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.
51 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
52 selectors, in both main and reject logs.
54 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
57 JH/07 Add malware type "sock" for talking to simple daemon.
59 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
61 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
62 routers/transports under cutthrough routing.
64 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
65 numbers. Touch up "bool" conditional to keep the same definition.
71 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
73 PP/02 Make -n do something, by making it not do something.
74 When combined with -bP, the name of an option is not output.
76 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
79 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
80 $sender_host_name and config options to manage this, and basic check
83 PP/05 DSCP support for outbound connections and control modifier for inbound.
85 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
86 (Only plugin which currently uses this is kerberos4, which nobody should
87 be using, but we should make it available and other future plugins might
88 conceivably use it, even though it would break NAT; stuff *should* be
89 using channel bindings instead).
91 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
92 name; added for Sendmail compatibility; requires admin caller.
93 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
94 just ignore it); requires trusted caller.
95 Also parse but ignore: -Ac -Am -X<logfile>
98 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
100 TL/02 Add +smtp_confirmation as a default logging option.
102 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
103 Patch by Magnus Holmgren from 2007-02-20.
105 TL/04 Bugzilla 1281 - Spec typo.
106 Bugzilla 1283 - Spec typo.
107 Bugzilla 1290 - Spec grammar fixes.
109 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
111 TL/06 Add Experimental DMARC support using libopendmarc libraries.
113 TL/07 Fix an out of order global option causing a segfault. Reported to dev
114 mailing list by by Dmitry Isaikin.
116 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
118 JH/02 Support "G" suffix to numbers in ${if comparisons.
120 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
122 NM/01 Bugzilla 1197 - Spec typo
123 Bugzilla 1196 - Spec examples corrections
125 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
127 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
128 gnutls_enable_pkcs11, but renamed to more accurately indicate its
131 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
132 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
134 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
135 "acl {{name}{arg}...}", and optional args on acl condition
138 JH/05 Permit multiple router/transport headers_add/remove lines.
140 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
142 JH/07 Avoid using a waiting database for a single-message-only transport.
143 Performance patch from Paul Fisher. Bugzilla 1262.
145 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
148 JH/09 Add $headers_added variable, with content from use of ACL modifier
149 add_header (but not yet added to the message). Bugzilla 199.
151 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
152 Pulled from Bugzilla 817 by Wolfgang Breyha.
154 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
156 (nb: this is the same fix as in Exim 4.80.1)
158 JH/11 Add A= logging on delivery lines, and a client_set_id option on
161 JH/12 Add optional authenticated_sender logging to A= and a log_selector
164 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
166 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
167 advertise SMTP AUTH mechanism to us, instead of a generic
168 protocol violation error. Also, make Exim more robust to bad
169 data from the Dovecot auth socket.
171 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
173 When a queue runner is handling a message, Exim first routes the
174 recipient addresses, during which it prunes them based on the retry
175 hints database. After that it attempts to deliver the message to
176 any remaining recipients. It then updates the hints database using
179 So if a recipient address works intermittently, it can get repeatedly
180 deferred at routing time. The retry hints record remains fresh so the
181 address never reaches the final cutoff time.
183 This is a fairly common occurrence when a user is bumping up against
184 their storage quota. Exim had some logic in its local delivery code
185 to deal with this. However it did not apply to per-recipient defers
186 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
188 This change adds a proper retry rule check during routing so that the
189 final cutoff time is checked against the message's age. We only do
190 this check if there is an address retry record and there is not a
191 domain retry record; this implies that previous attempts to handle
192 the address had the retry_use_local_parts option turned on. We use
193 this as an approximation for the destination being like a local
194 delivery, as in LMTP.
196 I suspect this new check makes the old local delivery cutoff check
197 redundant, but I have not verified this so I left the code in place.
199 TF/02 Correct gecos expansion when From: is a prefix of the username.
201 Test 0254 submits a message to Exim with the header
205 When I ran the test suite under the user fanf2, Exim expanded
206 the header to contain my full name, whereas it should have added
207 a Resent-Sender: header. It erroneously treats any prefix of the
208 username as equal to the username.
210 This change corrects that bug.
212 GF/01 DCC debug and logging tidyup
213 Error conditions log to paniclog rather than rejectlog.
214 Debug lines prefixed by "DCC: " to remove any ambiguity.
216 TF/03 Avoid unnecessary rebuilds of lookup-related code.
218 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
219 Bug spotted by Jeremy Harris; was flawed since initial commit.
220 Would have resulted in OCSP responses post-SNI triggering an Exim
221 NULL dereference and crash.
223 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
225 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
226 Bug detection, analysis and fix by Samuel Thibault.
227 Bugzilla 1331, Debian bug #698092.
229 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
231 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
232 Server implementation by Todd Lyons, client by JH.
233 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
234 config variable "prdr_enable" controls whether the server
235 advertises the facility. If the client requests PRDR a new
236 acl_data_smtp_prdr ACL is called once for each recipient, after
237 the body content is received and before the acl_smtp_data ACL.
238 The client is controlled by bolth of: a hosts_try_prdr option
239 on the smtp transport, and the server advertisement.
240 Default client logging of deliveries and rejections involving
241 PRDR are flagged with the string "PRDR".
243 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
244 fclose(). Diagnosis by Todd Lyons.
246 PP/17 Update configure.default to handle IPv6 localhost better.
247 Patch by Alain Williams (plus minor tweaks).
250 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
251 This is now consistent with GnuTLS, and is now documented: the
252 previous undocumented portable approach to treating the option as
253 unset was to force an expansion failure. That still works, and
254 an empty string is now equivalent.
256 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
257 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
258 not performing validation itself.
260 PP/20 Added force_command boolean option to pipe transport.
261 Patch from Nick Koston, of cPanel Inc.
263 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
266 TF/04 Added udpsend ACL modifer and hexquote expansion operator
268 PP/21 Fix eximon continuous updating with timestamped log-files.
269 Broken in a format-string cleanup in 4.80, missed when I repaired the
270 other false fix of the same issue.
271 Report and fix from Heiko Schlichting.
274 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
275 Report from Prashanth Katuri.
277 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
278 It's SecureTransport, so affects any MacOS clients which use the
279 system-integrated TLS libraries, including email clients.
281 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
282 using a MIME ACL for non-SMTP local injection.
283 Report and assistance in diagnosis by Warren Baker.
285 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
287 JH/16 Fix comparisons for 64b. Bugzilla 1385.
289 TL/09 Add expansion variable $authenticated_fail_id to keep track of
290 last id that failed so it may be referenced in subsequent ACL's.
292 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
295 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
296 ldap library initialization, allowing self-signed CA's to be
297 used. Also properly sets require_cert option later in code by
298 using NULL (global ldap config) instead of ldap handle (per
299 session). Bug diagnosis and testing by alxgomz.
301 TL/12 Enhanced documentation in the ratelimit.pl script provided in
302 the src/util/ subdirectory.
304 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
305 renamed to Transport Post Delivery Action by Jeremy Harris, as
308 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
309 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
310 redis_servers = needs to be configured which will be used by the redis
311 lookup. Patch from Warren Baker, of The Packet Hub.
313 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
315 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
316 hostname or reverse DNS when processing a host list. Used suggestions
317 from multiple comments on this bug.
319 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
321 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
322 Missed a few lines, added it to make the runtest require no keyboard
325 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
326 contains upper case chars. Make router use caseful_local_part.
328 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
329 support when GnuTLS has been built with p11-kit.
335 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
337 This, or similar/improved, will also be change PP/11 of 4.82.
343 PP/01 Handle short writes when writing local log-files.
344 In practice, only affects FreeBSD (8 onwards).
345 Bugzilla 1053, with thanks to Dmitry Isaikin.
347 NM/01 Bugzilla 949 - Documentation tweak
349 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
352 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
354 PP/02 Implemented gsasl authenticator.
356 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
358 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
359 `pkg-config foo` for cflags/libs.
361 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
362 with rest of GSASL and with heimdal_gssapi.
364 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
365 `pkg-config foo` for cflags/libs for the TLS implementation.
367 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
368 properties get this fed in as external SSF. A number of robustness
369 and debugging improvements to the cyrus_sasl authenticator.
371 PP/08 cyrus_sasl server now expands the server_realm option.
373 PP/09 Bugzilla 1214 - Log authentication information in reject log.
374 Patch by Jeremy Harris.
376 PP/10 Added dbmjz lookup type.
378 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
380 PP/12 MAIL args handles TAB as well as SP, for better interop with
381 non-compliant senders.
382 Analysis and variant patch by Todd Lyons.
384 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
385 Bug report from Lars Müller <lars@samba.org> (via SUSE),
386 Patch from Dirk Mueller <dmueller@suse.com>
388 PP/13 tls_peerdn now print-escaped for spool files.
389 Observed some $tls_peerdn in wild which contained \n, which resulted
390 in spool file corruption.
392 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
393 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
394 or write after TLS renegotiation, which otherwise led to messages
397 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
398 as a tracking header (ie: a signed header comes before the signature).
399 Patch from Wolfgang Breyha.
401 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
402 comma-sep list; embedded commas doubled.
404 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
406 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
408 Report and patch from Dmitry Banschikov.
410 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
411 Removed SSL_clear() after SSL_new() which led to protocol negotiation
412 failures. We appear to now support TLS1.1+ with Exim.
414 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
415 lets Exim select keys and certificates based upon TLS SNI from client.
416 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
417 before an outbound SMTP session. New log_selector, +tls_sni.
419 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
420 NULL dereference. Report and patch from Alun Jones.
422 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
423 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
424 Not seeing resolver debug output on NetBSD, but suspect this is a
425 resolver implementation change.
427 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
428 Left warnings. Added "eximon gdb" invocation mode.
430 PP/21 Defaulting "accept_8bitmime" to true, not false.
432 PP/22 Added -bw for inetd wait mode support.
434 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
435 locate the relevant includes and libraries. Made this the default.
437 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
438 Bugzilla 1246, report and most of solution from Tomasz Kusy.
440 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
441 This may cause build issues on older platforms.
443 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
444 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
445 gnutls_require_mac & gnutls_require_protocols (no longer supported).
446 Added SNI support via GnuTLS too.
447 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
449 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
451 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
452 Applied second patch from Janne, implementing suggestion to default
453 multiple-strings-in-record handling to match SPF spec.
455 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
457 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
458 read-only, out of scope).
459 Patch from Wolfgang Breyha, report from Stuart Northfield.
461 PP/29 Fix three issues highlighted by clang analyser static analysis.
462 Only crash-plausible issue would require the Cambridge-specific
463 iplookup router and a misconfiguration.
464 Report from Marcin Mirosław.
466 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
468 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
469 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
470 As part of this, removing so much warning spew let me fix some minor
471 real issues in debug logging.
473 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
474 assignment on my part. Fixed.
476 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
477 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
478 Janne Snabb (who went above and beyond: thank you).
480 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
481 string otherwise requires a connection and a bunch more work and it's
482 relatively easy to get wrong. Should also expose TLS library linkage
485 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
486 64-bit ${eval} (JH/03).
488 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
489 GNU libc to support some of the 64-bit stuff, should not lead to
490 conflicts. Defined before os.h is pulled in, so if a given platform
491 needs to override this, it can.
493 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
494 protection layer was required, which is not implemented.
495 Bugzilla 1254, patch from Wolfgang Breyha.
497 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
498 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
499 tls_dhparam take prime identifiers. Also unbreak combination of
500 OpenSSL+DH_params+TLSSNI.
502 PP/39 Disable SSLv2 by default in OpenSSL support.
508 PP/01 Solaris build fix for Oracle's LDAP libraries.
509 Bugzilla 1109, patch from Stephen Usher.
511 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
513 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
516 TF/02 Fix a couple more cases where we did not log the error message
517 when unlink() failed. See also change 4.74-TF/03.
519 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
520 lock up or crash if it happened to be inside a call to libc when it
521 got a SIGUSR1 from exiwhat.
523 The SIGUSR1 handler appends the current process status to the process
524 log which is later printed by exiwhat. It used to use the general
525 purpose logging code to do this, but several functions it calls are
526 not safe for signals.
528 The new output code in the SIGUSR1 handler is specific to the process
529 log, and simple enough that it's easy to inspect for signal safety.
530 Removing some special cases also simplifies the general logging code.
531 Removing the spurious timestamps from the process log simplifies
534 TF/04 Improved ratelimit ACL condition.
536 The /noupdate option has been deprecated in favour of /readonly which
537 has clearer semantics. The /leaky, /strict, and /readonly update modes
538 are mutually exclusive. The update mode is no longer included in the
539 database key; it just determines when the database is updated. (This
540 means that when you upgrde Exim will forget old rate measurements.)
542 Exim now checks that the per_* options are used with an update mode that
543 makes sense for the current ACL. For example, when Exim is processing a
544 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
545 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
546 must specify per_mail/readonly. If you omit the update mode it defaults to
547 /leaky where that makes sense (as before) or /readonly where required.
549 The /noupdate option is now undocumented but still supported for
550 backwards compatibility. It is equivalent to /readonly except that in
551 ACLs where /readonly is required you may specify /leaky/noupdate or
552 /strict/noupdate which are treated the same as /readonly.
554 A useful new feature is the /count= option. This is a generalization
555 of the per_byte option, so that you can measure the throughput of other
556 aggregate values. For example, the per_byte option is now equivalent
557 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
559 The per_rcpt option has been generalized using the /count= mechanism
560 (though it's more complicated than the per_byte equivalence). When it is
561 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
562 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
563 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
564 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
565 Note that using per_rcpt with a non-readonly update mode in more than
566 one ACL will cause the recipients to be double-counted. (The per_mail
567 and per_byte options don't have this problem.)
569 The handling of very low rates has changed slightly. If the computed rate
570 is less than the event's count (usually one) then this event is the first
571 after a long gap. In this case the rate is set to the same as this event's
572 count, so that the first message of a spam run is counted properly.
574 The major new feature is a mechanism for counting the rate of unique
575 events. The new per_addr option counts the number of different
576 recipients that someone has sent messages to in the last time period. It
577 behaves like per_rcpt if all the recipient addresses are different, but
578 duplicate recipient addresses do not increase the measured rate. Like
579 the /count= option this is a general mechanism, so the per_addr option
580 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
581 example, measure the rate that a client uses different sender addresses
582 with the options per_mail/unique=$sender_address. There are further
583 details in the main documentation.
585 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
587 TF/06 Removed a few PCRE remnants.
589 TF/07 Automatically extract Exim's version number from tags in the git
590 repository when doing development or release builds.
592 PP/02 Raise smtp_cmd_buffer_size to 16kB.
593 Bugzilla 879. Patch from Paul Fisher.
595 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
596 Heavily based on revision 40f9a89a from Simon Arlott's tree.
599 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
601 PP/05 Variable $av_failed, true if the AV scanner deferred.
602 Bugzilla 1078. Patch from John Horne.
604 PP/06 Stop make process more reliably on build failure.
605 Bugzilla 1087. Patch from Heiko Schlittermann.
607 PP/07 Make maildir_use_size_file an _expandable_ boolean.
608 Bugzilla 1089. Patch from Heiko Schlittermann.
610 PP/08 Handle ${run} returning more data than OS pipe buffer size.
611 Bugzilla 1131. Patch from Holger Weiß.
613 PP/09 Handle IPv6 addresses with SPF.
614 Bugzilla 860. Patch from Wolfgang Breyha.
616 PP/10 GnuTLS: support TLS 1.2 & 1.1.
618 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
621 PP/11 match_* no longer expand right-hand-side by default.
622 New compile-time build option, EXPAND_LISTMATCH_RHS.
623 New expansion conditions, "inlist", "inlisti".
625 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
627 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
629 PP/14 fix log_write() format string regression from TF/03.
630 Bugzilla 1152. Patch from Dmitry Isaikin.
636 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
638 PP/02 Harmonised TLS library version reporting; only show if debugging.
639 Layout now matches that introduced for other libraries in 4.74 PP/03.
641 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
643 PP/04 New "dns_use_edns0" global option.
645 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
648 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
649 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
651 TK/01 Updated PolarSSL code to 0.14.2.
652 Bugzilla 1097. Patch from Andreas Metzler.
654 PP/07 Catch divide-by-zero in ${eval:...}.
657 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
660 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
661 format-string attack -- SECURITY: remote arbitrary code execution.
663 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
664 time unintentionally subject to list matching rules, letting the header
665 cause arbitrary Exim lookups (of items which can occur in lists, *not*
666 arbitrary string expansion). This allowed for information disclosure.
668 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
669 INT_MIN/-1 -- value coerced to INT_MAX.
675 NM/01 Workround for PCRE version dependency in version reporting
678 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
679 This fixes portability to compilers other than gcc, notably
680 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
682 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
683 makefiles for portability to HP-UX and POSIX correctness.
685 PP/01 Permit LOOKUP_foo enabling on the make command-line.
686 Also via indented variable definition in the Makefile.
687 (Debugging by Oliver Heesakkers).
689 PP/02 Restore caching of spamd results with expanded spamd_address.
690 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
692 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
693 Improves build reliability. Fix from: Frank Elsner
695 NM/02 Fix wide character breakage in the rfc2047 coding
696 Fixes bug 1064. Patch from Andrey N. Oktyabrski
698 NM/03 Allow underscore in dnslist lookups
699 Fixes bug 1026. Patch from Graeme Fowler
701 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
702 Code patches from Adam Ciarcinski of NetBSD.
704 NM/04 Fixed exiqgrep to cope with mailq missing size issue
707 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
708 is logged, to avoid truncation. Patch from John Horne.
710 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
711 Patch from Jakob Hirsch.
713 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
714 SQL string expansion failure details.
715 Patch from Andrey Oktyabrski.
717 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
718 Patch from Simon Arlott.
720 PP/09 New lookups functionality failed to compile on old gcc which rejects
721 extern declarations in function scope.
722 Patch from Oliver Fleischmann
724 PP/10 Use sig_atomic_t for flags set from signal handlers.
725 Check getgroups() return and improve debugging.
726 Fixed developed for diagnosis in bug 927 (which turned out to be
729 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
730 Patch from Mark Zealey.
732 PP/12 Bugzilla 1056: Improved spamd server selection.
733 Patch from Mark Zealey.
735 PP/13 Bugzilla 1086: Deal with maildir quota file races.
736 Based on patch from Heiko Schlittermann.
738 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
739 Patch from Uwe Doering, sign-off by Michael Haardt.
741 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
742 variable declaration deep within a block. Bug and patch from
745 PP/15 lookups-Makefile IRIX compatibilty coercion.
747 PP/16 Make DISABLE_DKIM build knob functional.
749 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
750 Patch by Simon Arlott
752 TF/03 Fix valgrind.h portability to C89 compilers that do not support
753 variable argument macros. Our copy now differs from upstream.
759 TF/01 Failure to get a lock on a hints database can have serious
760 consequences so log it to the panic log.
762 TF/02 Log LMTP confirmation messages in the same way as SMTP,
763 controlled using the smtp_confirmation log selector.
765 TF/03 Include the error message when we fail to unlink a spool file.
767 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
768 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
769 for maintaining out-of-tree patches for some time.
771 PP/01 Bugzilla 139: Documentation and portability issues.
772 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
773 Handle per-OS dynamic-module compilation flags.
775 PP/02 Let /dev/null have normal permissions.
776 The 4.73 fixes were a little too stringent and complained about the
777 permissions on /dev/null. Exempt it from some checks.
778 Reported by Andreas M. Kirchwitz.
780 PP/03 Report version information for many libraries, including
781 Exim version information for dynamically loaded libraries. Created
782 version.h, now support a version extension string for distributors
783 who patch heavily. Dynamic module ABI change.
785 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
786 privilege escalation vulnerability whereby the Exim run-time user
787 can cause root to append content of the attacker's choosing to
790 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
793 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
794 If dropping privileges for untrusted macros, we disabled normal logging
795 on the basis that it would fail; for the Exim run-time user, this is not
796 the case, and it resulted in successful deliveries going unlogged.
797 Fixed. Reported by Andreas Metzler.
803 PP/01 Date: & Message-Id: revert to normally being appended to a message,
804 only prepend for the Resent-* case. Fixes regression introduced in
805 Exim 4.70 by NM/22 for Bugzilla 607.
807 PP/02 Include check_rfc2047_length in configure.default because we're seeing
808 increasing numbers of administrators be bitten by this.
810 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
812 PP/03 Bugzilla 994: added openssl_options main configuration option.
814 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
816 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
818 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
820 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
821 without a peer certificate, leading to a segfault because of an
822 assumption that peers always have certificates. Be a little more
823 paranoid. Problem reported by Martin Tscholak.
825 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
826 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
827 NB: ClamAV planning to remove STREAM in "middle of 2010".
828 CL also introduces -bmalware, various -d+acl logging additions and
829 more caution in buffer sizes.
831 PP/09 Implemented reverse_ip expansion operator.
833 PP/10 Bugzilla 937: provide a "debug" ACL control.
835 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
837 PP/12 Bugzilla 973: Implement --version.
839 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
841 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
843 PP/15 Bugzilla 816: support multiple condition rules on Routers.
845 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
846 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
847 ignore trailing whitespace.
849 JJ/02 prevent non-panic DKIM error from being sent to paniclog
851 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
854 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
855 Notification from Dr Andrew Aitchison.
857 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
858 ExtendedDetectionInfo response format.
859 Notification from John Horne.
861 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
864 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
865 XSL and documented dependency on system catalogs, with examples of how
868 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
871 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
872 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
873 configuration file which is writeable by the Exim user or group.
875 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
876 of configuration files to cover files specified with the -C option if
877 they are going to be used with root privileges, not just the default
880 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
881 option (effectively making it always true).
883 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
884 files to be used while preserving root privileges.
886 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
887 that rogue child processes cannot use them.
889 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
890 run-time user, instead of root.
892 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
893 Exim run-time user without dropping privileges.
895 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
896 result string, instead of calling string_vformat() twice with the same
899 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
900 for other users. Others should always drop root privileges if they use
901 -C on the command line, even for a whitelisted configure file.
903 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
905 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
911 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
912 $data_path, and $header_path variables; fixed documentation bugs and
915 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
916 exipick to access non-standard spools, including the "frozen" queue
919 NM/01 Bugzilla 965: Support mysql stored procedures.
920 Patch from Alain Williams
922 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
924 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
925 Patch from Andreas Metzler
927 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
928 Patch from Kirill Miazine
930 NM/05 Bugzilla 671: Added umask to procmail example.
932 JJ/03 installed exipick 20100323.0, fixing doc bug
934 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
935 directory. Notification and patch from Dan Rosenberg.
