmail compatibility; requires admin caller.
76 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
77 just ignore it); requires trusted caller.
78 Also parse but ignore: -Ac -Am -X<logfile>
81 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
83 TL/02 Add +smtp_confirmation as a default logging option.
85 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
86 Patch by Magnus Holmgren from 2007-02-20.
88 TL/04 Bugzilla 1281 - Spec typo.
89 Bugzilla 1283 - Spec typo.
90 Bugzilla 1290 - Spec grammar fixes.
92 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
94 TL/06 Add Experimental DMARC support using libopendmarc libraries.
96 TL/07 Fix an out of order global option causing a segfault. Reported to dev
97 mailing list by by Dmitry Isaikin.
99 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
101 JH/02 Support "G" suffix to numbers in ${if comparisons.
103 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
105 NM/01 Bugzilla 1197 - Spec typo
106 Bugzilla 1196 - Spec examples corrections
108 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
110 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
111 gnutls_enable_pkcs11, but renamed to more accurately indicate its
114 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
115 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
117 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
118 "acl {{name}{arg}...}", and optional args on acl condition
121 JH/05 Permit multiple router/transport headers_add/remove lines.
123 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
125 JH/07 Avoid using a waiting database for a single-message-only transport.
126 Performance patch from Paul Fisher. Bugzilla 1262.
128 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
131 JH/09 Add $headers_added variable, with content from use of ACL modifier
132 add_header (but not yet added to the message). Bugzilla 199.
134 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
135 Pulled from Bugzilla 817 by Wolfgang Breyha.
137 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
139 (nb: this is the same fix as in Exim 4.80.1)
141 JH/11 Add A= logging on delivery lines, and a client_set_id option on
144 JH/12 Add optional authenticated_sender logging to A= and a log_selector
147 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
149 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
150 advertise SMTP AUTH mechanism to us, instead of a generic
151 protocol violation error. Also, make Exim more robust to bad
152 data from the Dovecot auth socket.
154 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
156 When a queue runner is handling a message, Exim first routes the
157 recipient addresses, during which it prunes them based on the retry
158 hints database. After that it attempts to deliver the message to
159 any remaining recipients. It then updates the hints database using
162 So if a recipient address works intermittently, it can get repeatedly
163 deferred at routing time. The retry hints record remains fresh so the
164 address never reaches the final cutoff time.
166 This is a fairly common occurrence when a user is bumping up against
167 their storage quota. Exim had some logic in its local delivery code
168 to deal with this. However it did not apply to per-recipient defers
169 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
171 This change adds a proper retry rule check during routing so that the
172 final cutoff time is checked against the message's age. We only do
173 this check if there is an address retry record and there is not a
174 domain retry record; this implies that previous attempts to handle
175 the address had the retry_use_local_parts option turned on. We use
176 this as an approximation for the destination being like a local
177 delivery, as in LMTP.
179 I suspect this new check makes the old local delivery cutoff check
180 redundant, but I have not verified this so I left the code in place.
182 TF/02 Correct gecos expansion when From: is a prefix of the username.
184 Test 0254 submits a message to Exim with the header
188 When I ran the test suite under the user fanf2, Exim expanded
189 the header to contain my full name, whereas it should have added
190 a Resent-Sender: header. It erroneously treats any prefix of the
191 username as equal to the username.
193 This change corrects that bug.
195 GF/01 DCC debug and logging tidyup
196 Error conditions log to paniclog rather than rejectlog.
197 Debug lines prefixed by "DCC: " to remove any ambiguity.
199 TF/03 Avoid unnecessary rebuilds of lookup-related code.
201 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
202 Bug spotted by Jeremy Harris; was flawed since initial commit.
203 Would have resulted in OCSP responses post-SNI triggering an Exim
204 NULL dereference and crash.
206 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
208 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
209 Bug detection, analysis and fix by Samuel Thibault.
210 Bugzilla 1331, Debian bug #698092.
212 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
214 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
215 Server implementation by Todd Lyons, client by JH.
216 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
217 config variable "prdr_enable" controls whether the server
218 advertises the facility. If the client requests PRDR a new
219 acl_data_smtp_prdr ACL is called once for each recipient, after
220 the body content is received and before the acl_smtp_data ACL.
221 The client is controlled by bolth of: a hosts_try_prdr option
222 on the smtp transport, and the server advertisement.
223 Default client logging of deliveries and rejections involving
224 PRDR are flagged with the string "PRDR".
226 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
227 fclose(). Diagnosis by Todd Lyons.
229 PP/17 Update configure.default to handle IPv6 localhost better.
230 Patch by Alain Williams (plus minor tweaks).
233 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
234 This is now consistent with GnuTLS, and is now documented: the
235 previous undocumented portable approach to treating the option as
236 unset was to force an expansion failure. That still works, and
237 an empty string is now equivalent.
239 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
240 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
241 not performing validation itself.
243 PP/20 Added force_command boolean option to pipe transport.
244 Patch from Nick Koston, of cPanel Inc.
246 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
249 TF/04 Added udpsend ACL modifer and hexquote expansion operator
251 PP/21 Fix eximon continuous updating with timestamped log-files.
252 Broken in a format-string cleanup in 4.80, missed when I repaired the
253 other false fix of the same issue.
254 Report and fix from Heiko Schlichting.
257 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
258 Report from Prashanth Katuri.
260 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
261 It's SecureTransport, so affects any MacOS clients which use the
262 system-integrated TLS libraries, including email clients.
264 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
265 using a MIME ACL for non-SMTP local injection.
266 Report and assistance in diagnosis by Warren Baker.
268 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
270 JH/16 Fix comparisons for 64b. Bugzilla 1385.
272 TL/09 Add expansion variable $authenticated_fail_id to keep track of
273 last id that failed so it may be referenced in subsequent ACL's.
275 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
278 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
279 ldap library initialization, allowing self-signed CA's to be
280 used. Also properly sets require_cert option later in code by
281 using NULL (global ldap config) instead of ldap handle (per
282 session). Bug diagnosis and testing by alxgomz.
284 TL/12 Enhanced documentation in the ratelimit.pl script provided in
285 the src/util/ subdirectory.
287 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
288 renamed to Transport Post Delivery Action by Jeremy Harris, as
291 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
292 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
293 redis_servers = needs to be configured which will be used by the redis
294 lookup. Patch from Warren Baker, of The Packet Hub.
296 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
298 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
299 hostname or reverse DNS when processing a host list. Used suggestions
300 from multiple comments on this bug.
302 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
304 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
305 Missed a few lines, added it to make the runtest require no keyboard
308 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
309 contains upper case chars. Make router use caseful_local_part.
311 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
312 support when GnuTLS has been built with p11-kit.
318 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
320 This, or similar/improved, will also be change PP/11 of 4.82.
326 PP/01 Handle short writes when writing local log-files.
327 In practice, only affects FreeBSD (8 onwards).
328 Bugzilla 1053, with thanks to Dmitry Isaikin.
330 NM/01 Bugzilla 949 - Documentation tweak
332 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
335 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
337 PP/02 Implemented gsasl authenticator.
339 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
341 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
342 `pkg-config foo` for cflags/libs.
344 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
345 with rest of GSASL and with heimdal_gssapi.
347 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
348 `pkg-config foo` for cflags/libs for the TLS implementation.
350 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
351 properties get this fed in as external SSF. A number of robustness
352 and debugging improvements to the cyrus_sasl authenticator.
354 PP/08 cyrus_sasl server now expands the server_realm option.
356 PP/09 Bugzilla 1214 - Log authentication information in reject log.
357 Patch by Jeremy Harris.
359 PP/10 Added dbmjz lookup type.
361 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
363 PP/12 MAIL args handles TAB as well as SP, for better interop with
364 non-compliant senders.
365 Analysis and variant patch by Todd Lyons.
367 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
368 Bug report from Lars Müller <lars@samba.org> (via SUSE),
369 Patch from Dirk Mueller <dmueller@suse.com>
371 PP/13 tls_peerdn now print-escaped for spool files.
372 Observed some $tls_peerdn in wild which contained \n, which resulted
373 in spool file corruption.
375 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
376 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
377 or write after TLS renegotiation, which otherwise led to messages
380 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
381 as a tracking header (ie: a signed header comes before the signature).
382 Patch from Wolfgang Breyha.
384 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
385 comma-sep list; embedded commas doubled.
387 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
389 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
391 Report and patch from Dmitry Banschikov.
393 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
394 Removed SSL_clear() after SSL_new() which led to protocol negotiation
395 failures. We appear to now support TLS1.1+ with Exim.
397 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
398 lets Exim select keys and certificates based upon TLS SNI from client.
399 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
400 before an outbound SMTP session. New log_selector, +tls_sni.
402 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
403 NULL dereference. Report and patch from Alun Jones.
405 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
406 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
407 Not seeing resolver debug output on NetBSD, but suspect this is a
408 resolver implementation change.
410 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
411 Left warnings. Added "eximon gdb" invocation mode.
413 PP/21 Defaulting "accept_8bitmime" to true, not false.
415 PP/22 Added -bw for inetd wait mode support.
417 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
418 locate the relevant includes and libraries. Made this the default.
420 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
421 Bugzilla 1246, report and most of solution from Tomasz Kusy.
423 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
424 This may cause build issues on older platforms.
426 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
427 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
428 gnutls_require_mac & gnutls_require_protocols (no longer supported).
429 Added SNI support via GnuTLS too.
430 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
432 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
434 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
435 Applied second patch from Janne, implementing suggestion to default
436 multiple-strings-in-record handling to match SPF spec.
438 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
440 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
441 read-only, out of scope).
442 Patch from Wolfgang Breyha, report from Stuart Northfield.
444 PP/29 Fix three issues highlighted by clang analyser static analysis.
445 Only crash-plausible issue would require the Cambridge-specific
446 iplookup router and a misconfiguration.
447 Report from Marcin Mirosław.
449 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
451 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
452 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
453 As part of this, removing so much warning spew let me fix some minor
454 real issues in debug logging.
456 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
457 assignment on my part. Fixed.
459 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
460 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
461 Janne Snabb (who went above and beyond: thank you).
463 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
464 string otherwise requires a connection and a bunch more work and it's
465 relatively easy to get wrong. Should also expose TLS library linkage
468 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
469 64-bit ${eval} (JH/03).
471 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
472 GNU libc to support some of the 64-bit stuff, should not lead to
473 conflicts. Defined before os.h is pulled in, so if a given platform
474 needs to override this, it can.
476 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
477 protection layer was required, which is not implemented.
478 Bugzilla 1254, patch from Wolfgang Breyha.
480 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
481 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
482 tls_dhparam take prime identifiers. Also unbreak combination of
483 OpenSSL+DH_params+TLSSNI.
485 PP/39 Disable SSLv2 by default in OpenSSL support.
491 PP/01 Solaris build fix for Oracle's LDAP libraries.
492 Bugzilla 1109, patch from Stephen Usher.
494 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
496 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
499 TF/02 Fix a couple more cases where we did not log the error message
500 when unlink() failed. See also change 4.74-TF/03.
502 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
503 lock up or crash if it happened to be inside a call to libc when it
504 got a SIGUSR1 from exiwhat.
506 The SIGUSR1 handler appends the current process status to the process
507 log which is later printed by exiwhat. It used to use the general
508 purpose logging code to do this, but several functions it calls are
509 not safe for signals.
511 The new output code in the SIGUSR1 handler is specific to the process
512 log, and simple enough that it's easy to inspect for signal safety.
513 Removing some special cases also simplifies the general logging code.
514 Removing the spurious timestamps from the process log simplifies
517 TF/04 Improved ratelimit ACL condition.
519 The /noupdate option has been deprecated in favour of /readonly which
520 has clearer semantics. The /leaky, /strict, and /readonly update modes
521 are mutually exclusive. The update mode is no longer included in the
522 database key; it just determines when the database is updated. (This
523 means that when you upgrde Exim will forget old rate measurements.)
525 Exim now checks that the per_* options are used with an update mode that
526 makes sense for the current ACL. For example, when Exim is processing a
527 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
528 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
529 must specify per_mail/readonly. If you omit the update mode it defaults to
530 /leaky where that makes sense (as before) or /readonly where required.
532 The /noupdate option is now undocumented but still supported for
533 backwards compatibility. It is equivalent to /readonly except that in
534 ACLs where /readonly is required you may specify /leaky/noupdate or
535 /strict/noupdate which are treated the same as /readonly.
537 A useful new feature is the /count= option. This is a generalization
538 of the per_byte option, so that you can measure the throughput of other
539 aggregate values. For example, the per_byte option is now equivalent
540 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
542 The per_rcpt option has been generalized using the /count= mechanism
543 (though it's more complicated than the per_byte equivalence). When it is
544 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
545 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
546 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
547 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
548 Note that using per_rcpt with a non-readonly update mode in more than
549 one ACL will cause the recipients to be double-counted. (The per_mail
550 and per_byte options don't have this problem.)
552 The handling of very low rates has changed slightly. If the computed rate
553 is less than the event's count (usually one) then this event is the first
554 after a long gap. In this case the rate is set to the same as this event's
555 count, so that the first message of a spam run is counted properly.
557 The major new feature is a mechanism for counting the rate of unique
558 events. The new per_addr option counts the number of different
559 recipients that someone has sent messages to in the last time period. It
560 behaves like per_rcpt if all the recipient addresses are different, but
561 duplicate recipient addresses do not increase the measured rate. Like
562 the /count= option this is a general mechanism, so the per_addr option
563 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
564 example, measure the rate that a client uses different sender addresses
565 with the options per_mail/unique=$sender_address. There are further
566 details in the main documentation.
568 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
570 TF/06 Removed a few PCRE remnants.
572 TF/07 Automatically extract Exim's version number from tags in the git
573 repository when doing development or release builds.
575 PP/02 Raise smtp_cmd_buffer_size to 16kB.
576 Bugzilla 879. Patch from Paul Fisher.
578 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
579 Heavily based on revision 40f9a89a from Simon Arlott's tree.
582 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
584 PP/05 Variable $av_failed, true if the AV scanner deferred.
585 Bugzilla 1078. Patch from John Horne.
587 PP/06 Stop make process more reliably on build failure.
588 Bugzilla 1087. Patch from Heiko Schlittermann.
590 PP/07 Make maildir_use_size_file an _expandable_ boolean.
591 Bugzilla 1089. Patch from Heiko Schlittermann.
593 PP/08 Handle ${run} returning more data than OS pipe buffer size.
594 Bugzilla 1131. Patch from Holger Weiß.
596 PP/09 Handle IPv6 addresses with SPF.
597 Bugzilla 860. Patch from Wolfgang Breyha.
599 PP/10 GnuTLS: support TLS 1.2 & 1.1.
601 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
604 PP/11 match_* no longer expand right-hand-side by default.
605 New compile-time build option, EXPAND_LISTMATCH_RHS.
606 New expansion conditions, "inlist", "inlisti".
608 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
610 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
612 PP/14 fix log_write() format string regression from TF/03.
613 Bugzilla 1152. Patch from Dmitry Isaikin.
619 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
621 PP/02 Harmonised TLS library version reporting; only show if debugging.
622 Layout now matches that introduced for other libraries in 4.74 PP/03.
624 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
626 PP/04 New "dns_use_edns0" global option.
628 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
631 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
632 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
634 TK/01 Updated PolarSSL code to 0.14.2.
635 Bugzilla 1097. Patch from Andreas Metzler.
637 PP/07 Catch divide-by-zero in ${eval:...}.
640 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
643 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
644 format-string attack -- SECURITY: remote arbitrary code execution.
646 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
647 time unintentionally subject to list matching rules, letting the header
648 cause arbitrary Exim lookups (of items which can occur in lists, *not*
649 arbitrary string expansion). This allowed for information disclosure.
651 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
652 INT_MIN/-1 -- value coerced to INT_MAX.
658 NM/01 Workround for PCRE version dependency in version reporting
661 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
662 This fixes portability to compilers other than gcc, notably
663 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
665 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
666 makefiles for portability to HP-UX and POSIX correctness.
668 PP/01 Permit LOOKUP_foo enabling on the make command-line.
669 Also via indented variable definition in the Makefile.
670 (Debugging by Oliver Heesakkers).
672 PP/02 Restore caching of spamd results with expanded spamd_address.
673 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
675 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
676 Improves build reliability. Fix from: Frank Elsner
678 NM/02 Fix wide character breakage in the rfc2047 coding
679 Fixes bug 1064. Patch from Andrey N. Oktyabrski
681 NM/03 Allow underscore in dnslist lookups
682 Fixes bug 1026. Patch from Graeme Fowler
684 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
685 Code patches from Adam Ciarcinski of NetBSD.
687 NM/04 Fixed exiqgrep to cope with mailq missing size issue
690 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
691 is logged, to avoid truncation. Patch from John Horne.
693 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
694 Patch from Jakob Hirsch.
696 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
697 SQL string expansion failure details.
698 Patch from Andrey Oktyabrski.
700 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
701 Patch from Simon Arlott.
703 PP/09 New lookups functionality failed to compile on old gcc which rejects
704 extern declarations in function scope.
705 Patch from Oliver Fleischmann
707 PP/10 Use sig_atomic_t for flags set from signal handlers.
708 Check getgroups() return and improve debugging.
709 Fixed developed for diagnosis in bug 927 (which turned out to be
712 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
713 Patch from Mark Zealey.
715 PP/12 Bugzilla 1056: Improved spamd server selection.
716 Patch from Mark Zealey.
718 PP/13 Bugzilla 1086: Deal with maildir quota file races.
719 Based on patch from Heiko Schlittermann.
721 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
722 Patch from Uwe Doering, sign-off by Michael Haardt.
724 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
725 variable declaration deep within a block. Bug and patch from
728 PP/15 lookups-Makefile IRIX compatibilty coercion.
730 PP/16 Make DISABLE_DKIM build knob functional.
732 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
733 Patch by Simon Arlott
735 TF/03 Fix valgrind.h portability to C89 compilers that do not support
736 variable argument macros. Our copy now differs from upstream.
742 TF/01 Failure to get a lock on a hints database can have serious
743 consequences so log it to the panic log.
745 TF/02 Log LMTP confirmation messages in the same way as SMTP,
746 controlled using the smtp_confirmation log selector.
748 TF/03 Include the error message when we fail to unlink a spool file.
750 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
751 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
752 for maintaining out-of-tree patches for some time.
754 PP/01 Bugzilla 139: Documentation and portability issues.
755 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
756 Handle per-OS dynamic-module compilation flags.
758 PP/02 Let /dev/null have normal permissions.
759 The 4.73 fixes were a little too stringent and complained about the
760 permissions on /dev/null. Exempt it from some checks.
761 Reported by Andreas M. Kirchwitz.
763 PP/03 Report version information for many libraries, including
764 Exim version information for dynamically loaded libraries. Created
765 version.h, now support a version extension string for distributors
766 who patch heavily. Dynamic module ABI change.
768 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
769 privilege escalation vulnerability whereby the Exim run-time user
770 can cause root to append content of the attacker's choosing to
773 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
776 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
777 If dropping privileges for untrusted macros, we disabled normal logging
778 on the basis that it would fail; for the Exim run-time user, this is not
779 the case, and it resulted in successful deliveries going unlogged.
780 Fixed. Reported by Andreas Metzler.
786 PP/01 Date: & Message-Id: revert to normally being appended to a message,
787 only prepend for the Resent-* case. Fixes regression introduced in
788 Exim 4.70 by NM/22 for Bugzilla 607.
790 PP/02 Include check_rfc2047_length in configure.default because we're seeing
791 increasing numbers of administrators be bitten by this.
793 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
795 PP/03 Bugzilla 994: added openssl_options main configuration option.
797 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
799 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
801 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
803 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
804 without a peer certificate, leading to a segfault because of an
805 assumption that peers always have certificates. Be a little more
806 paranoid. Problem reported by Martin Tscholak.
808 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
809 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
810 NB: ClamAV planning to remove STREAM in "middle of 2010".
811 CL also introduces -bmalware, various -d+acl logging additions and
812 more caution in buffer sizes.
814 PP/09 Implemented reverse_ip expansion operator.
816 PP/10 Bugzilla 937: provide a "debug" ACL control.
818 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
820 PP/12 Bugzilla 973: Implement --version.
822 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
824 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
826 PP/15 Bugzilla 816: support multiple condition rules on Routers.
828 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
829 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
830 ignore trailing whitespace.
832 JJ/02 prevent non-panic DKIM error from being sent to paniclog
834 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
837 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
838 Notification from Dr Andrew Aitchison.
840 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
841 ExtendedDetectionInfo response format.
842 Notification from John Horne.
844 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
847 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
848 XSL and documented dependency on system catalogs, with examples of how
851 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
854 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
855 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
856 configuration file which is writeable by the Exim user or group.
858 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
859 of configuration files to cover files specified with the -C option if
860 they are going to be used with root privileges, not just the default
863 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
864 option (effectively making it always true).
866 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
867 files to be used while preserving root privileges.
869 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
870 that rogue child processes cannot use them.
872 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
873 run-time user, instead of root.
875 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
876 Exim run-time user without dropping privileges.
878 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
879 result string, instead of calling string_vformat() twice with the same
882 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
883 for other users. Others should always drop root privileges if they use
884 -C on the command line, even for a whitelisted configure file.
886 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
888 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
894 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
895 $data_path, and $header_path variables; fixed documentation bugs and
898 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
899 exipick to access non-standard spools, including the "frozen" queue
902 NM/01 Bugzilla 965: Support mysql stored procedures.
903 Patch from Alain Williams
905 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
907 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
908 Patch from Andreas Metzler
910 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
911 Patch from Kirill Miazine
913 NM/05 Bugzilla 671: Added umask to procmail example.
915 JJ/03 installed exipick 20100323.0, fixing doc bug
917 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
918 directory. Notification and patch from Dan Rosenberg.
920 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
922 TK/02 Improve log output when DKIM signing operation fails.
924 MH/01 Treat the transport option dkim_domain as a colon separated
925 list, not as a single string, and sign the message with each element,
926 omitting multiple occurences of the same signer.
928 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
929 Bugzilla 985, 986. Patch by Simon Arlott
931 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
932 Patch by Simon Arlott
934 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
935 MBX locking. Notification from Dan Rosenberg.
941 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
943 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
945 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
947 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
949 NM/04 Bugzilla 915: Flag broken perl installation during build.
955 TK/01 Added patch by Johannes Berg that expands the main option
956 "spamd_address" if it starts with a dollar sign.
958 TK/02 Write list of recipients to X-Envelope-Sender header when building
959 the mbox-format spool file for content scanning (suggested by Jakob
962 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
963 (http://www.dcc-servers.net/) support via dccifd. Activated by
964 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
966 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
967 by Mark Daniel Reidel <mr@df.eu>.
969 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
970 When building exim an external PCRE library is now needed -
971 PCRE is a system library on the majority of modern systems.
972 See entry on PCRE_LIBS in EDITME file.
974 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
975 conversation. Added nologin parameter to request.
976 Patch contributed by Kirill Miazine.
978 TF/01 Do not log submission mode rewrites if they do not change the address.
980 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
982 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
983 log files in place. Contributed by Roberto Lima.
985 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
987 TF/03 Bugzilla 615: When checking the local_parts router precondition
988 after a local_part_suffix or local_part_prefix option, Exim now
989 does not use the address's named list lookup cache, since this
990 contains cached lookups for the whole local part.
992 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
993 Robert Millan. Documentation is in experimental-spec.txt.
995 TF/04 Bugzilla 668: Fix parallel build (make -j).
997 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
999 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1000 Patch provided by Jan Srzednicki.
1002 TF/05 Leading white space used to be stripped from $spam_report which
1003 wrecked the formatting. Now it is preserved.
1005 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1006 that they are available at delivery time.
1008 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1010 TF/08 TLS error reporting now respects the incoming_interface and
1011 incoming_port log selectors.
1013 TF/09 Produce a more useful error message if an SMTP transport's hosts
1014 setting expands to an empty string.
1016 NM/06 Bugzilla 744: EXPN did not work under TLS.
1017 Patch provided by Phil Pennock.
1019 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1020 Patch provided by Richard Godbee.
1022 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1023 acl_smtp_notquit, added index entry.
1025 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1026 Patch provided by Eugene Bujak.
1028 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1029 accept(). Patch provided by Maxim Dounin.
1031 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1032 Patch provided by Phil Pennock.
1034 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1036 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1037 Patch provided by Brad "anomie" Jorsch.
1039 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1040 Patch provided by Dean Brooks.
1042 TK/05 Add native DKIM support (does not depend on external libraries).
1044 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1045 Patch provided by Graeme Fowler.
1047 NM/16 Bugzilla 851: Documentation example syntax fix.
1049 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1051 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1054 NM/19 Bugzilla 745: TLS version reporting.
1055 Patch provided by Phil Pennock.
1057 NM/20 Bugzilla 167: bool: condition support.
1058 Patch provided by Phil Pennock.
1060 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1061 clients. Patch provided by Phil Pennock.
1063 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1064 Patch provided by Brad "anomie" Jorsch.
1066 NM/23 Bugzilla 687: Fix misparses in eximstats.
1067 Patch provided by Heiko Schlittermann.
1069 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1070 Patch provided by Heiko Schlittermann.
1072 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1073 plus update to original patch.
1075 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1077 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1078 Patch provided by David Brownlee.
1080 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1082 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1084 NM/30 Bugzilla 888: TLS documentation bugfixes.
1086 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1088 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1089 Unlike the original bugzilla I have changed all shell scripts in src tree.
1091 NM/33 Bugzilla 898: Transport filter timeout fix.
1092 Patch by Todd Rinaldo.
1094 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1095 Patch by Serge Demonchaux.
1097 NM/35 Bugzilla 39: Base64 decode bug fixes.
1098 Patch by Jakob Hirsch.
1100 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1102 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1104 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1106 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1112 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1113 ALT-N's libdkim that I have put here:
1114 http://duncanthrax.net/exim-experimental/
1116 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1117 were called 'true' and it seems that C99 defines that as a reserved
1118 keyword to be used with 'bool' variable types. That means you could
1119 not include C99-style headers which use bools without triggering
1120 build errors in sieve.c.
1122 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1123 as mailq or other aliases. Changed the --help handling significantly
1124 to do whats expected. exim_usage() emits usage/help information.
1126 SC/01 Added the -bylocaldomain option to eximstats.
1128 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1130 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1132 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1138 PH/01 Another patch from the Sieve maintainer.
1140 PH/02 When an IPv6 address is converted to a string for single-key lookup
1141 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1142 dots are used instead of colons so that keys in lsearch files need not
1143 contain colons. This was done some time before quoting was made available
1144 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1145 (notated using the quote facility) so as to distinguish them from IPv4
1146 keys. This meant that lookups for IP addresses in host lists did not work
1147 for iplsearch lookups.
1149 This has been fixed by arranging for IPv6 addresses to be expressed with
1150 colons if the lookup type is iplsearch. This is not incompatible, because
1151 previously such lookups could never work.
1153 The situation is now rather anomolous, since one *can* have colons in
1154 ordinary lsearch keys. However, making the change in all cases is
1155 incompatible and would probably break a number of configurations.
1157 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1160 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1161 conversion specification without a maximum field width, thereby enabling
1162 a rogue spamd server to cause a buffer overflow. While nobody in their
1163 right mind would setup Exim to query an untrusted spamd server, an
1164 attacker that gains access to a server running spamd could potentially
1165 exploit this vulnerability to run arbitrary code as the Exim user.
1167 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1168 $primary_hostname instead of what libspf2 thinks the hosts name is.
1170 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1171 a directory entry by the name of the lookup key. Previously, if a
1172 symlink pointed to a non-existing file or a file in a directory that
1173 Exim lacked permissions to read, a lookup for a key matching that
1174 symlink would fail. Now it is enough that a matching directory entry
1175 exists, symlink or not. (Bugzilla 503.)
1177 PH/03 The body_linecount and body_zerocount variables are now exported in the
1180 PH/04 Added the $dnslist_matched variable.
1182 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1183 This means they are set thereafter only if the connection becomes
1186 PH/06 Added the client_condition to authenticators so that some can be skipped
1187 by clients under certain conditions.
1189 PH/07 The error message for a badly-placed control=no_multiline_responses left
1190 "_responses" off the end of the name.
1192 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1194 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1195 (without spaces) instead of just copying the configuration text.
1197 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1199 PH/11 Added $max_received_linelength.
1201 PH/12 Added +ignore_defer and +include_defer to host lists.
1203 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1204 way in which PCRE > 7.0 is built.
1206 PH/14 Implemented queue_only_load_latch.
1208 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1209 MAIL command. The effect was to mangle the value on 64-bit systems.
1211 PH/16 Another patch from the Sieve maintainer.
1213 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1215 PH/18 If a system quota error occurred while trying to create the file for
1216 a maildir delivery, the message "Mailbox is full" was not appended to the
1217 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1218 only to a quota excession during the actual writing of the file.
1220 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1221 characters?) which causes problems in log lines. The DN values are now
1222 passed through string_printing() before being added to log lines.
1224 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1225 and InterBase are left for another time.)
1227 PH/21 Added message_body_newlines option.
1229 PH/22 Guard against possible overflow in moan_check_errorcopy().
1231 PH/23 POSIX allows open() to be a macro; guard against that.
1233 PH/24 If the recipient of an error message contained an @ in the local part
1234 (suitably quoted, of course), incorrect values were put in $domain and
1235 $local_part during the evaluation of errors_copy.
1241 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1242 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1245 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1246 issue a MAIL command.
1248 PH/02 In an ACL statement such as
1250 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1252 if a client was not listed at all, or was listed with a value other than
1253 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1254 the condition was not true (as it should be), so access was not denied.
1255 The bug was that the ! inversion was incorrectly passed on to the second
1256 item. This has been fixed.
1258 PH/03 Added additional dnslists conditions == and =& which are different from
1259 = and & when the dns lookup returns more than one IP address.
1261 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1262 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1264 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1265 FSYNC, which compiles an option called disable_fsync that allows for
1266 bypassing fsync(). The documentation is heavily laced with warnings.
1268 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1270 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1271 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1272 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1273 including adding "make clean"; (3) Added -fPIC when compiling the test
1274 dynamically loaded module, to get rid of a warning.
1276 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1277 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1278 = 0s. The bug is otherwise harmless.
1280 PH/07 There was a bug in the dovecot authenticator such that the value of
1281 $auth1 could be overwritten, and so not correctly preserved, after a
1282 successful authentication. This usually meant that the value preserved by
1283 the server_setid option was incorrect.
1285 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1287 PH/09 Installed PCRE release 7.0.
1289 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1290 run for batched SMTP input. It is now run at the start of every message
1291 in the batch. While fixing this I discovered that the process information
1292 (output by running exiwhat) was not always getting set for -bs and -bS
1293 input. This is fixed, and it now also says "batched" for BSMTP.
1295 PH/11 Added control=no_pipelining.
1297 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1298 patch, slightly modified), and move the expansion of helo_data till after
1299 the connection is made in the smtp transport (so it can use these
1302 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1304 PH/14 Added log_selector = +pid.
1306 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1308 PH/16 Add ${if forany and ${if forall.
1310 PH/17 Added dsn_from option to vary the From: line in DSNs.
1312 PH/18 Flush SMTP output before performing a callout, unless control =
1313 no_callout_flush is set.
1315 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1316 was true (the default) a successful delivery failed to delete the retry
1317 item, thus causing premature timeout of the address. The bug is now
1320 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1322 PH/21 Long custom messages for fakedefer and fakereject are now split up
1323 into multiline reponses in the same way that messages for "deny" and
1324 other ACL rejections are.
1326 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1327 with slight modification.
1329 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1330 draft, changing the syntax and factoring some duplicate code".
1332 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1333 for deliveries of the second and subsequent messages over the same SMTP
1336 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1337 ${reduce, with only minor "tidies".
1339 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1341 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1342 expansion side effects.
1344 PH/27 When a message times out after an over-quota error from an Exim-imposed
1345 quota, the bounce message says "mailbox is full". This message was not
1346 being given when it was a system quota that was exceeded. It now should
1349 MH/03 Made $recipients available in local_scan(). local_scan() already has
1350 better access to the recipient list through recipients_list[], but
1351 $recipients can be useful in postmaster-provided expansion strings.
1353 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1354 in the case of a MAIL command with additional options following the
1355 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1356 were accidentally chopped off.
1358 PH/29 SMTP synchronization checks are implemented when a command is read -
1359 there is a check that no more input is waiting when there shouldn't be
1360 any. However, for some commands, a delay in an ACL can mean that it is
1361 some time before the response is written. In this time, more input might
1362 arrive, invalidly. So now there are extra checks after an ACL has run for
1363 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1364 pipelining has not been advertised.
1366 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1368 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1369 returned in the SMTP response when smtp_return_error_details was false.
1370 This has been fixed.
1372 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1373 instead of the C I/O that was originally supplied, because problems were
1374 reported on Solaris.
1376 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1377 Exim which did not show up earlier: it was assuming that a call to
1378 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1379 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1380 was a macro that became an assignment, so it seemed to work. This has
1381 changed to a proper function call with a void return, hence the compile
1382 error. Exim's code has been fixed.
1384 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1387 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1389 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1391 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1392 because it turns out to be too big or there is a timeout) there is no
1393 "Completed" line in the log. When some messages of this type were
1394 selected by exigrep, they were listed as "not completed". Others were
1395 picked up by some special patterns. I have improved the selection
1396 criteria to be more general.
1398 PH/38 The host_find_failed option in the manualroute router can now be set
1399 to "ignore", to completely ignore a host whose IP address cannot be
1400 found. If all hosts are ignored, the behaviour is controlled by the new
1401 host_all_ignored option.
1403 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1404 homing or because of multiple MX records with /mx) generated more than
1405 one IP address, and the following item turned out to be the local host,
1406 all the secondary addresses of the first item were incorrectly removed
1407 from the list, along with the local host and any following hosts (which
1408 is what is supposed to happen).
1410 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1411 whoever called Exim into the -H file. In the case of the daemon it was
1412 behaving confusingly. When first started, it used values for whoever
1413 started the daemon, but after a SIGHUP it used the Exim user (because it
1414 calls itself on a restart). I have changed the code so that it now always
1417 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1418 message are rejected with the same error (e.g. no authentication or bad
1419 sender address), and a DATA command is nevertheless sent (as can happen
1420 with PIPELINING or a stupid MUA), the error message that was given to the
1421 RCPT commands is included in the rejection of the DATA command. This is
1422 intended to be helpful for MUAs that show only the final error to their
1425 PH/42 Another patch from the Sieve maintainer.
1427 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1428 Eximstats - Fixed some broken HTML links and added missing column headers
1430 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1431 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1433 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1435 PH/43 Yet another patch from the Sieve maintainer.
1437 PH/44 I found a way to check for a TCP/IP connection going away before sending
1438 the response to the final '.' that terminates a message, but only in the
1439 case where the client has not sent further data following the '.'
1440 (unfortunately, this is allowed). However, in many cases there won't be
1441 any further data because there won't be any more messages to send. A call
1442 to select() can be used: if it shows that the input is "ready", there is
1443 either input waiting, or the socket has been closed. An attempt to read
1444 the next input character can distinguish the two cases. Previously, Exim
1445 would have sent an OK response which the client would never have see.
1446 This could lead to message repetition. This fix should cure that, at
1447 least in a lot of common cases.
1449 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1450 advertised in response to EHLO.
1456 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1457 fixed by 4.65/MH/01 (is this a record?) are fixed:
1459 (i) An empty string was always treated as zero by the numeric comparison
1460 operators. This behaviour has been restored.
1462 (ii) It is documented that the numeric comparison operators always treat
1463 their arguments as decimal numbers. This was broken in that numbers
1464 starting with 0 were being interpreted as octal.
1466 While fixing these problems I realized that there was another issue that
1467 hadn't been noticed. Values of message_size_limit (both the global option
1468 and the transport option) were treated as octal if they started with 0.
1469 The documentation was vague. These values are now always treated as
1470 decimal, and I will make that clear in the documentation.
1476 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1477 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1480 MH/01 Don't check that the operands of numeric comparison operators are
1481 integers when their expansion is in "skipping" mode (fixes bug
1482 introduced by 4.64-PH/07).
1484 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1485 child addresses, Exim now panics and dies. Previously, because the count
1486 is held in a short int, deliveries were likely to be lost. As such a
1487 large number of recipients for a single message is ridiculous
1488 (performance will be very, very poor), I have chosen to impose a limit
1489 rather than extend the field.
1495 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1496 leftover -K file (the existence of which was triggered by #402).
1497 While we were at it, introduced process PID as part of the -K
1498 filename. This should rule out race conditions when creating
1501 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1502 processing considerably. Previous code took too long for large mails,
1503 triggering a timeout which in turn triggers #401.
1505 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1506 in the DK code in transports.c. sendfile() is not really portable,
1507 hence the _LINUX specificness.
1509 TF/01 In the add_headers option to the mail command in an Exim filter,
1510 there was a bug that Exim would claim a syntax error in any
1511 header after the first one which had an odd number of characters
1514 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1515 callout verification, Exim cached a "reject" for the entire domain. This
1516 is correct for most verifications, but it is not correct for a recipient
1517 verification with use_sender or use_postmaster set, because in that case
1518 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1519 case of MAIL FROM:<> rejection from other early rejections (e.g.
1520 rejection of HELO). When verifying a recipient using a non-null MAIL
1521 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1522 Whatever the result of the callout, the value of the domain cache is
1523 left unchanged (for any other kind of callout, getting as far as trying
1524 RCPT means that the domain itself is ok).
1526 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1529 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1530 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1531 session, the daemon ignores ECONNECT errors and logs others; it now
1532 ignores EPIPE as well.
1534 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1535 (quoted-printable decoding).
1537 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1538 later the small subsequent patch to fix an introduced bug.
1540 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1542 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1544 PH/08 An error is now given if message_size_limit is specified negative.
1546 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1547 to be given (somewhat) arbitrary names.
1549 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1552 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1553 miscellaneous code fixes
1555 PH/10 Added the log_reject_target ACL modifier to specify where to log
1558 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1559 hostname. This is wrong, because it relates to the incoming message (and
1560 probably the interface on which it is arriving) and not to the outgoing
1561 callout (which could be using a different interface). This has been
1562 changed to use the value of the helo_data option from the smtp transport
1563 instead - this is what is used when a message is actually being sent. If
1564 there is no remote transport (possible with a router that sets up host
1565 addresses), $smtp_active_hostname is used.
1567 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1568 tweaks were necessary in order to get it to work (see also 21 below):
1569 (a) The code assumed that strncpy() returns a negative number on buffer
1570 overflow, which isn't the case. Replaced with Exim's string_format()
1572 (b) There were several signed/unsigned issues. I just did the minimum
1573 hacking in of casts. There is scope for a larger refactoring.
1574 (c) The code used strcasecmp() which is not a standard C function.
1575 Replaced with Exim's strcmpic() function.
1576 (d) The code set only $1; it now sets $auth1 as well.
1577 (e) A simple test gave the error "authentication client didn't specify
1578 service in request". It would seem that Dovecot has changed its
1579 interface. Fortunately there's a specification; I followed it and
1580 changed what the client sends and it appears to be working now.
1582 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1585 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1586 address A is aliased to B and C, where B exists and C does not. Without
1587 -v the output is "A verified" because verification stops after a
1588 successful redirection if more than one address is generated. However,
1589 with -v the child addresses are also verified. Exim was outputting "A
1590 failed to verify" and then showing the successful verification for C,
1591 with its parentage. It now outputs "B failed to verify", showing B's
1592 parentage before showing the successful verification of C.
1594 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1595 look up a TXT record in a specific list after matching in a combined
1598 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1599 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1600 they consult the DNS. I had assumed they would set it the way they
1601 wanted; and indeed my experiments on Linux seem to show that in some
1602 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1603 To be on the safe side, however, I have now made the interface to
1604 host_find_byname() similar to host_find_bydns(), with an argument
1605 containing the DNS resolver options. The host_find_byname() function now
1606 sets these options at its start, just as host_find_bydns() does. The smtp
1607 transport options dns_qualify_single and dns_search_parents are passed to
1608 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1609 of host_find_byname() use the default settings of RES_DEFNAMES
1610 (qualify_single) but not RES_DNSRCH (search_parents).
1612 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1613 spool_read_header() do less string testing, by means of a preliminary
1614 switch on the second character of optional "-foo" lines. (This is
1615 overdue, caused by the large number of possibilities that now exist.
1616 Originally there were few.) While I was there, I also converted the
1617 str(n)cmp tests so they don't re-test the leading "-" and the first
1618 character, in the hope this might squeeze out yet more improvement.
1620 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1621 flag allowing group syntax was set by the header_syntax check but not
1622 turned off, possible causing trouble later; (2) The flag was not being
1623 set at all for the header_verify test, causing "group"-style headers to
1624 be rejected. I have now set it in this case, and also caused header_
1625 verify to ignore an empty address taken from a group. While doing this, I
1626 came across some other cases where the code for allowing group syntax
1627 while scanning a header line wasn't quite right (mostly, not resetting
1628 the flag correctly in the right place). These bugs could have caused
1629 trouble for malformed header lines. I hope it is now all correct.
1631 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1632 with the "reply" argument non-NULL. The code, however (which originally
1633 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1634 but it didn't always do it. This confused somebody who was copying the
1635 code for some other use. I have removed all the tests.
1637 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1638 feature that was used to support insecure browsers during the U.S. crypto
1639 embargo. It requires special client support, and Exim is probably the
1640 only MTA that supported it -- and would never use it because real RSA is
1641 always available. This code has been removed, because it had the bad
1642 effect of slowing Exim down by computing (never used) parameters for the
1643 RSA_EXPORT functionality.
1645 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1646 authenticator to fail if there's a tab character in the incoming data
1647 (there should never be unless someone is messing about, as it's supposed
1648 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1649 if the connection is using TLS or if the remote IP is the same as the
1650 local IP, and the "valid-client-cert option" if a client certificate has
1653 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1654 authenticators. This can be used for authorization after authentication
1655 succeeds. (In the case of plaintext, it servers for both authentication
1658 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1659 if any retry times were supplied.
1661 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1662 connection, where there is no client IP address to check. In this
1663 situation, the verify now always succeeds.
1665 PH/25 Applied John Jetmore's -Mset patch.
1667 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1669 PH/27 In a string expansion for a processed (not raw) header when multiple
1670 headers of the same name were present, leading whitespace was being
1671 removed from all of them, but trailing whitespace was being removed only
1672 from the last one. Now trailing whitespace is removed from each header
1673 before concatenation. Completely empty headers in a concatenation (as
1674 before) are ignored.
1676 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1677 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1679 PH/29 [Removed. This was a change that I later backed out, and forgot to
1680 correct the ChangeLog entry (that I had efficiently created) before
1681 committing the later change.]
1683 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1684 address errors (4xx response to RCPT) over the same connection as other
1685 messages routed to the same hosts. Such deliveries are always "forced",
1686 so retry times are not inspected. This resulted in far too many retries
1687 for the affected addresses. The effect occurred only when there were more
1688 hosts than the hosts_max_try setting in the smtp transport when it had
1689 the 4xx errors. Those hosts that it had tried were not added to the list
1690 of hosts for which the message was waiting, so if all were tried, there
1691 was no problem. Two fixes have been applied:
1693 (i) If there are any address or message errors in an SMTP delivery, none
1694 of the hosts (tried or untried) are now added to the list of hosts
1695 for which the message is waiting, so the message should not be a
1696 candidate for sending over the same connection that was used for a
1697 successful delivery of some other message. This seems entirely
1698 reasonable: after all the message is NOT "waiting for some host".
1699 This is so "obvious" that I'm not sure why it wasn't done
1700 previously. Hope I haven't missed anything, but it can't do any
1701 harm, as the worst effect is to miss an optimization.
1703 (ii) If, despite (i), such a delivery is accidentally attempted, the
1704 routing retry time is respected, so at least it doesn't keep
1705 hammering the server.
1707 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1708 in ${readsocket because some servers need this prod.
1710 PH/32 Added some extra debug output when updating a wait-xxx database.
1712 PH/33 The hint "could be header name not terminated by colon", which has been
1713 given for certain expansion errors for a long time, was not being given
1714 for the ${if def:h_colon_omitted{... case.
1716 PH/34 The spec says: "With one important exception, whenever a domain list is
1717 being scanned, $domain contains the subject domain." There was at least
1718 one case where this was not true.
1720 PH/35 The error "getsockname() failed: connection reset by peer" was being
1721 written to the panic log as well as the main log, but it isn't really
1722 panic-worthy as it just means the connection died rather early on. I have
1723 removed the panic log writing for the ECONNRESET error when getsockname()
1726 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1727 runs only) independently of the message's sender address. This meant
1728 that, if the 4xx error was in fact related to the sender, a different
1729 message to the same recipient with a different sender could confuse
1730 things. In particualar, this can happen when sending to a greylisting
1731 server, but other circumstances could also provoke similar problems.
1732 I have changed the default so that the retry time for these errors is now
1733 based a combination of the sender and recipient addresses. This change
1734 can be overridden by setting address_retry_include_sender=false in the
1737 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1738 remote server are returned as part of bounce messages. This was not
1739 happening for LMTP over a pipe (the lmtp transport), but now it is the
1740 same for both kinds of LMTP.
1742 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1743 in header lines that were in fact CNAMEs. This is no longer the case.
1745 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1746 and queue runs started by the daemon processed all messages. This has
1747 been fixed so that -R and -S can now usefully be given with -q<time>.
1749 PH/40 Import PCRE release 6.7 (fixes some bugs).
1751 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1753 PH/42 Give an error if -q is specified more than once.