937 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
939 TK/02 Improve log output when DKIM signing operation fails.
941 MH/01 Treat the transport option dkim_domain as a colon separated
942 list, not as a single string, and sign the message with each element,
943 omitting multiple occurences of the same signer.
945 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
946 Bugzilla 985, 986. Patch by Simon Arlott
948 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
949 Patch by Simon Arlott
951 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
952 MBX locking. Notification from Dan Rosenberg.
958 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
960 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
962 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
964 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
966 NM/04 Bugzilla 915: Flag broken perl installation during build.
972 TK/01 Added patch by Johannes Berg that expands the main option
973 "spamd_address" if it starts with a dollar sign.
975 TK/02 Write list of recipients to X-Envelope-Sender header when building
976 the mbox-format spool file for content scanning (suggested by Jakob
979 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
980 (http://www.dcc-servers.net/) support via dccifd. Activated by
981 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
983 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
984 by Mark Daniel Reidel <mr@df.eu>.
986 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
987 When building exim an external PCRE library is now needed -
988 PCRE is a system library on the majority of modern systems.
989 See entry on PCRE_LIBS in EDITME file.
991 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
992 conversation. Added nologin parameter to request.
993 Patch contributed by Kirill Miazine.
995 TF/01 Do not log submission mode rewrites if they do not change the address.
997 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
999 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1000 log files in place. Contributed by Roberto Lima.
1002 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1004 TF/03 Bugzilla 615: When checking the local_parts router precondition
1005 after a local_part_suffix or local_part_prefix option, Exim now
1006 does not use the address's named list lookup cache, since this
1007 contains cached lookups for the whole local part.
1009 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1010 Robert Millan. Documentation is in experimental-spec.txt.
1012 TF/04 Bugzilla 668: Fix parallel build (make -j).
1014 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1016 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1017 Patch provided by Jan Srzednicki.
1019 TF/05 Leading white space used to be stripped from $spam_report which
1020 wrecked the formatting. Now it is preserved.
1022 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1023 that they are available at delivery time.
1025 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1027 TF/08 TLS error reporting now respects the incoming_interface and
1028 incoming_port log selectors.
1030 TF/09 Produce a more useful error message if an SMTP transport's hosts
1031 setting expands to an empty string.
1033 NM/06 Bugzilla 744: EXPN did not work under TLS.
1034 Patch provided by Phil Pennock.
1036 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1037 Patch provided by Richard Godbee.
1039 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1040 acl_smtp_notquit, added index entry.
1042 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1043 Patch provided by Eugene Bujak.
1045 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1046 accept(). Patch provided by Maxim Dounin.
1048 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1049 Patch provided by Phil Pennock.
1051 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1053 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1054 Patch provided by Brad "anomie" Jorsch.
1056 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1057 Patch provided by Dean Brooks.
1059 TK/05 Add native DKIM support (does not depend on external libraries).
1061 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1062 Patch provided by Graeme Fowler.
1064 NM/16 Bugzilla 851: Documentation example syntax fix.
1066 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1068 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1071 NM/19 Bugzilla 745: TLS version reporting.
1072 Patch provided by Phil Pennock.
1074 NM/20 Bugzilla 167: bool: condition support.
1075 Patch provided by Phil Pennock.
1077 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1078 clients. Patch provided by Phil Pennock.
1080 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1081 Patch provided by Brad "anomie" Jorsch.
1083 NM/23 Bugzilla 687: Fix misparses in eximstats.
1084 Patch provided by Heiko Schlittermann.
1086 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1087 Patch provided by Heiko Schlittermann.
1089 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1090 plus update to original patch.
1092 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1094 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1095 Patch provided by David Brownlee.
1097 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1099 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1101 NM/30 Bugzilla 888: TLS documentation bugfixes.
1103 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1105 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1106 Unlike the original bugzilla I have changed all shell scripts in src tree.
1108 NM/33 Bugzilla 898: Transport filter timeout fix.
1109 Patch by Todd Rinaldo.
1111 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1112 Patch by Serge Demonchaux.
1114 NM/35 Bugzilla 39: Base64 decode bug fixes.
1115 Patch by Jakob Hirsch.
1117 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1119 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1121 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1123 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1129 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1130 ALT-N's libdkim that I have put here:
1131 http://duncanthrax.net/exim-experimental/
1133 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1134 were called 'true' and it seems that C99 defines that as a reserved
1135 keyword to be used with 'bool' variable types. That means you could
1136 not include C99-style headers which use bools without triggering
1137 build errors in sieve.c.
1139 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1140 as mailq or other aliases. Changed the --help handling significantly
1141 to do whats expected. exim_usage() emits usage/help information.
1143 SC/01 Added the -bylocaldomain option to eximstats.
1145 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1147 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1149 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1155 PH/01 Another patch from the Sieve maintainer.
1157 PH/02 When an IPv6 address is converted to a string for single-key lookup
1158 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1159 dots are used instead of colons so that keys in lsearch files need not
1160 contain colons. This was done some time before quoting was made available
1161 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1162 (notated using the quote facility) so as to distinguish them from IPv4
1163 keys. This meant that lookups for IP addresses in host lists did not work
1164 for iplsearch lookups.
1166 This has been fixed by arranging for IPv6 addresses to be expressed with
1167 colons if the lookup type is iplsearch. This is not incompatible, because
1168 previously such lookups could never work.
1170 The situation is now rather anomolous, since one *can* have colons in
1171 ordinary lsearch keys. However, making the change in all cases is
1172 incompatible and would probably break a number of configurations.
1174 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1177 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1178 conversion specification without a maximum field width, thereby enabling
1179 a rogue spamd server to cause a buffer overflow. While nobody in their
1180 right mind would setup Exim to query an untrusted spamd server, an
1181 attacker that gains access to a server running spamd could potentially
1182 exploit this vulnerability to run arbitrary code as the Exim user.
1184 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1185 $primary_hostname instead of what libspf2 thinks the hosts name is.
1187 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1188 a directory entry by the name of the lookup key. Previously, if a
1189 symlink pointed to a non-existing file or a file in a directory that
1190 Exim lacked permissions to read, a lookup for a key matching that
1191 symlink would fail. Now it is enough that a matching directory entry
1192 exists, symlink or not. (Bugzilla 503.)
1194 PH/03 The body_linecount and body_zerocount variables are now exported in the
1197 PH/04 Added the $dnslist_matched variable.
1199 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1200 This means they are set thereafter only if the connection becomes
1203 PH/06 Added the client_condition to authenticators so that some can be skipped
1204 by clients under certain conditions.
1206 PH/07 The error message for a badly-placed control=no_multiline_responses left
1207 "_responses" off the end of the name.
1209 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1211 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1212 (without spaces) instead of just copying the configuration text.
1214 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1216 PH/11 Added $max_received_linelength.
1218 PH/12 Added +ignore_defer and +include_defer to host lists.
1220 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1221 way in which PCRE > 7.0 is built.
1223 PH/14 Implemented queue_only_load_latch.
1225 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1226 MAIL command. The effect was to mangle the value on 64-bit systems.
1228 PH/16 Another patch from the Sieve maintainer.
1230 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1232 PH/18 If a system quota error occurred while trying to create the file for
1233 a maildir delivery, the message "Mailbox is full" was not appended to the
1234 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1235 only to a quota excession during the actual writing of the file.
1237 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1238 characters?) which causes problems in log lines. The DN values are now
1239 passed through string_printing() before being added to log lines.
1241 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1242 and InterBase are left for another time.)
1244 PH/21 Added message_body_newlines option.
1246 PH/22 Guard against possible overflow in moan_check_errorcopy().
1248 PH/23 POSIX allows open() to be a macro; guard against that.
1250 PH/24 If the recipient of an error message contained an @ in the local part
1251 (suitably quoted, of course), incorrect values were put in $domain and
1252 $local_part during the evaluation of errors_copy.
1258 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1259 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1262 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1263 issue a MAIL command.
1265 PH/02 In an ACL statement such as
1267 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1269 if a client was not listed at all, or was listed with a value other than
1270 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1271 the condition was not true (as it should be), so access was not denied.
1272 The bug was that the ! inversion was incorrectly passed on to the second
1273 item. This has been fixed.
1275 PH/03 Added additional dnslists conditions == and =& which are different from
1276 = and & when the dns lookup returns more than one IP address.
1278 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1279 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1281 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1282 FSYNC, which compiles an option called disable_fsync that allows for
1283 bypassing fsync(). The documentation is heavily laced with warnings.
1285 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1287 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1288 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1289 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1290 including adding "make clean"; (3) Added -fPIC when compiling the test
1291 dynamically loaded module, to get rid of a warning.
1293 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1294 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1295 = 0s. The bug is otherwise harmless.
1297 PH/07 There was a bug in the dovecot authenticator such that the value of
1298 $auth1 could be overwritten, and so not correctly preserved, after a
1299 successful authentication. This usually meant that the value preserved by
1300 the server_setid option was incorrect.
1302 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1304 PH/09 Installed PCRE release 7.0.
1306 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1307 run for batched SMTP input. It is now run at the start of every message
1308 in the batch. While fixing this I discovered that the process information
1309 (output by running exiwhat) was not always getting set for -bs and -bS
1310 input. This is fixed, and it now also says "batched" for BSMTP.
1312 PH/11 Added control=no_pipelining.
1314 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1315 patch, slightly modified), and move the expansion of helo_data till after
1316 the connection is made in the smtp transport (so it can use these
1319 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1321 PH/14 Added log_selector = +pid.
1323 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1325 PH/16 Add ${if forany and ${if forall.
1327 PH/17 Added dsn_from option to vary the From: line in DSNs.
1329 PH/18 Flush SMTP output before performing a callout, unless control =
1330 no_callout_flush is set.
1332 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1333 was true (the default) a successful delivery failed to delete the retry
1334 item, thus causing premature timeout of the address. The bug is now
1337 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1339 PH/21 Long custom messages for fakedefer and fakereject are now split up
1340 into multiline reponses in the same way that messages for "deny" and
1341 other ACL rejections are.
1343 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1344 with slight modification.
1346 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1347 draft, changing the syntax and factoring some duplicate code".
1349 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1350 for deliveries of the second and subsequent messages over the same SMTP
1353 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1354 ${reduce, with only minor "tidies".
1356 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1358 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1359 expansion side effects.
1361 PH/27 When a message times out after an over-quota error from an Exim-imposed
1362 quota, the bounce message says "mailbox is full". This message was not
1363 being given when it was a system quota that was exceeded. It now should
1366 MH/03 Made $recipients available in local_scan(). local_scan() already has
1367 better access to the recipient list through recipients_list[], but
1368 $recipients can be useful in postmaster-provided expansion strings.
1370 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1371 in the case of a MAIL command with additional options following the
1372 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1373 were accidentally chopped off.
1375 PH/29 SMTP synchronization checks are implemented when a command is read -
1376 there is a check that no more input is waiting when there shouldn't be
1377 any. However, for some commands, a delay in an ACL can mean that it is
1378 some time before the response is written. In this time, more input might
1379 arrive, invalidly. So now there are extra checks after an ACL has run for
1380 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1381 pipelining has not been advertised.
1383 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1385 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1386 returned in the SMTP response when smtp_return_error_details was false.
1387 This has been fixed.
1389 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1390 instead of the C I/O that was originally supplied, because problems were
1391 reported on Solaris.
1393 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1394 Exim which did not show up earlier: it was assuming that a call to
1395 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1396 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1397 was a macro that became an assignment, so it seemed to work. This has
1398 changed to a proper function call with a void return, hence the compile
1399 error. Exim's code has been fixed.
1401 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1404 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1406 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1408 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1409 because it turns out to be too big or there is a timeout) there is no
1410 "Completed" line in the log. When some messages of this type were
1411 selected by exigrep, they were listed as "not completed". Others were
1412 picked up by some special patterns. I have improved the selection
1413 criteria to be more general.
1415 PH/38 The host_find_failed option in the manualroute router can now be set
1416 to "ignore", to completely ignore a host whose IP address cannot be
1417 found. If all hosts are ignored, the behaviour is controlled by the new
1418 host_all_ignored option.
1420 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1421 homing or because of multiple MX records with /mx) generated more than
1422 one IP address, and the following item turned out to be the local host,
1423 all the secondary addresses of the first item were incorrectly removed
1424 from the list, along with the local host and any following hosts (which
1425 is what is supposed to happen).
1427 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1428 whoever called Exim into the -H file. In the case of the daemon it was
1429 behaving confusingly. When first started, it used values for whoever
1430 started the daemon, but after a SIGHUP it used the Exim user (because it
1431 calls itself on a restart). I have changed the code so that it now always
1434 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1435 message are rejected with the same error (e.g. no authentication or bad
1436 sender address), and a DATA command is nevertheless sent (as can happen
1437 with PIPELINING or a stupid MUA), the error message that was given to the
1438 RCPT commands is included in the rejection of the DATA command. This is
1439 intended to be helpful for MUAs that show only the final error to their
1442 PH/42 Another patch from the Sieve maintainer.
1444 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1445 Eximstats - Fixed some broken HTML links and added missing column headers
1447 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1448 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1450 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1452 PH/43 Yet another patch from the Sieve maintainer.
1454 PH/44 I found a way to check for a TCP/IP connection going away before sending
1455 the response to the final '.' that terminates a message, but only in the
1456 case where the client has not sent further data following the '.'
1457 (unfortunately, this is allowed). However, in many cases there won't be
1458 any further data because there won't be any more messages to send. A call
1459 to select() can be used: if it shows that the input is "ready", there is
1460 either input waiting, or the socket has been closed. An attempt to read
1461 the next input character can distinguish the two cases. Previously, Exim
1462 would have sent an OK response which the client would never have see.
1463 This could lead to message repetition. This fix should cure that, at
1464 least in a lot of common cases.
1466 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1467 advertised in response to EHLO.
1473 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1474 fixed by 4.65/MH/01 (is this a record?) are fixed:
1476 (i) An empty string was always treated as zero by the numeric comparison
1477 operators. This behaviour has been restored.
1479 (ii) It is documented that the numeric comparison operators always treat
1480 their arguments as decimal numbers. This was broken in that numbers
1481 starting with 0 were being interpreted as octal.
1483 While fixing these problems I realized that there was another issue that
1484 hadn't been noticed. Values of message_size_limit (both the global option
1485 and the transport option) were treated as octal if they started with 0.
1486 The documentation was vague. These values are now always treated as
1487 decimal, and I will make that clear in the documentation.
1493 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1494 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1497 MH/01 Don't check that the operands of numeric comparison operators are
1498 integers when their expansion is in "skipping" mode (fixes bug
1499 introduced by 4.64-PH/07).
1501 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1502 child addresses, Exim now panics and dies. Previously, because the count
1503 is held in a short int, deliveries were likely to be lost. As such a
1504 large number of recipients for a single message is ridiculous
1505 (performance will be very, very poor), I have chosen to impose a limit
1506 rather than extend the field.
1512 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1513 leftover -K file (the existence of which was triggered by #402).
1514 While we were at it, introduced process PID as part of the -K
1515 filename. This should rule out race conditions when creating
1518 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1519 processing considerably. Previous code took too long for large mails,
1520 triggering a timeout which in turn triggers #401.
1522 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1523 in the DK code in transports.c. sendfile() is not really portable,
1524 hence the _LINUX specificness.
1526 TF/01 In the add_headers option to the mail command in an Exim filter,
1527 there was a bug that Exim would claim a syntax error in any
1528 header after the first one which had an odd number of characters
1531 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1532 callout verification, Exim cached a "reject" for the entire domain. This
1533 is correct for most verifications, but it is not correct for a recipient
1534 verification with use_sender or use_postmaster set, because in that case
1535 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1536 case of MAIL FROM:<> rejection from other early rejections (e.g.
1537 rejection of HELO). When verifying a recipient using a non-null MAIL
1538 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1539 Whatever the result of the callout, the value of the domain cache is
1540 left unchanged (for any other kind of callout, getting as far as trying
1541 RCPT means that the domain itself is ok).
1543 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1546 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1547 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1548 session, the daemon ignores ECONNECT errors and logs others; it now
1549 ignores EPIPE as well.
1551 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1552 (quoted-printable decoding).
1554 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1555 later the small subsequent patch to fix an introduced bug.
1557 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1559 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1561 PH/08 An error is now given if message_size_limit is specified negative.
1563 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1564 to be given (somewhat) arbitrary names.
1566 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1569 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1570 miscellaneous code fixes
1572 PH/10 Added the log_reject_target ACL modifier to specify where to log
1575 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1576 hostname. This is wrong, because it relates to the incoming message (and
1577 probably the interface on which it is arriving) and not to the outgoing
1578 callout (which could be using a different interface). This has been
1579 changed to use the value of the helo_data option from the smtp transport
1580 instead - this is what is used when a message is actually being sent. If
1581 there is no remote transport (possible with a router that sets up host
1582 addresses), $smtp_active_hostname is used.
1584 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1585 tweaks were necessary in order to get it to work (see also 21 below):
1586 (a) The code assumed that strncpy() returns a negative number on buffer
1587 overflow, which isn't the case. Replaced with Exim's string_format()
1589 (b) There were several signed/unsigned issues. I just did the minimum
1590 hacking in of casts. There is scope for a larger refactoring.
1591 (c) The code used strcasecmp() which is not a standard C function.
1592 Replaced with Exim's strcmpic() function.
1593 (d) The code set only $1; it now sets $auth1 as well.
1594 (e) A simple test gave the error "authentication client didn't specify
1595 service in request". It would seem that Dovecot has changed its
1596 interface. Fortunately there's a specification; I followed it and
1597 changed what the client sends and it appears to be working now.
1599 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1602 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1603 address A is aliased to B and C, where B exists and C does not. Without
1604 -v the output is "A verified" because verification stops after a
1605 successful redirection if more than one address is generated. However,
1606 with -v the child addresses are also verified. Exim was outputting "A
1607 failed to verify" and then showing the successful verification for C,
1608 with its parentage. It now outputs "B failed to verify", showing B's
1609 parentage before showing the successful verification of C.
1611 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1612 look up a TXT record in a specific list after matching in a combined
1615 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1616 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1617 they consult the DNS. I had assumed they would set it the way they
1618 wanted; and indeed my experiments on Linux seem to show that in some
1619 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1620 To be on the safe side, however, I have now made the interface to
1621 host_find_byname() similar to host_find_bydns(), with an argument
1622 containing the DNS resolver options. The host_find_byname() function now
1623 sets these options at its start, just as host_find_bydns() does. The smtp
1624 transport options dns_qualify_single and dns_search_parents are passed to
1625 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1626 of host_find_byname() use the default settings of RES_DEFNAMES
1627 (qualify_single) but not RES_DNSRCH (search_parents).
1629 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1630 spool_read_header() do less string testing, by means of a preliminary
1631 switch on the second character of optional "-foo" lines. (This is
1632 overdue, caused by the large number of possibilities that now exist.
1633 Originally there were few.) While I was there, I also converted the
1634 str(n)cmp tests so they don't re-test the leading "-" and the first
1635 character, in the hope this might squeeze out yet more improvement.
1637 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1638 flag allowing group syntax was set by the header_syntax check but not
1639 turned off, possible causing trouble later; (2) The flag was not being
1640 set at all for the header_verify test, causing "group"-style headers to
1641 be rejected. I have now set it in this case, and also caused header_
1642 verify to ignore an empty address taken from a group. While doing this, I
1643 came across some other cases where the code for allowing group syntax
1644 while scanning a header line wasn't quite right (mostly, not resetting
1645 the flag correctly in the right place). These bugs could have caused
1646 trouble for malformed header lines. I hope it is now all correct.
1648 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1649 with the "reply" argument non-NULL. The code, however (which originally
1650 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1651 but it didn't always do it. This confused somebody who was copying the
1652 code for some other use. I have removed all the tests.
1654 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1655 feature that was used to support insecure browsers during the U.S. crypto
1656 embargo. It requires special client support, and Exim is probably the
1657 only MTA that supported it -- and would never use it because real RSA is
1658 always available. This code has been removed, because it had the bad
1659 effect of slowing Exim down by computing (never used) parameters for the
1660 RSA_EXPORT functionality.
1662 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1663 authenticator to fail if there's a tab character in the incoming data
1664 (there should never be unless someone is messing about, as it's supposed
1665 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1666 if the connection is using TLS or if the remote IP is the same as the
1667 local IP, and the "valid-client-cert option" if a client certificate has
1670 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1671 authenticators. This can be used for authorization after authentication
1672 succeeds. (In the case of plaintext, it servers for both authentication
1675 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1676 if any retry times were supplied.
1678 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1679 connection, where there is no client IP address to check. In this
1680 situation, the verify now always succeeds.
1682 PH/25 Applied John Jetmore's -Mset patch.
1684 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1686 PH/27 In a string expansion for a processed (not raw) header when multiple
1687 headers of the same name were present, leading whitespace was being
1688 removed from all of them, but trailing whitespace was being removed only
1689 from the last one. Now trailing whitespace is removed from each header
1690 before concatenation. Completely empty headers in a concatenation (as
1691 before) are ignored.
1693 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1694 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1696 PH/29 [Removed. This was a change that I later backed out, and forgot to
1697 correct the ChangeLog entry (that I had efficiently created) before
1698 committing the later change.]
1700 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1701 address errors (4xx response to RCPT) over the same connection as other
1702 messages routed to the same hosts. Such deliveries are always "forced",
1703 so retry times are not inspected. This resulted in far too many retries
1704 for the affected addresses. The effect occurred only when there were more
1705 hosts than the hosts_max_try setting in the smtp transport when it had
1706 the 4xx errors. Those hosts that it had tried were not added to the list
1707 of hosts for which the message was waiting, so if all were tried, there
1708 was no problem. Two fixes have been applied:
1710 (i) If there are any address or message errors in an SMTP delivery, none
1711 of the hosts (tried or untried) are now added to the list of hosts
1712 for which the message is waiting, so the message should not be a
1713 candidate for sending over the same connection that was used for a
1714 successful delivery of some other message. This seems entirely
1715 reasonable: after all the message is NOT "waiting for some host".
1716 This is so "obvious" that I'm not sure why it wasn't done
1717 previously. Hope I haven't missed anything, but it can't do any
1718 harm, as the worst effect is to miss an optimization.
1720 (ii) If, despite (i), such a delivery is accidentally attempted, the
1721 routing retry time is respected, so at least it doesn't keep
1722 hammering the server.
1724 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1725 in ${readsocket because some servers need this prod.
1727 PH/32 Added some extra debug output when updating a wait-xxx database.
1729 PH/33 The hint "could be header name not terminated by colon", which has been
1730 given for certain expansion errors for a long time, was not being given
1731 for the ${if def:h_colon_omitted{... case.
1733 PH/34 The spec says: "With one important exception, whenever a domain list is
1734 being scanned, $domain contains the subject domain." There was at least
1735 one case where this was not true.