1755 PH/43 Renamed the variables $interface_address and $interface_port as
1756 $received_ip_address and $received_port, to make it clear that these
1757 values apply to message reception, and not to the outgoing interface when
1758 a message is delivered. (The old names remain recognized, of course.)
1760 PH/44 There was no timeout on the connect() call when using a Unix domain
1761 socket in the ${readsocket expansion. There now is.
1763 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1764 be meaningful with "accept".
1766 SC/01 Eximstats V1.43
1767 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1769 SC/02 Eximstats V1.44
1770 Use a glob alias rather than an array ref in the generated
1771 parser. This improves both readability and performance.
1773 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1774 Collect SpamAssassin and rejection statistics.
1775 Don't display local sender or destination tables unless
1776 there is data to show.
1777 Added average volumes into the top table text output.
1779 SC/04 Eximstats V1.46
1780 Collect data on the number of addresses (recipients)
1781 as well as the number of messages.
1783 SC/05 Eximstats V1.47
1784 Added 'Message too big' to the list of mail rejection
1785 reasons (thanks to Marco Gaiarin).
1787 SC/06 Eximstats V1.48
1788 Mainlog lines which have GMT offsets and are too short to
1789 have a flag are now skipped.
1791 SC/07 Eximstats V1.49 (Alain Williams)
1792 Added the -emptyok flag.
1794 SC/08 Eximstats V1.50
1795 Fixes for obtaining the IP address from reject messages.
1797 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1798 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1799 whitesspace changes from 4.64-PH/27
1801 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1804 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1805 are found, allow negative numbers in numeric criteria)
1807 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1809 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1810 to match changes made in 4.64-PH/43
1812 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1814 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1815 "domain" is actually an IP address, and force a failure. This locks out
1816 those revolvers/nameservers that support "A-for-A" lookups, in
1817 contravention of the specifications.
1819 PH/48 When a host name was looked up from an IP address, and the subsequent
1820 forward lookup of the name timed out, the host name was left in
1821 $sender_host_name, contrary to the specification.
1823 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1824 restricted to single-key lookups, Exim was not diagnosing an error if
1825 * or *@ was used with a query-style lookup.
1827 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1829 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1830 long ago, but noone interested enough thought of it. Let's just say that
1831 the "1.1" means that there are some new functions that weren't there at
1832 some point in the past.
1834 PH/51 Error processing for expansion failure of helo_data from an smtp
1835 transport during callout processing was broken.
1837 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1838 tested/used via the -bh/-bhc/-bs options.
1840 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1841 bug, fixed in subsequent PCRE releases).
1843 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1844 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1846 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1852 SC/01 Use a glob alias rather than an array ref in eximstats generated
1853 parser. This improves both readability and performance.
1855 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1856 Don't display local sender or destination tables in eximstats unless
1857 there is data to show.
1858 Added average volumes into the eximstats top table text output.
1860 SC/03 Collect data on the number of addresses (recipients) as well
1861 as the number of messages in eximstats.
1863 TF/01 Correct an error in the documentation for the redirect router. Exim
1864 does (usually) call initgroups() when daemonizing.
1866 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1867 with consistent privilege compared to when running as a daemon.
1869 TF/03 Note in the spec that $authenticated_id is not set for local
1870 submissions from trusted users.
1872 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1873 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1875 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1876 by adding some example configuration directives to the default
1877 configuration file. A little bit of work is required to uncomment the
1878 directives and define how usernames and passwords are checked, but
1879 there is now a framework to start from.
1881 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1882 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1883 without this. I don't know how relevant this is to other LDAP libraries.
1885 PH/02 Add the verb name to the "unknown ACL verb" error.
1887 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1889 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1891 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1892 directory not expanded when it should be if an expanded home directory
1893 was set for the address (which is overridden by the transport).
1895 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1898 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1899 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1900 because it is too late at that time, and has no effect.
1902 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1903 security issue with \' (bugzilla #107). I could not use the
1904 PQescapeStringConn() function, because it needs a PGconn value as one of
1907 PH/08 When testing addresses using -bt, indicate those final addresses that
1908 are duplicates that would not cause an additional delivery. At least one
1909 person was confused, thinking that -bt output corresponded to deliveries.
1910 (Suppressing duplicates isn't a good idea as you lose the information
1911 about possibly different redirections that led to the duplicates.)
1913 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1914 systems where poll() doesn't work, in particular OS X.
1916 PH/10 Added more information to debugging output for retry time not reached.
1918 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1919 operations in malware.c.
1921 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1924 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1925 syslog_duplication set false, log lines that would normally be written
1926 both the the main log and to the reject log were not written to syslog at
1929 PH/14 In the default configuration, change the use of "message" in ACL warn
1930 statements to "add_header".
1932 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1933 not followed by a command (e.g. "seen endif").
1935 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1936 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1939 PH/17 Added extra conditions to the default value of delay_warning_condition
1943 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1944 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1945 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1948 The Auto-Submitted: and various List- headers are standardised, whereas I
1949 don't think Precedence: ever was.
1951 PH/18 Refactored debugging code in route_finduser() to show more information,
1952 in particular, the error code if getpwnam() issues one.
1954 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1955 This is apparently needed in addition to the PH/07 change above to avoid
1956 any possible encoding problems.
1958 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1959 but not after initializing Perl.
1961 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1962 output them only if debugging. By default they are written stderr,
1963 apparently, which is not desirable.
1965 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1968 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1971 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1973 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1974 authenticated or an ident call has been made. Suppress the default
1975 values for $authenticated_id and $authenticated_sender (but permit -oMai
1976 and -oMas) when testing with -bh.
1978 PH/24 Re-jigged the order of the tests in the default configuration so that the
1979 tests for valid domains and recipients precede the DNS black list and CSA
1980 tests, on the grounds that those ones are more expensive.
1982 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1983 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1984 This bug exists in every version of Exim that I still have, right back to
1987 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1988 However, an attempt to turn on case-sensitivity in a regex key by
1989 including (?-i) didn't work because the subject string was already
1990 lowercased, and the effects were non-intuitive. It turns out that a
1991 one-line patch can be used to allow (?-i) to work as expected.
1997 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1998 other effects) broke the use of negated acl sub-conditions.
2000 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2003 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2004 "Deny" causes Exim to reject the incoming connection with a 554 error.
2005 Unfortunately, if there is a major crisis, such as a disk failure,
2006 tcp-wrappers gives "deny", whereas what one would like would be some
2007 kind of temporary error. A kludge has been added to help with this.
2008 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2009 554 error is used if errno is still zero or contains ENOENT (which occurs
2010 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2013 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2015 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2016 errors. Otherwise a message that provokes a temporary error (when other
2017 messages do not) can cause a whole host to time out.
2019 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2020 addresses were routed directly to files or pipes from a redirect router.
2021 File deliveries just didn't batch; pipe deliveries might have suffered
2024 PH/06 A failure to get a lock for a hints database would erroneously always say
2025 "Failed to get write lock", even when it was really a read lock.
2027 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2028 of 0600. This has been changed to use the value of the lockfile_mode
2029 option (which defaults to 0600).
2031 PH/08 Applied small patch from the Sieve maintainer.
2033 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2034 folder from quota calculations, a direct delivery into this folder messed
2035 up the contents of the maildirsize file. This was because the regex was
2036 used only to exclude .Trash (or whatever) when the size of the mailbox
2037 was calculated. There was no check that a delivery was happening into an
2038 excluded directory. This bug has been fixed by ignoring all quota
2039 processing for deliveries into excluded directories.
2041 PH/10 Added the maildirfolder_create_regex option to appendfile.
2047 PH/01 The code for finding all the local interface addresses on a FreeBSD
2048 system running IPv6 was broken. This may well have applied to all BSD
2049 systems, as well as to others that have similar system calls. The broken
2050 code found IPv4 interfaces correctly, but gave incorrect values for the
2051 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2052 that it would not match correctly against @[] and not recognize the IPv6
2055 PH/02 The ipliteral router was not recognizing addresses of the form user@
2056 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2058 PH/03 Added disable_ipv6.
2060 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2061 decoded form, because it is most often used to construct To: headers
2062 lines in autoreplies, and the decoded form may well be syntactically
2063 invalid. However, $reply_address has leading white space removed, and all
2064 newlines turned into spaces so that the autoreply transport does not
2067 PH/05 If group was specified without a user on a router, and no group or user
2068 was specified on a transport, the group from the router was ignored.
2070 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2071 for visible compile-time settings that can be used to change these
2072 numbers, for those that want even more. Backwards compatibility with old
2073 spool files has been maintained. However, going back to a previous Exim
2074 release will lost any variables that are in spool files.
2076 PH/07 Two small changes when running in the test harness: increase delay when
2077 passing a TCP/IP connection to a new process, in case the original
2078 process has to generate a bounce, and remove special handling of
2079 127.0.0.2 (sic), which is no longer necessary.
2081 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2082 be the same on different OS.
2084 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2087 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2088 whether --show-vars was specified or not
2090 JJ/02 exipick: Added support for new ACL variable spool format introduced
2093 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2094 syntactically invalid From: or Reply-to: line, and a filter used this to
2095 generate an autoreply, and therefore failed to obtain an address for the
2096 autoreply, Exim could try to deliver to a non-existent relative file
2097 name, causing unrelated and misleading errors. What now happens is that
2098 it logs this as a hard delivery error, but does not attempt to create a
2101 PH/11 The exinext utility has a -C option for testing purposes, but although
2102 the given file was scanned by exinext itself; it wasn't being passed on
2103 when Exim was called.
2105 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2106 an end-of-file indication when reading a command response.
2108 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2109 compiled. In many other places in Exim, IPv6 addresses are always
2110 recognized, so I have changed this. It also means that IPv4 domain
2111 literals of the form [IPV4:n.n.n.n] are now always recognized.
2113 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2114 used if the router is not running as root, for example, when verifying at
2115 ACL time, or when using -bh. The debugging output from this situation was
2116 non-existent - all you got was a failure to exec. I have made two
2119 (a) Failures to set uid/gid, the current directory, or a process leader
2120 in a subprocess such as that created by queryprogram now generate
2121 suitable debugging ouput when -d is set.
2123 (b) The queryprogram router detects when it is not running as root,
2124 outputs suitable debugging information if -d is set, and then runs
2125 the subprocess without attempting to change uid/gid.
2127 PH/15 Minor change to Makefile for building test_host (undocumented testing
2130 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2131 additional section of a DNS packet that returns MX or SRV records.
2132 Instead, it always explicitly searches for A/AAAA records. This avoids
2133 major problems that occur when a DNS server includes only records of one
2134 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2135 fixed another bug: if SRV records were looked up and the corresponding
2136 address records were *not* found in the additional section, the port
2137 values from the SRV records were lost.
2139 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2140 using the correct key (the original address) when searching the retry
2141 rules in order to find which one to use for generating the retry hint.
2143 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2144 adding the default one. Similarly, if it contains a Reply-To: header, the
2145 errors_reply_to option, if set, is not used.
2147 PH/19 When calculating a retry time, Exim used to measure the "time since
2148 failure" by looking at the "first failed" field in the retry record. Now
2149 it does not use this if it is later than than the arrival time of the
2150 message. Instead it uses the arrival time. This makes for better
2151 behaviour in cases where some deliveries succeed, thus re-setting the
2152 "first failed" field. An example is a quota failure for a huge message
2153 when small messages continue to be delivered. Without this change, the
2154 "time since failure" will always be short, possible causing more frequent
2155 delivery attempts for the huge message than are intended.
2156 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2158 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2159 $1, $2, $3) because the numerical variables can be reset during some
2160 expansion items (e.g. "match"), thereby losing the authentication data.
2162 PH/21 Make -bV show the size of off_t variables so that the test suite can
2163 decide whether to run tests for quotas > 2G.
2165 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2166 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2167 filecount value is greater than 2G or if a quota value is greater than 2G
2168 on a system where the size of off_t is not greater than 4, a panic error
2171 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2172 never match. The debug and -bh output now contains an explicit error
2173 message indicating a malformed IPv4 address or mask.
2175 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2176 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2177 PH/23 above applies.
2179 PH/25 Do not write to syslog when running in the test harness. The only
2180 occasion when this arises is a failure to open the main or panic logs
2181 (for which there is an explicit test).
2183 PH/26 Added the /no_tell option to "control=freeze".
2185 PH/27 If a host name lookup failed very early in a connection, for example, if
2186 the IP address matched host_lookup and the reverse lookup yielded a name
2187 that did not have a forward lookup, an error message of the form "no IP
2188 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2189 could be logged. Now it outputs the IP address instead of "NULL".
2191 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2192 allows the sender and the authenticated sender to be set when
2193 submitting a message from within Exim. Since child_open_exim() is
2194 documented for local_scan(), the new function should be too.
2196 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2197 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2198 results in an empty string is now treated as unset.
2200 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2202 PH/31 Added sender_verify_fail logging option.
2204 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2205 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2206 tidied the source and removed it altogether.
2208 PH/33 When a queue run was abandoned because the load average was too high, a
2209 log line was always written; now it is written only if the queue_run log
2210 selector is set. In addition, the log line for abandonment now contains
2211 information about the queue run such as the pid. This is always present
2212 in "start" and "stop" lines but was omitted from the "abandon" line.
2214 PH/34 Omit spaces between a header name and the colon in the error message that
2215 is given when verify = headers_syntax fails (if there are lots of them,
2216 the message gets confusing).
2218 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2219 names, as there are now some PTR records that contain slashes. This check
2220 is only to protect against broken name servers that fall over on strange
2221 characters, so the fact that it applies to all lookups doesn't matter.
2223 PH/36 Now that the new test suite is complete, we can remove some of the
2224 special code in Exim that was needed for the old test suite. For example,
2225 sorting DNS records because real resolvers return them in an arbitrary
2226 order. The new test suite's fake resolver always returns records in the
2229 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2230 bounces) except when queue_only is set, to avoid logging races between
2231 the different processes.
2233 PH/38 Panic-die if .include specifies a non-absolute path.
2235 PH/39 A tweak to the "H" retry rule from its user.
2237 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2238 a label. They prevented compilation on older perls.
2240 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2241 a warning to be raised on newish perls.
2243 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2244 on queue. Changes to match documented behaviour of showing count of
2245 messages matching specified criteria.
2247 PH/40 Changed the default ident timeout from 30s to 5s.
2249 PH/41 Added support for the use of login_cap features, on those BSD systems
2250 that have them, for controlling the resources used by pipe deliveries.
2252 PH/42 The content-scanning code uses fopen() to create files in which to put
2253 message data. Previously it was not paying any attention to the mode of
2254 the files. Exim runs with umask(0) because the rest of the code creates
2255 files with open(), and sets the required mode explicitly. Thus, these
2256 files were ending up world-writeable. This was not a big issue, because,
2257 being within the spool directory, they were not world-accessible. I have
2258 created a function called modefopen, which takes an additional mode
2259 argument. It sets umask(777), creates the file, chmods it to the required
2260 mode, then resets the umask. All the relevant calls to fopen() in the
2261 content scanning code have been changed to use this function.
2263 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2264 to 24 hours. This avoids potential overflow problems when processing G
2265 and H retry rules. I suspect nobody ever tinkers with this value.
2267 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2269 PH/45 When the plaintext authenticator is running as a client, the server's
2270 challenges are checked to ensure they are valid base64 strings. By
2271 default, the authentication attempt is cancelled if an invalid string is
2272 received. Setting client_ignore_invalid_base64 true ignores these errors.
2273 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2274 they are received. Thus, the responses can be made to depend on the
2275 challenges. If an invalid string is ignored, an empty string is placed in
2278 PH/46 Messages that are created by the autoreply transport now contains a
2279 References: header, in accordance with RFCs 2822 and 3834.
2281 PH/47 Added authenticated_sender_force to the smtp transport.
2283 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2285 PH/49 Installed latest patch from the Sieve maintainer.
2287 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2288 was also set, the appendfile transport was unnecessarily scanning a
2289 directory of message files (e.g. for maildir delivery) to find the count
2290 of files (along with the size), even though it did not need this
2291 information. It now does the scan only if it needs to find either the
2292 size of the count of files.
2294 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2296 PH/52 Two bugs concerned with error handling when the smtp transport is
2299 (i) Exim was not creating retry information for temporary errors given
2300 for individual recipients after the DATA command when the smtp transport
2301 was used in LMTP mode. This meant that they could be retried too
2302 frequently, and not timed out correctly.
2304 (ii) Exim was setting the flag that allows error details to be returned
2305 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2306 recipients that were returned after the DATA command.
2308 PH/53 This is related to PH/52, but is more general: for any failing address,
2309 when detailed error information was permitted to be returned to the
2310 sender, but the error was temporary, then after the final timeout, only
2311 "retry timeout exceeded" was returned. Now it returns the full error as
2312 well as "retry timeout exceeded".
2314 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2315 do this, and (what is worse) MTAs that accept it.
2317 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2318 will now be deprecated.
2320 PH/56 New os.c-cygwin from the Cygwin maintainer.
2322 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2323 formats (previously only available in exim formats via -bpr, -bpru,
2324 and -bpra. Now also available in native and exiqgrep formats)
2326 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2327 with very large, slow to parse queues
2329 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2331 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2333 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2334 responses to authentication challenges, though it was showing the
2335 challenges; (ii) I've removed the CR characters from the debug output for
2338 PH/58 Allow for the insertion of a newline as well as a space when a string
2339 is turned into more than one encoded-word during RFC 2047 encoding. The
2340 Sieve code now uses this.
2342 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2343 data_4xx, lost_connection, tls_required.
2345 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2346 message was being sent as an SMTP response.
2348 PH/61 Add -l and -k options to exicyclog.
2350 PH/62 When verifying, if an address was redirected to one new address, so that
2351 verification continued, and the new address failed or deferred after
2352 having set something in $address_data, the value of $address_data was not
2353 passed back to the ACL. This was different to the case when no
2354 redirection occurred. The value is now passed back in both cases.
2356 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2357 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2358 use login_cap.h, so on its own it isn't the distinguishing feature. The
2359 new name refers directly to the setclassresources() function.
2361 PH/65 Added configuration files for NetBSD3.
2363 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2365 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2366 is preferred over IPv4.
2368 PH/68 The bounce_return_message and bounce_return_body options were not being
2369 honoured for bounces generated during the reception of non-SMTP messages.
2370 In particular, this applied to messages rejected by the ACL. This bug has
2371 been fixed. However, if bounce_return_message is true and bounce_return_
2372 body is false, the headers that are returned for a non-SMTP message
2373 include only those that have been read before the error was detected.
2374 (In the case of an ACL rejection, they have all been read.)
2376 PH/69 The HTML version of the specification is now built in a directory called
2377 spec_html instead of spec.html, because the latter looks like a path with
2378 a MIME-type, and this confuses some software.
2380 PH/70 Catch two compiler warnings in sieve.c.
2382 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2383 function verify_get_ident() calls ip_connect() to connect a socket, but
2384 if the "connect()" function timed out, ip_connect() used to close the
2385 socket. However, verify_get_ident() also closes the socket later, and in
2386 between Exim writes to the log, which may get opened at this point. When
2387 the socket was closed in ip_connect(), the log could get the same file
2388 descriptor number as the socket. This naturally causes chaos. The fix is
2389 not to close the socket in ip_connect(); the socket should be closed by
2390 the function that creates it. There was only one place in the code where
2391 this was missing, in the iplookup router, which I don't think anybody now
2392 uses, but I've fixed it anyway.
2394 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2395 well as to direct DNS lookups. Otherwise the handling of names in host
2396 lists is inconsistent and therefore confusing.
2402 PH/01 Two changes to the default runtime configuration:
2404 (1) Move the checks for relay_from_hosts and authenticated clients from
2405 after to before the (commented out) DNS black list checks.
2407 (2) Add control=submission to the relay_from_hosts and authenticated
2408 clients checks, on the grounds that messages accepted by these
2409 statements are most likely to be submissions.
2411 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2413 (1) Generate an error if the third argument for the ${prvs expansion is
2416 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2419 (3) Reset the variables that are obtained from the first argument of
2420 ${prvscheck and used in the second argument before leaving the code,
2421 because their memory is reclaimed, so using them afterwards may do
2424 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2425 one (it's much easier than Tom thought :-).
2427 (5) Because of (4), we can now allow for the use of $prvscheck_result
2428 inside the third argument.
2430 PH/03 For some reason, the default setting of PATH when running a command from
2431 a pipe transport was just "/usr/bin". I have changed it to
2434 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2435 anything to be listed in the output from -bV.
2437 PH/05 When a filter generated an autoreply, the entire To: header line was
2438 quoted in the delivery log line, like this:
2440 => >A.N.Other <ano@some.domain> <original@ddress> ...
2442 This has been changed so that it extracts the operative address. There
2443 may be more than one such address. If so, they are comma-separated, like
2446 => >ano@some.domain,ona@other.domain <original@ddress> ...
2448 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2449 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2450 not being looked up in the rDNS to get a host name, Exim was showing the
2451 IP address twice in Received: lines, even though the IP addresses were
2452 identical. For example:
2454 Received: from [1.2.3.4] (helo=[1.2.3.4])
2456 However, if the real host name was known, it was omitting the HELO data
2457 if it matched the actual IP address. This has been tidied up so that it
2458 doesn't show the same IP address twice.
2460 PH/07 When both +timestamp and +memory debugging was on, the value given by
2461 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2462 called by the debug printing, thereby overwriting the timestamp buffer.
2463 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2465 PH/08 When the original message was included in an autoreply transport, it
2466 always said "this is a copy of the message, including all the headers",
2467 even if body_only or headers_only was set. It now gives an appropriate
2470 PH/09 Applied a patch from the Sieve maintainer which:
2472 o fixes some comments
2473 o adds the (disabled) notify extension core
2474 o adds some debug output for the result of if/elsif tests
2475 o points to the current vacation draft in the documentation
2476 and documents the missing references header update
2480 o fixes a bug in processing the envelope test (when testing
2481 multiple envelope elements, the last element determinted the
2484 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2485 Electronic Mail") by including:
2487 Auto-submitted: auto-generated
2489 in the messages that it generates (bounce messages and others, such as
2490 warnings). In the case of bounce messages for non-SMTP mesages, there was
2491 also a typo: it was using "Auto_submitted" (underscore instead of
2492 hyphen). Since every message generated by Exim is necessarily in response
2493 to another message, thes have all been changed to:
2495 Auto-Submitted: auto-replied
2497 in accordance with these statements in the RFC:
2499 The auto-replied keyword:
2501 - SHOULD be used on messages sent in direct response to another
2502 message by an automatic process,
2504 - MUST NOT be used on manually-generated messages,
2506 - MAY be used on Delivery Status Notifications (DSNs) and Message
2507 Disposition Notifications (MDNs),
2509 - MUST NOT be used on messages generated by automatic or periodic
2510 processes, except for messages which are automatic responses to
2513 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2514 to the default Received: header definition.