1737 PH/35 The error "getsockname() failed: connection reset by peer" was being
1738 written to the panic log as well as the main log, but it isn't really
1739 panic-worthy as it just means the connection died rather early on. I have
1740 removed the panic log writing for the ECONNRESET error when getsockname()
1743 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1744 runs only) independently of the message's sender address. This meant
1745 that, if the 4xx error was in fact related to the sender, a different
1746 message to the same recipient with a different sender could confuse
1747 things. In particualar, this can happen when sending to a greylisting
1748 server, but other circumstances could also provoke similar problems.
1749 I have changed the default so that the retry time for these errors is now
1750 based a combination of the sender and recipient addresses. This change
1751 can be overridden by setting address_retry_include_sender=false in the
1754 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1755 remote server are returned as part of bounce messages. This was not
1756 happening for LMTP over a pipe (the lmtp transport), but now it is the
1757 same for both kinds of LMTP.
1759 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1760 in header lines that were in fact CNAMEs. This is no longer the case.
1762 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1763 and queue runs started by the daemon processed all messages. This has
1764 been fixed so that -R and -S can now usefully be given with -q<time>.
1766 PH/40 Import PCRE release 6.7 (fixes some bugs).
1768 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1770 PH/42 Give an error if -q is specified more than once.
1772 PH/43 Renamed the variables $interface_address and $interface_port as
1773 $received_ip_address and $received_port, to make it clear that these
1774 values apply to message reception, and not to the outgoing interface when
1775 a message is delivered. (The old names remain recognized, of course.)
1777 PH/44 There was no timeout on the connect() call when using a Unix domain
1778 socket in the ${readsocket expansion. There now is.
1780 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1781 be meaningful with "accept".
1783 SC/01 Eximstats V1.43
1784 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1786 SC/02 Eximstats V1.44
1787 Use a glob alias rather than an array ref in the generated
1788 parser. This improves both readability and performance.
1790 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1791 Collect SpamAssassin and rejection statistics.
1792 Don't display local sender or destination tables unless
1793 there is data to show.
1794 Added average volumes into the top table text output.
1796 SC/04 Eximstats V1.46
1797 Collect data on the number of addresses (recipients)
1798 as well as the number of messages.
1800 SC/05 Eximstats V1.47
1801 Added 'Message too big' to the list of mail rejection
1802 reasons (thanks to Marco Gaiarin).
1804 SC/06 Eximstats V1.48
1805 Mainlog lines which have GMT offsets and are too short to
1806 have a flag are now skipped.
1808 SC/07 Eximstats V1.49 (Alain Williams)
1809 Added the -emptyok flag.
1811 SC/08 Eximstats V1.50
1812 Fixes for obtaining the IP address from reject messages.
1814 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1815 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1816 whitesspace changes from 4.64-PH/27
1818 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1821 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1822 are found, allow negative numbers in numeric criteria)
1824 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1826 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1827 to match changes made in 4.64-PH/43
1829 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1831 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1832 "domain" is actually an IP address, and force a failure. This locks out
1833 those revolvers/nameservers that support "A-for-A" lookups, in
1834 contravention of the specifications.
1836 PH/48 When a host name was looked up from an IP address, and the subsequent
1837 forward lookup of the name timed out, the host name was left in
1838 $sender_host_name, contrary to the specification.
1840 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1841 restricted to single-key lookups, Exim was not diagnosing an error if
1842 * or *@ was used with a query-style lookup.
1844 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1846 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1847 long ago, but noone interested enough thought of it. Let's just say that
1848 the "1.1" means that there are some new functions that weren't there at
1849 some point in the past.
1851 PH/51 Error processing for expansion failure of helo_data from an smtp
1852 transport during callout processing was broken.
1854 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1855 tested/used via the -bh/-bhc/-bs options.
1857 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1858 bug, fixed in subsequent PCRE releases).
1860 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1861 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1863 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1869 SC/01 Use a glob alias rather than an array ref in eximstats generated
1870 parser. This improves both readability and performance.
1872 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1873 Don't display local sender or destination tables in eximstats unless
1874 there is data to show.
1875 Added average volumes into the eximstats top table text output.
1877 SC/03 Collect data on the number of addresses (recipients) as well
1878 as the number of messages in eximstats.
1880 TF/01 Correct an error in the documentation for the redirect router. Exim
1881 does (usually) call initgroups() when daemonizing.
1883 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1884 with consistent privilege compared to when running as a daemon.
1886 TF/03 Note in the spec that $authenticated_id is not set for local
1887 submissions from trusted users.
1889 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1890 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1892 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1893 by adding some example configuration directives to the default
1894 configuration file. A little bit of work is required to uncomment the
1895 directives and define how usernames and passwords are checked, but
1896 there is now a framework to start from.
1898 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1899 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1900 without this. I don't know how relevant this is to other LDAP libraries.
1902 PH/02 Add the verb name to the "unknown ACL verb" error.
1904 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1906 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1908 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1909 directory not expanded when it should be if an expanded home directory
1910 was set for the address (which is overridden by the transport).
1912 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1915 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1916 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1917 because it is too late at that time, and has no effect.
1919 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1920 security issue with \' (bugzilla #107). I could not use the
1921 PQescapeStringConn() function, because it needs a PGconn value as one of
1924 PH/08 When testing addresses using -bt, indicate those final addresses that
1925 are duplicates that would not cause an additional delivery. At least one
1926 person was confused, thinking that -bt output corresponded to deliveries.
1927 (Suppressing duplicates isn't a good idea as you lose the information
1928 about possibly different redirections that led to the duplicates.)
1930 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1931 systems where poll() doesn't work, in particular OS X.
1933 PH/10 Added more information to debugging output for retry time not reached.
1935 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1936 operations in malware.c.
1938 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1941 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1942 syslog_duplication set false, log lines that would normally be written
1943 both the the main log and to the reject log were not written to syslog at
1946 PH/14 In the default configuration, change the use of "message" in ACL warn
1947 statements to "add_header".
1949 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1950 not followed by a command (e.g. "seen endif").
1952 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1953 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1956 PH/17 Added extra conditions to the default value of delay_warning_condition
1960 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1961 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1962 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1965 The Auto-Submitted: and various List- headers are standardised, whereas I
1966 don't think Precedence: ever was.
1968 PH/18 Refactored debugging code in route_finduser() to show more information,
1969 in particular, the error code if getpwnam() issues one.
1971 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1972 This is apparently needed in addition to the PH/07 change above to avoid
1973 any possible encoding problems.
1975 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1976 but not after initializing Perl.
1978 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1979 output them only if debugging. By default they are written stderr,
1980 apparently, which is not desirable.
1982 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1985 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1988 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1990 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1991 authenticated or an ident call has been made. Suppress the default
1992 values for $authenticated_id and $authenticated_sender (but permit -oMai
1993 and -oMas) when testing with -bh.
1995 PH/24 Re-jigged the order of the tests in the default configuration so that the
1996 tests for valid domains and recipients precede the DNS black list and CSA
1997 tests, on the grounds that those ones are more expensive.
1999 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2000 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2001 This bug exists in every version of Exim that I still have, right back to
2004 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2005 However, an attempt to turn on case-sensitivity in a regex key by
2006 including (?-i) didn't work because the subject string was already
2007 lowercased, and the effects were non-intuitive. It turns out that a
2008 one-line patch can be used to allow (?-i) to work as expected.
2014 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2015 other effects) broke the use of negated acl sub-conditions.
2017 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2020 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2021 "Deny" causes Exim to reject the incoming connection with a 554 error.
2022 Unfortunately, if there is a major crisis, such as a disk failure,
2023 tcp-wrappers gives "deny", whereas what one would like would be some
2024 kind of temporary error. A kludge has been added to help with this.
2025 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2026 554 error is used if errno is still zero or contains ENOENT (which occurs
2027 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2030 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2032 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2033 errors. Otherwise a message that provokes a temporary error (when other
2034 messages do not) can cause a whole host to time out.
2036 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2037 addresses were routed directly to files or pipes from a redirect router.
2038 File deliveries just didn't batch; pipe deliveries might have suffered
2041 PH/06 A failure to get a lock for a hints database would erroneously always say
2042 "Failed to get write lock", even when it was really a read lock.
2044 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2045 of 0600. This has been changed to use the value of the lockfile_mode
2046 option (which defaults to 0600).
2048 PH/08 Applied small patch from the Sieve maintainer.
2050 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2051 folder from quota calculations, a direct delivery into this folder messed
2052 up the contents of the maildirsize file. This was because the regex was
2053 used only to exclude .Trash (or whatever) when the size of the mailbox
2054 was calculated. There was no check that a delivery was happening into an
2055 excluded directory. This bug has been fixed by ignoring all quota
2056 processing for deliveries into excluded directories.
2058 PH/10 Added the maildirfolder_create_regex option to appendfile.
2064 PH/01 The code for finding all the local interface addresses on a FreeBSD
2065 system running IPv6 was broken. This may well have applied to all BSD
2066 systems, as well as to others that have similar system calls. The broken
2067 code found IPv4 interfaces correctly, but gave incorrect values for the
2068 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2069 that it would not match correctly against @[] and not recognize the IPv6
2072 PH/02 The ipliteral router was not recognizing addresses of the form user@
2073 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2075 PH/03 Added disable_ipv6.
2077 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2078 decoded form, because it is most often used to construct To: headers
2079 lines in autoreplies, and the decoded form may well be syntactically
2080 invalid. However, $reply_address has leading white space removed, and all
2081 newlines turned into spaces so that the autoreply transport does not
2084 PH/05 If group was specified without a user on a router, and no group or user
2085 was specified on a transport, the group from the router was ignored.
2087 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2088 for visible compile-time settings that can be used to change these
2089 numbers, for those that want even more. Backwards compatibility with old
2090 spool files has been maintained. However, going back to a previous Exim
2091 release will lost any variables that are in spool files.
2093 PH/07 Two small changes when running in the test harness: increase delay when
2094 passing a TCP/IP connection to a new process, in case the original
2095 process has to generate a bounce, and remove special handling of
2096 127.0.0.2 (sic), which is no longer necessary.
2098 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2099 be the same on different OS.
2101 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2104 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2105 whether --show-vars was specified or not
2107 JJ/02 exipick: Added support for new ACL variable spool format introduced
2110 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2111 syntactically invalid From: or Reply-to: line, and a filter used this to
2112 generate an autoreply, and therefore failed to obtain an address for the
2113 autoreply, Exim could try to deliver to a non-existent relative file
2114 name, causing unrelated and misleading errors. What now happens is that
2115 it logs this as a hard delivery error, but does not attempt to create a
2118 PH/11 The exinext utility has a -C option for testing purposes, but although
2119 the given file was scanned by exinext itself; it wasn't being passed on
2120 when Exim was called.
2122 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2123 an end-of-file indication when reading a command response.
2125 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2126 compiled. In many other places in Exim, IPv6 addresses are always
2127 recognized, so I have changed this. It also means that IPv4 domain
2128 literals of the form [IPV4:n.n.n.n] are now always recognized.
2130 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2131 used if the router is not running as root, for example, when verifying at
2132 ACL time, or when using -bh. The debugging output from this situation was
2133 non-existent - all you got was a failure to exec. I have made two
2136 (a) Failures to set uid/gid, the current directory, or a process leader
2137 in a subprocess such as that created by queryprogram now generate
2138 suitable debugging ouput when -d is set.
2140 (b) The queryprogram router detects when it is not running as root,
2141 outputs suitable debugging information if -d is set, and then runs
2142 the subprocess without attempting to change uid/gid.
2144 PH/15 Minor change to Makefile for building test_host (undocumented testing
2147 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2148 additional section of a DNS packet that returns MX or SRV records.
2149 Instead, it always explicitly searches for A/AAAA records. This avoids
2150 major problems that occur when a DNS server includes only records of one
2151 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2152 fixed another bug: if SRV records were looked up and the corresponding
2153 address records were *not* found in the additional section, the port
2154 values from the SRV records were lost.
2156 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2157 using the correct key (the original address) when searching the retry
2158 rules in order to find which one to use for generating the retry hint.
2160 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2161 adding the default one. Similarly, if it contains a Reply-To: header, the
2162 errors_reply_to option, if set, is not used.
2164 PH/19 When calculating a retry time, Exim used to measure the "time since
2165 failure" by looking at the "first failed" field in the retry record. Now
2166 it does not use this if it is later than than the arrival time of the
2167 message. Instead it uses the arrival time. This makes for better
2168 behaviour in cases where some deliveries succeed, thus re-setting the
2169 "first failed" field. An example is a quota failure for a huge message
2170 when small messages continue to be delivered. Without this change, the
2171 "time since failure" will always be short, possible causing more frequent
2172 delivery attempts for the huge message than are intended.
2173 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2175 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2176 $1, $2, $3) because the numerical variables can be reset during some
2177 expansion items (e.g. "match"), thereby losing the authentication data.
2179 PH/21 Make -bV show the size of off_t variables so that the test suite can
2180 decide whether to run tests for quotas > 2G.
2182 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2183 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2184 filecount value is greater than 2G or if a quota value is greater than 2G
2185 on a system where the size of off_t is not greater than 4, a panic error
2188 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2189 never match. The debug and -bh output now contains an explicit error
2190 message indicating a malformed IPv4 address or mask.
2192 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2193 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2194 PH/23 above applies.
2196 PH/25 Do not write to syslog when running in the test harness. The only
2197 occasion when this arises is a failure to open the main or panic logs
2198 (for which there is an explicit test).
2200 PH/26 Added the /no_tell option to "control=freeze".
2202 PH/27 If a host name lookup failed very early in a connection, for example, if
2203 the IP address matched host_lookup and the reverse lookup yielded a name
2204 that did not have a forward lookup, an error message of the form "no IP
2205 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2206 could be logged. Now it outputs the IP address instead of "NULL".
2208 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2209 allows the sender and the authenticated sender to be set when
2210 submitting a message from within Exim. Since child_open_exim() is
2211 documented for local_scan(), the new function should be too.
2213 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2214 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2215 results in an empty string is now treated as unset.
2217 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2219 PH/31 Added sender_verify_fail logging option.
2221 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2222 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2223 tidied the source and removed it altogether.
2225 PH/33 When a queue run was abandoned because the load average was too high, a
2226 log line was always written; now it is written only if the queue_run log
2227 selector is set. In addition, the log line for abandonment now contains
2228 information about the queue run such as the pid. This is always present
2229 in "start" and "stop" lines but was omitted from the "abandon" line.
2231 PH/34 Omit spaces between a header name and the colon in the error message that
2232 is given when verify = headers_syntax fails (if there are lots of them,
2233 the message gets confusing).
2235 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2236 names, as there are now some PTR records that contain slashes. This check
2237 is only to protect against broken name servers that fall over on strange
2238 characters, so the fact that it applies to all lookups doesn't matter.
2240 PH/36 Now that the new test suite is complete, we can remove some of the
2241 special code in Exim that was needed for the old test suite. For example,
2242 sorting DNS records because real resolvers return them in an arbitrary
2243 order. The new test suite's fake resolver always returns records in the
2246 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2247 bounces) except when queue_only is set, to avoid logging races between
2248 the different processes.
2250 PH/38 Panic-die if .include specifies a non-absolute path.
2252 PH/39 A tweak to the "H" retry rule from its user.
2254 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2255 a label. They prevented compilation on older perls.
2257 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2258 a warning to be raised on newish perls.
2260 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2261 on queue. Changes to match documented behaviour of showing count of
2262 messages matching specified criteria.
2264 PH/40 Changed the default ident timeout from 30s to 5s.
2266 PH/41 Added support for the use of login_cap features, on those BSD systems
2267 that have them, for controlling the resources used by pipe deliveries.
2269 PH/42 The content-scanning code uses fopen() to create files in which to put
2270 message data. Previously it was not paying any attention to the mode of
2271 the files. Exim runs with umask(0) because the rest of the code creates
2272 files with open(), and sets the required mode explicitly. Thus, these
2273 files were ending up world-writeable. This was not a big issue, because,
2274 being within the spool directory, they were not world-accessible. I have
2275 created a function called modefopen, which takes an additional mode
2276 argument. It sets umask(777), creates the file, chmods it to the required
2277 mode, then resets the umask. All the relevant calls to fopen() in the
2278 content scanning code have been changed to use this function.
2280 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2281 to 24 hours. This avoids potential overflow problems when processing G
2282 and H retry rules. I suspect nobody ever tinkers with this value.
2284 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2286 PH/45 When the plaintext authenticator is running as a client, the server's
2287 challenges are checked to ensure they are valid base64 strings. By
2288 default, the authentication attempt is cancelled if an invalid string is
2289 received. Setting client_ignore_invalid_base64 true ignores these errors.
2290 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2291 they are received. Thus, the responses can be made to depend on the
2292 challenges. If an invalid string is ignored, an empty string is placed in
2295 PH/46 Messages that are created by the autoreply transport now contains a
2296 References: header, in accordance with RFCs 2822 and 3834.
2298 PH/47 Added authenticated_sender_force to the smtp transport.
2300 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2302 PH/49 Installed latest patch from the Sieve maintainer.
2304 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2305 was also set, the appendfile transport was unnecessarily scanning a
2306 directory of message files (e.g. for maildir delivery) to find the count
2307 of files (along with the size), even though it did not need this
2308 information. It now does the scan only if it needs to find either the
2309 size of the count of files.
2311 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2313 PH/52 Two bugs concerned with error handling when the smtp transport is
2316 (i) Exim was not creating retry information for temporary errors given
2317 for individual recipients after the DATA command when the smtp transport
2318 was used in LMTP mode. This meant that they could be retried too
2319 frequently, and not timed out correctly.
2321 (ii) Exim was setting the flag that allows error details to be returned
2322 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2323 recipients that were returned after the DATA command.
2325 PH/53 This is related to PH/52, but is more general: for any failing address,
2326 when detailed error information was permitted to be returned to the
2327 sender, but the error was temporary, then after the final timeout, only
2328 "retry timeout exceeded" was returned. Now it returns the full error as
2329 well as "retry timeout exceeded".
2331 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2332 do this, and (what is worse) MTAs that accept it.
2334 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2335 will now be deprecated.
2337 PH/56 New os.c-cygwin from the Cygwin maintainer.
2339 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2340 formats (previously only available in exim formats via -bpr, -bpru,
2341 and -bpra. Now also available in native and exiqgrep formats)
2343 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2344 with very large, slow to parse queues
2346 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2348 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2350 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2351 responses to authentication challenges, though it was showing the
2352 challenges; (ii) I've removed the CR characters from the debug output for
2355 PH/58 Allow for the insertion of a newline as well as a space when a string
2356 is turned into more than one encoded-word during RFC 2047 encoding. The
2357 Sieve code now uses this.
2359 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2360 data_4xx, lost_connection, tls_required.
2362 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2363 message was being sent as an SMTP response.
2365 PH/61 Add -l and -k options to exicyclog.
2367 PH/62 When verifying, if an address was redirected to one new address, so that
2368 verification continued, and the new address failed or deferred after
2369 having set something in $address_data, the value of $address_data was not
2370 passed back to the ACL. This was different to the case when no
2371 redirection occurred. The value is now passed back in both cases.
2373 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2374 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2375 use login_cap.h, so on its own it isn't the distinguishing feature. The
2376 new name refers directly to the setclassresources() function.
2378 PH/65 Added configuration files for NetBSD3.
2380 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2382 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2383 is preferred over IPv4.
2385 PH/68 The bounce_return_message and bounce_return_body options were not being
2386 honoured for bounces generated during the reception of non-SMTP messages.
2387 In particular, this applied to messages rejected by the ACL. This bug has
2388 been fixed. However, if bounce_return_message is true and bounce_return_
2389 body is false, the headers that are returned for a non-SMTP message
2390 include only those that have been read before the error was detected.
2391 (In the case of an ACL rejection, they have all been read.)
2393 PH/69 The HTML version of the specification is now built in a directory called
2394 spec_html instead of spec.html, because the latter looks like a path with
2395 a MIME-type, and this confuses some software.
2397 PH/70 Catch two compiler warnings in sieve.c.
2399 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2400 function verify_get_ident() calls ip_connect() to connect a socket, but
2401 if the "connect()" function timed out, ip_connect() used to close the
2402 socket. However, verify_get_ident() also closes the socket later, and in
2403 between Exim writes to the log, which may get opened at this point. When
2404 the socket was closed in ip_connect(), the log could get the same file
2405 descriptor number as the socket. This naturally causes chaos. The fix is
2406 not to close the socket in ip_connect(); the socket should be closed by
2407 the function that creates it. There was only one place in the code where
2408 this was missing, in the iplookup router, which I don't think anybody now
2409 uses, but I've fixed it anyway.
2411 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2412 well as to direct DNS lookups. Otherwise the handling of names in host
2413 lists is inconsistent and therefore confusing.
2419 PH/01 Two changes to the default runtime configuration:
2421 (1) Move the checks for relay_from_hosts and authenticated clients from
2422 after to before the (commented out) DNS black list checks.
2424 (2) Add control=submission to the relay_from_hosts and authenticated
2425 clients checks, on the grounds that messages accepted by these
2426 statements are most likely to be submissions.
2428 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2430 (1) Generate an error if the third argument for the ${prvs expansion is
2433 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2436 (3) Reset the variables that are obtained from the first argument of
2437 ${prvscheck and used in the second argument before leaving the code,
2438 because their memory is reclaimed, so using them afterwards may do
2441 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2442 one (it's much easier than Tom thought :-).
2444 (5) Because of (4), we can now allow for the use of $prvscheck_result
2445 inside the third argument.
2447 PH/03 For some reason, the default setting of PATH when running a command from
2448 a pipe transport was just "/usr/bin". I have changed it to
2451 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2452 anything to be listed in the output from -bV.
2454 PH/05 When a filter generated an autoreply, the entire To: header line was
2455 quoted in the delivery log line, like this:
2457 => >A.N.Other <ano@some.domain> <original@ddress> ...
2459 This has been changed so that it extracts the operative address. There
2460 may be more than one such address. If so, they are comma-separated, like
2463 => >ano@some.domain,ona@other.domain <original@ddress> ...
2465 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2466 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2467 not being looked up in the rDNS to get a host name, Exim was showing the
2468 IP address twice in Received: lines, even though the IP addresses were
2469 identical. For example:
2471 Received: from [1.2.3.4] (helo=[1.2.3.4])
2473 However, if the real host name was known, it was omitting the HELO data
2474 if it matched the actual IP address. This has been tidied up so that it
2475 doesn't show the same IP address twice.
2477 PH/07 When both +timestamp and +memory debugging was on, the value given by
2478 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2479 called by the debug printing, thereby overwriting the timestamp buffer.
2480 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2482 PH/08 When the original message was included in an autoreply transport, it
2483 always said "this is a copy of the message, including all the headers",
2484 even if body_only or headers_only was set. It now gives an appropriate
2487 PH/09 Applied a patch from the Sieve maintainer which:
2489 o fixes some comments
2490 o adds the (disabled) notify extension core
2491 o adds some debug output for the result of if/elsif tests
2492 o points to the current vacation draft in the documentation
2493 and documents the missing references header update
2497 o fixes a bug in processing the envelope test (when testing
2498 multiple envelope elements, the last element determinted the
2501 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2502 Electronic Mail") by including:
2504 Auto-submitted: auto-generated
2506 in the messages that it generates (bounce messages and others, such as
2507 warnings). In the case of bounce messages for non-SMTP mesages, there was
2508 also a typo: it was using "Auto_submitted" (underscore instead of
2509 hyphen). Since every message generated by Exim is necessarily in response
2510 to another message, thes have all been changed to:
2512 Auto-Submitted: auto-replied
2514 in accordance with these statements in the RFC:
2516 The auto-replied keyword:
2518 - SHOULD be used on messages sent in direct response to another
2519 message by an automatic process,
2521 - MUST NOT be used on manually-generated messages,
2523 - MAY be used on Delivery Status Notifications (DSNs) and Message
2524 Disposition Notifications (MDNs),
2526 - MUST NOT be used on messages generated by automatic or periodic
2527 processes, except for messages which are automatic responses to
2530 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2531 to the default Received: header definition.
2533 PH/12 Added log selector acl_warn_skipped (default on).
2535 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2536 variables because (a) they are in the wrong storage pool and (b) even if
2537 they were copied, it wouldn't work properly because of the caching.
2539 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2540 checking when decoding. Apparently there are clients that generate
2541 overlong encoded strings. Why am I not surprised?
2543 PH/15 If the first argument of "${if match_address" was not empty, but did not
2544 contain an "@" character, Exim crashed. Now it writes a panic log message
2545 and treats the condition as false.
2547 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2549 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2550 extension "envelope-auth". The code is finished and in agreement with
2551 other implementations, but there is no documentation so far and in fact,
2552 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2553 not changing the active code.
2555 Print executed "if" and "elsif" statements when debugging is used. This
2556 helps a great deal to understand what a filter does.
2558 Document more things not specified clearly in RFC3028. I had all this
2559 sorted out, when out of a sudden new issues came to my mind. Oops."
2561 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2564 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2565 canonical form (as documented). However, after a host name lookup from
2566 the IP address, check_host() was doing a simple string comparison with
2567 addresses acquired from the DNS when checking that the found name did
2568 have the original IP as one of its addresses. Since any found IPv6
2569 addresses are likely to be in abbreviated form, the comparison could
2570 fail. Luckily, there already exists a function for doing the comparison
2571 by converting both addresses to binary, so now that is used instead of
2572 the text comparison.
2574 PH/20 There was another similar case to PH/19, when a complete host name was
2575 given in a host list; looking up its IP address could give an abbreviated
2576 form, whereas the current host's name might or might not be abbreviated.
2577 The same fix has been applied.
2583 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2584 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2587 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2588 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2590 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2592 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2593 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2594 Also, for "Auto-Submitted", treat anything other than "no" as
2595 non-personal, in accordance with RFC 3834. (Previously it treated
2596 anything starting "auto-" as non-personal.)
2598 TF/01 The control=submission/name=... option had a problem with syntax
2599 errors if the name included a slash character. The /name= option
2600 now slurps the rest of the string, so it can include any characters
2601 but it must come last in the list of options (after /sender_retain
2604 PH/05 Some modifications to the interface to the fake nameserver for the new
2612 TK/01 Added the "success_on_redirect" address verification option. See
2613 NewStuff for rationale and an example.
2615 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2617 PH/02 Patch to exigrep to allow it to work on syslog lines.
2619 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2620 fread() to skip over the body file's header line, because in Cygwin the
2621 header line is locked and is inaccessible.
2623 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2624 co-exist for some time) to make it clear that it is the Exim ID that is
2625 referenced, not the Message-ID: header line.
2627 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2628 string_format() function, because snprintf() does not exist on all
2631 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2632 ${stat: expansion item.
2634 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2635 protocol synchronization error", to keep the pedants happy.
2637 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2638 well as for IRIX systems, when gcc is being used. See the host.c source
2641 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2643 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2646 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2647 transport and to the smtp transport in LMTP mode.
2649 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2651 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2652 run a filter in a subprocess. This could lead to confusion in subsequent
2653 lookups in the parent process. There should also be a search_tidyup() at
2654 the end of the subprocess.
2656 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2657 only if the host matched helo_try_verify_hosts, which caused the
2658 verification to occur when the EHLO/HELO command was issued. The ACL just
2659 tested the remembered result. Now, if a previous verification attempt has
2660 not happened, "verify = helo" does it there and then.
2662 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2664 TK/03 Fix log output including CR from clamd.
2666 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2667 exist provoked a memory error which could cause a segfault.
2669 PH/15 Installed PCRE 6.2
2671 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2673 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2674 of the problem. Specifically, suggested +O2 rather than +O1 for the
2677 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2679 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2680 record, the port was not correctly logged when the outgoing_port log
2681 selector was set (it logged the transort's default port).
2683 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2684 fallback_hosts, and "hosts" in the smtp transport.
2686 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2687 host errors such as "Connection refused".
2689 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2690 authentication with radiusclient 0.4.9:
2692 - Error returned from rc_read_config was caught wrongly
2693 - Username/password not passed on to radius server due to wrong length.
2695 The presumption is that some radiusclient API changes for 4.51/PH/17
2696 were not taken care of correctly. The code is still untested by me (my
2697 Linux distribution still has 0.3.2 of radiusclient), but it was
2698 contributed by a Radius user.
2700 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2701 expanding the "port" option of the smtp transport.
2703 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2704 while reading a MIME header. Thanks to Tom Hughes for a patch.
2706 PH/24 Include config.h inside local_scan.h so that configuration settings are
2709 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2710 that in an ACL for RCPT (for example), you can examine exactly what was
2713 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2714 commands, but it was not correctly comparing the address with the actual
2715 client host address. Thus, it would show the EHLO address in Received:
2716 header lines when this was not necessary.
2718 PH/27 Added the % operator to ${eval:}.
2720 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2721 it should be ignoring failures (because with -C, for example, it has lost
2722 privilege). It wasn't ignoring creation failures other than "already
2725 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2728 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2729 because an input line was too long, either on its own, or by virtue of
2730 too many continuations, the temporary file was not being removed, and the
2731 return code was incorrect.
2733 PH/31 Missing "BOOL" in function definition in filtertest.c.
2735 PH/32 Applied Sieve patches from the maintainer.
2737 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2739 PH/33 Added "verify = not_blind".
2741 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2742 Local/Makefile (with some defaults set). These are used in built scripts
2743 such as exicyclog, but they have never been used in the exim_install
2744 script (though there are many overriding facilities there). I have
2745 arranged that the exim_install script now takes note of these two
2748 PH/35 Installed configuration files for Dragonfly.
2750 PH/36 When a locally submitted message by a trusted user did not contain a
2751 From: header, and the sender address was obtained from -f or from an SMTP
2752 MAIL command, and the trusted user did not use -F to supply a sender
2753 name, $originator_name was incorrectly used when constructing a From:
2754 header. Furthermore, $originator_name was used for submission mode
2755 messages from external hosts without From: headers in a similar way,
2756 which is clearly wrong.
2758 PH/37 Added control=suppress_local_fixups.
2760 PH/38 When log_selector = +received_sender was set, and the addition of the
2761 sender made the log line's construction buffer exactly full, or one byte
2762 less than full, an overflow happened when the terminating "\n" was
2765 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2766 when the result of a list match is failure because a DNS lookup failed.
2768 PH/40 RM_COMMAND is now used in the building process.
2770 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2771 the "build-* directories that it finds.
2773 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2774 address in a domain literal was a prefix of an interface address.
2776 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2777 when verifying a sender address, unless rewrite_headers is false.
2779 PH/44 Wrote a long comment about why errors_to addresses are verified as
2780 recipients, not senders.
2782 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2783 the ratelimit ACL was added.
2785 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2787 PH/46 Added extra information about PostgreSQL errors to the error string.
2789 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2790 avoiding the need to install special zones in a real server. This is
2791 backwards compatible; if it can't find the fake resolver, it drops back.
2792 Thus, both old and new test suites can be run.
2794 TF/02 Added util/ratelimit.pl
2796 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2797 clock is set back in time.
2799 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2800 Candler <B.Candler@pobox.com>.
2802 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2803 OK for addresses that are the result of redirections.
2805 PH/48 A number of further additions for the benefit of the new test suite,
2806 including a fake gethostbyname() that interfaces to the fake DNS resolver
2809 TF/06 The fix for widen_domains has also been applied to qualify_single and
2810 search_parents which are the other dnslookup options that can cause
2813 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2816 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2818 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2819 service" if it encountered an MX record with an empty target hostname.
2820 The message is now "an MX or SRV record indicated no SMTP service".
2822 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2823 if the DNS of the sending site is misconfigured. This is quite a
2824 common situation. This change restores the behaviour of treating a
2825 helo verification defer as a failure.
2827 PH/51 If self=fail was set on a router, the bounce message did not include the
2828 actual error message.
2834 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2836 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2837 command itself ended in error, the underlying message about the transport
2838 filter timeout was being overwritten with the pipe command error. Now the
2839 underlying error message should be appended to the second error message.
2841 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2843 PH/02 Reduce the amount of output that "make" produces by default. Full output
2844 can still be requested.
2846 PH/03 The warning log line about a condition test deferring for a "warn" verb
2847 was being output only once per connection, rather than after each
2848 occurrence (because it was using the same function as for successful
2849 "warn" verbs). This seems wrong, so I have changed it.
2851 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2852 it should not have, which might have caused a crash in the right
2853 circumstances, but probably never did.
2855 PH/04 Installed a modified version of Tony Finch's patch to make submission
2856 mode fix the return path as well as the Sender: header line, and to
2857 add a /name= option so that you can make the user's friendly name appear
2860 TF/03 Added the control = fakedefer ACL modifier.
2862 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2863 Mark Lowes for thorough testing.
2865 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2867 TK/03 Merged latest SRS patch from Miles Wilton.
2869 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2870 with the definition in sysexits.h (which is #included earlier).
2871 Fortunately, Exim does not actually use EX_OK. The code used to try to
2872 preserve the sysexits.h value, by assumimg that macro definitions were
2873 scanned for macro replacements. I have been disabused of this notion,
2874 so now the code just undefines EX_OK before #including unistd.h.
2876 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2877 in the smtp transport. When a block could not be written in a single
2878 write() function, the timeout was being re-applied to each part-write.
2879 This seems wrong - if the receiver was accepting one byte at a time it
2880 would take for ever. The timeout is now adjusted when this happens. It
2881 doesn't have to be particularly precise.
2883 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2884 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2886 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2887 without a domain if the check to <postmaster@domain> fails.
2889 SC/01 Eximstats: added -xls and the ability to specify output files
2890 (patch written by Frank Heydlauf).
2892 SC/02 Eximstats: use FileHandles for outputing results.
2894 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2896 SC/04 Eximstats: fixed display of large numbers with -nvr option
2898 SC/05 Eximstats: fixed merging of reports with empty tables.
2900 SC/06 Eximstats: added the -include_original_destination flag
2902 SC/07 Eximstats: removed tabs and trailing whitespace.
2904 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2906 TK/06 MBOX spool code: Add real "From " MBOX separator line
2907 so the .eml file is really in mbox format (even though
2908 most programs do not really care). Patch from Alex Miller.
2910 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2911 The latter is generated from $received_to and is only set if the
2912 message has one envelope recipient. SA can use these headers,
2913 obviously out-of-the-box. Patch from Alex Miller.
2915 PH/08 The ${def test on a variable was returning false if the variable's
2916 value was "0", contrary to what the specification has always said!
2917 The result should be true unless the variable is empty.
2919 PH/09 The syntax error of a character other than { following "${if
2920 def:variable_name" (after optional whitespace) was not being diagnosed.
2921 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2922 accidental colon was present, for example, could give incorrect results.
2924 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2925 result is used (not including appendfile, where other changes are about
2928 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2929 This involved changing a lot of size variables from int to off_t. It
2930 should work with maildirs and everything.
2932 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2933 spamd dying while we are connected to it.
2935 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2938 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2939 function for BDB 4.3.
2941 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2943 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2944 the sources for PCRE have been re-arranged and more files are now
2947 PH/15 The code I had for printing potentially long long variables in PH/11
2948 above was not the best (it lost precision). The length of off_t variables
2949 is now inspected at build time, and an appropriate printing format (%ld
2950 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2951 to be "long long int" or "long int". This is needed for the internal
2952 formatting function string_vformat().
2954 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2955 the configuration file to be ":syslog", then the script "guesses" where
2956 the logs files are, rather than using the compiled in default. In our
2957 case the guess is not the same as the compiled default, so the script
2958 suddenly stopped working when I started to use syslog. The patch checks
2959 to see if log_file_path is "". If so, it attempts to read it from exim
2960 with no configuration file to get the compiled in version, before it
2961 falls back to the previous guessing code."
2963 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2964 implementing BATV in an Exim configuration. See NewStuff for the gory
2967 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2968 Makefile that are specific to HP-UX.
2970 PH/18 If the "use_postmaster" option was set for a recipient callout together
2971 with the "random" option, the postmaster address was used as the MAIL
2972 FROM address for the random test, but not for the subsequent recipient
2973 test. It is now used for both.
2975 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2976 patch removes a few documentation additions to RFC 3028, because the
2977 latest draft now contains them. It adds the new en;ascii-case comparator
2978 and a new error check for 8bit text in MIME parts. Comparator and
2979 require names are now matched exactly. I enabled the subaddress
2980 extension, but it is not well tested yet (read: it works for me)."
2982 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2983 rework some of the code of TK/09 above to avoid the hardwired use of
2984 "%lld" and "long long". Replaced the call to snprintf() with a call to
2987 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2988 records point to non-existent hosts", "retry timeout exceeded", and
2989 "retry time not reached for any host after a long failure period".
2991 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2992 experimental DomainKeys support:
2994 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2995 (2) On an error such as an illegally used "control", the wrong name for
2996 the control was given.
2998 These problems did NOT occur unless DomainKeys support was compiled.
3000 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3002 PH/24 Added ${if match_ip condition.
3004 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3005 databases so that it will be absolutely obvious if a crash occurs in the
3006 DB library. This is a regular occurrence (often caused by mis-matched
3009 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3010 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3011 picked up on a user's system that detects such things. There doesn't seem
3012 to be a gcc warning option for this - only an attribute that has to be
3013 put on the function's prototype. It seems that in Fedora Core 4 they have
3014 set this on a number of new functions. No doubt there will be more in due
3017 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3018 specify a transport. However, if an address that was verified by such a
3019 router was the subject of a callout, Exim crashed because it tried to
3020 read the rcpt_include_affixes from the non-existent transport. Now it
3021 just assumes that the setting of that option is false. This bug was
3022 introduced by 4.51/PH/31.
3024 PH/28 Changed -d+all to exclude +memory, because that information is very
3025 rarely of interest, but it makes the output a lot bigger. People tend to
3026 do -d+all out of habit.
3028 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3029 code in os-type was giving problems when libc.so lives in lib64, like on
3032 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3033 aren't the modern standard, and it seems that some systems' include files
3034 don't always have them. Exim was already checking for some of the newer
3035 ones like T_AAAA, and defining it itself. I've added checks for all the
3036 record types that Exim uses.
3038 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3039 not automatically generating a new one, as it is supposed to. This
3040 prevented TLS from working. If the file did exist, but contained invalid
3041 data, a new version was generated, as expected. It was only the case of a
3042 non-existent file that was broken.
3044 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3045 with a change in libdomainkeys > 0.64.
3047 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3048 from DNS. If the selector record carries the flag, it now has
3049 precedence over the domain-wide flag.
3051 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3053 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3054 the use of a port name that isn't defined in /etc/services) occurred, the
3055 message was deferred as in a normal delivery, and thus remained on the
3056 spool, instead of being failed because of the mua_wrapper setting. This
3057 is now fixed, and I tidied up some of the mua_wrapper messages at the
3060 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3061 the messages in a hash of arrays rather than using individual hashes.
3062 This is a bit cleaner and results in dramatic memory savings, albeit
3063 at a slight CPU cost.
3065 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3066 as requested by Marc Sherman.
3068 SC/10 Eximstats: added histograms for user specified patterns as requested
3071 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3073 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3074 fopen() in the content-scanning modules that did not already have it.
3080 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3081 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3083 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3085 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3087 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3088 to test Sieve filters that use "vacation".
3090 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3091 that changes the way the GnuTLS parameters are stored in the cache file.
3092 The new format can be generated externally. For backward compatibility,
3093 if the data in the cache doesn't make sense, Exim assumes it has read an
3094 old-format file, and it generates new data and writes a new file. This
3095 means that you can't go back to an older release without removing the
3098 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3099 work if there are any delivery delays because "one_time" forces the
3100 parent to be marked "delivered", so its unseen clone is never tried
3101 again. For this reason, Exim now forbids the simultaneous setting of
3104 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3105 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3106 entry if you want to know the details. The fix, however, neglected to
3107 consider the case when local delivery batching is involved. The test for
3108 "previously delivered" was not happening when checking to see if an
3109 address could be batched with a previous (undelivered) one; under
3110 certain circumstances this could lead to multiple deliveries to the same
3113 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3114 in its include files, and this causes problems building Exim.
3116 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3117 header_syntax) but Exim was just ignoring anything given after a slash.
3118 In particular, this caused confusion with an attempt to use "verify =
3119 reverse_host_lookup/defer_ok". An error is now given when options are
3120 supplied for verify items that do not have them. (Maybe reverse_host_
3121 lookup should have a defer_ok option, but that's a different point.)
3123 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3124 defined by RFC 821) to 2048, because there were problems with some AUTH
3125 commands, and RFC 1869 says the size should be increased for extended
3126 SMTP commands that take arguments.
3128 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3131 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3132 "unknown" error; now it says that the functionality isn't in the binary.
3134 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3135 an address' error message when a string expansion fails (syntax or
3136 whatever). Otherwise the password may appear in the log. Following change
3137 PH/42 below, there is no longer a chance of it appearing in a bounce
3140 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3142 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3143 of its addresses was ever tried. (Bugzilla bug #2.)
3145 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3146 the result incorrectly in the debug output. (It correctly added a newline
3147 to what was transported.)
3149 TF/01 Added $received_time.
3151 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3152 commented out examples of how to interface to a virus scanner and to
3153 SpamAssassin. Also added commented examples of av_scanner and
3154 spamd_address settings.
3156 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3157 and controls are allowed in which ACLs. There were a couple of minor
3158 errors. Some of the entries in the conditions table (which is a table of
3159 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3160 negation of where the condition IS allowed.
3162 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3164 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3165 header file does not have a version number, so I've had to invent a new
3166 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3167 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3168 radiusclient), but it was contributed by a Radius user.
3170 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3171 files or directories, for OpenSSL.
3173 PH/19 When an Exim process that is running as root has to create an Exim log
3174 file, it does so in a subprocess that runs as exim:exim so as to get the
3175 ownership right at creation (otherwise, other Exim processes might see
3176 the file with the wrong ownership). There was no test for failure of this
3177 fork() call, which would lead to the process getting stuck as it waited
3178 for a non-existent subprocess. Forks do occasionally fail when resources
3179 run out. I reviewed all the other calls to fork(); they all seem to check
3182 PH/20 When checking for unexpected SMTP input at connect time (before writing
3183 the banner), Exim was not dealing correctly with a non-positive return
3184 from the read() function. If the client had disconnected by this time,
3185 the result was a log entry for a synchronization error with an empty
3186 string after "input=" when read() returned zero. If read() returned -1
3187 (an event I could not check), uninitialized data bytes were printed.
3188 There were reports of junk text (parts of files, etc) appearing after
3191 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3193 PH/22 Added support for macro redefinition, and (re)definition in between
3194 driver and ACL definitions.
3196 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3197 forgetting to use the resulting value; it was using the unexpanded value.
3199 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3200 hadn't been configured. The fix is from Juergen Kreileder, who
3201 understands it better than I do:
3203 "Here's what I see happening with three configured cyrus_sasl
3204 authenticators configured (plain, login, cram-md5):
3206 On startup auth_cyrus_sasl_init() gets called for each of these.
3207 This means three calls to sasl_listmech() without a specified mech_list.
3208 => SASL tests which mechs of all available mechs actually work
3209 => three warnings about OTP not working
3210 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3212 With the patch, sasl_listmech() also gets called three times. But now
3213 SASL's mech_list option is set to the server_mech specified in the the
3214 authenticator. Or in other words, the answer from sasl_listmech()
3215 gets limited to just the mech you're testing for (which is different
3217 => the return list contains just 'plain' or 'login', 'cram-md5' or
3218 nothing depending on the value of ob->server_mech.
3220 I've just tested the patch: Authentication still works fine,
3221 unavailable mechs specified in the exim configuration are still
3222 caught, and the auth.log warnings about OTP are gone."
3224 PH/25 When debugging is enabled, the contents of the command line are added
3225 to the debugging output, even when log_selector=+arguments is not
3228 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3229 answer is "GNU", and only if the return is "GNU/something" is the answer
3232 PH/27 $acl_verify_message is now set immediately after the failure of a
3233 verification in an ACL, and so is available in subsequent modifiers. In
3234 particular, the message can be preserved by coding like this:
3236 warn !verify = sender
3237 set acl_m0 = $acl_verify_message
3239 Previously, $acl_verify_message was set only while expanding "message"
3240 and "log_message" when a very denied access.
3242 PH/28 Modified OS/os.c-Linux with
3244 -#ifndef OS_LOAD_AVERAGE
3245 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3247 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3248 nomenclature these days.)
3250 PH/29 Installed patch from the Sieve maintainer that adds the options
3251 sieve_useraddress and sieve_subaddress to the redirect router.
3253 PH/30 In these circumstances:
3254 . Two addresses routed to the same list of hosts;
3255 . First host does not offer TLS;
3256 . First host accepts first address;
3257 . First host gives temporary error to second address;
3258 . Second host offers TLS and a TLS session is established;
3259 . Second host accepts second address.
3260 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3261 and peerdn, if requested) that were in fact used only for the second
3264 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3265 attention to any local part prefix or suffix that was matched by the
3266 router that accepted the address. It now behaves in the same way as it
3267 does for delivery: the affixes are removed from the local part unless
3268 rcpt_include_affixes is set on the transport.