2516 PH/12 Added log selector acl_warn_skipped (default on).
2518 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2519 variables because (a) they are in the wrong storage pool and (b) even if
2520 they were copied, it wouldn't work properly because of the caching.
2522 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2523 checking when decoding. Apparently there are clients that generate
2524 overlong encoded strings. Why am I not surprised?
2526 PH/15 If the first argument of "${if match_address" was not empty, but did not
2527 contain an "@" character, Exim crashed. Now it writes a panic log message
2528 and treats the condition as false.
2530 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2532 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2533 extension "envelope-auth". The code is finished and in agreement with
2534 other implementations, but there is no documentation so far and in fact,
2535 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2536 not changing the active code.
2538 Print executed "if" and "elsif" statements when debugging is used. This
2539 helps a great deal to understand what a filter does.
2541 Document more things not specified clearly in RFC3028. I had all this
2542 sorted out, when out of a sudden new issues came to my mind. Oops."
2544 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2547 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2548 canonical form (as documented). However, after a host name lookup from
2549 the IP address, check_host() was doing a simple string comparison with
2550 addresses acquired from the DNS when checking that the found name did
2551 have the original IP as one of its addresses. Since any found IPv6
2552 addresses are likely to be in abbreviated form, the comparison could
2553 fail. Luckily, there already exists a function for doing the comparison
2554 by converting both addresses to binary, so now that is used instead of
2555 the text comparison.
2557 PH/20 There was another similar case to PH/19, when a complete host name was
2558 given in a host list; looking up its IP address could give an abbreviated
2559 form, whereas the current host's name might or might not be abbreviated.
2560 The same fix has been applied.
2566 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2567 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2570 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2571 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2573 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2575 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2576 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2577 Also, for "Auto-Submitted", treat anything other than "no" as
2578 non-personal, in accordance with RFC 3834. (Previously it treated
2579 anything starting "auto-" as non-personal.)
2581 TF/01 The control=submission/name=... option had a problem with syntax
2582 errors if the name included a slash character. The /name= option
2583 now slurps the rest of the string, so it can include any characters
2584 but it must come last in the list of options (after /sender_retain
2587 PH/05 Some modifications to the interface to the fake nameserver for the new
2595 TK/01 Added the "success_on_redirect" address verification option. See
2596 NewStuff for rationale and an example.
2598 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2600 PH/02 Patch to exigrep to allow it to work on syslog lines.
2602 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2603 fread() to skip over the body file's header line, because in Cygwin the
2604 header line is locked and is inaccessible.
2606 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2607 co-exist for some time) to make it clear that it is the Exim ID that is
2608 referenced, not the Message-ID: header line.
2610 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2611 string_format() function, because snprintf() does not exist on all
2614 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2615 ${stat: expansion item.
2617 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2618 protocol synchronization error", to keep the pedants happy.
2620 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2621 well as for IRIX systems, when gcc is being used. See the host.c source
2624 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2626 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2629 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2630 transport and to the smtp transport in LMTP mode.
2632 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2634 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2635 run a filter in a subprocess. This could lead to confusion in subsequent
2636 lookups in the parent process. There should also be a search_tidyup() at
2637 the end of the subprocess.
2639 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2640 only if the host matched helo_try_verify_hosts, which caused the
2641 verification to occur when the EHLO/HELO command was issued. The ACL just
2642 tested the remembered result. Now, if a previous verification attempt has
2643 not happened, "verify = helo" does it there and then.
2645 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2647 TK/03 Fix log output including CR from clamd.
2649 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2650 exist provoked a memory error which could cause a segfault.
2652 PH/15 Installed PCRE 6.2
2654 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2656 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2657 of the problem. Specifically, suggested +O2 rather than +O1 for the
2660 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2662 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2663 record, the port was not correctly logged when the outgoing_port log
2664 selector was set (it logged the transort's default port).
2666 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2667 fallback_hosts, and "hosts" in the smtp transport.
2669 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2670 host errors such as "Connection refused".
2672 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2673 authentication with radiusclient 0.4.9:
2675 - Error returned from rc_read_config was caught wrongly
2676 - Username/password not passed on to radius server due to wrong length.
2678 The presumption is that some radiusclient API changes for 4.51/PH/17
2679 were not taken care of correctly. The code is still untested by me (my
2680 Linux distribution still has 0.3.2 of radiusclient), but it was
2681 contributed by a Radius user.
2683 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2684 expanding the "port" option of the smtp transport.
2686 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2687 while reading a MIME header. Thanks to Tom Hughes for a patch.
2689 PH/24 Include config.h inside local_scan.h so that configuration settings are
2692 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2693 that in an ACL for RCPT (for example), you can examine exactly what was
2696 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2697 commands, but it was not correctly comparing the address with the actual
2698 client host address. Thus, it would show the EHLO address in Received:
2699 header lines when this was not necessary.
2701 PH/27 Added the % operator to ${eval:}.
2703 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2704 it should be ignoring failures (because with -C, for example, it has lost
2705 privilege). It wasn't ignoring creation failures other than "already
2708 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2711 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2712 because an input line was too long, either on its own, or by virtue of
2713 too many continuations, the temporary file was not being removed, and the
2714 return code was incorrect.
2716 PH/31 Missing "BOOL" in function definition in filtertest.c.
2718 PH/32 Applied Sieve patches from the maintainer.
2720 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2722 PH/33 Added "verify = not_blind".
2724 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2725 Local/Makefile (with some defaults set). These are used in built scripts
2726 such as exicyclog, but they have never been used in the exim_install
2727 script (though there are many overriding facilities there). I have
2728 arranged that the exim_install script now takes note of these two
2731 PH/35 Installed configuration files for Dragonfly.
2733 PH/36 When a locally submitted message by a trusted user did not contain a
2734 From: header, and the sender address was obtained from -f or from an SMTP
2735 MAIL command, and the trusted user did not use -F to supply a sender
2736 name, $originator_name was incorrectly used when constructing a From:
2737 header. Furthermore, $originator_name was used for submission mode
2738 messages from external hosts without From: headers in a similar way,
2739 which is clearly wrong.
2741 PH/37 Added control=suppress_local_fixups.
2743 PH/38 When log_selector = +received_sender was set, and the addition of the
2744 sender made the log line's construction buffer exactly full, or one byte
2745 less than full, an overflow happened when the terminating "\n" was
2748 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2749 when the result of a list match is failure because a DNS lookup failed.
2751 PH/40 RM_COMMAND is now used in the building process.
2753 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2754 the "build-* directories that it finds.
2756 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2757 address in a domain literal was a prefix of an interface address.
2759 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2760 when verifying a sender address, unless rewrite_headers is false.
2762 PH/44 Wrote a long comment about why errors_to addresses are verified as
2763 recipients, not senders.
2765 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2766 the ratelimit ACL was added.
2768 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2770 PH/46 Added extra information about PostgreSQL errors to the error string.
2772 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2773 avoiding the need to install special zones in a real server. This is
2774 backwards compatible; if it can't find the fake resolver, it drops back.
2775 Thus, both old and new test suites can be run.
2777 TF/02 Added util/ratelimit.pl
2779 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2780 clock is set back in time.
2782 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2783 Candler <B.Candler@pobox.com>.
2785 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2786 OK for addresses that are the result of redirections.
2788 PH/48 A number of further additions for the benefit of the new test suite,
2789 including a fake gethostbyname() that interfaces to the fake DNS resolver
2792 TF/06 The fix for widen_domains has also been applied to qualify_single and
2793 search_parents which are the other dnslookup options that can cause
2796 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2799 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2801 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2802 service" if it encountered an MX record with an empty target hostname.
2803 The message is now "an MX or SRV record indicated no SMTP service".
2805 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2806 if the DNS of the sending site is misconfigured. This is quite a
2807 common situation. This change restores the behaviour of treating a
2808 helo verification defer as a failure.
2810 PH/51 If self=fail was set on a router, the bounce message did not include the
2811 actual error message.
2817 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2819 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2820 command itself ended in error, the underlying message about the transport
2821 filter timeout was being overwritten with the pipe command error. Now the
2822 underlying error message should be appended to the second error message.
2824 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2826 PH/02 Reduce the amount of output that "make" produces by default. Full output
2827 can still be requested.
2829 PH/03 The warning log line about a condition test deferring for a "warn" verb
2830 was being output only once per connection, rather than after each
2831 occurrence (because it was using the same function as for successful
2832 "warn" verbs). This seems wrong, so I have changed it.
2834 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2835 it should not have, which might have caused a crash in the right
2836 circumstances, but probably never did.
2838 PH/04 Installed a modified version of Tony Finch's patch to make submission
2839 mode fix the return path as well as the Sender: header line, and to
2840 add a /name= option so that you can make the user's friendly name appear
2843 TF/03 Added the control = fakedefer ACL modifier.
2845 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2846 Mark Lowes for thorough testing.
2848 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2850 TK/03 Merged latest SRS patch from Miles Wilton.
2852 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2853 with the definition in sysexits.h (which is #included earlier).
2854 Fortunately, Exim does not actually use EX_OK. The code used to try to
2855 preserve the sysexits.h value, by assumimg that macro definitions were
2856 scanned for macro replacements. I have been disabused of this notion,
2857 so now the code just undefines EX_OK before #including unistd.h.
2859 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2860 in the smtp transport. When a block could not be written in a single
2861 write() function, the timeout was being re-applied to each part-write.
2862 This seems wrong - if the receiver was accepting one byte at a time it
2863 would take for ever. The timeout is now adjusted when this happens. It
2864 doesn't have to be particularly precise.
2866 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2867 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2869 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2870 without a domain if the check to <postmaster@domain> fails.
2872 SC/01 Eximstats: added -xls and the ability to specify output files
2873 (patch written by Frank Heydlauf).
2875 SC/02 Eximstats: use FileHandles for outputing results.
2877 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2879 SC/04 Eximstats: fixed display of large numbers with -nvr option
2881 SC/05 Eximstats: fixed merging of reports with empty tables.
2883 SC/06 Eximstats: added the -include_original_destination flag
2885 SC/07 Eximstats: removed tabs and trailing whitespace.
2887 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2889 TK/06 MBOX spool code: Add real "From " MBOX separator line
2890 so the .eml file is really in mbox format (even though
2891 most programs do not really care). Patch from Alex Miller.
2893 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2894 The latter is generated from $received_to and is only set if the
2895 message has one envelope recipient. SA can use these headers,
2896 obviously out-of-the-box. Patch from Alex Miller.
2898 PH/08 The ${def test on a variable was returning false if the variable's
2899 value was "0", contrary to what the specification has always said!
2900 The result should be true unless the variable is empty.
2902 PH/09 The syntax error of a character other than { following "${if
2903 def:variable_name" (after optional whitespace) was not being diagnosed.
2904 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2905 accidental colon was present, for example, could give incorrect results.
2907 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2908 result is used (not including appendfile, where other changes are about
2911 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2912 This involved changing a lot of size variables from int to off_t. It
2913 should work with maildirs and everything.
2915 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2916 spamd dying while we are connected to it.
2918 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2921 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2922 function for BDB 4.3.
2924 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2926 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2927 the sources for PCRE have been re-arranged and more files are now
2930 PH/15 The code I had for printing potentially long long variables in PH/11
2931 above was not the best (it lost precision). The length of off_t variables
2932 is now inspected at build time, and an appropriate printing format (%ld
2933 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2934 to be "long long int" or "long int". This is needed for the internal
2935 formatting function string_vformat().
2937 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2938 the configuration file to be ":syslog", then the script "guesses" where
2939 the logs files are, rather than using the compiled in default. In our
2940 case the guess is not the same as the compiled default, so the script
2941 suddenly stopped working when I started to use syslog. The patch checks
2942 to see if log_file_path is "". If so, it attempts to read it from exim
2943 with no configuration file to get the compiled in version, before it
2944 falls back to the previous guessing code."
2946 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2947 implementing BATV in an Exim configuration. See NewStuff for the gory
2950 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2951 Makefile that are specific to HP-UX.
2953 PH/18 If the "use_postmaster" option was set for a recipient callout together
2954 with the "random" option, the postmaster address was used as the MAIL
2955 FROM address for the random test, but not for the subsequent recipient
2956 test. It is now used for both.
2958 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2959 patch removes a few documentation additions to RFC 3028, because the
2960 latest draft now contains them. It adds the new en;ascii-case comparator
2961 and a new error check for 8bit text in MIME parts. Comparator and
2962 require names are now matched exactly. I enabled the subaddress
2963 extension, but it is not well tested yet (read: it works for me)."
2965 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2966 rework some of the code of TK/09 above to avoid the hardwired use of
2967 "%lld" and "long long". Replaced the call to snprintf() with a call to
2970 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2971 records point to non-existent hosts", "retry timeout exceeded", and
2972 "retry time not reached for any host after a long failure period".
2974 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2975 experimental DomainKeys support:
2977 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2978 (2) On an error such as an illegally used "control", the wrong name for
2979 the control was given.
2981 These problems did NOT occur unless DomainKeys support was compiled.
2983 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2985 PH/24 Added ${if match_ip condition.
2987 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2988 databases so that it will be absolutely obvious if a crash occurs in the
2989 DB library. This is a regular occurrence (often caused by mis-matched
2992 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2993 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2994 picked up on a user's system that detects such things. There doesn't seem
2995 to be a gcc warning option for this - only an attribute that has to be
2996 put on the function's prototype. It seems that in Fedora Core 4 they have
2997 set this on a number of new functions. No doubt there will be more in due
3000 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3001 specify a transport. However, if an address that was verified by such a
3002 router was the subject of a callout, Exim crashed because it tried to
3003 read the rcpt_include_affixes from the non-existent transport. Now it
3004 just assumes that the setting of that option is false. This bug was
3005 introduced by 4.51/PH/31.
3007 PH/28 Changed -d+all to exclude +memory, because that information is very
3008 rarely of interest, but it makes the output a lot bigger. People tend to
3009 do -d+all out of habit.
3011 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3012 code in os-type was giving problems when libc.so lives in lib64, like on
3015 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3016 aren't the modern standard, and it seems that some systems' include files
3017 don't always have them. Exim was already checking for some of the newer
3018 ones like T_AAAA, and defining it itself. I've added checks for all the
3019 record types that Exim uses.
3021 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3022 not automatically generating a new one, as it is supposed to. This
3023 prevented TLS from working. If the file did exist, but contained invalid
3024 data, a new version was generated, as expected. It was only the case of a
3025 non-existent file that was broken.
3027 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3028 with a change in libdomainkeys > 0.64.
3030 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3031 from DNS. If the selector record carries the flag, it now has
3032 precedence over the domain-wide flag.
3034 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3036 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3037 the use of a port name that isn't defined in /etc/services) occurred, the
3038 message was deferred as in a normal delivery, and thus remained on the
3039 spool, instead of being failed because of the mua_wrapper setting. This
3040 is now fixed, and I tidied up some of the mua_wrapper messages at the
3043 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3044 the messages in a hash of arrays rather than using individual hashes.
3045 This is a bit cleaner and results in dramatic memory savings, albeit
3046 at a slight CPU cost.
3048 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3049 as requested by Marc Sherman.
3051 SC/10 Eximstats: added histograms for user specified patterns as requested
3054 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3056 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3057 fopen() in the content-scanning modules that did not already have it.
3063 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3064 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3066 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3068 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3070 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3071 to test Sieve filters that use "vacation".
3073 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3074 that changes the way the GnuTLS parameters are stored in the cache file.
3075 The new format can be generated externally. For backward compatibility,
3076 if the data in the cache doesn't make sense, Exim assumes it has read an
3077 old-format file, and it generates new data and writes a new file. This
3078 means that you can't go back to an older release without removing the
3081 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3082 work if there are any delivery delays because "one_time" forces the
3083 parent to be marked "delivered", so its unseen clone is never tried
3084 again. For this reason, Exim now forbids the simultaneous setting of
3087 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3088 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3089 entry if you want to know the details. The fix, however, neglected to
3090 consider the case when local delivery batching is involved. The test for
3091 "previously delivered" was not happening when checking to see if an
3092 address could be batched with a previous (undelivered) one; under
3093 certain circumstances this could lead to multiple deliveries to the same
3096 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3097 in its include files, and this causes problems building Exim.
3099 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3100 header_syntax) but Exim was just ignoring anything given after a slash.
3101 In particular, this caused confusion with an attempt to use "verify =
3102 reverse_host_lookup/defer_ok". An error is now given when options are
3103 supplied for verify items that do not have them. (Maybe reverse_host_
3104 lookup should have a defer_ok option, but that's a different point.)
3106 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3107 defined by RFC 821) to 2048, because there were problems with some AUTH
3108 commands, and RFC 1869 says the size should be increased for extended
3109 SMTP commands that take arguments.
3111 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3114 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3115 "unknown" error; now it says that the functionality isn't in the binary.
3117 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3118 an address' error message when a string expansion fails (syntax or
3119 whatever). Otherwise the password may appear in the log. Following change
3120 PH/42 below, there is no longer a chance of it appearing in a bounce
3123 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3125 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3126 of its addresses was ever tried. (Bugzilla bug #2.)
3128 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3129 the result incorrectly in the debug output. (It correctly added a newline
3130 to what was transported.)
3132 TF/01 Added $received_time.
3134 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3135 commented out examples of how to interface to a virus scanner and to
3136 SpamAssassin. Also added commented examples of av_scanner and
3137 spamd_address settings.
3139 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3140 and controls are allowed in which ACLs. There were a couple of minor
3141 errors. Some of the entries in the conditions table (which is a table of
3142 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3143 negation of where the condition IS allowed.
3145 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3147 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3148 header file does not have a version number, so I've had to invent a new
3149 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3150 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3151 radiusclient), but it was contributed by a Radius user.
3153 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3154 files or directories, for OpenSSL.
3156 PH/19 When an Exim process that is running as root has to create an Exim log
3157 file, it does so in a subprocess that runs as exim:exim so as to get the
3158 ownership right at creation (otherwise, other Exim processes might see
3159 the file with the wrong ownership). There was no test for failure of this
3160 fork() call, which would lead to the process getting stuck as it waited
3161 for a non-existent subprocess. Forks do occasionally fail when resources
3162 run out. I reviewed all the other calls to fork(); they all seem to check
3165 PH/20 When checking for unexpected SMTP input at connect time (before writing
3166 the banner), Exim was not dealing correctly with a non-positive return
3167 from the read() function. If the client had disconnected by this time,
3168 the result was a log entry for a synchronization error with an empty
3169 string after "input=" when read() returned zero. If read() returned -1
3170 (an event I could not check), uninitialized data bytes were printed.
3171 There were reports of junk text (parts of files, etc) appearing after
3174 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3176 PH/22 Added support for macro redefinition, and (re)definition in between
3177 driver and ACL definitions.
3179 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3180 forgetting to use the resulting value; it was using the unexpanded value.
3182 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3183 hadn't been configured. The fix is from Juergen Kreileder, who
3184 understands it better than I do:
3186 "Here's what I see happening with three configured cyrus_sasl
3187 authenticators configured (plain, login, cram-md5):
3189 On startup auth_cyrus_sasl_init() gets called for each of these.
3190 This means three calls to sasl_listmech() without a specified mech_list.
3191 => SASL tests which mechs of all available mechs actually work
3192 => three warnings about OTP not working
3193 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3195 With the patch, sasl_listmech() also gets called three times. But now
3196 SASL's mech_list option is set to the server_mech specified in the the
3197 authenticator. Or in other words, the answer from sasl_listmech()
3198 gets limited to just the mech you're testing for (which is different
3200 => the return list contains just 'plain' or 'login', 'cram-md5' or
3201 nothing depending on the value of ob->server_mech.
3203 I've just tested the patch: Authentication still works fine,
3204 unavailable mechs specified in the exim configuration are still
3205 caught, and the auth.log warnings about OTP are gone."
3207 PH/25 When debugging is enabled, the contents of the command line are added
3208 to the debugging output, even when log_selector=+arguments is not
3211 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3212 answer is "GNU", and only if the return is "GNU/something" is the answer
3215 PH/27 $acl_verify_message is now set immediately after the failure of a
3216 verification in an ACL, and so is available in subsequent modifiers. In
3217 particular, the message can be preserved by coding like this:
3219 warn !verify = sender
3220 set acl_m0 = $acl_verify_message
3222 Previously, $acl_verify_message was set only while expanding "message"
3223 and "log_message" when a very denied access.
3225 PH/28 Modified OS/os.c-Linux with
3227 -#ifndef OS_LOAD_AVERAGE
3228 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3230 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3231 nomenclature these days.)
3233 PH/29 Installed patch from the Sieve maintainer that adds the options
3234 sieve_useraddress and sieve_subaddress to the redirect router.
3236 PH/30 In these circumstances:
3237 . Two addresses routed to the same list of hosts;
3238 . First host does not offer TLS;
3239 . First host accepts first address;
3240 . First host gives temporary error to second address;
3241 . Second host offers TLS and a TLS session is established;
3242 . Second host accepts second address.
3243 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3244 and peerdn, if requested) that were in fact used only for the second
3247 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3248 attention to any local part prefix or suffix that was matched by the
3249 router that accepted the address. It now behaves in the same way as it
3250 does for delivery: the affixes are removed from the local part unless
3251 rcpt_include_affixes is set on the transport.
3253 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3254 timeout during the DATA phase of an incoming message.
3256 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3257 applied a patch sanctioned by the Sieve maintainer.
3259 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3260 the uid or gid is negative. A case of a negative gid caused this to be
3261 noticed. The fix allows for either to be negative.
3263 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3264 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3267 PH/36 The change PH/12 above was broken. Fixed it.
3269 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3270 the grounds that routing the same address twice would always produce the
3271 same answer. This might have been true once, but it is certainly no
3272 longer true now. Routing a child address may depend on the previous
3273 routing that produced that child. Some complicated redirection strategies
3274 went wrong when messages had multiple recipients, and made Exim's
3275 behaviour dependent on the order in which the addresses were given.
3277 I have moved the duplicate checking until after the routing is complete.
3278 Exim scans the addresses that are assigned to local and remote
3279 transports, and removes any duplicates. This means that more work will be
3280 done, as duplicates will always all be routed, but duplicates are
3281 presumably rare, so I don't expect this is of any significance.
3283 For deliveries to pipes, files, and autoreplies, the duplicate checking
3284 still happens during the routing process, since they are not going to be
3287 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3288 It corrects a timeout issue with spamd. This is Ian's comment: "The
3289 background is that sometimes spamd either never reads data from a
3290 connection it has accepted, or it never writes response data. The exiscan
3291 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3292 blindly assumes that writes won't block so it may never time out."
3294 PH/39 Allow G after quota size as well as K and M.
3296 PH/40 The value set for $authenticated_id in an authenticator may not contain
3297 binary zeroes or newlines because the value is written to log lines and
3298 to spool files. There was no check on this. Now the value is run through
3299 the string_printing() function so that such characters are converted to
3300 printable escape sequences.