3270 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3271 timeout during the DATA phase of an incoming message.
3273 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3274 applied a patch sanctioned by the Sieve maintainer.
3276 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3277 the uid or gid is negative. A case of a negative gid caused this to be
3278 noticed. The fix allows for either to be negative.
3280 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3281 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3284 PH/36 The change PH/12 above was broken. Fixed it.
3286 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3287 the grounds that routing the same address twice would always produce the
3288 same answer. This might have been true once, but it is certainly no
3289 longer true now. Routing a child address may depend on the previous
3290 routing that produced that child. Some complicated redirection strategies
3291 went wrong when messages had multiple recipients, and made Exim's
3292 behaviour dependent on the order in which the addresses were given.
3294 I have moved the duplicate checking until after the routing is complete.
3295 Exim scans the addresses that are assigned to local and remote
3296 transports, and removes any duplicates. This means that more work will be
3297 done, as duplicates will always all be routed, but duplicates are
3298 presumably rare, so I don't expect this is of any significance.
3300 For deliveries to pipes, files, and autoreplies, the duplicate checking
3301 still happens during the routing process, since they are not going to be
3304 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3305 It corrects a timeout issue with spamd. This is Ian's comment: "The
3306 background is that sometimes spamd either never reads data from a
3307 connection it has accepted, or it never writes response data. The exiscan
3308 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3309 blindly assumes that writes won't block so it may never time out."
3311 PH/39 Allow G after quota size as well as K and M.
3313 PH/40 The value set for $authenticated_id in an authenticator may not contain
3314 binary zeroes or newlines because the value is written to log lines and
3315 to spool files. There was no check on this. Now the value is run through
3316 the string_printing() function so that such characters are converted to
3317 printable escape sequences.
3319 PH/41 $message_linecount is a new variable that contains the total number of
3320 lines in the message. Compare $body_linecount, which is the count for the
3323 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3324 bounce and delay warning messages, except in certain special cases, which
3327 (a) An SMTP error message from a remote host;
3328 (b) A message specified in a :fail: redirection;
3329 (c) A message specified in a "fail" command in a system filter;
3330 (d) A message specified in a FAIL return from the queryprogram router;
3331 (e) A message specified by the cannot_route_message router option.
3333 In these cases only, Exim does include the error details in bounce and
3334 warning messages. There are also a few cases where bland messages such
3335 as "unrouteable address" or "local delivery error" are given.
3337 PH/43 $value is now also set for the "else" part of a ${run expansion.
3339 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3340 being worked on, but at least Exim now implements the latest version to
3343 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3344 process to complete was treated as a delivery failure, a timeout while
3345 writing the message to the pipe was logged, but erroneously treated as a
3346 successful delivery. Such timeouts include transport filter timeouts. For
3347 consistency with the overall process timeout, these timeouts are now
3348 treated as errors, giving rise to delivery failures by default. However,
3349 there is now a new Boolean option for the pipe transport called
3350 timeout_defer, which, if set TRUE, converts the failures into defers for
3351 both kinds of timeout. A transport filter timeout is now identified in
3354 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3355 systems where "make" and "gmake" are different, calling "gmake" at top
3356 level broke things. I've arranged for the value of $(MAKE) to be passed
3357 from the Makefile to this script so that it can call the same version of
3361 A note about Exim versions 4.44 and 4.50
3362 ----------------------------------------
3364 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3365 changes of various kinds. As a consequence, a big documentation update was
3366 needed. This delayed the release for rather longer than seemed good, especially
3367 in the light of a couple of (minor) security issues. Therefore, the changes
3368 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3369 release. So 4.44 and 4.50 are in effect two different branches that both start
3372 I have left the 4.50 change log unchanged; it contains all the changes since
3373 4.43. The change log for 4.44 is below; many of its items are identical to
3374 those for 4.50. This seems to be the most sensible way to preserve the
3375 historical information.
3381 1. Minor wording change to the doc/README.SIEVE file.
3383 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3384 computation of the current number of files was incorrect.
3386 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3387 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3390 4. Give more explanation in the error message when the command for a transport
3391 filter fails to execute.
3393 5. There are several places where Exim runs a non-Exim command in a
3394 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3395 was being done only for the command run by the queryprogram router. It is
3396 now done for all such subprocesses. The other cases are: ${run, transport
3397 filters, and the commands run by the lmtp and pipe transports.
3399 6. Added CONFIGURE_GROUP build-time option.
3401 7. Some older OS have a limit of 256 on the maximum number of file
3402 descriptors. Exim was using setrlimit() to set 1000 as a large value
3403 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3404 systems. I've change it so that if it can't get 1000, it tries for 256.
3406 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3407 was an oversight, and furthermore, ever since the addition of extra
3408 controls (e.g. 4.43/32), the checks on when to allow different forms of
3409 "control" were broken. There should now be diagnostics for all cases when a
3410 control that does not make sense is encountered.
3412 9. Added the /retain_sender option to "control=submission".
3414 10. $recipients is now available in the predata ACL (oversight).
3416 11. Tidy the search cache before the fork to do a delivery from a message
3417 received from the command line. Otherwise the child will trigger a lookup
3418 failure and thereby defer the delivery if it tries to use (for example) a
3419 cached ldap connection that the parent has called unbind on.
3421 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3422 of $address_data from the recipient verification was clobbered by the
3423 sender verification.
3425 13. The value of address_data from a sender verification is now available in
3426 $sender_address_data in subsequent conditions in the ACL statement.
3428 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3430 15. Added a new option "connect=<time>" to callout options, to set a different
3433 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3434 was its contents. (It was OK if the option was not defined at all.)
3436 17. A "Completed" log line is now written for messages that are removed from
3437 the spool by the -Mrm option.
3439 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3440 information about exactly what failed.
3442 19. Added -dd to debug only the daemon process.
3444 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3445 handles timeouts, both on the server side and network timeouts. Renamed the
3446 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3448 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3449 It is now set to "smtps".
3451 22. $host_address is now set to the target address during the checking of
3452 ignore_target_hosts.
3454 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3455 being passed; this would have caused $sender_host_name to have been used if
3456 matching the list had actually called for a host name (not very likely,
3457 since this list is usually IP addresses). A host name is now passed as
3460 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3461 code that specifies a non-restarting handler (typically sigaction() in
3462 modern systems) in an attempt to fix a rare and obscure crash bug.
3464 25. Narrowed the window for a race in the daemon that could cause it to ignore
3465 SIGCHLD signals. This is not a major problem, because they are used only to
3466 wake it up if nothing else does.
3468 26. A malformed maildirsize file could cause Exim to calculate negative values
3469 for the mailbox size or file count. Odd effects could occur as a result.
3470 The maildirsize information is now recalculated if the size or filecount
3473 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3474 support for a long time. Removed HAVE_SYS_VFS_H.
3476 28. Installed the latest version of exipick from John Jetmore.
3478 29. In an address list, if the pattern was not a regular expression, an empty
3479 subject address (from a bounce message) matched only if the pattern was an
3480 empty string. Non-empty patterns were not even tested. This was the wrong
3481 because it is perfectly reasonable to use an empty address as part of a
3482 database query. An empty address is now tested by patterns that are
3483 lookups. However, all the other forms of pattern expect the subject to
3484 contain a local part and a domain, and therefore, for them, an empty
3485 address still always fails if the pattern is not itself empty.
3487 30. Exim went into a mad DNS loop when attempting to do a callout where the
3488 host was specified on an smtp transport, and looking it up yielded more
3489 than one IP address.
3491 31. Re-factored the code for checking spool and log partition space into a
3492 function that finds that data and another that does the check. The former
3493 is then used to implement four new variables: $spool_space, $log_space,
3494 $spool_inodes, and $log_inodes.
3496 32. The RFC2047 encoding function was originally intended for short strings
3497 such as real names; it was not keeping to the 75-character limit for
3498 encoded words that the RFC imposes. It now respects the limit, and
3499 generates multiple encoded words if necessary. To be on the safe side, I
3500 have increased the buffer size for the ${rfc2047: expansion operator from
3503 33. It is now permitted to omit both strings after an "if" condition; if the
3504 condition is true, the result is "true". As before, when the second string
3505 is omitted, a false condition yields an empty string. This makes it less
3506 cumbersome to write custom ACL and router conditions.
3508 34. Failure to deliver a bounce message always caused it to be frozen, even if
3509 there was an errors_to setting on the router. The errors_to setting is now
3512 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3513 canonical form (fully expanded) before being placed in
3514 $sender_host_address.
3516 36. The table in the code that translates DNS record types into text (T_A to
3517 "A" for instance) was missing entries for NS and CNAME. It is just possible
3518 that this could have caused confusion if both these types were looked up
3519 for the same domain, because the text type is used as part of Exim's
3520 per-process caching. But the chance of anyone hitting this buglet seems
3523 37. The dnsdb lookup has been extended in a number of ways.
3525 (1) There is a new type, "zns", which walks up the domain tree until it
3526 finds some nameserver records. It should be used with care.
3528 (2) There is a new type, "mxh", which is like "mx" except that it returns
3529 just the host names, not the priorities.
3531 (3) It is now possible to give a list of domains (or IP addresses) to be
3532 looked up. The behaviour when one of the lookups defers can be
3533 controlled by a keyword.
3535 (4) It is now possible to specify the separator character for use when
3536 multiple records are returned.
3538 38. The dnslists ACL condition has been extended: it is now possible to supply
3539 a list of IP addresses and/or domains to be looked up in a particular DNS
3542 39. Added log_selector=+queue_time_overall.
3544 40. When running the queue in the test harness, wait just a tad after forking a
3545 delivery process, to get repeatability of debugging output.
3547 41. Include certificate and key file names in error message when GnuTLS fails
3548 to set them up, because the GnuTLS error message doesn't include the name
3549 of the failing file when there is a problem reading it.
3551 42. Allow both -bf and -bF in the same test run.
3553 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3555 44. The "Exiscan patch" is now merged into the mainline Exim source.
3557 45. Sometimes the final signoff response after QUIT could fail to get
3558 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3559 < 0 before doing a fflush(). This bug looks as though it goes back to the
3560 introduction of TLS in release 3.20, but "sometimes" must have been rare
3561 because the tests only now provoked it.
3563 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3564 (this can affect the format of dates).
3566 47. exim_tidydb, when checking for the continued existence of a message for
3567 which it has found a message-specific retry record, was not finding
3568 messages that were in split spool directories. Consequently, it was
3569 deleting retry records that should have stayed in existence.
3571 48. Steve fixed some bugs in eximstats.
3573 49. The SPA authentication driver was not abandoning authentication and moving
3574 on to the next authenticator when an expansion was forced to fail,
3575 contradicting the general specification for all authenticators. Instead it
3576 was generating a temporary error. It now behaves as specified.
3578 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3579 (the order specifies the preference for clients). The order is now AES256,
3580 AES128, 3DES, ARCFOUR128.
3582 51. Small patch to Sieve code - explicitly set From: when generating an
3585 52. Exim crashed if a remote delivery caused a very long error message to be
3586 recorded - for instance if somebody sent an entire SpamAssassin report back
3587 as a large number of 550 error lines. This bug was coincidentally fixed by
3588 increasing the size of one of Exim's internal buffers (big_buffer) that
3589 happened as part of the Exiscan merge. However, to be on the safe side, I
3590 have made the code more robust (and fixed the comments that describe what
3593 53. Now that there can be additional text after "Completed" in log lines (if
3594 the queue_time_overall log selector is set), a one-byte patch to exigrep
3595 was needed to allow it to recognize "Completed" as not the last thing in
3598 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3599 patch that reportedly fixes this has been added. I am not expert enough to
3600 create a test for it. This is what the patch creator wrote:
3602 "I found a little strange behaviour of ldap code when working with
3603 Windows 2003 AD Domain, where users was placed in more than one
3604 Organization Units. When I tried to give exim partial DN, the exit code
3605 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3606 But simultaneously result of request was absolutely normal ldap result,
3607 so I produce this patch..."
3609 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3610 REFERENCE, so I have modified the code to exclude the patch when that macro
3613 55. Some experimental protocols are using DNS PTR records for new purposes. The
3614 keys for these records are domain names, not reversed IP addresses. The
3615 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3616 leaves it alone. Component reversal etc. now happens only for IP addresses.
3619 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3621 57. Double the size of the debug message buffer (to 2048) so that more of very
3622 long debug lines gets shown.
3624 58. The exicyclog utility now does better if the number of log files to keep
3625 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3627 59. Two changes related to the smtp_active_hostname option:
3629 (1) $smtp_active_hostname is now available as a variable.
3630 (2) The default for smtp_banner uses $smtp_active_hostname instead
3631 of $primary_hostname.
3633 60. The host_aton() function is supposed to be passed a string that is known
3634 to be a valid IP address. However, in the case of IPv6 addresses, it was
3635 not checking this. This is a hostage to fortune. Exim now panics and dies
3636 if the condition is not met. A case was found where this could be provoked
3637 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3638 components; fortuitously, this particular loophole had already been fixed
3639 by change 4.50/55 above.
3641 If there are any other similar loopholes, the new check in host_aton()
3642 itself should stop them being exploited. The report I received stated that
3643 data on the command line could provoke the exploit when Exim was running as
3644 exim, but did not say which command line option was involved. All I could
3645 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3646 running as the user.
3649 61. There was a buffer overflow vulnerability in the SPA authentication code
3650 (which came originally from the Samba project). I have added a test to the
3651 spa_base64_to_bits() function which I hope fixes it.
3654 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3655 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3657 63. The daemon start-up calls getloadavg() while still root for those OS that
3658 need the first call to be done as root, but it missed one case: when
3659 deliver_queue_load_max is set with deliver_drop_privilege. This is
3660 necessary for the benefit of the queue runner, because there is no re-exec
3661 when deliver_drop_privilege is set.
3663 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3664 This has been fixed.
3666 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3667 was in use, was not putting the data itself into the right store pool;
3668 consequently, it could be overwritten for a subsequent message in the same
3669 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3672 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3674 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3675 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3676 the function were treating the return as a boolean value, which happened to
3677 work because 0=false and not-0=true, but is not correct code.
3679 68. The host_aton() function was not handling scoped IPv6 addresses (those
3680 with, for example, "%eth0" on the end) correctly.
3682 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3683 negated items (that is, ~something) in unsigned ints. Some compilers
3684 apparently mutter when there is no cast.
3686 70. If an address verification called from an ACL failed, and did not produce a
3687 user-specific message (i.e. there was only a "system" message), nothing was
3688 put in $acl_verify_message. In this situation, it now puts the system
3691 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3692 session; change 4.31/43 added the unwanted input to the log line - except
3693 that it did not do this in the start of session case. It now does.
3695 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3696 This is wrong and can cause the other end to generate a synchronization
3697 error if it is another Exim or anything else that does the synchronization
3698 check. A QUIT command is no longer sent after a timeout.
3700 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3701 during host lookups.
3703 74. The defer_ok option of callout verification was not working if it was used
3704 when verifying addresses in header lines, that is, for this case:
3706 verify = header_sender/callout=defer_ok
3708 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3709 those file descriptors could be used for SMTP connections. If anything
3710 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3711 could be sent to the SMTP client, causing chaos. The daemon now opens
3712 stdin, stdout, and stderr to /dev/null when it puts itself into the
3715 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3716 log by default. The user can override this with suitable Perl magic.
3718 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3719 the log message when discard triggers, was not working for the DATA ACL or
3720 for the non-SMTP ACL.
3722 78. Error message wording change in sieve.c.
3724 79. If smtp_accept_max_per_host was set, the number of connections could be
3725 restricted to fewer than expected, because the daemon was trying to set up
3726 a new connection before checking whether the processes handling previous
3727 connections had finished. The check for completed processes is now done
3728 earlier. On busy systems, this bug wouldn't be noticed because something
3729 else would have woken the daemon, and it would have reaped the completed
3732 80. If a message was submitted locally by a user whose login name contained one
3733 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3734 It caused a spool format error. I have fixed the spool reading code. A
3735 related problem was that the "from" clause in the Received: line became
3736 illegal because of the space(s). It is now covered by ${quote_local_part.
3738 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3741 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3743 83. Patch from Sieve maintainer to fix unterminated string problem in
3744 "vacation" handling.
3746 84. Some minor changes to the Linux configuration files to help with other
3747 OS variants using glibc.
3749 85. One more patch for Sieve to update vacation handling to latest spec.
3752 ----------------------------------------------------
3753 See the note above about the 4.44 and 4.50 releases.
3754 ----------------------------------------------------
3760 1. Change 4.43/35 introduced a bug that caused file counts to be
3761 incorrectly computed when quota_filecount was set in an appendfile
3764 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3765 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3768 3. Give more explanation in the error message when the command for a transport
3769 filter fails to execute.
3771 4. There are several places where Exim runs a non-Exim command in a
3772 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3773 was being done only for the command run by the queryprogram router. It is
3774 now done for all such subprocesses. The other cases are: ${run, transport
3775 filters, and the commands run by the lmtp and pipe transports.
3777 5. Some older OS have a limit of 256 on the maximum number of file
3778 descriptors. Exim was using setrlimit() to set 1000 as a large value
3779 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3780 systems. I've change it so that if it can't get 1000, it tries for 256.
3782 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3783 was an oversight, and furthermore, ever since the addition of extra
3784 controls (e.g. 4.43/32), the checks on when to allow different forms of
3785 "control" were broken. There should now be diagnostics for all cases when a
3786 control that does not make sense is encountered.
3788 7. $recipients is now available in the predata ACL (oversight).
3790 8. Tidy the search cache before the fork to do a delivery from a message
3791 received from the command line. Otherwise the child will trigger a lookup
3792 failure and thereby defer the delivery if it tries to use (for example) a
3793 cached ldap connection that the parent has called unbind on.
3795 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3796 of $address_data from the recipient verification was clobbered by the
3797 sender verification.
3799 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3800 was its contents. (It was OK if the option was not defined at all.)
3802 11. A "Completed" log line is now written for messages that are removed from
3803 the spool by the -Mrm option.
3805 12. $host_address is now set to the target address during the checking of
3806 ignore_target_hosts.
3808 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3809 being passed; this would have caused $sender_host_name to have been used if
3810 matching the list had actually called for a host name (not very likely,
3811 since this list is usually IP addresses). A host name is now passed as
3814 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3815 code that specifies a non-restarting handler (typically sigaction() in
3816 modern systems) in an attempt to fix a rare and obscure crash bug.
3818 15. Narrowed the window for a race in the daemon that could cause it to ignore
3819 SIGCHLD signals. This is not a major problem, because they are used only to
3820 wake it up if nothing else does.
3822 16. A malformed maildirsize file could cause Exim to calculate negative values
3823 for the mailbox size or file count. Odd effects could occur as a result.
3824 The maildirsize information is now recalculated if the size or filecount
3827 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3828 support for a long time. Removed HAVE_SYS_VFS_H.
3830 18. Updated exipick to current release from John Jetmore.
3832 19. Allow an empty sender to be matched against a lookup in an address list.
3833 Previously the only cases considered were a regular expression, or an
3836 20. Exim went into a mad DNS lookup loop when doing a callout where the
3837 host was specified on the transport, if the DNS lookup yielded more than
3840 21. The RFC2047 encoding function was originally intended for short strings
3841 such as real names; it was not keeping to the 75-character limit for
3842 encoded words that the RFC imposes. It now respects the limit, and
3843 generates multiple encoded words if necessary. To be on the safe side, I
3844 have increased the buffer size for the ${rfc2047: expansion operator from
3847 22. Failure to deliver a bounce message always caused it to be frozen, even if
3848 there was an errors_to setting on the router. The errors_to setting is now
3851 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3852 canonical form (fully expanded) before being placed in
3853 $sender_host_address.
3855 24. Updated eximstats to version 1.33
3857 25. Include certificate and key file names in error message when GnuTLS fails
3858 to set them up, because the GnuTLS error message doesn't include the name
3859 of the failing file when there is a problem reading it.
3861 26. Expand error message when OpenSSL has problems setting up cert/key files.
3864 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3865 (this can affect the format of dates).
3867 28. exim_tidydb, when checking for the continued existence of a message for
3868 which it has found a message-specific retry record, was not finding
3869 messages that were in split spool directories. Consequently, it was
3870 deleting retry records that should have stayed in existence.
3872 29. eximstats updated to version 1.35
3873 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3874 1.35 - bugfix such that pie charts by volume are generated correctly
3876 30. The SPA authentication driver was not abandoning authentication and moving
3877 on to the next authenticator when an expansion was forced to fail,
3878 contradicting the general specification for all authenticators. Instead it
3879 was generating a temporary error. It now behaves as specified.
3881 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3882 (the order specifies the preference for clients). The order is now AES256,
3883 AES128, 3DES, ARCFOUR128.
3885 31. Small patch to Sieve code - explicitly set From: when generating an
3888 32. Exim crashed if a remote delivery caused a very long error message to be
3889 recorded - for instance if somebody sent an entire SpamAssassin report back
3890 as a large number of 550 error lines. This bug was coincidentally fixed by
3891 increasing the size of one of Exim's internal buffers (big_buffer) that
3892 happened as part of the Exiscan merge. However, to be on the safe side, I
3893 have made the code more robust (and fixed the comments that describe what
3896 33. Some experimental protocols are using DNS PTR records for new purposes. The
3897 keys for these records are domain names, not reversed IP addresses. The
3898 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3899 leaves it alone. Component reversal etc. now happens only for IP addresses.
3902 34. The host_aton() function is supposed to be passed a string that is known
3903 to be a valid IP address. However, in the case of IPv6 addresses, it was
3904 not checking this. This is a hostage to fortune. Exim now panics and dies
3905 if the condition is not met. A case was found where this could be provoked
3906 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3907 components; fortuitously, this particular loophole had already been fixed
3908 by change 4.50/55 or 4.44/33 above.
3910 If there are any other similar loopholes, the new check in host_aton()
3911 itself should stop them being exploited. The report I received stated that
3912 data on the command line could provoke the exploit when Exim was running as
3913 exim, but did not say which command line option was involved. All I could
3914 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3915 running as the user.
3918 35. There was a buffer overflow vulnerability in the SPA authentication code
3919 (which came originally from the Samba project). I have added a test to the
3920 spa_base64_to_bits() function which I hope fixes it.
3923 36. The daemon start-up calls getloadavg() while still root for those OS that
3924 need the first call to be done as root, but it missed one case: when
3925 deliver_queue_load_max is set with deliver_drop_privilege. This is
3926 necessary for the benefit of the queue runner, because there is no re-exec
3927 when deliver_drop_privilege is set.