3302 PH/41 $message_linecount is a new variable that contains the total number of
3303 lines in the message. Compare $body_linecount, which is the count for the
3306 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3307 bounce and delay warning messages, except in certain special cases, which
3310 (a) An SMTP error message from a remote host;
3311 (b) A message specified in a :fail: redirection;
3312 (c) A message specified in a "fail" command in a system filter;
3313 (d) A message specified in a FAIL return from the queryprogram router;
3314 (e) A message specified by the cannot_route_message router option.
3316 In these cases only, Exim does include the error details in bounce and
3317 warning messages. There are also a few cases where bland messages such
3318 as "unrouteable address" or "local delivery error" are given.
3320 PH/43 $value is now also set for the "else" part of a ${run expansion.
3322 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3323 being worked on, but at least Exim now implements the latest version to
3326 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3327 process to complete was treated as a delivery failure, a timeout while
3328 writing the message to the pipe was logged, but erroneously treated as a
3329 successful delivery. Such timeouts include transport filter timeouts. For
3330 consistency with the overall process timeout, these timeouts are now
3331 treated as errors, giving rise to delivery failures by default. However,
3332 there is now a new Boolean option for the pipe transport called
3333 timeout_defer, which, if set TRUE, converts the failures into defers for
3334 both kinds of timeout. A transport filter timeout is now identified in
3337 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3338 systems where "make" and "gmake" are different, calling "gmake" at top
3339 level broke things. I've arranged for the value of $(MAKE) to be passed
3340 from the Makefile to this script so that it can call the same version of
3344 A note about Exim versions 4.44 and 4.50
3345 ----------------------------------------
3347 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3348 changes of various kinds. As a consequence, a big documentation update was
3349 needed. This delayed the release for rather longer than seemed good, especially
3350 in the light of a couple of (minor) security issues. Therefore, the changes
3351 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3352 release. So 4.44 and 4.50 are in effect two different branches that both start
3355 I have left the 4.50 change log unchanged; it contains all the changes since
3356 4.43. The change log for 4.44 is below; many of its items are identical to
3357 those for 4.50. This seems to be the most sensible way to preserve the
3358 historical information.
3364 1. Minor wording change to the doc/README.SIEVE file.
3366 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3367 computation of the current number of files was incorrect.
3369 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3370 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3373 4. Give more explanation in the error message when the command for a transport
3374 filter fails to execute.
3376 5. There are several places where Exim runs a non-Exim command in a
3377 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3378 was being done only for the command run by the queryprogram router. It is
3379 now done for all such subprocesses. The other cases are: ${run, transport
3380 filters, and the commands run by the lmtp and pipe transports.
3382 6. Added CONFIGURE_GROUP build-time option.
3384 7. Some older OS have a limit of 256 on the maximum number of file
3385 descriptors. Exim was using setrlimit() to set 1000 as a large value
3386 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3387 systems. I've change it so that if it can't get 1000, it tries for 256.
3389 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3390 was an oversight, and furthermore, ever since the addition of extra
3391 controls (e.g. 4.43/32), the checks on when to allow different forms of
3392 "control" were broken. There should now be diagnostics for all cases when a
3393 control that does not make sense is encountered.
3395 9. Added the /retain_sender option to "control=submission".
3397 10. $recipients is now available in the predata ACL (oversight).
3399 11. Tidy the search cache before the fork to do a delivery from a message
3400 received from the command line. Otherwise the child will trigger a lookup
3401 failure and thereby defer the delivery if it tries to use (for example) a
3402 cached ldap connection that the parent has called unbind on.
3404 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3405 of $address_data from the recipient verification was clobbered by the
3406 sender verification.
3408 13. The value of address_data from a sender verification is now available in
3409 $sender_address_data in subsequent conditions in the ACL statement.
3411 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3413 15. Added a new option "connect=<time>" to callout options, to set a different
3416 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3417 was its contents. (It was OK if the option was not defined at all.)
3419 17. A "Completed" log line is now written for messages that are removed from
3420 the spool by the -Mrm option.
3422 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3423 information about exactly what failed.
3425 19. Added -dd to debug only the daemon process.
3427 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3428 handles timeouts, both on the server side and network timeouts. Renamed the
3429 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3431 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3432 It is now set to "smtps".
3434 22. $host_address is now set to the target address during the checking of
3435 ignore_target_hosts.
3437 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3438 being passed; this would have caused $sender_host_name to have been used if
3439 matching the list had actually called for a host name (not very likely,
3440 since this list is usually IP addresses). A host name is now passed as
3443 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3444 code that specifies a non-restarting handler (typically sigaction() in
3445 modern systems) in an attempt to fix a rare and obscure crash bug.
3447 25. Narrowed the window for a race in the daemon that could cause it to ignore
3448 SIGCHLD signals. This is not a major problem, because they are used only to
3449 wake it up if nothing else does.
3451 26. A malformed maildirsize file could cause Exim to calculate negative values
3452 for the mailbox size or file count. Odd effects could occur as a result.
3453 The maildirsize information is now recalculated if the size or filecount
3456 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3457 support for a long time. Removed HAVE_SYS_VFS_H.
3459 28. Installed the latest version of exipick from John Jetmore.
3461 29. In an address list, if the pattern was not a regular expression, an empty
3462 subject address (from a bounce message) matched only if the pattern was an
3463 empty string. Non-empty patterns were not even tested. This was the wrong
3464 because it is perfectly reasonable to use an empty address as part of a
3465 database query. An empty address is now tested by patterns that are
3466 lookups. However, all the other forms of pattern expect the subject to
3467 contain a local part and a domain, and therefore, for them, an empty
3468 address still always fails if the pattern is not itself empty.
3470 30. Exim went into a mad DNS loop when attempting to do a callout where the
3471 host was specified on an smtp transport, and looking it up yielded more
3472 than one IP address.
3474 31. Re-factored the code for checking spool and log partition space into a
3475 function that finds that data and another that does the check. The former
3476 is then used to implement four new variables: $spool_space, $log_space,
3477 $spool_inodes, and $log_inodes.
3479 32. The RFC2047 encoding function was originally intended for short strings
3480 such as real names; it was not keeping to the 75-character limit for
3481 encoded words that the RFC imposes. It now respects the limit, and
3482 generates multiple encoded words if necessary. To be on the safe side, I
3483 have increased the buffer size for the ${rfc2047: expansion operator from
3486 33. It is now permitted to omit both strings after an "if" condition; if the
3487 condition is true, the result is "true". As before, when the second string
3488 is omitted, a false condition yields an empty string. This makes it less
3489 cumbersome to write custom ACL and router conditions.
3491 34. Failure to deliver a bounce message always caused it to be frozen, even if
3492 there was an errors_to setting on the router. The errors_to setting is now
3495 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3496 canonical form (fully expanded) before being placed in
3497 $sender_host_address.
3499 36. The table in the code that translates DNS record types into text (T_A to
3500 "A" for instance) was missing entries for NS and CNAME. It is just possible
3501 that this could have caused confusion if both these types were looked up
3502 for the same domain, because the text type is used as part of Exim's
3503 per-process caching. But the chance of anyone hitting this buglet seems
3506 37. The dnsdb lookup has been extended in a number of ways.
3508 (1) There is a new type, "zns", which walks up the domain tree until it
3509 finds some nameserver records. It should be used with care.
3511 (2) There is a new type, "mxh", which is like "mx" except that it returns
3512 just the host names, not the priorities.
3514 (3) It is now possible to give a list of domains (or IP addresses) to be
3515 looked up. The behaviour when one of the lookups defers can be
3516 controlled by a keyword.
3518 (4) It is now possible to specify the separator character for use when
3519 multiple records are returned.
3521 38. The dnslists ACL condition has been extended: it is now possible to supply
3522 a list of IP addresses and/or domains to be looked up in a particular DNS
3525 39. Added log_selector=+queue_time_overall.
3527 40. When running the queue in the test harness, wait just a tad after forking a
3528 delivery process, to get repeatability of debugging output.
3530 41. Include certificate and key file names in error message when GnuTLS fails
3531 to set them up, because the GnuTLS error message doesn't include the name
3532 of the failing file when there is a problem reading it.
3534 42. Allow both -bf and -bF in the same test run.
3536 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3538 44. The "Exiscan patch" is now merged into the mainline Exim source.
3540 45. Sometimes the final signoff response after QUIT could fail to get
3541 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3542 < 0 before doing a fflush(). This bug looks as though it goes back to the
3543 introduction of TLS in release 3.20, but "sometimes" must have been rare
3544 because the tests only now provoked it.
3546 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3547 (this can affect the format of dates).
3549 47. exim_tidydb, when checking for the continued existence of a message for
3550 which it has found a message-specific retry record, was not finding
3551 messages that were in split spool directories. Consequently, it was
3552 deleting retry records that should have stayed in existence.
3554 48. Steve fixed some bugs in eximstats.
3556 49. The SPA authentication driver was not abandoning authentication and moving
3557 on to the next authenticator when an expansion was forced to fail,
3558 contradicting the general specification for all authenticators. Instead it
3559 was generating a temporary error. It now behaves as specified.
3561 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3562 (the order specifies the preference for clients). The order is now AES256,
3563 AES128, 3DES, ARCFOUR128.
3565 51. Small patch to Sieve code - explicitly set From: when generating an
3568 52. Exim crashed if a remote delivery caused a very long error message to be
3569 recorded - for instance if somebody sent an entire SpamAssassin report back
3570 as a large number of 550 error lines. This bug was coincidentally fixed by
3571 increasing the size of one of Exim's internal buffers (big_buffer) that
3572 happened as part of the Exiscan merge. However, to be on the safe side, I
3573 have made the code more robust (and fixed the comments that describe what
3576 53. Now that there can be additional text after "Completed" in log lines (if
3577 the queue_time_overall log selector is set), a one-byte patch to exigrep
3578 was needed to allow it to recognize "Completed" as not the last thing in
3581 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3582 patch that reportedly fixes this has been added. I am not expert enough to
3583 create a test for it. This is what the patch creator wrote:
3585 "I found a little strange behaviour of ldap code when working with
3586 Windows 2003 AD Domain, where users was placed in more than one
3587 Organization Units. When I tried to give exim partial DN, the exit code
3588 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3589 But simultaneously result of request was absolutely normal ldap result,
3590 so I produce this patch..."
3592 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3593 REFERENCE, so I have modified the code to exclude the patch when that macro
3596 55. Some experimental protocols are using DNS PTR records for new purposes. The
3597 keys for these records are domain names, not reversed IP addresses. The
3598 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3599 leaves it alone. Component reversal etc. now happens only for IP addresses.
3602 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3604 57. Double the size of the debug message buffer (to 2048) so that more of very
3605 long debug lines gets shown.
3607 58. The exicyclog utility now does better if the number of log files to keep
3608 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3610 59. Two changes related to the smtp_active_hostname option:
3612 (1) $smtp_active_hostname is now available as a variable.
3613 (2) The default for smtp_banner uses $smtp_active_hostname instead
3614 of $primary_hostname.
3616 60. The host_aton() function is supposed to be passed a string that is known
3617 to be a valid IP address. However, in the case of IPv6 addresses, it was
3618 not checking this. This is a hostage to fortune. Exim now panics and dies
3619 if the condition is not met. A case was found where this could be provoked
3620 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3621 components; fortuitously, this particular loophole had already been fixed
3622 by change 4.50/55 above.
3624 If there are any other similar loopholes, the new check in host_aton()
3625 itself should stop them being exploited. The report I received stated that
3626 data on the command line could provoke the exploit when Exim was running as
3627 exim, but did not say which command line option was involved. All I could
3628 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3629 running as the user.
3632 61. There was a buffer overflow vulnerability in the SPA authentication code
3633 (which came originally from the Samba project). I have added a test to the
3634 spa_base64_to_bits() function which I hope fixes it.
3637 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3638 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3640 63. The daemon start-up calls getloadavg() while still root for those OS that
3641 need the first call to be done as root, but it missed one case: when
3642 deliver_queue_load_max is set with deliver_drop_privilege. This is
3643 necessary for the benefit of the queue runner, because there is no re-exec
3644 when deliver_drop_privilege is set.
3646 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3647 This has been fixed.
3649 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3650 was in use, was not putting the data itself into the right store pool;
3651 consequently, it could be overwritten for a subsequent message in the same
3652 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3655 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3657 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3658 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3659 the function were treating the return as a boolean value, which happened to
3660 work because 0=false and not-0=true, but is not correct code.
3662 68. The host_aton() function was not handling scoped IPv6 addresses (those
3663 with, for example, "%eth0" on the end) correctly.
3665 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3666 negated items (that is, ~something) in unsigned ints. Some compilers
3667 apparently mutter when there is no cast.
3669 70. If an address verification called from an ACL failed, and did not produce a
3670 user-specific message (i.e. there was only a "system" message), nothing was
3671 put in $acl_verify_message. In this situation, it now puts the system
3674 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3675 session; change 4.31/43 added the unwanted input to the log line - except
3676 that it did not do this in the start of session case. It now does.
3678 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3679 This is wrong and can cause the other end to generate a synchronization
3680 error if it is another Exim or anything else that does the synchronization
3681 check. A QUIT command is no longer sent after a timeout.
3683 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3684 during host lookups.
3686 74. The defer_ok option of callout verification was not working if it was used
3687 when verifying addresses in header lines, that is, for this case:
3689 verify = header_sender/callout=defer_ok
3691 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3692 those file descriptors could be used for SMTP connections. If anything
3693 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3694 could be sent to the SMTP client, causing chaos. The daemon now opens
3695 stdin, stdout, and stderr to /dev/null when it puts itself into the
3698 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3699 log by default. The user can override this with suitable Perl magic.
3701 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3702 the log message when discard triggers, was not working for the DATA ACL or
3703 for the non-SMTP ACL.
3705 78. Error message wording change in sieve.c.
3707 79. If smtp_accept_max_per_host was set, the number of connections could be
3708 restricted to fewer than expected, because the daemon was trying to set up
3709 a new connection before checking whether the processes handling previous
3710 connections had finished. The check for completed processes is now done
3711 earlier. On busy systems, this bug wouldn't be noticed because something
3712 else would have woken the daemon, and it would have reaped the completed
3715 80. If a message was submitted locally by a user whose login name contained one
3716 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3717 It caused a spool format error. I have fixed the spool reading code. A
3718 related problem was that the "from" clause in the Received: line became
3719 illegal because of the space(s). It is now covered by ${quote_local_part.
3721 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3724 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3726 83. Patch from Sieve maintainer to fix unterminated string problem in
3727 "vacation" handling.
3729 84. Some minor changes to the Linux configuration files to help with other
3730 OS variants using glibc.
3732 85. One more patch for Sieve to update vacation handling to latest spec.
3735 ----------------------------------------------------
3736 See the note above about the 4.44 and 4.50 releases.
3737 ----------------------------------------------------
3743 1. Change 4.43/35 introduced a bug that caused file counts to be
3744 incorrectly computed when quota_filecount was set in an appendfile
3747 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3748 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3751 3. Give more explanation in the error message when the command for a transport
3752 filter fails to execute.
3754 4. There are several places where Exim runs a non-Exim command in a
3755 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3756 was being done only for the command run by the queryprogram router. It is
3757 now done for all such subprocesses. The other cases are: ${run, transport
3758 filters, and the commands run by the lmtp and pipe transports.
3760 5. Some older OS have a limit of 256 on the maximum number of file
3761 descriptors. Exim was using setrlimit() to set 1000 as a large value
3762 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3763 systems. I've change it so that if it can't get 1000, it tries for 256.
3765 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3766 was an oversight, and furthermore, ever since the addition of extra
3767 controls (e.g. 4.43/32), the checks on when to allow different forms of
3768 "control" were broken. There should now be diagnostics for all cases when a
3769 control that does not make sense is encountered.
3771 7. $recipients is now available in the predata ACL (oversight).
3773 8. Tidy the search cache before the fork to do a delivery from a message
3774 received from the command line. Otherwise the child will trigger a lookup
3775 failure and thereby defer the delivery if it tries to use (for example) a
3776 cached ldap connection that the parent has called unbind on.
3778 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3779 of $address_data from the recipient verification was clobbered by the
3780 sender verification.
3782 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3783 was its contents. (It was OK if the option was not defined at all.)
3785 11. A "Completed" log line is now written for messages that are removed from
3786 the spool by the -Mrm option.
3788 12. $host_address is now set to the target address during the checking of
3789 ignore_target_hosts.
3791 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3792 being passed; this would have caused $sender_host_name to have been used if
3793 matching the list had actually called for a host name (not very likely,
3794 since this list is usually IP addresses). A host name is now passed as
3797 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3798 code that specifies a non-restarting handler (typically sigaction() in
3799 modern systems) in an attempt to fix a rare and obscure crash bug.
3801 15. Narrowed the window for a race in the daemon that could cause it to ignore
3802 SIGCHLD signals. This is not a major problem, because they are used only to
3803 wake it up if nothing else does.
3805 16. A malformed maildirsize file could cause Exim to calculate negative values
3806 for the mailbox size or file count. Odd effects could occur as a result.
3807 The maildirsize information is now recalculated if the size or filecount
3810 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3811 support for a long time. Removed HAVE_SYS_VFS_H.
3813 18. Updated exipick to current release from John Jetmore.
3815 19. Allow an empty sender to be matched against a lookup in an address list.
3816 Previously the only cases considered were a regular expression, or an
3819 20. Exim went into a mad DNS lookup loop when doing a callout where the
3820 host was specified on the transport, if the DNS lookup yielded more than
3823 21. The RFC2047 encoding function was originally intended for short strings
3824 such as real names; it was not keeping to the 75-character limit for
3825 encoded words that the RFC imposes. It now respects the limit, and
3826 generates multiple encoded words if necessary. To be on the safe side, I
3827 have increased the buffer size for the ${rfc2047: expansion operator from
3830 22. Failure to deliver a bounce message always caused it to be frozen, even if
3831 there was an errors_to setting on the router. The errors_to setting is now
3834 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3835 canonical form (fully expanded) before being placed in
3836 $sender_host_address.
3838 24. Updated eximstats to version 1.33
3840 25. Include certificate and key file names in error message when GnuTLS fails
3841 to set them up, because the GnuTLS error message doesn't include the name
3842 of the failing file when there is a problem reading it.
3844 26. Expand error message when OpenSSL has problems setting up cert/key files.
3847 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3848 (this can affect the format of dates).
3850 28. exim_tidydb, when checking for the continued existence of a message for
3851 which it has found a message-specific retry record, was not finding
3852 messages that were in split spool directories. Consequently, it was
3853 deleting retry records that should have stayed in existence.
3855 29. eximstats updated to version 1.35
3856 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3857 1.35 - bugfix such that pie charts by volume are generated correctly
3859 30. The SPA authentication driver was not abandoning authentication and moving
3860 on to the next authenticator when an expansion was forced to fail,
3861 contradicting the general specification for all authenticators. Instead it
3862 was generating a temporary error. It now behaves as specified.
3864 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3865 (the order specifies the preference for clients). The order is now AES256,
3866 AES128, 3DES, ARCFOUR128.
3868 31. Small patch to Sieve code - explicitly set From: when generating an
3871 32. Exim crashed if a remote delivery caused a very long error message to be
3872 recorded - for instance if somebody sent an entire SpamAssassin report back
3873 as a large number of 550 error lines. This bug was coincidentally fixed by
3874 increasing the size of one of Exim's internal buffers (big_buffer) that
3875 happened as part of the Exiscan merge. However, to be on the safe side, I
3876 have made the code more robust (and fixed the comments that describe what
3879 33. Some experimental protocols are using DNS PTR records for new purposes. The
3880 keys for these records are domain names, not reversed IP addresses. The
3881 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3882 leaves it alone. Component reversal etc. now happens only for IP addresses.
3885 34. The host_aton() function is supposed to be passed a string that is known
3886 to be a valid IP address. However, in the case of IPv6 addresses, it was
3887 not checking this. This is a hostage to fortune. Exim now panics and dies
3888 if the condition is not met. A case was found where this could be provoked
3889 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3890 components; fortuitously, this particular loophole had already been fixed
3891 by change 4.50/55 or 4.44/33 above.
3893 If there are any other similar loopholes, the new check in host_aton()
3894 itself should stop them being exploited. The report I received stated that
3895 data on the command line could provoke the exploit when Exim was running as
3896 exim, but did not say which command line option was involved. All I could
3897 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3898 running as the user.
3901 35. There was a buffer overflow vulnerability in the SPA authentication code
3902 (which came originally from the Samba project). I have added a test to the
3903 spa_base64_to_bits() function which I hope fixes it.
3906 36. The daemon start-up calls getloadavg() while still root for those OS that
3907 need the first call to be done as root, but it missed one case: when
3908 deliver_queue_load_max is set with deliver_drop_privilege. This is
3909 necessary for the benefit of the queue runner, because there is no re-exec
3910 when deliver_drop_privilege is set.
3912 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3913 was in use, was not putting the data itself into the right store pool;
3914 consequently, it could be overwritten for a subsequent message in the same
3915 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3918 38. Sometimes the final signoff response after QUIT could fail to get
3919 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3920 < 0 before doing a fflush(). This bug looks as though it goes back to the
3921 introduction of TLS in release 3.20, but "sometimes" must have been rare
3922 because the tests only now provoked it.
3928 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3929 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3930 Now it returns an integer. A number of calls were still expecting a T/F
3931 return. Fortuitously, in all cases, the tests worked in OK situations,
3932 which is the norm. However, things would have gone wrong on any write
3933 failures on the smtp file descriptor. This function is used when sending
3934 messages over SMTP and also when doing verify callouts.
3936 2. When Exim is called to do synchronous delivery of a locally submitted
3937 message (the -odf or -odi options), it no longer closes stderr before doing
3940 3. Implemented the mua_wrapper option.
3942 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3944 5. Implemented the functions header_remove(), header_testname(),
3945 header_add_at_position(), and receive_remove_recipient(), and exported them
3948 6. If an ACL "warn" statement specified the addition of headers, Exim already
3949 inserted X-ACL-Warn: at the start if there was no header name. However, it
3950 was not making this test for the second and subsequent header lines if
3951 there were newlines in the string. This meant that an invalid header could
3952 be inserted if Exim was badly configured.
3954 7. Allow an ACL "warn" statement to add header lines at the start or after all
3955 the Received: headers, as well as at the end.
3957 8. Added the rcpt_4xx retry error code.
3959 9. Added postmaster_mailfrom=xxx to callout verification option.
3961 10. Added mailfrom=xxxx to the callout verification option, for verify=
3964 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3965 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3967 12. Inserted some casts to stop certain compilers warning when using pointer
3968 differences as field lengths or precisions in printf-type calls (mostly
3969 affecting debugging statements).
3971 13. Added optional readline() support for -be (dynamically loaded).