3929 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3930 was in use, was not putting the data itself into the right store pool;
3931 consequently, it could be overwritten for a subsequent message in the same
3932 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3935 38. Sometimes the final signoff response after QUIT could fail to get
3936 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3937 < 0 before doing a fflush(). This bug looks as though it goes back to the
3938 introduction of TLS in release 3.20, but "sometimes" must have been rare
3939 because the tests only now provoked it.
3945 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3946 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3947 Now it returns an integer. A number of calls were still expecting a T/F
3948 return. Fortuitously, in all cases, the tests worked in OK situations,
3949 which is the norm. However, things would have gone wrong on any write
3950 failures on the smtp file descriptor. This function is used when sending
3951 messages over SMTP and also when doing verify callouts.
3953 2. When Exim is called to do synchronous delivery of a locally submitted
3954 message (the -odf or -odi options), it no longer closes stderr before doing
3957 3. Implemented the mua_wrapper option.
3959 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3961 5. Implemented the functions header_remove(), header_testname(),
3962 header_add_at_position(), and receive_remove_recipient(), and exported them
3965 6. If an ACL "warn" statement specified the addition of headers, Exim already
3966 inserted X-ACL-Warn: at the start if there was no header name. However, it
3967 was not making this test for the second and subsequent header lines if
3968 there were newlines in the string. This meant that an invalid header could
3969 be inserted if Exim was badly configured.
3971 7. Allow an ACL "warn" statement to add header lines at the start or after all
3972 the Received: headers, as well as at the end.
3974 8. Added the rcpt_4xx retry error code.
3976 9. Added postmaster_mailfrom=xxx to callout verification option.
3978 10. Added mailfrom=xxxx to the callout verification option, for verify=
3981 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3982 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3984 12. Inserted some casts to stop certain compilers warning when using pointer
3985 differences as field lengths or precisions in printf-type calls (mostly
3986 affecting debugging statements).
3988 13. Added optional readline() support for -be (dynamically loaded).
3990 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3991 same clock tick as a message's arrival, so that its received time was the
3992 same as the "first fail" time on the retry record, and that message
3993 remained on the queue past the ultimate address timeout, every queue runner
3994 would try a delivery (because it was past the ultimate address timeout) but
3995 after another failure, the ultimate address timeout, which should have then
3996 bounced the address, did not kick in. This was a "< instead of <=" error;
3997 in most cases the first failure would have been in the next clock tick
3998 after the received time, and all would be well.
4000 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4001 being recognized when the domain list was tested by the match_domain
4002 condition in an expansion string.
4004 16. Added the ${str2b64: operator.
4006 17. Exim was always calling setrlimit() to set a large limit for the number of
4007 processes, without checking whether the existing limit was already
4008 adequate. (It did check for the limit on file descriptors.) Furthermore,
4009 errors from getrlimit() and setrlimit() were being ignored. Now they are
4010 logged to the main and panic logs, but Exim does carry on, to try to do its
4011 job under whatever limits there are.
4013 18. Imported PCRE 5.0.
4015 19. Trivial typo in log message " temporarily refused connection" (the leading
4018 20. If the log selector return_path_on_delivery was set and an address was
4019 redirected to /dev/null, the delivery process crashed because it assumed
4020 that a return path would always be set for a "successful" delivery. In this
4021 case, the whole delivery is bypassed as an optimization, and therefore no
4024 21. Internal re-arrangement: the function for sending a challenge and reading
4025 a response while authentication was assuming a zero-terminated challenge
4026 string. It's now changed to take a pointer and a length, to allow for
4027 binary data in such strings.
4029 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4031 23. Exim was not respecting finduser_retries when seeking the login of the
4032 uid under which it was called; it was always trying 10 times. (The default
4033 setting of finduser_retries is zero.) Also, it was sleeping after the final
4034 failure, which is pointless.
4036 24. Implemented tls_on_connect_ports.
4038 25. Implemented acl_smtp_predata.
4040 26. If the domain in control=submission is set empty, Exim assumes that the
4041 authenticated id is a complete email address when it generates From: or
4042 Sender: header lines.
4044 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4045 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4046 chown and chgrp in /bin and hostname in /usr/bin.
4048 28. Exim was keeping the "process log" file open after each use, just as it
4049 does for the main log. This opens the possibility of it remaining open for
4050 long periods when the USR1 signal hits a daemon. Occasional processlog
4051 errors were reported, that could have been caused by this. Anyway, it seems
4052 much more sensible not to leave this file open at all, so that is what now
4055 29. The long-running daemon process does not normally write to the log once it
4056 has entered its main loop, and it closes the log before doing so. This is
4057 so that log files can straightforwardly be renamed and moved. However,
4058 there are a couple of unusual error situations where the daemon does write
4059 log entries, and I had neglected to close the log afterwards.
4061 30. The text of an SMTP error response that was received during a remote
4062 delivery was being truncated at 512 bytes. This is too short for some of
4063 the long messages that one sometimes sees. I've increased the limit to
4066 31. It is now possible to make retry rules that apply only when a message has a
4067 specific sender, in particular, an empty sender.
4069 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4070 it possible to be selective about when SMTP synchronization is enforced.
4072 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4074 32. Implemented hosts_connection_nolog.
4076 33. Added an ACL for QUIT.
4078 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4081 35. Added mailbox_size and mailbox_filecount to appendfile.
4083 36. Added control = no_multiline_responses to ACLs.
4085 37. There was a bug in the logic of the code that waits for the clock to tick
4086 in the case where the clock went backwards by a substantial amount such
4087 that the microsecond fraction of "now" was more than the microsecond
4088 fraction of "then" (but the whole seconds number was less).
4090 38. Added support for the libradius Radius client library this is found on
4091 FreeBSD (previously only the radiusclient library was supported).
4097 1. When certain lookups returned multiple values in the form name=value, the
4098 quoting of the values was not always being done properly. Specifically:
4099 (a) If the value started with a double quote, but contained no whitespace,
4101 (b) If the value contained whitespace other than a space character (i.e.
4102 tabs or newlines or carriage returns) it was not quoted.
4103 This fix has been applied to the mysql and pgsql lookups by writing a
4104 separate quoting function and calling it from the lookup code. The fix
4105 should probably also be applied to nisplus, ibase and oracle lookups, but
4106 since I cannot test any of those, I have not disturbed their existing code.
4108 2. A hit in the callout cache for a specific address caused a log line with no
4109 reason for rejecting RCPT. Now it says "Previous (cached) callout
4110 verification failure".
4112 3. There was an off-by-one bug in the queryprogram router. An over-long
4113 return line was truncated at 256 instead of 255 characters, thereby
4114 overflowing its buffer with the terminating zero. As well as fixing this, I
4115 have increased the buffer size to 1024 (and made a note to document this).
4117 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4118 when Exim is waiting for an SMTP response from a remote server, Exim
4119 restarts its select() call on the socket, thereby resetting its timeout.
4120 This is not a problem when such interrupts are rare. Somebody set up a cron
4121 job to run exiwhat every 2 minutes, which is less than the normal select()
4122 timeout (5 or 10 minutes). This meant that the select() timeout never
4123 kicked in because it was always reset. I have fixed this by comparing the
4124 time when an interrupt arrives with the time at the start of the first call
4125 to select(). If more time than the timeout has elapsed, the interrupt is
4126 treated as a timeout.
4128 5. Some internal re-factoring in preparation for the addition of Sieve
4129 extensions (by MH). In particular, the "personal" test is moved to a
4130 separate function, and given an option for scanning Cc: and Bcc: (which is
4131 not set for Exim filters).
4133 6. When Exim created an email address using the login of the caller as the
4134 local part (e.g. when creating a From: or Sender: header line), it was not
4135 quoting the local part when it contained special characters such as @.
4137 7. Installed new OpenBSD configuration files.
4139 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4140 try to make them clearer.
4142 9. Callout options, other than the timeout value, were being ignored when
4143 verifying sender addresses in header lines. For example, when using
4145 verify = header_sender/callout=no_cache
4147 the cache was (incorrectly) being used.
4149 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4150 only the Cygwin environment.
4152 11. When return_path_on_delivery was set as a log selector, if different remote
4153 addresses in the same message used different return paths and parallel
4154 remote delivery occurred, the wrong values would sometimes be logged.
4155 (Whenever a remote delivery process finished, the return path value from
4156 the most recently started remote delivery process was logged.)
4158 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4159 lines when AUTH and/or TLS are in use. This is the "received protocol"
4160 field. Exim used to use "asmtp" for authenticated SMTP, without any
4161 indication (in the protocol name) for TLS use. Now it follows the RFC and
4162 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4163 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4164 names appear in log lines as well as in Received: header lines.
4166 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4167 extensions, and comparison tests, and to fix some bugs.
4169 14. Changes to the "personal" filter test:
4171 (1) The test was buggy in that it was just doing the equivalent of
4172 "contains" tests on header lines. For example, if a user's address was
4173 anne@some.where, the "personal" test would incorrectly be true for
4175 To: susanne@some.where
4177 This test is now done by extracting each address from the header in turn,
4178 and checking the entire address. Other tests that are part of "personal"
4179 are now done using regular expressions (for example, to check local parts
4180 of addresses in From: header lines).
4182 (2) The list of non-personal local parts in From: addresses has been
4183 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4184 taken from the Sieve specification recommendations.
4186 (3) If the message contains any header line starting with "List-" it is
4187 treated as non-personal.
4189 (4) The test for "circular" in the Subject: header line has been removed
4190 because it now seems ill-conceived.
4192 15. Minor typos in src/EDITME comments corrected.
4194 16. Installed latest exipick from John Jetmore.
4196 17. If headers_add on a router specified a text string that was too long for
4197 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4198 of string_sprintf() is now avoided.
4200 18. $message_body_size was not set (it was always zero) when running the DATA
4201 ACL and the local_scan() function.
4203 19. For the "mail" command in an Exim filter, no default was being set for
4204 the once_repeat time, causing a random time value to be used if "once" was
4205 specified. (If the value happened to be <= 0, no repeat happened.) The
4206 default is now 0s, meaning "never repeat". The "vacation" command was OK
4207 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4208 (I found it when inspecting the code).
4210 20. There is now an overall timeout for performing a callout verification. It
4211 defaults to 4 times the callout timeout, which applies to individual SMTP
4212 commands during the callout. The overall timeout applies when there is more
4213 than one host that can be tried. The timeout is checked before trying the
4214 next host. This prevents very long delays if there are a large number of
4215 hosts and all are timing out (e.g. when the network connections are timing
4216 out). The value of the overall timeout can be changed by specifying an
4217 additional sub-option for "callout", called "maxwait". For example:
4219 verify = sender/callout=5s,maxwait=20s
4221 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4222 to the end before writing, but this should make it even safer).
4224 22. Exim was forgetting that it had advertised PIPELINING for the second and
4225 subsequent messages on an SMTP connection. It was also not resetting its
4226 memory on STARTTLS and an internal HELO.
4228 23. When Exim logs an SMTP synchronization error within a session, it now
4229 records whether PIPELINING has been advertised or not.
4231 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4232 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4233 rather than long int.
4235 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4237 26. Added the never_mail option to autoreply.
4243 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4244 crash if the getsockname() call failed; this can happen if a connection is
4245 closed very soon after it is established. The problem was simply in the
4246 order in which certain operations were done, causing Exim to try to write
4247 to the SMTP stream before it had set up the file descriptor. The bug has
4248 been fixed by making things happen in the correct order.
4254 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4255 before the connection was closed, thus losing the rejection response.
4257 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4258 some early Solaris releases, but causes trouble in current releases where
4259 socklen_t is defined.
4261 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4264 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4267 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4268 that releasing the top of it at the end releases what was used for sub-
4269 expansions (unless the block got too big). However, discard this block if
4270 the first thing is a variable or header, so that we can use its block when
4271 it is dynamic (useful for very large $message_headers, for example).
4273 6. Lookups now cache *every* query, not just the most recent. A new, separate
4274 store pool is used for this. It can be recovered when all lookup caches are
4275 flushed. Lookups now release memory at the end of their result strings.
4276 This has involved some general refactoring of the lookup sources.
4278 7. Some code has been added to the store_xxx() functions to reduce the amount
4279 of flapping under certain conditions.
4281 8. log_incoming_interface used to affect only the <= reception log lines. Now
4282 it causes the local interface and port to be added to several more SMTP log
4283 lines, for example "SMTP connection from", and rejection lines.
4285 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4287 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4289 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4290 could be overwritten at the end of the current message (or the start of a
4291 new message if it was set in a HELO ACL). The value is now preserved for
4292 the duration of the SMTP connection.
4294 12. If a transport had a headers_rewrite setting, and a matching header line
4295 contained an unqualified address, that address was qualified, even if it
4296 did not match any rewriting rules. The underlying bug was that the values
4297 of the flags that permit the existence of unqualified sender and recipient
4298 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4299 non-local messages, and by -bnq for local messages) were not being
4300 preserved with the message after it was received.
4302 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4303 "next input=" as part of the text comprising the host identity instead of
4304 the correct text. The code was using the same buffer for two different
4305 strings. However, depending on which order the printing function evaluated
4306 its arguments, the bug did not always show up. Under Linux, for example, my
4307 test suite worked just fine.
4309 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4310 doesn't work with some older versions of Perl. It has been changed to "my",
4311 which in any case is probably the better facility to use.
4313 15. A really picky compiler found some instances of statements for creating
4314 error messages that either had too many or two few arguments for the format
4317 16. The size of the buffer for calls to the DNS resolver has been increased
4318 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4319 for addresses that have a lot of PTR records. This alleviates a problem; it
4320 does not fully solve it.
4322 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4323 buffer now truncates the list and logs the incident, which is the same
4324 action as happens when Exim is looking up a host name and its aliases.
4325 Previously in this situation something unpredictable would happen;
4326 sometimes it was "internal error: store_reset failed".
4328 18. If a server dropped the connection unexpectedly when an Exim client was
4329 using GnuTLS and trying to read a response, the client delivery process
4330 crashed while trying to generate an error log message.
4332 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4333 string, for example:
4335 warn message = H1: something\nH2: something
4337 the text was added as a single header line from Exim's point of view
4338 though it ended up OK in the delivered message. However, searching for the
4339 second and subsequent header lines using $h_h2: did not work. This has been
4340 fixed. Similarly, if a system filter added multiple headers in this way,
4341 the routers could not see them.
4343 20. Expanded the error message when iplsearch is called with an invalid key to
4344 suggest using net-iplsearch in a host list.
4346 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4347 ACLs are no longer actually imposed (and a message to that effect is
4350 22. If a "gecos" field in a passwd entry contained escaped characters, in
4351 particular, if it contained a \" sequence, Exim got it wrong when building
4352 a From: or a Sender: header from that name. A second bug also caused
4353 incorrect handling when an unquoted " was present following a character
4354 that needed quoting.
4356 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4357 was not being matched caselessly.
4359 24. Arranged for all hyphens in the exim.8 source to be escaped with
4362 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4363 properly. Recipient callout cache records were still being keyed to include
4364 the sender, even when use_sender was set false. This led to far more
4365 callouts that were necessary. The sender is no longer included in the key
4366 when use_sender is false.
4368 26. Added "control = submission" modifier to ACLs.
4370 27. Added the ${base62d: operator to decode base 62 numbers.
4372 28. dnsdb lookups can now access SRV records.
4374 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4375 the configuration file.
4377 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4378 (-v) mode. This makes the output for a verbose queue run more intelligible.
4380 31. Added a use_postmaster feature to recipient callouts.
4382 32. Added the $body_zerocount variable, containing the number of binary zero
4383 bytes in the message body.
4385 33. The time of last modification of the "new" subdirectory is now used as the
4386 "mailbox time last read" when there is a quota error for a maildir
4389 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4391 35. Added +ignore_unknown as a special item in host lists.
4393 36. Code for decoding IPv6 addresses in host lists is now included, even if
4394 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4395 address was recognized as an IP address, but was then not correctly decoded
4396 into binary, causing unexpected and incorrect effects when compared with
4403 1. Very minor rewording of debugging text in manualroute to say "list of
4404 hosts" instead of "hostlist".
4406 2. If verify=header_syntax was set, and a header line with an unqualified
4407 address (no domain) and a large number of spaces between the end of the
4408 name and the colon was received, the reception process suffered a buffer
4409 overflow, and (when I tested it) crashed. This was caused by some obsolete
4410 code that should have been removed. The fix is to remove it!
4412 3. When running in the test harness, delay a bit after writing a bounce
4413 message to get a bit more predictability in the log output.
4415 4. Added a call to search_tidyup() just before forking a reception process. In
4416 theory, someone could use a lookup in the expansion of smtp_accept_max_
4417 per_host which, without the tidyup, could leave open a database connection.
4419 5. Added the variables $recipient_data and $sender_data which get set from a
4420 lookup success in an ACL "recipients" or "senders" condition, or a router
4421 "senders" option, similar to $domain_data and $local_part_data.
4423 6. Moved the writing of debug_print from before to after the "senders" test
4426 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4427 problems for message scanning, either using a data ACL, or using
4428 local_scan() because the Received: header was not generated till after they
4429 were called (in order to set the time as the time of reception completion).
4430 I have revised the way this works. The header is now generated after the
4431 body is received, but before the ACL or local_scan() are called. After they
4432 are run, the timestamp in the header is updated.
4438 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4439 before starting a queue runner without re-exec. This happened only when
4440 deliver_drop_privilege was set or when the Exim user was set to root. The
4441 effect of the bug was that timeouts during subsequent deliveries caused
4442 crashes instead of being properly handled. The handler is now left at its
4443 default (and expected) setting.
4445 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4446 message, again when deliver_drop_privilege is set or Exim is run as root.
4447 The bug described in (1) was not present in this case, but the tidying up
4448 of the other signals was missing. I have made the two cases consistent.
4450 3. The ignore_target_hosts setting on a manualroute router was being ignored
4451 for hosts that were looked up using the /MX notation.
4453 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4456 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4457 operated on the sender address. After changing the $sender_address to <>
4458 for the sender address verify, Exim was re-instated it as the original
4459 (before rewriting) address, but remembering that it had rewritten it, so it
4460 wasn't rewriting it again. This bug also had the effect of breaking the
4461 sender address verification caching when the sender address was rewritten.
4463 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4464 This has been changed so that if the ip literal address matches
4465 ignore_target_hosts, the router declines.
4467 7. Added expansion conditions match_domain, match_address, and match_local_
4468 part (NOT match_host).
4470 8. The placeholder for the Received: header didn't have a length field set.
4472 9. Added code to Exim itself and to exim_lock to test for a specific race
4473 condition that could lead to file corruption when using MBX delivery. The
4474 issue is with the lockfile that is created in /tmp. If this file is removed
4475 after a process has opened it but before that process has acquired a lock,
4476 there is the potential for a second process to recreate the file and also
4477 acquire a lock. This could lead to two Exim processes writing to the file
4478 at the same time. The added code performs the same test as UW imapd; it
4479 checks after acquiring the lock that its file descriptor still refers to
4480 the same named file.
4482 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4483 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4486 11. Added the smtp_active_hostname option. If used, this will typically be made
4487 to depend on the incoming interface address. Because $interface_address is
4488 not set up until the daemon has forked a reception process, error responses
4489 that can happen earlier (such as "too many connections") no longer contain
4492 12. If an expansion in a condition on a "warn" statement fails because a lookup
4493 defers, the "warn" statement is abandoned, and the next ACL statement is
4494 processed. Previously this caused the whole ACL to be aborted.
4496 13. Added the iplsearch lookup type.
4498 14. Added ident_timeout as a log selector.
4500 15. Added tls_certificate_verified as a log selector.
4502 16. Added a global option tls_require_ciphers (compare the smtp transport
4503 option of the same name). This controls incoming TLS connections.
4505 17. I finally figured out how to make tls_require_ciphers do a similar thing
4506 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4507 before starting the TLS session.
4509 18. Tabs are now shown as \t in -bP output.
4511 19. If the log selector return_path_on_delivery was set, Exim crashed when
4512 bouncing a message because it had too many Received: header lines.
4514 20. If two routers both had headers_remove settings, and the first one included
4515 a superfluous trailing colon, the final name in the first list and the
4516 first name in the second list were incorrectly joined into one item (with a
4517 colon in the middle).
4523 1. Added -C and -D options to the exinext utility, mainly to make it easier
4524 to include in the automated testing, but these could be helpful when
4525 multiple configurations are in use.
4527 2. The exinext utility was not formatting the output nicely when there was
4528 an alternate port involved in the retry record key, nor when there was a
4529 message id as well (for retries that were specific to a specific message
4530 and a specific host). It was also confused by IPv6 addresses, because of
4531 the additional colons they contain. I have fixed the IPv4 problem, and
4532 patched it up to do a reasonable job for IPv6.
4534 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4535 delivery, the log line now contains "pipelined" if PIPELINING was used.
4537 4. An SMTP transport process used to panic and die if the bind() call to set
4538 an explicit outgoing interface failed. This has been changed; it is now
4539 treated in the same way as a connect() failure.
4541 5. A reference to $sender_host_name in the part of a conditional expansion
4542 that was being skipped was still causing a DNS lookup. This no longer
4545 6. The def: expansion condition was not recognizing references to header lines
4546 that used bh_ and bheader_.
4548 7. Added the _cache feature to named lists.
4550 8. The code for checking quota_filecount in the appendfile transport was
4551 allowing one more file than it should have been.
4553 9. For compatibility with Sendmail, the command line option
4561 and sets the incoming protocol and host name (for trusted callers). The
4562 host name and its colon can be omitted when only the protocol is to be set.
4563 Note the Exim already has two private options, -pd and -ps, that refer to
4564 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4565 "s", but I don't think that's a major issue.
4567 10. A number of refactoring changes to the code, none of which should affect
4570 (a) The number of logging options was getting close to filling up the
4571 32-bit word that was used as a bit map. I have split them into two classes:
4572 those that are passed in the argument to log_write(), and those that are
4573 only ever tested independently outside of that function. These are now in
4574 separate 32-bit words, so there is plenty of room for expansion again.
4575 There is no change in the user interface or the logging behaviour.
4577 (b) When building, for example, log lines, the code previously used a
4578 macro that called string_cat() twice, in order to add two strings. This is
4579 not really sufficiently general. Furthermore, there was one instance where
4580 it was actually wrong because one of the argument was used twice, and in
4581 one call a function was used. (As it happened, calling the function twice
4582 did not affect the overall behaviour.) The macro has been replaced by a
4583 function that can join an arbitrary number of extra strings onto a growing
4586 (c) The code for expansion conditions now uses a table and a binary chop
4587 instead of a serial search (which was left over from when there were very
4588 few conditions). Also, it now recognizes conditions like "pam" even when
4589 the relevant support is not compiled in: a suitably worded error message is
4590 given if an attempt is made to use such a condition.