3973 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3974 same clock tick as a message's arrival, so that its received time was the
3975 same as the "first fail" time on the retry record, and that message
3976 remained on the queue past the ultimate address timeout, every queue runner
3977 would try a delivery (because it was past the ultimate address timeout) but
3978 after another failure, the ultimate address timeout, which should have then
3979 bounced the address, did not kick in. This was a "< instead of <=" error;
3980 in most cases the first failure would have been in the next clock tick
3981 after the received time, and all would be well.
3983 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3984 being recognized when the domain list was tested by the match_domain
3985 condition in an expansion string.
3987 16. Added the ${str2b64: operator.
3989 17. Exim was always calling setrlimit() to set a large limit for the number of
3990 processes, without checking whether the existing limit was already
3991 adequate. (It did check for the limit on file descriptors.) Furthermore,
3992 errors from getrlimit() and setrlimit() were being ignored. Now they are
3993 logged to the main and panic logs, but Exim does carry on, to try to do its
3994 job under whatever limits there are.
3996 18. Imported PCRE 5.0.
3998 19. Trivial typo in log message " temporarily refused connection" (the leading
4001 20. If the log selector return_path_on_delivery was set and an address was
4002 redirected to /dev/null, the delivery process crashed because it assumed
4003 that a return path would always be set for a "successful" delivery. In this
4004 case, the whole delivery is bypassed as an optimization, and therefore no
4007 21. Internal re-arrangement: the function for sending a challenge and reading
4008 a response while authentication was assuming a zero-terminated challenge
4009 string. It's now changed to take a pointer and a length, to allow for
4010 binary data in such strings.
4012 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4014 23. Exim was not respecting finduser_retries when seeking the login of the
4015 uid under which it was called; it was always trying 10 times. (The default
4016 setting of finduser_retries is zero.) Also, it was sleeping after the final
4017 failure, which is pointless.
4019 24. Implemented tls_on_connect_ports.
4021 25. Implemented acl_smtp_predata.
4023 26. If the domain in control=submission is set empty, Exim assumes that the
4024 authenticated id is a complete email address when it generates From: or
4025 Sender: header lines.
4027 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4028 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4029 chown and chgrp in /bin and hostname in /usr/bin.
4031 28. Exim was keeping the "process log" file open after each use, just as it
4032 does for the main log. This opens the possibility of it remaining open for
4033 long periods when the USR1 signal hits a daemon. Occasional processlog
4034 errors were reported, that could have been caused by this. Anyway, it seems
4035 much more sensible not to leave this file open at all, so that is what now
4038 29. The long-running daemon process does not normally write to the log once it
4039 has entered its main loop, and it closes the log before doing so. This is
4040 so that log files can straightforwardly be renamed and moved. However,
4041 there are a couple of unusual error situations where the daemon does write
4042 log entries, and I had neglected to close the log afterwards.
4044 30. The text of an SMTP error response that was received during a remote
4045 delivery was being truncated at 512 bytes. This is too short for some of
4046 the long messages that one sometimes sees. I've increased the limit to
4049 31. It is now possible to make retry rules that apply only when a message has a
4050 specific sender, in particular, an empty sender.
4052 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4053 it possible to be selective about when SMTP synchronization is enforced.
4055 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4057 32. Implemented hosts_connection_nolog.
4059 33. Added an ACL for QUIT.
4061 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4064 35. Added mailbox_size and mailbox_filecount to appendfile.
4066 36. Added control = no_multiline_responses to ACLs.
4068 37. There was a bug in the logic of the code that waits for the clock to tick
4069 in the case where the clock went backwards by a substantial amount such
4070 that the microsecond fraction of "now" was more than the microsecond
4071 fraction of "then" (but the whole seconds number was less).
4073 38. Added support for the libradius Radius client library this is found on
4074 FreeBSD (previously only the radiusclient library was supported).
4080 1. When certain lookups returned multiple values in the form name=value, the
4081 quoting of the values was not always being done properly. Specifically:
4082 (a) If the value started with a double quote, but contained no whitespace,
4084 (b) If the value contained whitespace other than a space character (i.e.
4085 tabs or newlines or carriage returns) it was not quoted.
4086 This fix has been applied to the mysql and pgsql lookups by writing a
4087 separate quoting function and calling it from the lookup code. The fix
4088 should probably also be applied to nisplus, ibase and oracle lookups, but
4089 since I cannot test any of those, I have not disturbed their existing code.
4091 2. A hit in the callout cache for a specific address caused a log line with no
4092 reason for rejecting RCPT. Now it says "Previous (cached) callout
4093 verification failure".
4095 3. There was an off-by-one bug in the queryprogram router. An over-long
4096 return line was truncated at 256 instead of 255 characters, thereby
4097 overflowing its buffer with the terminating zero. As well as fixing this, I
4098 have increased the buffer size to 1024 (and made a note to document this).
4100 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4101 when Exim is waiting for an SMTP response from a remote server, Exim
4102 restarts its select() call on the socket, thereby resetting its timeout.
4103 This is not a problem when such interrupts are rare. Somebody set up a cron
4104 job to run exiwhat every 2 minutes, which is less than the normal select()
4105 timeout (5 or 10 minutes). This meant that the select() timeout never
4106 kicked in because it was always reset. I have fixed this by comparing the
4107 time when an interrupt arrives with the time at the start of the first call
4108 to select(). If more time than the timeout has elapsed, the interrupt is
4109 treated as a timeout.
4111 5. Some internal re-factoring in preparation for the addition of Sieve
4112 extensions (by MH). In particular, the "personal" test is moved to a
4113 separate function, and given an option for scanning Cc: and Bcc: (which is
4114 not set for Exim filters).
4116 6. When Exim created an email address using the login of the caller as the
4117 local part (e.g. when creating a From: or Sender: header line), it was not
4118 quoting the local part when it contained special characters such as @.
4120 7. Installed new OpenBSD configuration files.
4122 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4123 try to make them clearer.
4125 9. Callout options, other than the timeout value, were being ignored when
4126 verifying sender addresses in header lines. For example, when using
4128 verify = header_sender/callout=no_cache
4130 the cache was (incorrectly) being used.
4132 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4133 only the Cygwin environment.
4135 11. When return_path_on_delivery was set as a log selector, if different remote
4136 addresses in the same message used different return paths and parallel
4137 remote delivery occurred, the wrong values would sometimes be logged.
4138 (Whenever a remote delivery process finished, the return path value from
4139 the most recently started remote delivery process was logged.)
4141 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4142 lines when AUTH and/or TLS are in use. This is the "received protocol"
4143 field. Exim used to use "asmtp" for authenticated SMTP, without any
4144 indication (in the protocol name) for TLS use. Now it follows the RFC and
4145 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4146 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4147 names appear in log lines as well as in Received: header lines.
4149 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4150 extensions, and comparison tests, and to fix some bugs.
4152 14. Changes to the "personal" filter test:
4154 (1) The test was buggy in that it was just doing the equivalent of
4155 "contains" tests on header lines. For example, if a user's address was
4156 anne@some.where, the "personal" test would incorrectly be true for
4158 To: susanne@some.where
4160 This test is now done by extracting each address from the header in turn,
4161 and checking the entire address. Other tests that are part of "personal"
4162 are now done using regular expressions (for example, to check local parts
4163 of addresses in From: header lines).
4165 (2) The list of non-personal local parts in From: addresses has been
4166 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4167 taken from the Sieve specification recommendations.
4169 (3) If the message contains any header line starting with "List-" it is
4170 treated as non-personal.
4172 (4) The test for "circular" in the Subject: header line has been removed
4173 because it now seems ill-conceived.
4175 15. Minor typos in src/EDITME comments corrected.
4177 16. Installed latest exipick from John Jetmore.
4179 17. If headers_add on a router specified a text string that was too long for
4180 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4181 of string_sprintf() is now avoided.
4183 18. $message_body_size was not set (it was always zero) when running the DATA
4184 ACL and the local_scan() function.
4186 19. For the "mail" command in an Exim filter, no default was being set for
4187 the once_repeat time, causing a random time value to be used if "once" was
4188 specified. (If the value happened to be <= 0, no repeat happened.) The
4189 default is now 0s, meaning "never repeat". The "vacation" command was OK
4190 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4191 (I found it when inspecting the code).
4193 20. There is now an overall timeout for performing a callout verification. It
4194 defaults to 4 times the callout timeout, which applies to individual SMTP
4195 commands during the callout. The overall timeout applies when there is more
4196 than one host that can be tried. The timeout is checked before trying the
4197 next host. This prevents very long delays if there are a large number of
4198 hosts and all are timing out (e.g. when the network connections are timing
4199 out). The value of the overall timeout can be changed by specifying an
4200 additional sub-option for "callout", called "maxwait". For example:
4202 verify = sender/callout=5s,maxwait=20s
4204 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4205 to the end before writing, but this should make it even safer).
4207 22. Exim was forgetting that it had advertised PIPELINING for the second and
4208 subsequent messages on an SMTP connection. It was also not resetting its
4209 memory on STARTTLS and an internal HELO.
4211 23. When Exim logs an SMTP synchronization error within a session, it now
4212 records whether PIPELINING has been advertised or not.
4214 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4215 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4216 rather than long int.
4218 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4220 26. Added the never_mail option to autoreply.
4226 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4227 crash if the getsockname() call failed; this can happen if a connection is
4228 closed very soon after it is established. The problem was simply in the
4229 order in which certain operations were done, causing Exim to try to write
4230 to the SMTP stream before it had set up the file descriptor. The bug has
4231 been fixed by making things happen in the correct order.
4237 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4238 before the connection was closed, thus losing the rejection response.
4240 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4241 some early Solaris releases, but causes trouble in current releases where
4242 socklen_t is defined.
4244 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4247 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4250 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4251 that releasing the top of it at the end releases what was used for sub-
4252 expansions (unless the block got too big). However, discard this block if
4253 the first thing is a variable or header, so that we can use its block when
4254 it is dynamic (useful for very large $message_headers, for example).
4256 6. Lookups now cache *every* query, not just the most recent. A new, separate
4257 store pool is used for this. It can be recovered when all lookup caches are
4258 flushed. Lookups now release memory at the end of their result strings.
4259 This has involved some general refactoring of the lookup sources.
4261 7. Some code has been added to the store_xxx() functions to reduce the amount
4262 of flapping under certain conditions.
4264 8. log_incoming_interface used to affect only the <= reception log lines. Now
4265 it causes the local interface and port to be added to several more SMTP log
4266 lines, for example "SMTP connection from", and rejection lines.
4268 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4270 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4272 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4273 could be overwritten at the end of the current message (or the start of a
4274 new message if it was set in a HELO ACL). The value is now preserved for
4275 the duration of the SMTP connection.
4277 12. If a transport had a headers_rewrite setting, and a matching header line
4278 contained an unqualified address, that address was qualified, even if it
4279 did not match any rewriting rules. The underlying bug was that the values
4280 of the flags that permit the existence of unqualified sender and recipient
4281 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4282 non-local messages, and by -bnq for local messages) were not being
4283 preserved with the message after it was received.
4285 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4286 "next input=" as part of the text comprising the host identity instead of
4287 the correct text. The code was using the same buffer for two different
4288 strings. However, depending on which order the printing function evaluated
4289 its arguments, the bug did not always show up. Under Linux, for example, my
4290 test suite worked just fine.
4292 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4293 doesn't work with some older versions of Perl. It has been changed to "my",
4294 which in any case is probably the better facility to use.
4296 15. A really picky compiler found some instances of statements for creating
4297 error messages that either had too many or two few arguments for the format
4300 16. The size of the buffer for calls to the DNS resolver has been increased
4301 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4302 for addresses that have a lot of PTR records. This alleviates a problem; it
4303 does not fully solve it.
4305 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4306 buffer now truncates the list and logs the incident, which is the same
4307 action as happens when Exim is looking up a host name and its aliases.
4308 Previously in this situation something unpredictable would happen;
4309 sometimes it was "internal error: store_reset failed".
4311 18. If a server dropped the connection unexpectedly when an Exim client was
4312 using GnuTLS and trying to read a response, the client delivery process
4313 crashed while trying to generate an error log message.
4315 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4316 string, for example:
4318 warn message = H1: something\nH2: something
4320 the text was added as a single header line from Exim's point of view
4321 though it ended up OK in the delivered message. However, searching for the
4322 second and subsequent header lines using $h_h2: did not work. This has been
4323 fixed. Similarly, if a system filter added multiple headers in this way,
4324 the routers could not see them.
4326 20. Expanded the error message when iplsearch is called with an invalid key to
4327 suggest using net-iplsearch in a host list.
4329 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4330 ACLs are no longer actually imposed (and a message to that effect is
4333 22. If a "gecos" field in a passwd entry contained escaped characters, in
4334 particular, if it contained a \" sequence, Exim got it wrong when building
4335 a From: or a Sender: header from that name. A second bug also caused
4336 incorrect handling when an unquoted " was present following a character
4337 that needed quoting.
4339 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4340 was not being matched caselessly.
4342 24. Arranged for all hyphens in the exim.8 source to be escaped with
4345 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4346 properly. Recipient callout cache records were still being keyed to include
4347 the sender, even when use_sender was set false. This led to far more
4348 callouts that were necessary. The sender is no longer included in the key
4349 when use_sender is false.
4351 26. Added "control = submission" modifier to ACLs.
4353 27. Added the ${base62d: operator to decode base 62 numbers.
4355 28. dnsdb lookups can now access SRV records.
4357 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4358 the configuration file.
4360 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4361 (-v) mode. This makes the output for a verbose queue run more intelligible.
4363 31. Added a use_postmaster feature to recipient callouts.
4365 32. Added the $body_zerocount variable, containing the number of binary zero
4366 bytes in the message body.
4368 33. The time of last modification of the "new" subdirectory is now used as the
4369 "mailbox time last read" when there is a quota error for a maildir
4372 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4374 35. Added +ignore_unknown as a special item in host lists.
4376 36. Code for decoding IPv6 addresses in host lists is now included, even if
4377 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4378 address was recognized as an IP address, but was then not correctly decoded
4379 into binary, causing unexpected and incorrect effects when compared with
4386 1. Very minor rewording of debugging text in manualroute to say "list of
4387 hosts" instead of "hostlist".
4389 2. If verify=header_syntax was set, and a header line with an unqualified
4390 address (no domain) and a large number of spaces between the end of the
4391 name and the colon was received, the reception process suffered a buffer
4392 overflow, and (when I tested it) crashed. This was caused by some obsolete
4393 code that should have been removed. The fix is to remove it!
4395 3. When running in the test harness, delay a bit after writing a bounce
4396 message to get a bit more predictability in the log output.
4398 4. Added a call to search_tidyup() just before forking a reception process. In
4399 theory, someone could use a lookup in the expansion of smtp_accept_max_
4400 per_host which, without the tidyup, could leave open a database connection.
4402 5. Added the variables $recipient_data and $sender_data which get set from a
4403 lookup success in an ACL "recipients" or "senders" condition, or a router
4404 "senders" option, similar to $domain_data and $local_part_data.
4406 6. Moved the writing of debug_print from before to after the "senders" test
4409 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4410 problems for message scanning, either using a data ACL, or using
4411 local_scan() because the Received: header was not generated till after they
4412 were called (in order to set the time as the time of reception completion).
4413 I have revised the way this works. The header is now generated after the
4414 body is received, but before the ACL or local_scan() are called. After they
4415 are run, the timestamp in the header is updated.
4421 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4422 before starting a queue runner without re-exec. This happened only when
4423 deliver_drop_privilege was set or when the Exim user was set to root. The
4424 effect of the bug was that timeouts during subsequent deliveries caused
4425 crashes instead of being properly handled. The handler is now left at its
4426 default (and expected) setting.
4428 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4429 message, again when deliver_drop_privilege is set or Exim is run as root.
4430 The bug described in (1) was not present in this case, but the tidying up
4431 of the other signals was missing. I have made the two cases consistent.
4433 3. The ignore_target_hosts setting on a manualroute router was being ignored
4434 for hosts that were looked up using the /MX notation.
4436 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4439 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4440 operated on the sender address. After changing the $sender_address to <>
4441 for the sender address verify, Exim was re-instated it as the original
4442 (before rewriting) address, but remembering that it had rewritten it, so it
4443 wasn't rewriting it again. This bug also had the effect of breaking the
4444 sender address verification caching when the sender address was rewritten.
4446 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4447 This has been changed so that if the ip literal address matches
4448 ignore_target_hosts, the router declines.
4450 7. Added expansion conditions match_domain, match_address, and match_local_
4451 part (NOT match_host).
4453 8. The placeholder for the Received: header didn't have a length field set.
4455 9. Added code to Exim itself and to exim_lock to test for a specific race
4456 condition that could lead to file corruption when using MBX delivery. The
4457 issue is with the lockfile that is created in /tmp. If this file is removed
4458 after a process has opened it but before that process has acquired a lock,
4459 there is the potential for a second process to recreate the file and also
4460 acquire a lock. This could lead to two Exim processes writing to the file
4461 at the same time. The added code performs the same test as UW imapd; it
4462 checks after acquiring the lock that its file descriptor still refers to
4463 the same named file.
4465 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4466 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4469 11. Added the smtp_active_hostname option. If used, this will typically be made
4470 to depend on the incoming interface address. Because $interface_address is
4471 not set up until the daemon has forked a reception process, error responses
4472 that can happen earlier (such as "too many connections") no longer contain
4475 12. If an expansion in a condition on a "warn" statement fails because a lookup
4476 defers, the "warn" statement is abandoned, and the next ACL statement is
4477 processed. Previously this caused the whole ACL to be aborted.
4479 13. Added the iplsearch lookup type.
4481 14. Added ident_timeout as a log selector.
4483 15. Added tls_certificate_verified as a log selector.
4485 16. Added a global option tls_require_ciphers (compare the smtp transport
4486 option of the same name). This controls incoming TLS connections.
4488 17. I finally figured out how to make tls_require_ciphers do a similar thing
4489 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4490 before starting the TLS session.
4492 18. Tabs are now shown as \t in -bP output.
4494 19. If the log selector return_path_on_delivery was set, Exim crashed when
4495 bouncing a message because it had too many Received: header lines.
4497 20. If two routers both had headers_remove settings, and the first one included
4498 a superfluous trailing colon, the final name in the first list and the
4499 first name in the second list were incorrectly joined into one item (with a
4500 colon in the middle).
4506 1. Added -C and -D options to the exinext utility, mainly to make it easier
4507 to include in the automated testing, but these could be helpful when
4508 multiple configurations are in use.
4510 2. The exinext utility was not formatting the output nicely when there was
4511 an alternate port involved in the retry record key, nor when there was a
4512 message id as well (for retries that were specific to a specific message
4513 and a specific host). It was also confused by IPv6 addresses, because of
4514 the additional colons they contain. I have fixed the IPv4 problem, and
4515 patched it up to do a reasonable job for IPv6.
4517 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4518 delivery, the log line now contains "pipelined" if PIPELINING was used.
4520 4. An SMTP transport process used to panic and die if the bind() call to set
4521 an explicit outgoing interface failed. This has been changed; it is now
4522 treated in the same way as a connect() failure.
4524 5. A reference to $sender_host_name in the part of a conditional expansion
4525 that was being skipped was still causing a DNS lookup. This no longer
4528 6. The def: expansion condition was not recognizing references to header lines
4529 that used bh_ and bheader_.
4531 7. Added the _cache feature to named lists.
4533 8. The code for checking quota_filecount in the appendfile transport was
4534 allowing one more file than it should have been.
4536 9. For compatibility with Sendmail, the command line option
4544 and sets the incoming protocol and host name (for trusted callers). The
4545 host name and its colon can be omitted when only the protocol is to be set.
4546 Note the Exim already has two private options, -pd and -ps, that refer to
4547 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4548 "s", but I don't think that's a major issue.
4550 10. A number of refactoring changes to the code, none of which should affect
4553 (a) The number of logging options was getting close to filling up the
4554 32-bit word that was used as a bit map. I have split them into two classes:
4555 those that are passed in the argument to log_write(), and those that are
4556 only ever tested independently outside of that function. These are now in
4557 separate 32-bit words, so there is plenty of room for expansion again.
4558 There is no change in the user interface or the logging behaviour.
4560 (b) When building, for example, log lines, the code previously used a
4561 macro that called string_cat() twice, in order to add two strings. This is
4562 not really sufficiently general. Furthermore, there was one instance where
4563 it was actually wrong because one of the argument was used twice, and in
4564 one call a function was used. (As it happened, calling the function twice
4565 did not affect the overall behaviour.) The macro has been replaced by a
4566 function that can join an arbitrary number of extra strings onto a growing
4569 (c) The code for expansion conditions now uses a table and a binary chop
4570 instead of a serial search (which was left over from when there were very
4571 few conditions). Also, it now recognizes conditions like "pam" even when
4572 the relevant support is not compiled in: a suitably worded error message is
4573 given if an attempt is made to use such a condition.
4575 11. Added ${time_interval:xxxxx}.
4577 12. A bug was causing one of the ddress fields not to be passed back correctly
4578 from remote delivery subprocesses. The field in question was not being
4579 subsequently used, so this caused to problems in practice.
4581 13. Added new log selectors queue_time and deliver_time.
4583 14. Might have fixed a bug in maildirsizefile handling that threw up
4584 "unexpected character" debug warnings, and recalculated the data
4585 unnecessarily. In any case, I expanded the warning message to give more
4588 15. Added the message "Restricted characters in address" to the statements in
4589 the default ACL that block characters like @ and % in local parts.
4591 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4592 Three changes have been made:
4594 (a) There was a serious bug; a negative response to MAIL caused the whole
4595 recipient domain to be cached as invalid, thereby blocking all messages
4596 to all local parts at the same domain, from all senders. This bug has
4597 been fixed. The domain is no longer cached after a negative response to
4598 MAIL if the sender used is not empty.
4600 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4603 (c) A new callout option, "use_sender" has been added for people who want
4604 the modified behaviour.
4610 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4613 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4614 indeed breaks things for older releases.
4616 3. Added additional logging to the case where there is a problem reading data
4617 from a filter that is running in a subprocess using a pipe, in order to
4618 try to track down a specific problem.
4620 4. Testing facility fudge: when running in the test harness and attempting
4621 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4622 getting "No route to host". Convert this to a timeout.
4624 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4627 6. Some OS don't have socklen_t but use size_t instead. This affects the
4628 fifth argument of getsockopt() amongst other things. This is now
4629 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4630 can be set for individual OS. I have set it for SunOS5, OSF1, and
4631 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4632 some earlier ones do not.
4634 7. Change 4.30/15 was not doing the test caselessly.
4636 8. The standard form for an IPv6 address literal was being rejected by address
4637 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4638 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4639 this, as well as the form without the "IPv6" on the front (but only when
4640 address literals are enabled, of course).
4642 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4644 10. Exim crashed if a message with an empty sender address specified by -f
4645 encountered a router with an errors_to setting. This could be provoked only
4646 by a command such as
4650 where an empty string was supplied; "<>" did not hit this bug.
4652 11. Installed PCRE release 4.5.
4654 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4655 remained set. It is now erased.
4657 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4658 times from message ids (which are base 36 rather than the normal 62).
4660 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4661 were being counted as actual protocol errors, and logged if the log
4662 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4663 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4664 and DATA following a set of rejected RCPTs do not count as protocol errors.
4665 In other words, Exim assumes they were pipelined, though this may not
4666 actually be the case. Of course, in all cases the client gets an
4667 appropriate error code.
4669 15. If a lookup fails in an ACL condition, a message about the failure may
4670 be available; it is used if testing the ACL cannot continue, because most
4671 such messages specify what the cause of the deferral is. However, some
4672 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4673 that caused an old message to be retained and used if a later statement
4674 caused a defer, replacing the real cause of the deferral.
4676 16. If an IP address had so many PTR records that the DNS lookup buffer
4677 was not large enough to hold them, Exim could crash while trying to process
4678 the truncated data. It now detects and logs this case.
4680 17. Further to 4.21/58, another change has been made: if (and only if) the
4681 first line of a message (the first header line) ends with CRLF, a bare LF
4682 in a subsequent header line has a space inserted after it, so as not to
4683 terminate the header.
4685 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4686 unnecessarily, used atoi() instead of strtol(), and didn't check the
4687 termination when getting file sizes from file names by regex.
4689 19. Completely re-implemented the support for maildirsize files, in the light
4690 of a number of problems with the previous contributed implementation
4691 (4.30/29). In particular:
4693 . If the quota is zero, the maildirsize file is maintained, but no quota is
4696 . If the maildir directory does not exist, it is created before any attempt
4697 to write a maildirsize file.
4699 . The quota value in the file is just a cache; if the quota is changed in
4700 the transport, the new value overrides.
4702 . A regular expression is available for excluding directories from the
4705 20. The autoreply transport checks the characters in options that define the
4706 message's headers; it allows continued headers, but it was checking with
4707 isspace() after an embedded newline instead of explicitly looking for a
4710 21. If all the "regular" hosts to which an address was routed had passed their
4711 expiry times, and had not reached their retry times, the address was
4712 bounced, even if fallback hosts were defined. Now Exim should go on to try
4715 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4716 equivalent code in the SMTP transport. Some hosts send humungous responses
4717 to HELO/EHLO, more than 1024 it seems.
4719 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4720 gives compiler warnings in some environments. I've now done it "properly",
4723 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4724 (because of problems with the built-in one) was declared to return uschar *
4725 instead of char *, causing compiler failure.
4727 25. Fixed a file descriptor leak when processing alias/forward files.
4729 26. Fixed a minor format string issue in dbfn.c.
4731 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4733 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4734 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4737 29. When a sender address is verified, it is cached, to save repeating the test
4738 when there is more than one recipient in a message. However, when the
4739 verification involves a callout, it is possible for different callout
4740 options to be set for different recipients. It is too complicated to keep
4741 track of this in the cache, so now Exim always runs a verification when a
4742 callout is required, relying on the callout cache for the optimization.
4743 The overhead is duplication of the address routing, but this should not be
4746 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4747 to be verified with callout=postmaster, and the main callout worked but the
4748 postmaster check failed, the verification correctly failed. However, if a
4749 subsequent RCPT command asked for sender verification *without* the
4750 postmaster check, incorrect caching caused this verification also to fail,
4753 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4754 it was not caching the DNS options (qualify_single, search_parents) that
4755 were used when the lookup failed. A subsequent lookup with different
4756 options therefore always gave the same answer, though there were cases
4757 where it should not have. (Example: a "domains = !$mx_any" option on a
4758 dnslookup router: the "domains" option is always processed without any
4759 widening, but the router might have qualify_single set.) Now Exim uses the
4760 cached value only when the same options are set.
4762 32. Added John Jetmore's "exipick" utility to the distribution.
4764 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4765 than a timeout (e.g. a certificate is required, and is not provided), an
4766 Exim server now closes the connection immediately. Previously it waited for
4767 the client to close - but if the client is SSL, it seems that they each
4768 wait for each other, leading to a delay before one of them times out.
4770 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4771 maintained 0.8.x compatibility because I don't think many are using it, and
4772 it is clearly obsolete.
4774 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4777 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4778 client certificate was expired. A simple patch fixes this, though I don't
4779 understand the full logic of why the verify callback is called multiple
4782 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4783 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4784 which causes problems with some clients (such as the Certicom SSL Plus
4785 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4786 disables the coutermeasure allowing Eudora to connect."
4788 38. Exim was not checking that a write() to a log file succeeded. This could
4789 lead to Bad Things if a log got too big, in particular if it hit a file
4790 size limit. Exim now panics and dies if it cannot write to a log file, just
4791 as it does if it cannot open a log file.
4793 39. Modified OS/Makefile-Linux so that it now contains
4795 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4797 The two -D definitions ensure that Exim is compiled with large file
4798 support, which makes it possible to handle log files that are bigger than
4801 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4802 instance) a domain was checked against a named list that involved a lookup,
4803 causing $domain_data to be set, then another domain was checked against the
4804 same list, then the first domain was re-checked, the value of $domain_data
4805 after the final check could be wrong. In particular, if the second check
4806 failed, it could be set empty. This bug probably also applied to
4809 41. The strip_trailing_dot option was not being applied to the address given
4810 with the -f command-line option.
4812 42. The code for reading a message's header from the spool was incrementing
4813 $received_count, but never initializing it. This meant that the value was
4814 incorrect (doubled) while delivering a message in the same process in which
4815 it was received. In the most common configuration of Exim, this never
4816 happens - a fresh exec is done - but it can happen when
4817 deliver_drop_privilege is set.
4819 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4820 it now includes up to 150 characters of the unexpected data in the log
4823 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4824 and building data strings. The size of both of these buffers was 10 000
4825 bytes - far larger than anybody would *ever* want, thought I. Needless to
4826 say, somebody hit the limit. I have increased the maximum line length to
4827 20 000 and the maximum data length of concatenated lines to 100 000. I have
4828 also fixed two bugs, because there was no checking on these buffers. Tsk,
4829 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4830 buffer is too small.
4832 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4833 lsearch lookups. Now it does.
4835 46. When parsing a route_list item in a manualroute router, a fixed-length
4836 buffer was used for the list of hosts. I made this 1024 bytes long,
4837 thinking that nobody would ever have a list of hosts that long. Wrong.
4838 Somebody had a whole pile of complicated expansion conditions, and the
4839 string was silently truncated, leading to an expansion error. It turns out
4840 that it is easier to change to an unlimited length (owing to other changes
4841 that have happened since this code was originally written) than to build
4842 structure for giving a limitation error. The length of the item that
4843 expands into the list of hosts is now unlimited.
4845 47. The lsearch lookup could not handle data where the length of text line was
4846 more than 4095 characters. Such lines were truncated, leading to shortened
4847 data being returned. It should now handle lines of any length.
4849 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4850 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4853 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4854 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4855 no longer happens in comment lines. A list of replacements is now placed
4856 at the head of all of the source files, except those whose only change is
4857 to replace PERL_COMMAND in the very first #! line.
4859 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4860 messages on the queue, with a bottom-up merge sort, using code contributed
4861 by Michael Haardt. This should make operations like -bp somewhat faster on
4862 large queues. It won't affect queue runners, except when queue_run_in_order
4865 51. Installed eximstats 1.31 in the distribution.
4867 52. Added support for SRV lookups to the dnslookup router.
4869 53. If an ACL referred to $message_body or $message_body_end, the value was not
4870 reset for any messages that followed in the same SMTP session.
4872 54. The store-handling optimization for building very long strings was not
4873 differentiating between the different store pools. I don't think this
4874 actually made any difference in practice, but I've tidied it.
4876 55. While running the routers to verify a sender address, $sender_address
4877 was still set to the sender address. This is wrong, because when routing to
4878 send a bounce to the sender, it would be empty. Therefore, I have changed
4879 it so that, while verifying a sender address, $sender_address is set to <>.
4880 (There is no change to what happens when verifying a recipient address.)
4882 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4883 target A or AAAA records (if not already returned) without resetting the
4884 qualify_single or search_parents options of the DNS resolver. These are
4885 inappropriate in this case because the targets of MX and SRV records must
4886 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4887 target that, when qualified, matched something in the local domain. These
4888 two options are now turned off when doing these lookups.
4890 57. It seems that at least some releases of Reiserfs (which does not have the
4891 concept of a fixed number of inodes) returns zero and not -1 for the
4892 number of available inodes. This interacted badly with check_spool_inodes,
4893 which assumed that -1 was the "no such thing" setting. What I have done is
4894 to check that the total number of inodes is greater than zero before doing
4895 the test of how many are available.
4897 58. When a "warn" ACL statement has a log_message modifier, the message is
4898 remembered, and not repeated. This is to avoid a lot of repetition when a
4899 message has many recipients that cause the same warning to be written.
4900 Howewer, Exim was preserving the list of already written lines for an
4901 entire SMTP session, which doesn't seem right. The memory is now reset if a
4902 new message is started.
4904 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4905 debugging output unless log_rewrite was also set.
4907 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4908 of (int)(handle) when we know that handle contains (void *)(-1).
4910 61. The Exim daemon panic-logs an error return when it closes the incoming
4911 connection. However "connection reset by peer" seems to be common, and
4912 isn't really an error worthy of noting specially, so that particular error
4915 62. When Exim is trying to find all the local interfaces, it used to panic and
4916 die if the ioctl to get the interface flags failed. However, it seems that
4917 on at least one OS (Solaris 9) it is possible to have an interface that is
4918 included in the list of interfaces, but for which you get a failure error
4919 for this call. This happens when the interface is not "plumbed" into a
4920 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4921 failure of the "get flags" call assumes that the interface is down.
4923 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4924 makes life easier for people who are doing arithmetic on fields extracted
4925 from dates, where you often get leading zeros that should not be
4926 interpreted as octal.
4928 64. Added qualify_domain to the redirect router, to override the global
4931 65. If a pathologically long header line contained very many addresses (the
4932 report of this problem mentioned 10 000) and each of them was rewritten,
4933 Exim could use up a very large amount of memory. (It kept on making new
4934 copies of the header line as it rewrote, and never released the old ones.)
4935 At the expense of a bit more processing, the header rewriting function has
4936 been changed so that it no longer eats memory in this way.
4938 66. The generation of the Received: header has been moved from the time that a
4939 message starts to be received, to the time that it finishes. The timestamp
4940 in the Received: header should now be very close to that of the <= log
4941 line. There are two side-effects of this change:
4943 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4944 logged header lines no longer include the local Received: line, because
4945 it has not yet been created. The same applies to a copy of the message
4946 that is returned to a non-SMTP sender when a message is rejected.
4948 (b) When a filter file is tested using -bf, no additional Received: header
4949 is added to the test message. After some thought, I decided that this
4952 This change does not affect the value of $received_for. It is still set
4953 after address rewriting, but before local_scan() is called.
4955 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4957 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4958 gave an unhelpful panic error message, and a defer error. I have managed to
4959 change this behaviour so that it now rejects any supplied certificate,
4960 which seems right, as the list of acceptable certificates is empty.
4962 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4963 gave an unhelpful defer error. I have not managed to make this reject any
4964 supplied certificates, but the error message it gives is "no certificate
4965 supplied", which is not helpful.
4967 70. exigrep's output now also includes lines that are not associated with any
4968 message, but which match the given pattern. Implemented by a patch from
4969 Martin Sluka, which also tidied up the Perl a bit.
4971 71. Recipient callout verification, like sender verification, was using <> in
4972 the MAIL FROM command. This isn't really the right thing, since the actual
4973 sender may affect whether the remote host accepts the recipient or not. I
4974 have changed it to use the actual sender in the callout; this means that
4975 the cache record is now keyed on a recipient/sender pair, not just the
4976 recipient address. There doesn't seem to be a real danger of callout loops,
4977 since a callout by the remote host to check the sender would use <>.
4978 [SEE ABOVE: changed after hitting problems.]
4980 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4981 temporary errors. However, in the case of such a code being given after
4982 the end of a data transmission (i.e. after ".") Exim was failing to write
4983 a retry record for the message. (Yes, there was some broken host that was
4984 actually sending 8xx at this point.)
4986 73. An unknown lookup type in a host list could cause Exim to panic-die when
4987 the list was checked. (An example that provoked this was putting <; in the
4988 middle of a list instead of at the start.) If this happened during a DATA
4989 ACL check, a -D file could be left lying around. This kind of configuration
4990 error no longer causes Exim to die; instead it causes a defer errror. The
4991 incident is still logged to the main and panic logs.
4993 74. Buglet left over from Exim 3 conversion. The message "too many messages
4994 in one connection" was written to the rejectlog but not the mainlog, except
4995 when address rewriting (yes!) was being logged.
4997 75. Added write_rejectlog option.
4999 76. When a system filter was run not as root (that is, when system_filter_user
5000 was set), the values of the $n variables were not being returned to the
5001 main process; thus, they were not subsequently available in the $sn
5004 77. Added +return_path_on_delivery log selector.
5006 78. A connection timeout was being treated differently from recipients deferred
5007 when testing hosts_max_try with a message that was older than the host's
5008 retry timeout. (The host should not be counted, thus allowing all hosts to
5009 be tried at least once before bouncing.) This may have been the cause of an
5010 occasionally reported bug whereby a message would remain on the queue
5011 longer than the retry timeout, but would be bounced if a delivery was
5012 forced. I say "may" because I never totally pinned down the problem;
5013 setting up timeout/retry tests is difficult. See also the next item.
5015 79. The ultimate address timeout was not being applied to errors that involved
5016 a combination of host plus message (for example, a timeout on a MAIL
5017 command). When an address resolved to a number of possible hosts, and they
5018 were not all tried for each delivery (e.g. because of hosts_max_try), a
5019 message could remain on the queue longer than the retry timeout.
5021 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5024 81. Fixed an obscure SMTP outgoing bug which required at least the following
5025 conditions: (a) there was another message waiting for the same server;
5026 (b) the server returned 5xx to all RCPT commands in the first message so
5027 that the message was not completed; (c) the server dropped the connection
5028 or gave a negative response to the RSET that Exim sends to abort the
5029 transaction. The observed case was a dropped connection after DATA that had
5030 been sent in pipelining mode. That is, the server had advertised PIPELINING
5031 but was not implementing it correctly. The effect of the bug was incorrect
5032 behaviour, such as trying another host, and this could lead to a crash.
5038 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5039 and daemon.c were passed as pointers to ints; they should have been
5040 pointers to socklen_t variables (which are typically unsigned ints).
5042 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5045 3. Fixed a really odd bug that affected only the testing scheme; patching a
5046 certain fixed string in the binary changed the value of another string that
5047 happened to be identical to the end of the original first string.
5049 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5050 name", it returns that address as the IP address. On some operating
5051 systems (e.g. Solaris), it also passes back the IP address string as the
5052 "host name". However, on others (e.g. Linux), it passes back an empty
5053 string. Exim wasn't checking for this, and was changing the host name to an
5054 empty string, assuming it had been canonicized.
5056 5. Although rare, it is permitted to have more than one PTR record for a given
5057 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5058 all the names associated with an address, because they do in Solaris.
5059 However, it seems that they do not in Linux for data that comes from the
5060 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5061 I found this out when I moved to a new Linux workstation and tried to run
5062 the Exim test suite.
5064 To get round this problem I have changed the code so that it now does its
5065 own call to the DNS to look up PTR records when searching for a host name.
5066 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5067 addresses that are only in /etc/hosts are still found.
5069 This behaviour is, however, controlled by an option called host_lookup_
5070 order, which defaults to "bydns:byaddr". If people want to use the other
5071 order, or indeed, just use one or the other means of lookup, they can
5072 specify it in this variable.
5074 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5075 some operating systems, this comes back from gethostbyaddr() as an empty
5076 string, and this is what Exim used to test for. However, it seems that in
5077 other systems, "." is yielded. Exim now tests for this case too.
5079 7. The values of check_spool_space and check_log_space are now held internally
5080 as a number of kilobytes instead of an absolute number of bytes. If a
5081 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5082 kilobyte. This means that much larger values can be stored.
5084 8. Exim monitor: an attempt to get the action menu when not actually pointing
5085 at a message produces an empty menu entitled "No message selected". This
5086 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5087 no entries in it ("Shell widget menu has zero width and/or height"). So I
5088 have added a single, blank menu entry in this case.
5090 9. Added ${quote_local_part.
5092 10. MIME decoding is now applied to the contents of Subject: header lines when
5095 11. Now that a reference to $sender_host_address automatically causes a reverse
5096 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5097 host lookup before query-style lookups in lists that might use this
5098 variable. This has therefore been abolished, and the "net-" prefix is no
5099 longer necessary for query-style lookups.
5101 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5102 have been a typo for LFLAGS, so it has been changed.
5104 13. The install script calls Exim with "-C /dev/null" in order to find the
5105 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5106 to be output. Howeve, since Exim outputs its version number before the
5107 error, it didn't break the script. It just looked ugly. I fixed this by
5108 always allowing "-C /dev/null" if the caller is root.
5110 14. Ignore overlarge ACL variable number when reading spool file - insurance
5111 against a later release with more variables having written the file.
5113 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5114 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5115 this, as well as the form without the "IPv6" on the front.
5117 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5118 OS/Makefile-Darwin file.
5120 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5121 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5123 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5124 variable. However, some OS use ints for inodes. I've added cast to long int
5125 to get rid of the compiler warning.
5127 19. I had forgotten to lock out "/../" in configuration file names when
5128 ALT_CONFIG_PREFIX was set.
5130 20. Routers used for verification do not need to specify transports. However,
5131 if such a router generated a host list, and callout was configured, Exim
5132 crashed, because it could not find a port number from the (non-existent)
5133 transport. It now assumes port 25 in this circumstance.
5135 21. Added the -t option to exigrep.
5137 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5138 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5139 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5140 them is set, LOOKUP_LSEARCH is forced.
5142 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5144 24. Added sender and host information to the "rejected by local_scan()" log
5145 line; previously there was no indication of these.
5147 25. Added .include_if_exists.
5149 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5150 Linux. It turns out that not all file systems support this. Apparently some
5151 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5152 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5153 not supported on the file descriptor, is now ignored when Exim is trying to
5154 sync a directory. This applies only to Linux.
5156 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5158 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5159 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5162 29. Added support for maildirsize files from supplied patch (modified a bit).
5164 30. The use of :fail: followed by an empty string could lead Exim to respond to
5165 sender verification failures with (e.g.):
5167 550 Verification failed for <xxx>
5168 550 Sender verify failed
5170 where the first response line was missing the '-' that indicates it is not
5171 the final line of the response.
5173 31. The loop for finding the name of the user that called Exim had a hardwired
5174 limit of 10; it now uses the value of finduser_retries, which is used for
5175 all other user lookups.
5177 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5180 33. Exim was neglecting to zero errno before one call of strtol() when
5181 expanding a string and expecting an integer value. On some systems this
5182 resulted in spurious "integer overflow" errors. Also, it was casting the
5183 result into an int without checking.
5185 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5186 did not work. The code looks as if it has *never* worked, though it appears
5187 to have been documented since at least releast 1.62. I have made it work.
5189 35. The "timeout_DNS" error in retry rules, also documented since at least
5190 1.62, also never worked. As it isn't clear exactly what this means, and
5191 clearly it isn't a major issue, I have abolished the feature by treating it
5192 as "timeout", and writing a warning to the main and panic logs.
5194 36. The display of retry rules for -brt wasn't always showing the error code
5197 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5198 timeout_connect_A, timeout_connect_MX.
5200 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5201 to the empty sender.
5203 39. The daemon was not analysing the content of -oX till after it had closed
5204 stderr and disconnected from the controlling terminal. This meant that any
5205 syntax errors were only noted on the panic log, and the return code from
5206 the command was 0. By re-arranging the code a little, I've made the
5207 decoding happen first, so such errors now appear on stderr, and the return
5208 code is 1. However, the actual setting up of the sockets still happens in
5209 the disconnected process, so errors there are still only recorded on the
5212 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5213 connections (as happens on some IP stacks) was logged at start up time as
5214 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5215 it from "IPv6 and IPv4", which means that two separate sockets are being
5218 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5219 says whether AF_INET or AF_INET6 was passed as an argument.
5221 42. Exiwhat output was messed up when time zones were included in log
5224 43. Exiwhat now gives more information about the daemon's listening ports,
5225 and whether -tls-on-connect was used.
5227 44. The "port" option of the smtp transport is now expanded.
5229 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5230 silently ignored. Now an error message is written to the main and panic
5233 46. There's a new ACL modifier called "logwrite" which writes to a log file
5234 as soon as it is encountered.
5236 47. Added $local_user_uid and $local_user_gid at routing time.
5238 48. Exim crashed when trying to verify a sender address that was being
5241 49. Exim was recognizing only a space character after ".include". It now also
5242 recognizes a tab character.
5244 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5245 extracting the relevant information from the specification. The man page no
5246 longer contains scrambled data for the -d option, and I've added a section
5247 at the front about calling Exim under different names.
5249 51. Added "extra_headers" argument to the "mail" command in filter files.
5251 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5254 53. Installed eximstats 1.29.
5256 54. Added transport_filter_timeout as a generic transport option.
5258 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5259 Cc: header lines. This was required by RFC 822, but it not required by RFC
5262 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5263 incoming messages that did not have them. Now it does so only if the
5264 message originates locally, that is, if there is no associated remote host
5265 address. When Resent- header lines are present, this applies to the Resent-
5266 lines rather than the non-Resent- lines.
5268 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5269 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5271 58. Messages for configuration errors now include the name of the main
5272 configuration file - useful now that there may be more than one file in a
5273 list (.included file names were always shown).
5275 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5276 for those rare installations that do not start the daemon as root or run it
5277 setuid root. I've cut out the call to initgroups() if the daemon is not
5280 60. The Exim user and group can now be bound into the binary as text strings
5281 that are looked up at the start of Exim's processing.
5283 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5285 62. Added $mailstore_basename variable.
5287 63. Installed patch to sieve.c from Michael Haardt.
5289 64. When Exim failed to open the panic log after failing to open the main log,
5290 the original message it was trying to log was written to stderr and debug
5291 output, but if they were not available (the usual case in production), it
5292 was lost. Now it is written to syslog before the two lines that record the
5293 failures to open the logs.
5295 65. Users' Exim filters run in subprocesses under the user's uid. It is
5296 possible for a "deliver" command or an alias in a "personal" command to
5297 provoke an address rewrite. If logging of address rewriting is configured,
5298 this fails because the process is not running as root or exim. There may be
5299 a better way of dealing with this, but for the moment (because 4.30 needs
5300 to be released), I have disabled address rewrite logging when running a
5301 filter in a non-root, non-exim process.