4592 11. Added ${time_interval:xxxxx}.
4594 12. A bug was causing one of the ddress fields not to be passed back correctly
4595 from remote delivery subprocesses. The field in question was not being
4596 subsequently used, so this caused to problems in practice.
4598 13. Added new log selectors queue_time and deliver_time.
4600 14. Might have fixed a bug in maildirsizefile handling that threw up
4601 "unexpected character" debug warnings, and recalculated the data
4602 unnecessarily. In any case, I expanded the warning message to give more
4605 15. Added the message "Restricted characters in address" to the statements in
4606 the default ACL that block characters like @ and % in local parts.
4608 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4609 Three changes have been made:
4611 (a) There was a serious bug; a negative response to MAIL caused the whole
4612 recipient domain to be cached as invalid, thereby blocking all messages
4613 to all local parts at the same domain, from all senders. This bug has
4614 been fixed. The domain is no longer cached after a negative response to
4615 MAIL if the sender used is not empty.
4617 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4620 (c) A new callout option, "use_sender" has been added for people who want
4621 the modified behaviour.
4627 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4630 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4631 indeed breaks things for older releases.
4633 3. Added additional logging to the case where there is a problem reading data
4634 from a filter that is running in a subprocess using a pipe, in order to
4635 try to track down a specific problem.
4637 4. Testing facility fudge: when running in the test harness and attempting
4638 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4639 getting "No route to host". Convert this to a timeout.
4641 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4644 6. Some OS don't have socklen_t but use size_t instead. This affects the
4645 fifth argument of getsockopt() amongst other things. This is now
4646 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4647 can be set for individual OS. I have set it for SunOS5, OSF1, and
4648 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4649 some earlier ones do not.
4651 7. Change 4.30/15 was not doing the test caselessly.
4653 8. The standard form for an IPv6 address literal was being rejected by address
4654 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4655 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4656 this, as well as the form without the "IPv6" on the front (but only when
4657 address literals are enabled, of course).
4659 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4661 10. Exim crashed if a message with an empty sender address specified by -f
4662 encountered a router with an errors_to setting. This could be provoked only
4663 by a command such as
4667 where an empty string was supplied; "<>" did not hit this bug.
4669 11. Installed PCRE release 4.5.
4671 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4672 remained set. It is now erased.
4674 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4675 times from message ids (which are base 36 rather than the normal 62).
4677 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4678 were being counted as actual protocol errors, and logged if the log
4679 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4680 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4681 and DATA following a set of rejected RCPTs do not count as protocol errors.
4682 In other words, Exim assumes they were pipelined, though this may not
4683 actually be the case. Of course, in all cases the client gets an
4684 appropriate error code.
4686 15. If a lookup fails in an ACL condition, a message about the failure may
4687 be available; it is used if testing the ACL cannot continue, because most
4688 such messages specify what the cause of the deferral is. However, some
4689 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4690 that caused an old message to be retained and used if a later statement
4691 caused a defer, replacing the real cause of the deferral.
4693 16. If an IP address had so many PTR records that the DNS lookup buffer
4694 was not large enough to hold them, Exim could crash while trying to process
4695 the truncated data. It now detects and logs this case.
4697 17. Further to 4.21/58, another change has been made: if (and only if) the
4698 first line of a message (the first header line) ends with CRLF, a bare LF
4699 in a subsequent header line has a space inserted after it, so as not to
4700 terminate the header.
4702 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4703 unnecessarily, used atoi() instead of strtol(), and didn't check the
4704 termination when getting file sizes from file names by regex.
4706 19. Completely re-implemented the support for maildirsize files, in the light
4707 of a number of problems with the previous contributed implementation
4708 (4.30/29). In particular:
4710 . If the quota is zero, the maildirsize file is maintained, but no quota is
4713 . If the maildir directory does not exist, it is created before any attempt
4714 to write a maildirsize file.
4716 . The quota value in the file is just a cache; if the quota is changed in
4717 the transport, the new value overrides.
4719 . A regular expression is available for excluding directories from the
4722 20. The autoreply transport checks the characters in options that define the
4723 message's headers; it allows continued headers, but it was checking with
4724 isspace() after an embedded newline instead of explicitly looking for a
4727 21. If all the "regular" hosts to which an address was routed had passed their
4728 expiry times, and had not reached their retry times, the address was
4729 bounced, even if fallback hosts were defined. Now Exim should go on to try
4732 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4733 equivalent code in the SMTP transport. Some hosts send humungous responses
4734 to HELO/EHLO, more than 1024 it seems.
4736 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4737 gives compiler warnings in some environments. I've now done it "properly",
4740 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4741 (because of problems with the built-in one) was declared to return uschar *
4742 instead of char *, causing compiler failure.
4744 25. Fixed a file descriptor leak when processing alias/forward files.
4746 26. Fixed a minor format string issue in dbfn.c.
4748 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4750 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4751 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4754 29. When a sender address is verified, it is cached, to save repeating the test
4755 when there is more than one recipient in a message. However, when the
4756 verification involves a callout, it is possible for different callout
4757 options to be set for different recipients. It is too complicated to keep
4758 track of this in the cache, so now Exim always runs a verification when a
4759 callout is required, relying on the callout cache for the optimization.
4760 The overhead is duplication of the address routing, but this should not be
4763 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4764 to be verified with callout=postmaster, and the main callout worked but the
4765 postmaster check failed, the verification correctly failed. However, if a
4766 subsequent RCPT command asked for sender verification *without* the
4767 postmaster check, incorrect caching caused this verification also to fail,
4770 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4771 it was not caching the DNS options (qualify_single, search_parents) that
4772 were used when the lookup failed. A subsequent lookup with different
4773 options therefore always gave the same answer, though there were cases
4774 where it should not have. (Example: a "domains = !$mx_any" option on a
4775 dnslookup router: the "domains" option is always processed without any
4776 widening, but the router might have qualify_single set.) Now Exim uses the
4777 cached value only when the same options are set.
4779 32. Added John Jetmore's "exipick" utility to the distribution.
4781 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4782 than a timeout (e.g. a certificate is required, and is not provided), an
4783 Exim server now closes the connection immediately. Previously it waited for
4784 the client to close - but if the client is SSL, it seems that they each
4785 wait for each other, leading to a delay before one of them times out.
4787 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4788 maintained 0.8.x compatibility because I don't think many are using it, and
4789 it is clearly obsolete.
4791 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4794 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4795 client certificate was expired. A simple patch fixes this, though I don't
4796 understand the full logic of why the verify callback is called multiple
4799 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4800 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4801 which causes problems with some clients (such as the Certicom SSL Plus
4802 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4803 disables the coutermeasure allowing Eudora to connect."
4805 38. Exim was not checking that a write() to a log file succeeded. This could
4806 lead to Bad Things if a log got too big, in particular if it hit a file
4807 size limit. Exim now panics and dies if it cannot write to a log file, just
4808 as it does if it cannot open a log file.
4810 39. Modified OS/Makefile-Linux so that it now contains
4812 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4814 The two -D definitions ensure that Exim is compiled with large file
4815 support, which makes it possible to handle log files that are bigger than
4818 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4819 instance) a domain was checked against a named list that involved a lookup,
4820 causing $domain_data to be set, then another domain was checked against the
4821 same list, then the first domain was re-checked, the value of $domain_data
4822 after the final check could be wrong. In particular, if the second check
4823 failed, it could be set empty. This bug probably also applied to
4826 41. The strip_trailing_dot option was not being applied to the address given
4827 with the -f command-line option.
4829 42. The code for reading a message's header from the spool was incrementing
4830 $received_count, but never initializing it. This meant that the value was
4831 incorrect (doubled) while delivering a message in the same process in which
4832 it was received. In the most common configuration of Exim, this never
4833 happens - a fresh exec is done - but it can happen when
4834 deliver_drop_privilege is set.
4836 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4837 it now includes up to 150 characters of the unexpected data in the log
4840 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4841 and building data strings. The size of both of these buffers was 10 000
4842 bytes - far larger than anybody would *ever* want, thought I. Needless to
4843 say, somebody hit the limit. I have increased the maximum line length to
4844 20 000 and the maximum data length of concatenated lines to 100 000. I have
4845 also fixed two bugs, because there was no checking on these buffers. Tsk,
4846 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4847 buffer is too small.
4849 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4850 lsearch lookups. Now it does.
4852 46. When parsing a route_list item in a manualroute router, a fixed-length
4853 buffer was used for the list of hosts. I made this 1024 bytes long,
4854 thinking that nobody would ever have a list of hosts that long. Wrong.
4855 Somebody had a whole pile of complicated expansion conditions, and the
4856 string was silently truncated, leading to an expansion error. It turns out
4857 that it is easier to change to an unlimited length (owing to other changes
4858 that have happened since this code was originally written) than to build
4859 structure for giving a limitation error. The length of the item that
4860 expands into the list of hosts is now unlimited.
4862 47. The lsearch lookup could not handle data where the length of text line was
4863 more than 4095 characters. Such lines were truncated, leading to shortened
4864 data being returned. It should now handle lines of any length.
4866 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4867 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4870 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4871 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4872 no longer happens in comment lines. A list of replacements is now placed
4873 at the head of all of the source files, except those whose only change is
4874 to replace PERL_COMMAND in the very first #! line.
4876 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4877 messages on the queue, with a bottom-up merge sort, using code contributed
4878 by Michael Haardt. This should make operations like -bp somewhat faster on
4879 large queues. It won't affect queue runners, except when queue_run_in_order
4882 51. Installed eximstats 1.31 in the distribution.
4884 52. Added support for SRV lookups to the dnslookup router.
4886 53. If an ACL referred to $message_body or $message_body_end, the value was not
4887 reset for any messages that followed in the same SMTP session.
4889 54. The store-handling optimization for building very long strings was not
4890 differentiating between the different store pools. I don't think this
4891 actually made any difference in practice, but I've tidied it.
4893 55. While running the routers to verify a sender address, $sender_address
4894 was still set to the sender address. This is wrong, because when routing to
4895 send a bounce to the sender, it would be empty. Therefore, I have changed
4896 it so that, while verifying a sender address, $sender_address is set to <>.
4897 (There is no change to what happens when verifying a recipient address.)
4899 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4900 target A or AAAA records (if not already returned) without resetting the
4901 qualify_single or search_parents options of the DNS resolver. These are
4902 inappropriate in this case because the targets of MX and SRV records must
4903 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4904 target that, when qualified, matched something in the local domain. These
4905 two options are now turned off when doing these lookups.
4907 57. It seems that at least some releases of Reiserfs (which does not have the
4908 concept of a fixed number of inodes) returns zero and not -1 for the
4909 number of available inodes. This interacted badly with check_spool_inodes,
4910 which assumed that -1 was the "no such thing" setting. What I have done is
4911 to check that the total number of inodes is greater than zero before doing
4912 the test of how many are available.
4914 58. When a "warn" ACL statement has a log_message modifier, the message is
4915 remembered, and not repeated. This is to avoid a lot of repetition when a
4916 message has many recipients that cause the same warning to be written.
4917 Howewer, Exim was preserving the list of already written lines for an
4918 entire SMTP session, which doesn't seem right. The memory is now reset if a
4919 new message is started.
4921 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4922 debugging output unless log_rewrite was also set.
4924 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4925 of (int)(handle) when we know that handle contains (void *)(-1).
4927 61. The Exim daemon panic-logs an error return when it closes the incoming
4928 connection. However "connection reset by peer" seems to be common, and
4929 isn't really an error worthy of noting specially, so that particular error
4932 62. When Exim is trying to find all the local interfaces, it used to panic and
4933 die if the ioctl to get the interface flags failed. However, it seems that
4934 on at least one OS (Solaris 9) it is possible to have an interface that is
4935 included in the list of interfaces, but for which you get a failure error
4936 for this call. This happens when the interface is not "plumbed" into a
4937 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4938 failure of the "get flags" call assumes that the interface is down.
4940 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4941 makes life easier for people who are doing arithmetic on fields extracted
4942 from dates, where you often get leading zeros that should not be
4943 interpreted as octal.
4945 64. Added qualify_domain to the redirect router, to override the global
4948 65. If a pathologically long header line contained very many addresses (the
4949 report of this problem mentioned 10 000) and each of them was rewritten,
4950 Exim could use up a very large amount of memory. (It kept on making new
4951 copies of the header line as it rewrote, and never released the old ones.)
4952 At the expense of a bit more processing, the header rewriting function has
4953 been changed so that it no longer eats memory in this way.
4955 66. The generation of the Received: header has been moved from the time that a
4956 message starts to be received, to the time that it finishes. The timestamp
4957 in the Received: header should now be very close to that of the <= log
4958 line. There are two side-effects of this change:
4960 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4961 logged header lines no longer include the local Received: line, because
4962 it has not yet been created. The same applies to a copy of the message
4963 that is returned to a non-SMTP sender when a message is rejected.
4965 (b) When a filter file is tested using -bf, no additional Received: header
4966 is added to the test message. After some thought, I decided that this
4969 This change does not affect the value of $received_for. It is still set
4970 after address rewriting, but before local_scan() is called.
4972 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4974 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4975 gave an unhelpful panic error message, and a defer error. I have managed to
4976 change this behaviour so that it now rejects any supplied certificate,
4977 which seems right, as the list of acceptable certificates is empty.
4979 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4980 gave an unhelpful defer error. I have not managed to make this reject any
4981 supplied certificates, but the error message it gives is "no certificate
4982 supplied", which is not helpful.
4984 70. exigrep's output now also includes lines that are not associated with any
4985 message, but which match the given pattern. Implemented by a patch from
4986 Martin Sluka, which also tidied up the Perl a bit.
4988 71. Recipient callout verification, like sender verification, was using <> in
4989 the MAIL FROM command. This isn't really the right thing, since the actual
4990 sender may affect whether the remote host accepts the recipient or not. I
4991 have changed it to use the actual sender in the callout; this means that
4992 the cache record is now keyed on a recipient/sender pair, not just the
4993 recipient address. There doesn't seem to be a real danger of callout loops,
4994 since a callout by the remote host to check the sender would use <>.
4995 [SEE ABOVE: changed after hitting problems.]
4997 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4998 temporary errors. However, in the case of such a code being given after
4999 the end of a data transmission (i.e. after ".") Exim was failing to write
5000 a retry record for the message. (Yes, there was some broken host that was
5001 actually sending 8xx at this point.)
5003 73. An unknown lookup type in a host list could cause Exim to panic-die when
5004 the list was checked. (An example that provoked this was putting <; in the
5005 middle of a list instead of at the start.) If this happened during a DATA
5006 ACL check, a -D file could be left lying around. This kind of configuration
5007 error no longer causes Exim to die; instead it causes a defer errror. The
5008 incident is still logged to the main and panic logs.
5010 74. Buglet left over from Exim 3 conversion. The message "too many messages
5011 in one connection" was written to the rejectlog but not the mainlog, except
5012 when address rewriting (yes!) was being logged.
5014 75. Added write_rejectlog option.
5016 76. When a system filter was run not as root (that is, when system_filter_user
5017 was set), the values of the $n variables were not being returned to the
5018 main process; thus, they were not subsequently available in the $sn
5021 77. Added +return_path_on_delivery log selector.
5023 78. A connection timeout was being treated differently from recipients deferred
5024 when testing hosts_max_try with a message that was older than the host's
5025 retry timeout. (The host should not be counted, thus allowing all hosts to
5026 be tried at least once before bouncing.) This may have been the cause of an
5027 occasionally reported bug whereby a message would remain on the queue
5028 longer than the retry timeout, but would be bounced if a delivery was
5029 forced. I say "may" because I never totally pinned down the problem;
5030 setting up timeout/retry tests is difficult. See also the next item.
5032 79. The ultimate address timeout was not being applied to errors that involved
5033 a combination of host plus message (for example, a timeout on a MAIL
5034 command). When an address resolved to a number of possible hosts, and they
5035 were not all tried for each delivery (e.g. because of hosts_max_try), a
5036 message could remain on the queue longer than the retry timeout.
5038 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5041 81. Fixed an obscure SMTP outgoing bug which required at least the following
5042 conditions: (a) there was another message waiting for the same server;
5043 (b) the server returned 5xx to all RCPT commands in the first message so
5044 that the message was not completed; (c) the server dropped the connection
5045 or gave a negative response to the RSET that Exim sends to abort the
5046 transaction. The observed case was a dropped connection after DATA that had
5047 been sent in pipelining mode. That is, the server had advertised PIPELINING
5048 but was not implementing it correctly. The effect of the bug was incorrect
5049 behaviour, such as trying another host, and this could lead to a crash.
5055 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5056 and daemon.c were passed as pointers to ints; they should have been
5057 pointers to socklen_t variables (which are typically unsigned ints).
5059 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5062 3. Fixed a really odd bug that affected only the testing scheme; patching a
5063 certain fixed string in the binary changed the value of another string that
5064 happened to be identical to the end of the original first string.
5066 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5067 name", it returns that address as the IP address. On some operating
5068 systems (e.g. Solaris), it also passes back the IP address string as the
5069 "host name". However, on others (e.g. Linux), it passes back an empty
5070 string. Exim wasn't checking for this, and was changing the host name to an
5071 empty string, assuming it had been canonicized.
5073 5. Although rare, it is permitted to have more than one PTR record for a given
5074 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5075 all the names associated with an address, because they do in Solaris.
5076 However, it seems that they do not in Linux for data that comes from the
5077 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5078 I found this out when I moved to a new Linux workstation and tried to run
5079 the Exim test suite.
5081 To get round this problem I have changed the code so that it now does its
5082 own call to the DNS to look up PTR records when searching for a host name.
5083 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5084 addresses that are only in /etc/hosts are still found.
5086 This behaviour is, however, controlled by an option called host_lookup_
5087 order, which defaults to "bydns:byaddr". If people want to use the other
5088 order, or indeed, just use one or the other means of lookup, they can
5089 specify it in this variable.
5091 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5092 some operating systems, this comes back from gethostbyaddr() as an empty
5093 string, and this is what Exim used to test for. However, it seems that in
5094 other systems, "." is yielded. Exim now tests for this case too.
5096 7. The values of check_spool_space and check_log_space are now held internally
5097 as a number of kilobytes instead of an absolute number of bytes. If a
5098 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5099 kilobyte. This means that much larger values can be stored.
5101 8. Exim monitor: an attempt to get the action menu when not actually pointing
5102 at a message produces an empty menu entitled "No message selected". This
5103 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5104 no entries in it ("Shell widget menu has zero width and/or height"). So I
5105 have added a single, blank menu entry in this case.
5107 9. Added ${quote_local_part.
5109 10. MIME decoding is now applied to the contents of Subject: header lines when
5112 11. Now that a reference to $sender_host_address automatically causes a reverse
5113 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5114 host lookup before query-style lookups in lists that might use this
5115 variable. This has therefore been abolished, and the "net-" prefix is no
5116 longer necessary for query-style lookups.
5118 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5119 have been a typo for LFLAGS, so it has been changed.
5121 13. The install script calls Exim with "-C /dev/null" in order to find the
5122 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5123 to be output. Howeve, since Exim outputs its version number before the
5124 error, it didn't break the script. It just looked ugly. I fixed this by
5125 always allowing "-C /dev/null" if the caller is root.
5127 14. Ignore overlarge ACL variable number when reading spool file - insurance
5128 against a later release with more variables having written the file.
5130 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5131 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5132 this, as well as the form without the "IPv6" on the front.
5134 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5135 OS/Makefile-Darwin file.
5137 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5138 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5140 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5141 variable. However, some OS use ints for inodes. I've added cast to long int
5142 to get rid of the compiler warning.
5144 19. I had forgotten to lock out "/../" in configuration file names when
5145 ALT_CONFIG_PREFIX was set.
5147 20. Routers used for verification do not need to specify transports. However,
5148 if such a router generated a host list, and callout was configured, Exim
5149 crashed, because it could not find a port number from the (non-existent)
5150 transport. It now assumes port 25 in this circumstance.
5152 21. Added the -t option to exigrep.
5154 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5155 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5156 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5157 them is set, LOOKUP_LSEARCH is forced.
5159 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5161 24. Added sender and host information to the "rejected by local_scan()" log
5162 line; previously there was no indication of these.
5164 25. Added .include_if_exists.
5166 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5167 Linux. It turns out that not all file systems support this. Apparently some
5168 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5169 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5170 not supported on the file descriptor, is now ignored when Exim is trying to
5171 sync a directory. This applies only to Linux.
5173 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5175 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5176 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5179 29. Added support for maildirsize files from supplied patch (modified a bit).
5181 30. The use of :fail: followed by an empty string could lead Exim to respond to
5182 sender verification failures with (e.g.):
5184 550 Verification failed for <xxx>
5185 550 Sender verify failed
5187 where the first response line was missing the '-' that indicates it is not
5188 the final line of the response.
5190 31. The loop for finding the name of the user that called Exim had a hardwired
5191 limit of 10; it now uses the value of finduser_retries, which is used for
5192 all other user lookups.
5194 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5197 33. Exim was neglecting to zero errno before one call of strtol() when
5198 expanding a string and expecting an integer value. On some systems this
5199 resulted in spurious "integer overflow" errors. Also, it was casting the
5200 result into an int without checking.
5202 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5203 did not work. The code looks as if it has *never* worked, though it appears
5204 to have been documented since at least releast 1.62. I have made it work.
5206 35. The "timeout_DNS" error in retry rules, also documented since at least
5207 1.62, also never worked. As it isn't clear exactly what this means, and
5208 clearly it isn't a major issue, I have abolished the feature by treating it
5209 as "timeout", and writing a warning to the main and panic logs.
5211 36. The display of retry rules for -brt wasn't always showing the error code
5214 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5215 timeout_connect_A, timeout_connect_MX.
5217 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5218 to the empty sender.
5220 39. The daemon was not analysing the content of -oX till after it had closed
5221 stderr and disconnected from the controlling terminal. This meant that any
5222 syntax errors were only noted on the panic log, and the return code from
5223 the command was 0. By re-arranging the code a little, I've made the
5224 decoding happen first, so such errors now appear on stderr, and the return
5225 code is 1. However, the actual setting up of the sockets still happens in
5226 the disconnected process, so errors there are still only recorded on the
5229 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5230 connections (as happens on some IP stacks) was logged at start up time as
5231 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5232 it from "IPv6 and IPv4", which means that two separate sockets are being
5235 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5236 says whether AF_INET or AF_INET6 was passed as an argument.
5238 42. Exiwhat output was messed up when time zones were included in log
5241 43. Exiwhat now gives more information about the daemon's listening ports,
5242 and whether -tls-on-connect was used.
5244 44. The "port" option of the smtp transport is now expanded.
5246 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5247 silently ignored. Now an error message is written to the main and panic
5250 46. There's a new ACL modifier called "logwrite" which writes to a log file
5251 as soon as it is encountered.
5253 47. Added $local_user_uid and $local_user_gid at routing time.
5255 48. Exim crashed when trying to verify a sender address that was being
5258 49. Exim was recognizing only a space character after ".include". It now also
5259 recognizes a tab character.
5261 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5262 extracting the relevant information from the specification. The man page no
5263 longer contains scrambled data for the -d option, and I've added a section
5264 at the front about calling Exim under different names.
5266 51. Added "extra_headers" argument to the "mail" command in filter files.
5268 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5271 53. Installed eximstats 1.29.
5273 54. Added transport_filter_timeout as a generic transport option.
5275 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5276 Cc: header lines. This was required by RFC 822, but it not required by RFC
5279 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5280 incoming messages that did not have them. Now it does so only if the
5281 message originates locally, that is, if there is no associated remote host
5282 address. When Resent- header lines are present, this applies to the Resent-
5283 lines rather than the non-Resent- lines.
5285 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5286 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5288 58. Messages for configuration errors now include the name of the main
5289 configuration file - useful now that there may be more than one file in a
5290 list (.included file names were always shown).
5292 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5293 for those rare installations that do not start the daemon as root or run it
5294 setuid root. I've cut out the call to initgroups() if the daemon is not
5297 60. The Exim user and group can now be bound into the binary as text strings
5298 that are looked up at the start of Exim's processing.
5300 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5302 62. Added $mailstore_basename variable.
5304 63. Installed patch to sieve.c from Michael Haardt.
5306 64. When Exim failed to open the panic log after failing to open the main log,
5307 the original message it was trying to log was written to stderr and debug
5308 output, but if they were not available (the usual case in production), it
5309 was lost. Now it is written to syslog before the two lines that record the
5310 failures to open the logs.
5312 65. Users' Exim filters run in subprocesses under the user's uid. It is
5313 possible for a "deliver" command or an alias in a "personal" command to
5314 provoke an address rewrite. If logging of address rewriting is configured,
5315 this fails because the process is not running as root or exim. There may be
5316 a better way of dealing with this, but for the moment (because 4.30 needs
5317 to be released), I have disabled address rewrite logging when running a
5318 filter in a non-root, non-exim process.
5324 1. The buildconfig auxiliary program wasn't quoting the value set for
5325 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5326 not defined. This bug was masked in 4.22 by the effect that was fixed in
5329 2. Some messages that were rejected after a message id was allocated were
5330 shown as "incomplete" by exigrep. It no longer does this for messages that
5331 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5333 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5334 have allow_domain_literals set, the ID did not get logged in the <= line.
5335 Domain literals are now always recognized in Message-ID: header lines.
5337 4. The first argument for a ${extract expansion item is the key name or field
5338 number. Leading and trailing spaces in this item were not being ignored,
5339 causing some misleading effects.
5341 5. When deliver_drop_privilege was set, single queue runner processes started
5342 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5343 same command in the process it spins off) were not dropping privilege.
5345 6. When the daemon running as "exim" started a queue runner, it always
5346 re-executed Exim in the spun-off process. This is a waste of effort when
5347 deliver_drop_privilege is set. The new process now just calls the
5348 queue-runner function directly.
5354 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5357 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5358 ignored. Though the use of -oP was forcing the writing of a pid file, it
5359 was always written to the default place.
5361 3. If the message "no IP address found for host xxxx" is generated during
5362 incoming verification, it is now followed by identification of the incoming
5363 connection (so you can more easily find what provoked it).
5365 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5367 5. Added some features to "harden" Exim a bit more against certain attacks:
5369 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5370 be put in Local/Makefile. This is like the never_users runtime option,
5371 but it cannot be overridden. The default setting is "root".
5373 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5374 prefix string with which any file named in a -C command line option
5377 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5378 is retained for -C and -D only if the caller of Exim is root. Without
5379 it, the exim user may also use -C and -D and retain privilege.
5381 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5382 command line option is disabled.
5384 6. Macro names set by the -D option must start with an upper case letter, just
5385 like macro names defined in the configuration file.
5387 7. Added "dereference=" facility to LDAP.
5389 8. Two instances of the typo "uknown" in the source files are fixed.
5391 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5392 the Configure-Makefile script screwed up while processing it.
5394 10. Incorporated PCRE 4.4.
5396 11. The SMTP synchronization check was not operating right at the start of an
5397 SMTP session. For example, it could not catch a HELO sent before the client
5398 waited for the greeting. There is now a check for outstanding input at the
5399 point when the greeting is written. Because of the duplex, asynchronous
5400 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5401 way, but not yet received, when the check is performed.
5403 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5404 on TCP/IP sockets, because this apparently causes some broken clients to
5407 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5408 unchanged) from the Cygwin maintainer.
5410 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5411 was supported instead of testing for mbx. Effectively a typo.
5413 15. The spa authenticator server code was not checking that the input it
5414 received was valid base64.
5416 16. The debug output line for the "set" modifier in ACLs was not showing the
5417 name of the variable that was being set.
5419 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5421 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5422 lookup on its own. Something else was needed to trigger the lookup. For
5423 example, a match in host_lookup or the need for a host name in a host list.
5424 Now, if $sender_host_name is referenced and the host name has not yet been
5425 looked up, a lookup is performed. If the lookup fails, the variable remains
5426 empty, and $host_lookup_failed is set to "1".
5428 19. Added "eqi" as a case-independent comparison operator.
5430 20. The saslauthd authentication condition could segfault if neither service
5431 nor realm was specified.
5433 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5434 error message that was logged was misleading, and incoming SMTP
5435 connections were dropped. The message is now more accurate, and temporary
5436 errors are given to SMTP connections.
5438 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5439 (except RSET) with a 421 error, until QUIT is received. However, it was
5440 failing to send a response to QUIT.
5442 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5443 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5444 ACL is now run after the helo_try_verify_hosts code.
5446 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5447 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5448 case-independent, so other case variants are also recognized). Apparently
5449 some systems use these upper case variants.
5451 25. If more than two messages were waiting for the same host, and a transport
5452 filter was specified for the transport, Exim sent two messages over the
5453 same TCP/IP connection, and then failed with "socket operation on non-
5454 socket" when it tried to send the third.
5456 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5458 27. The extern definition of crypt16() in expand.c was not being excluded when
5459 the OS had its own crypt16() function.
5461 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5462 as a preferred synonym for return_size_limit, both as an option and as an
5465 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5467 30. Changed the default configuration ACL to relax the local part checking rule
5468 for addresses that are not in any local domains. For these addresses,
5469 slashes and pipe symbols are allowed within local parts, but the sequence
5470 /../ is explicitly forbidden.
5472 31. SPA server authentication was not clearing the challenge buffer before
5475 32. log_message in a "warn" ACL statement was writing to the reject log as
5476 well as to the main log, which contradicts the documentation and doesn't
5477 seem right (because no rejection is happening). So I have stopped it.
5479 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5480 However, I am unable to do any testing of this.
5482 34. Fixed an infelicity in the appendfile transport. When checking directories
5483 for a mailbox, to see if any needed to be created, it was accidentally
5484 using path names with one or more superfluous leading slashes; tracing
5485 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5487 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5488 contents are added to the log line that is written for every discarded
5489 recipient. (Previously a log_message setting was ignored.)
5491 36. The ${quote: operator now quotes the string if it is empty.
5493 37. The install script runs exim in order to find its version number. If for
5494 some reason other than non-existence or emptiness, which it checks, it
5495 could not run './exim', it was installing it with an empty version number,
5496 i.e. as "exim-". This error state is now caught, and the installation is
5499 38. An argument was missing from the function that creates an error message
5500 when Exim fails to connect to the socket for saslauthd authentication.
5501 This could cause Exim to crash, or give a corrupted message.
5503 39. Added isip, isip4, and isip6 to ${if conditions.
5505 40. The ACL variables $acl_xx are now saved with the message, and can be
5506 accessed later in routers, transports, and filters.
5508 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5509 strings in the file are not string-expanded.
5511 42. If a MAIL command specified a SIZE value that was too large to fit into an
5512 int variable, the check against message_size_limit failed. Such values are
5513 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5514 day this will have to be increased, but I don't think I want to be around
5515 when emails are that large.
5522 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5523 iconv() is not standard in FreeBSD.
5525 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5526 IPv6 enabled. The observed symptom was a segmentation fault on return from
5527 the function os_common_find_running_interfaces() in src/os.c.
5529 3. In the check_special_case() function in daemon.c I had used "errno" as an
5530 argument name, which causes warnings on some systems. This was basically a
5531 typo, since it was named "eno" in the comments!
5533 4. The code that waits for the clock to tick (at a resolution of some fraction
5534 of a second) so as to ensure message-id uniqueness was always waiting for
5535 at least one whole tick, when it could have waited for less. [This is
5536 almost certainly not relevant at current processor speeds, where it is
5537 unlikely to ever wait at all. But we try to future-proof.]
5539 5. The function that sleeps for a time interval that includes fractions of a
5540 second contained a race. It did not block SIGALRM between setting the
5541 timer, and suspending (a couple of lines later). If the interval was short
5542 and the sigsuspend() was delayed until after it had expired, the suspension
5543 never ended. On busy systems this could lead to processes getting stuck for
5546 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5547 process, before it forks any delivery processes. The open lookup caching
5548 mechanism meant that the open file or database connection was passed into
5549 the delivery process. The problem was that delivery processes always tidy
5550 up cached lookup data. This could cause a problem for the next delivery
5551 process started by the queue runner, because the external queue runner
5552 process does not know about the closure. So the next delivery process
5553 still has data in the lookup cache. In the case of a file lookup, there was
5554 no problem because closing a file descriptor in a subprocess doesn't affect
5555 the parent. However, if the lookup was caching a connection to a database,
5556 the connection was closed, and the second delivery process was likely to
5557 see errors such as "PGSQL: query failed: server closed the connection
5558 unexpectedly". The problem has been fixed by closing all cached lookups
5559 in a queue runner before running a delivery process.
5561 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5562 seem to have the "const" qualifier which it has on other OS. I've
5565 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5566 *of the same type* (client or server) with the same public name that an
5567 error should be diagnosed.
5569 9. When Exim looked up a host name for an IP address, but failed to find the
5570 original IP address when looking up the host name (a safety check), it
5571 output the message "<ip address> does not match any IP for NULL", which was
5572 confusing, to say the least. The bug was that the host name should have
5573 appeared instead of "NULL".
5575 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5576 user that is built into the binary, and the -C or -D options is used, root
5577 privilege is dropped before the configuration file is read. In addition,
5578 logging is switched to stderr instead of the normal log files. If the
5579 configuration then re-defines the Exim user, the unprivileged environment
5580 is probably not what is expected, so Exim logs a panic warning message (but
5583 However, if deliver_drop_privilege is set, the unprivileged state may well
5584 be exactly what is intended, so the warning has been cut out in that case,
5585 and Exim is allowed to try to write to its normal log files.
5591 1. smtp_return_error_details was not giving details for temporary sender
5592 or receiver verification errors.
5594 2. Diagnose a configuration error if two authenticators have the same public
5597 3. Exim used not to create the message log file for a message until the first
5598 delivery attempt. This could be confusing when incoming messages were held
5599 for policy or load reasons. The message log file is now created at the time
5600 the message is received, and an initial "Received" line is written to it.
5602 4. The automatically generated man page for command line options had a minor
5603 bug that caused no ill effects; however, a more serious problem was that
5604 the procedure for building the man page automatically didn't always
5605 operate. Consequently, release 4.20 contains an out-of-date version. This
5606 shouldn't happen again.
5608 5. When building Exim with embedded Perl support, the script that builds the
5609 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5610 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5612 6. The freeze_tell option was not being used for messages that were frozen on
5613 arrival, either by an ACL or by local_scan().
5615 7. Added the smtp_incomplete_transaction log selector.
5617 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5618 was accepting AUTH without a new EHLO.
5620 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5621 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5624 10. Logging of TCP/IP connections (when configured) now happens in the main
5625 daemon process instead of the child process, so that the TCP/IP connection
5626 count is more accurate (but it can never be perfect).
5628 11. The use of "drop" in a nested ACL was not being handled correctly in the
5629 outer ACL. Now, if condition failure induced by the nested "drop" causes
5630 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5631 or "require"), the connection is dropped.
5633 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5634 that yield "discard" can now be used with an "accept" or a "discard" verb,
5635 but an error is generated for any others (because I can't see a useful way
5636 to define what should happen).
5638 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5639 are now processed in the same way as lines in the Exim configuration file.
5640 In particular, continuation lines are supported.
5642 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5644 15. Added -ti meaning -t -i.
5646 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5647 domains, and warn by logging if others are found.
5649 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5650 returned from GIFCONF when Exim is trying to find the list of interfaces on
5651 a host. The code in os.c has been modified to copy each ifreq to an aligned
5652 structure in all cases.
5654 Also, in some cases, the returned ifreq's were being copied to a 'struct
5655 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5656 means the last couple of bytes of an IPv6 address could be chopped if the
5657 ifreq contained only a normal sockaddr (14 bytes storage).
5659 18. Named domain lists were not supported in the hosts_treat_as_local option.
5660 An entry such as +xxxx was not recognized, and was treated as a literal
5663 19. Ensure that header lines added by a DATA ACL are included in the reject log
5664 if the ACL subsequently rejects the message.
5666 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5667 MD5 (which is deprecated).
5669 21. When testing a filter file using -bf, Exim was writing a message when it
5670 took the sender from a "From " line in the message, but it was not doing so
5671 when it took $return_path from a Return-Path: header line. It now does.
5673 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5674 with a valid header line field name (a series of printing characters
5675 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5677 23. Changed "disc" in the source to "disk" to conform to the documentation and
5678 the book and for uniformity.
5680 24. Ignore Sendmail's -Ooption=value command line item.
5682 25. When execve() failed while trying to run a command in a pipe transport,
5683 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5684 could be confused with a return value of 69 from the command itself. This
5685 has been changed to 127, the value the shell returns if it is asked to run
5686 a non-existent command. The wording for the related log line suggests a
5687 non-existent command as the problem.
5689 26. If received_header_text expands to an empty string, do not add a Received:
5690 header line to the message. (Well, it adds a token one on the spool, but
5691 marks it "old" so that it doesn't get used or transmitted.)
5693 27. Installed eximstats 1.28 (addition of -nt option).
5695 28. There was no check for failure on the call to getsockname() in the daemon
5696 code. This can fail if there is a shortage of resources on the system, with
5697 ENOMEM, for example. A temporary error is now given on failure.
5699 29. Contrary to the C standard, it seems that in some environments, the
5700 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5701 program. Exim now does this explicitly; it affects the formatting of
5702 timestamps using strftime().
5704 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5705 complaints. I've now initialized all the variables, to avoid this.
5707 32. Header lines added by a system filter were not being "seen" during
5708 transport-time rewrites.
5710 33. The info_callback() function passed to OpenSSL is set up with type void
5711 (*)(SSL *, int, int), as described somewhere. However, when calling the
5712 function (actually a macro) that sets it up, the type void(*)() is
5713 expected. I've put in a cast to prevent warnings from picky compilers.
5715 34. If a DNS black list lookup found a CNAME record, but there were no A
5716 records associated with the domain it pointed at, Exim crashed.
5718 35. If a DNS black list lookup returned more than one A record, Exim ignored
5719 all but the first. It now scans all returned addresses if a particular IP
5720 value is being sought. In this situation, the contents of the
5721 $dnslist_value variable are a list of all the addresses, separated by a
5724 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5725 to accept a host name and all its aliases if the forward lookup for any of
5726 them yielded the IP address of the incoming connection. Now it accepts only
5727 those names whose forward lookup yields the correct IP address. Any other
5728 names are discarded. This closes a loophole whereby a rogue DNS
5729 administrator could create reverse DNS records to break through a
5730 wildcarded host restriction in an ACL.
5732 37. If a user filter or a system filter that ran in a subprocess used any of
5733 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5734 the wrong values were passed to the pipe command ($thisaddress had the
5735 value of $0, $0 had the value of $1, etc). This bug was introduced by
5736 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5738 38. Improved the line breaking for long SMTP error messages from ACLs.
5739 Previously, if there was no break point between 40 and 75 characters, Exim
5740 left the rest of the message alone. Two changes have been made: (a) I've
5741 reduced the minimum length to 35 characters; (b) if it can't find a break
5742 point between 35 and 75 characters, it looks ahead and uses the first one
5743 that it finds. This may give the occasional overlong line, but at least the
5744 remaining text gets split now.
5746 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5747 file descriptors might be low, and that setting 1000 would always raise it.
5748 It turns out that in some environments, the limit is already over 1000 and
5749 that lowering it causes trouble. So now Exim takes care not to decrease it.
5751 40. When delivering a message, the value of $return_path is set to $sender_
5752 address at the start of routing (routers may change the value). By an
5753 oversight, this default was not being set up when an address was tested by
5754 -bt or -bv, which affected the outcome if any router or filter referred to
5757 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5758 log only when "message" or "log_message" are set. However, if one of the
5759 conditions was an address verification, or a call to a nested ACL, the
5760 messages generated by the underlying test were being passed through. This
5761 no longer happens. The underlying message is available in $acl_verify_
5762 message for both "message" and "log_message" expansions, so it can be
5763 passed through if needed.
5765 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5766 new expansion $bh_ to give the encoded byte string without charset
5767 translation. Translation happens only if iconv() is available; HAVE_ICONV
5768 indicates this at build time. HEADERS_CHARSET gives the charset to
5769 translate to; headers_charset can change it in the configuration, and
5770 "headers charset" can change it in an individual filter file.
5772 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5773 that creates RFC 2047 encoded "words" labels them as that charset instead
5774 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5775 expansion operator; (ii) when Exim creates a From: line for a local
5776 message; (iii) when a header line is rewritten to include a "phrase" part.
5778 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5779 buggy, causing it to skip the first lines of messages whose message ID
5780 ended in 'D'. This would not have bitten before Exim release 4.14, because
5781 message IDs were unlikely to end in 'D' before then. The effect was to have
5782 incorrect size information for certain domains.
5784 45. #include "config.h" was missing at the start of the crypt16.c module. This
5785 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5788 46. If there was a timeout during a "random" callout check, Exim treated it as
5789 a failure of the random address, and carried on sending RSET and the real
5790 address. If the delay was just some slowness somewhere, the response to the
5791 original RCPT would be taken as a response to RSET and so on, causing
5792 mayhem of various kinds.
5794 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5795 when I implemented it. It didn't allow for the fact that some option values
5796 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5797 the right test for positive values.
5799 48. Domain names in DNS records are case-independent. Exim always looks them up
5800 in lower case. Some resolvers return domain names in exactly the case they
5801 appear in the zone file, that is, they may contain uppercase letters. Not
5802 all resolvers do this - some return always lower case. Exim was treating a
5803 change of case by a resolver as a change of domain, similar to a widening
5804 of a domain abbreviation. This triggered its re-routing code and so it was
5805 trying to route what was effectively the same domain again. This normally
5806 caused routing to fail (because the router wouldn't handle the domain
5807 twice). Now Exim checks for this case specially, and just changes the
5808 casing of the domain that it ultimately uses when it transmits the message
5811 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5814 50. If a filter generated a file delivery with a non-absolute name (possible if
5815 no home directory exists for the router), the forbid_file option was not
5818 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5819 the existing equality matching.
5821 52. Exim was using ints instead of ino_t variables in some places where it was
5822 dealing with inode numbers.
5824 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5825 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5826 called TMPDIR, and if it finds it is different, it changes its value.
5828 54. The smtp_printf() function is now made available to local_scan() so
5829 additional output lines can be written before returning. There is also an
5830 smtp_fflush() function to enable the detection of a dropped connection.
5831 The variables smtp_input and smtp_batched_input are exported to
5834 55. Changed the default runtime configuration: the message "Unknown user"
5835 has been removed from the ACL, and instead placed on the localuser router,
5836 using the cannot_route_message feature. This means that any verification
5837 failures that generate their own messages won't get overridden. Similarly,
5838 the "Unrouteable address" message that was in the ACL for unverifiable
5839 relay addresses has also been removed.
5841 56. Added hosts_avoid_esmtp to the smtp transport.
5843 57. The exicyclog script was not checking for the esoteric option
5844 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5845 will work only if exicyclog is run under the appropriate euid.
5847 58. Following a discussion on the list, the rules by which Exim recognises line
5848 endings on incoming messages have been changed. The -dropcr and drop_cr
5849 options are now no-ops, retained only for backwards compatibility. The
5850 following line terminators are recognized: LF CRLF CR. However, special
5851 processing applies to CR:
5853 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5854 nor a local message in the state where . is a terminator.
5856 (ii) If a bare CR is encountered in a header line, an extra space is added
5857 after the line terminator so as not to end the header. The reasoning
5858 behind this is that bare CRs in header lines are most likely either
5859 to be mistakes, or people trying to play silly games.
5861 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5862 was being incorrectly given as 18 bytes larger than it should have been.
5863 This is a VOB (very old bug).
5865 60. This may never have affected anything current, but just in case it has:
5866 When the local host is found other than at the start of a list of hosts,
5867 the local host, those with the same MX, and any that follow, are discarded.
5868 When the list in question was part of a longer list of hosts, the following
5869 hosts (not currently being processed) were also being discarded. This no
5870 longer happens. I'm not sure if this situation could ever has previously
5873 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5876 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5877 header. This is something that is recommended in a new Internet Draft, and
5878 is something that is documented as being done by Sendmail. There are two
5879 possible values. For messages generated by the autoreply transport, Exim
5882 Auto-Submitted: auto-replied
5884 whereas for all other generated messages (e.g. bounces) it adds
5886 Auto-Submitted: auto-generated
5888 63. The "personal" condition in filters now includes a test for the
5889 Auto-Submitted: header. If it contains the string "auto-" the message it
5890 not considered personal.
5892 64. Added rcpt_include_affixes as a generic transport option.
5894 65. Added queue_only_override (default true).
5896 66. Added the syslog_duplication option.
5898 67. If what should have been the first header line of a message consisted of
5899 a space followed by a colon, Exim was mis-interpreting it as a header line.
5900 It isn't of course - it is syntactically invalid and should therefore be
5901 treated as the start of the message body. The misbehaviour could have
5902 caused a number of strange effects, including loss of data in subsequent
5903 header lines, and spool format errors.
5905 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5906 client host had authenticated. This control can now be exercised by an ACL
5907 for more flexibility.
5909 69. By default, callouts do not happen when testing with -bh. There is now a
5910 variant, -bhc, which does