5307 1. The buildconfig auxiliary program wasn't quoting the value set for
5308 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5309 not defined. This bug was masked in 4.22 by the effect that was fixed in
5312 2. Some messages that were rejected after a message id was allocated were
5313 shown as "incomplete" by exigrep. It no longer does this for messages that
5314 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5316 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5317 have allow_domain_literals set, the ID did not get logged in the <= line.
5318 Domain literals are now always recognized in Message-ID: header lines.
5320 4. The first argument for a ${extract expansion item is the key name or field
5321 number. Leading and trailing spaces in this item were not being ignored,
5322 causing some misleading effects.
5324 5. When deliver_drop_privilege was set, single queue runner processes started
5325 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5326 same command in the process it spins off) were not dropping privilege.
5328 6. When the daemon running as "exim" started a queue runner, it always
5329 re-executed Exim in the spun-off process. This is a waste of effort when
5330 deliver_drop_privilege is set. The new process now just calls the
5331 queue-runner function directly.
5337 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5340 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5341 ignored. Though the use of -oP was forcing the writing of a pid file, it
5342 was always written to the default place.
5344 3. If the message "no IP address found for host xxxx" is generated during
5345 incoming verification, it is now followed by identification of the incoming
5346 connection (so you can more easily find what provoked it).
5348 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5350 5. Added some features to "harden" Exim a bit more against certain attacks:
5352 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5353 be put in Local/Makefile. This is like the never_users runtime option,
5354 but it cannot be overridden. The default setting is "root".
5356 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5357 prefix string with which any file named in a -C command line option
5360 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5361 is retained for -C and -D only if the caller of Exim is root. Without
5362 it, the exim user may also use -C and -D and retain privilege.
5364 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5365 command line option is disabled.
5367 6. Macro names set by the -D option must start with an upper case letter, just
5368 like macro names defined in the configuration file.
5370 7. Added "dereference=" facility to LDAP.
5372 8. Two instances of the typo "uknown" in the source files are fixed.
5374 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5375 the Configure-Makefile script screwed up while processing it.
5377 10. Incorporated PCRE 4.4.
5379 11. The SMTP synchronization check was not operating right at the start of an
5380 SMTP session. For example, it could not catch a HELO sent before the client
5381 waited for the greeting. There is now a check for outstanding input at the
5382 point when the greeting is written. Because of the duplex, asynchronous
5383 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5384 way, but not yet received, when the check is performed.
5386 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5387 on TCP/IP sockets, because this apparently causes some broken clients to
5390 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5391 unchanged) from the Cygwin maintainer.
5393 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5394 was supported instead of testing for mbx. Effectively a typo.
5396 15. The spa authenticator server code was not checking that the input it
5397 received was valid base64.
5399 16. The debug output line for the "set" modifier in ACLs was not showing the
5400 name of the variable that was being set.
5402 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5404 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5405 lookup on its own. Something else was needed to trigger the lookup. For
5406 example, a match in host_lookup or the need for a host name in a host list.
5407 Now, if $sender_host_name is referenced and the host name has not yet been
5408 looked up, a lookup is performed. If the lookup fails, the variable remains
5409 empty, and $host_lookup_failed is set to "1".
5411 19. Added "eqi" as a case-independent comparison operator.
5413 20. The saslauthd authentication condition could segfault if neither service
5414 nor realm was specified.
5416 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5417 error message that was logged was misleading, and incoming SMTP
5418 connections were dropped. The message is now more accurate, and temporary
5419 errors are given to SMTP connections.
5421 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5422 (except RSET) with a 421 error, until QUIT is received. However, it was
5423 failing to send a response to QUIT.
5425 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5426 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5427 ACL is now run after the helo_try_verify_hosts code.
5429 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5430 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5431 case-independent, so other case variants are also recognized). Apparently
5432 some systems use these upper case variants.
5434 25. If more than two messages were waiting for the same host, and a transport
5435 filter was specified for the transport, Exim sent two messages over the
5436 same TCP/IP connection, and then failed with "socket operation on non-
5437 socket" when it tried to send the third.
5439 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5441 27. The extern definition of crypt16() in expand.c was not being excluded when
5442 the OS had its own crypt16() function.
5444 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5445 as a preferred synonym for return_size_limit, both as an option and as an
5448 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5450 30. Changed the default configuration ACL to relax the local part checking rule
5451 for addresses that are not in any local domains. For these addresses,
5452 slashes and pipe symbols are allowed within local parts, but the sequence
5453 /../ is explicitly forbidden.
5455 31. SPA server authentication was not clearing the challenge buffer before
5458 32. log_message in a "warn" ACL statement was writing to the reject log as
5459 well as to the main log, which contradicts the documentation and doesn't
5460 seem right (because no rejection is happening). So I have stopped it.
5462 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5463 However, I am unable to do any testing of this.
5465 34. Fixed an infelicity in the appendfile transport. When checking directories
5466 for a mailbox, to see if any needed to be created, it was accidentally
5467 using path names with one or more superfluous leading slashes; tracing
5468 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5470 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5471 contents are added to the log line that is written for every discarded
5472 recipient. (Previously a log_message setting was ignored.)
5474 36. The ${quote: operator now quotes the string if it is empty.
5476 37. The install script runs exim in order to find its version number. If for
5477 some reason other than non-existence or emptiness, which it checks, it
5478 could not run './exim', it was installing it with an empty version number,
5479 i.e. as "exim-". This error state is now caught, and the installation is
5482 38. An argument was missing from the function that creates an error message
5483 when Exim fails to connect to the socket for saslauthd authentication.
5484 This could cause Exim to crash, or give a corrupted message.
5486 39. Added isip, isip4, and isip6 to ${if conditions.
5488 40. The ACL variables $acl_xx are now saved with the message, and can be
5489 accessed later in routers, transports, and filters.
5491 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5492 strings in the file are not string-expanded.
5494 42. If a MAIL command specified a SIZE value that was too large to fit into an
5495 int variable, the check against message_size_limit failed. Such values are
5496 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5497 day this will have to be increased, but I don't think I want to be around
5498 when emails are that large.
5505 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5506 iconv() is not standard in FreeBSD.
5508 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5509 IPv6 enabled. The observed symptom was a segmentation fault on return from
5510 the function os_common_find_running_interfaces() in src/os.c.
5512 3. In the check_special_case() function in daemon.c I had used "errno" as an
5513 argument name, which causes warnings on some systems. This was basically a
5514 typo, since it was named "eno" in the comments!
5516 4. The code that waits for the clock to tick (at a resolution of some fraction
5517 of a second) so as to ensure message-id uniqueness was always waiting for
5518 at least one whole tick, when it could have waited for less. [This is
5519 almost certainly not relevant at current processor speeds, where it is
5520 unlikely to ever wait at all. But we try to future-proof.]
5522 5. The function that sleeps for a time interval that includes fractions of a
5523 second contained a race. It did not block SIGALRM between setting the
5524 timer, and suspending (a couple of lines later). If the interval was short
5525 and the sigsuspend() was delayed until after it had expired, the suspension
5526 never ended. On busy systems this could lead to processes getting stuck for
5529 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5530 process, before it forks any delivery processes. The open lookup caching
5531 mechanism meant that the open file or database connection was passed into
5532 the delivery process. The problem was that delivery processes always tidy
5533 up cached lookup data. This could cause a problem for the next delivery
5534 process started by the queue runner, because the external queue runner
5535 process does not know about the closure. So the next delivery process
5536 still has data in the lookup cache. In the case of a file lookup, there was
5537 no problem because closing a file descriptor in a subprocess doesn't affect
5538 the parent. However, if the lookup was caching a connection to a database,
5539 the connection was closed, and the second delivery process was likely to
5540 see errors such as "PGSQL: query failed: server closed the connection
5541 unexpectedly". The problem has been fixed by closing all cached lookups
5542 in a queue runner before running a delivery process.
5544 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5545 seem to have the "const" qualifier which it has on other OS. I've
5548 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5549 *of the same type* (client or server) with the same public name that an
5550 error should be diagnosed.
5552 9. When Exim looked up a host name for an IP address, but failed to find the
5553 original IP address when looking up the host name (a safety check), it
5554 output the message "<ip address> does not match any IP for NULL", which was
5555 confusing, to say the least. The bug was that the host name should have
5556 appeared instead of "NULL".
5558 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5559 user that is built into the binary, and the -C or -D options is used, root
5560 privilege is dropped before the configuration file is read. In addition,
5561 logging is switched to stderr instead of the normal log files. If the
5562 configuration then re-defines the Exim user, the unprivileged environment
5563 is probably not what is expected, so Exim logs a panic warning message (but
5566 However, if deliver_drop_privilege is set, the unprivileged state may well
5567 be exactly what is intended, so the warning has been cut out in that case,
5568 and Exim is allowed to try to write to its normal log files.
5574 1. smtp_return_error_details was not giving details for temporary sender
5575 or receiver verification errors.
5577 2. Diagnose a configuration error if two authenticators have the same public
5580 3. Exim used not to create the message log file for a message until the first
5581 delivery attempt. This could be confusing when incoming messages were held
5582 for policy or load reasons. The message log file is now created at the time
5583 the message is received, and an initial "Received" line is written to it.
5585 4. The automatically generated man page for command line options had a minor
5586 bug that caused no ill effects; however, a more serious problem was that
5587 the procedure for building the man page automatically didn't always
5588 operate. Consequently, release 4.20 contains an out-of-date version. This
5589 shouldn't happen again.
5591 5. When building Exim with embedded Perl support, the script that builds the
5592 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5593 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5595 6. The freeze_tell option was not being used for messages that were frozen on
5596 arrival, either by an ACL or by local_scan().
5598 7. Added the smtp_incomplete_transaction log selector.
5600 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5601 was accepting AUTH without a new EHLO.
5603 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5604 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5607 10. Logging of TCP/IP connections (when configured) now happens in the main
5608 daemon process instead of the child process, so that the TCP/IP connection
5609 count is more accurate (but it can never be perfect).
5611 11. The use of "drop" in a nested ACL was not being handled correctly in the
5612 outer ACL. Now, if condition failure induced by the nested "drop" causes
5613 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5614 or "require"), the connection is dropped.
5616 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5617 that yield "discard" can now be used with an "accept" or a "discard" verb,
5618 but an error is generated for any others (because I can't see a useful way
5619 to define what should happen).
5621 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5622 are now processed in the same way as lines in the Exim configuration file.
5623 In particular, continuation lines are supported.
5625 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5627 15. Added -ti meaning -t -i.
5629 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5630 domains, and warn by logging if others are found.
5632 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5633 returned from GIFCONF when Exim is trying to find the list of interfaces on
5634 a host. The code in os.c has been modified to copy each ifreq to an aligned
5635 structure in all cases.
5637 Also, in some cases, the returned ifreq's were being copied to a 'struct
5638 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5639 means the last couple of bytes of an IPv6 address could be chopped if the
5640 ifreq contained only a normal sockaddr (14 bytes storage).
5642 18. Named domain lists were not supported in the hosts_treat_as_local option.
5643 An entry such as +xxxx was not recognized, and was treated as a literal
5646 19. Ensure that header lines added by a DATA ACL are included in the reject log
5647 if the ACL subsequently rejects the message.
5649 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5650 MD5 (which is deprecated).
5652 21. When testing a filter file using -bf, Exim was writing a message when it
5653 took the sender from a "From " line in the message, but it was not doing so
5654 when it took $return_path from a Return-Path: header line. It now does.
5656 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5657 with a valid header line field name (a series of printing characters
5658 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5660 23. Changed "disc" in the source to "disk" to conform to the documentation and
5661 the book and for uniformity.
5663 24. Ignore Sendmail's -Ooption=value command line item.
5665 25. When execve() failed while trying to run a command in a pipe transport,
5666 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5667 could be confused with a return value of 69 from the command itself. This
5668 has been changed to 127, the value the shell returns if it is asked to run
5669 a non-existent command. The wording for the related log line suggests a
5670 non-existent command as the problem.
5672 26. If received_header_text expands to an empty string, do not add a Received:
5673 header line to the message. (Well, it adds a token one on the spool, but
5674 marks it "old" so that it doesn't get used or transmitted.)
5676 27. Installed eximstats 1.28 (addition of -nt option).
5678 28. There was no check for failure on the call to getsockname() in the daemon
5679 code. This can fail if there is a shortage of resources on the system, with
5680 ENOMEM, for example. A temporary error is now given on failure.
5682 29. Contrary to the C standard, it seems that in some environments, the
5683 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5684 program. Exim now does this explicitly; it affects the formatting of
5685 timestamps using strftime().
5687 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5688 complaints. I've now initialized all the variables, to avoid this.
5690 32. Header lines added by a system filter were not being "seen" during
5691 transport-time rewrites.
5693 33. The info_callback() function passed to OpenSSL is set up with type void
5694 (*)(SSL *, int, int), as described somewhere. However, when calling the
5695 function (actually a macro) that sets it up, the type void(*)() is
5696 expected. I've put in a cast to prevent warnings from picky compilers.
5698 34. If a DNS black list lookup found a CNAME record, but there were no A
5699 records associated with the domain it pointed at, Exim crashed.
5701 35. If a DNS black list lookup returned more than one A record, Exim ignored
5702 all but the first. It now scans all returned addresses if a particular IP
5703 value is being sought. In this situation, the contents of the
5704 $dnslist_value variable are a list of all the addresses, separated by a
5707 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5708 to accept a host name and all its aliases if the forward lookup for any of
5709 them yielded the IP address of the incoming connection. Now it accepts only
5710 those names whose forward lookup yields the correct IP address. Any other
5711 names are discarded. This closes a loophole whereby a rogue DNS
5712 administrator could create reverse DNS records to break through a
5713 wildcarded host restriction in an ACL.
5715 37. If a user filter or a system filter that ran in a subprocess used any of
5716 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5717 the wrong values were passed to the pipe command ($thisaddress had the
5718 value of $0, $0 had the value of $1, etc). This bug was introduced by
5719 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5721 38. Improved the line breaking for long SMTP error messages from ACLs.
5722 Previously, if there was no break point between 40 and 75 characters, Exim
5723 left the rest of the message alone. Two changes have been made: (a) I've
5724 reduced the minimum length to 35 characters; (b) if it can't find a break
5725 point between 35 and 75 characters, it looks ahead and uses the first one
5726 that it finds. This may give the occasional overlong line, but at least the
5727 remaining text gets split now.
5729 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5730 file descriptors might be low, and that setting 1000 would always raise it.
5731 It turns out that in some environments, the limit is already over 1000 and
5732 that lowering it causes trouble. So now Exim takes care not to decrease it.
5734 40. When delivering a message, the value of $return_path is set to $sender_
5735 address at the start of routing (routers may change the value). By an
5736 oversight, this default was not being set up when an address was tested by
5737 -bt or -bv, which affected the outcome if any router or filter referred to
5740 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5741 log only when "message" or "log_message" are set. However, if one of the
5742 conditions was an address verification, or a call to a nested ACL, the
5743 messages generated by the underlying test were being passed through. This
5744 no longer happens. The underlying message is available in $acl_verify_
5745 message for both "message" and "log_message" expansions, so it can be
5746 passed through if needed.
5748 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5749 new expansion $bh_ to give the encoded byte string without charset
5750 translation. Translation happens only if iconv() is available; HAVE_ICONV
5751 indicates this at build time. HEADERS_CHARSET gives the charset to
5752 translate to; headers_charset can change it in the configuration, and
5753 "headers charset" can change it in an individual filter file.
5755 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5756 that creates RFC 2047 encoded "words" labels them as that charset instead
5757 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5758 expansion operator; (ii) when Exim creates a From: line for a local
5759 message; (iii) when a header line is rewritten to include a "phrase" part.
5761 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5762 buggy, causing it to skip the first lines of messages whose message ID
5763 ended in 'D'. This would not have bitten before Exim release 4.14, because
5764 message IDs were unlikely to end in 'D' before then. The effect was to have
5765 incorrect size information for certain domains.
5767 45. #include "config.h" was missing at the start of the crypt16.c module. This
5768 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5771 46. If there was a timeout during a "random" callout check, Exim treated it as
5772 a failure of the random address, and carried on sending RSET and the real
5773 address. If the delay was just some slowness somewhere, the response to the
5774 original RCPT would be taken as a response to RSET and so on, causing
5775 mayhem of various kinds.
5777 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5778 when I implemented it. It didn't allow for the fact that some option values
5779 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5780 the right test for positive values.
5782 48. Domain names in DNS records are case-independent. Exim always looks them up
5783 in lower case. Some resolvers return domain names in exactly the case they
5784 appear in the zone file, that is, they may contain uppercase letters. Not
5785 all resolvers do this - some return always lower case. Exim was treating a
5786 change of case by a resolver as a change of domain, similar to a widening
5787 of a domain abbreviation. This triggered its re-routing code and so it was
5788 trying to route what was effectively the same domain again. This normally
5789 caused routing to fail (because the router wouldn't handle the domain
5790 twice). Now Exim checks for this case specially, and just changes the
5791 casing of the domain that it ultimately uses when it transmits the message
5794 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5797 50. If a filter generated a file delivery with a non-absolute name (possible if
5798 no home directory exists for the router), the forbid_file option was not
5801 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5802 the existing equality matching.
5804 52. Exim was using ints instead of ino_t variables in some places where it was
5805 dealing with inode numbers.
5807 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5808 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5809 called TMPDIR, and if it finds it is different, it changes its value.
5811 54. The smtp_printf() function is now made available to local_scan() so
5812 additional output lines can be written before returning. There is also an
5813 smtp_fflush() function to enable the detection of a dropped connection.
5814 The variables smtp_input and smtp_batched_input are exported to
5817 55. Changed the default runtime configuration: the message "Unknown user"
5818 has been removed from the ACL, and instead placed on the localuser router,
5819 using the cannot_route_message feature. This means that any verification
5820 failures that generate their own messages won't get overridden. Similarly,
5821 the "Unrouteable address" message that was in the ACL for unverifiable
5822 relay addresses has also been removed.
5824 56. Added hosts_avoid_esmtp to the smtp transport.
5826 57. The exicyclog script was not checking for the esoteric option
5827 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5828 will work only if exicyclog is run under the appropriate euid.
5830 58. Following a discussion on the list, the rules by which Exim recognises line
5831 endings on incoming messages have been changed. The -dropcr and drop_cr
5832 options are now no-ops, retained only for backwards compatibility. The
5833 following line terminators are recognized: LF CRLF CR. However, special
5834 processing applies to CR:
5836 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5837 nor a local message in the state where . is a terminator.
5839 (ii) If a bare CR is encountered in a header line, an extra space is added
5840 after the line terminator so as not to end the header. The reasoning
5841 behind this is that bare CRs in header lines are most likely either
5842 to be mistakes, or people trying to play silly games.
5844 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5845 was being incorrectly given as 18 bytes larger than it should have been.
5846 This is a VOB (very old bug).
5848 60. This may never have affected anything current, but just in case it has:
5849 When the local host is found other than at the start of a list of hosts,
5850 the local host, those with the same MX, and any that follow, are discarded.
5851 When the list in question was part of a longer list of hosts, the following
5852 hosts (not currently being processed) were also being discarded. This no
5853 longer happens. I'm not sure if this situation could ever has previously
5856 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5859 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5860 header. This is something that is recommended in a new Internet Draft, and
5861 is something that is documented as being done by Sendmail. There are two
5862 possible values. For messages generated by the autoreply transport, Exim
5865 Auto-Submitted: auto-replied
5867 whereas for all other generated messages (e.g. bounces) it adds
5869 Auto-Submitted: auto-generated
5871 63. The "personal" condition in filters now includes a test for the
5872 Auto-Submitted: header. If it contains the string "auto-" the message it
5873 not considered personal.
5875 64. Added rcpt_include_affixes as a generic transport option.
5877 65. Added queue_only_override (default true).
5879 66. Added the syslog_duplication option.
5881 67. If what should have been the first header line of a message consisted of
5882 a space followed by a colon, Exim was mis-interpreting it as a header line.
5883 It isn't of course - it is syntactically invalid and should therefore be
5884 treated as the start of the message body. The misbehaviour could have
5885 caused a number of strange effects, including loss of data in subsequent
5886 header lines, and spool format errors.
5888 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5889 client host had authenticated. This control can now be exercised by an ACL
5890 for more flexibility.
5892 69. By default, callouts do not happen when testing with -bh. There is now a
5893 variant, -bhc, which does actually run the callout code, including
5894 consulting and updating the callout cache.
5896 70. Added support for saslauthd authentication, courtesy of Alexander
5899 71. If statvfs() failed on the spool or log directories while checking their
5900 size for availability, Exim confusingly gave the error "space shortage".
5901 Furthermore, in debugging mode it crashed with a floating point exception.
5902 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5903 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5904 serious problem, Exim now writes to the main and panic logs when this
5905 happens, with details of the failure. It then refuses to accept the
5906 incoming message, giving the message "spool directory problem" or "log
5907 directory problem" with a 421 code for SMTP messages.
5909 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5910 0, 1, and 2 exist, because some OS complain for execs without them (see
5911 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5912 descriptors. However, the code omitted to check that the open succeeded,
5913 causing mysterious errors if for some reason the permissions on /dev/null
5914 got screwed. Now Exim writes a message to the main and panic logs, and
5915 bombs out if it can't open /dev/null.
5917 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5918 interact so that it is all more flexible. It is supposed to remain
5919 backwards compatible. Also added extra_local_interfaces.
5921 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5922 to bomb out with an assertion failure - to the client this appears as a
5923 connection drop. This problem occurs in the part of the code that was taken
5924 from the Samba project. Fortunately, the assertion is in a very simple
5925 function, so I have fixed this by reproducing the function inline in the
5926 one place where it is called, and arranging for authentication to fail
5927 instead of killing the process with assert().
5929 75. The SPA client code was not working when the server requested OEM rather
5930 than Unicode encoding.
5932 76. Added code to make require_files with a specific uid setting more usable in
5933 the case where statting the file as root fails - usually a non-root-mounted
5934 NFS file system. When this happens and the failure is EACCES, Exim now
5935 forks a subprocess and does the per-uid checking as the relevant uid.
5937 77. Added process_log_path.
5939 78. If log_file_path was not explicitly set, a setting of check_log_space or
5940 check_log_inodes was ignored.
5942 79. If a space check for the spool or log partitions fails, the incident is now
5943 logged. Of course, in the latter case the data may get lost...
5945 80. Added the %p formatting code to string_format() so that it can be used to
5946 print addresses in debug_print(). Adjusted all the address printing in the
5947 debugging in store.c to use %p rather than %d.
5949 81. There was a concern that a line of code in smtp_in.c could overflow a
5950 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5951 initially expressed, the concern was not well-founded, because trailing
5952 spaces are removed early. However, if the trailing spaces were followed by
5953 a NULL, they did not get removed, so the overflow was possible. Two fixes
5956 (a) I re-wrote the offending code in a cleaner fashion.
5957 (b) If an incoming SMTP command contains a NULL character, it is rejected
5960 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5961 runs initgroups(), so that if the Exim user is in any additional groups,
5962 they will be used during message reception.
5968 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob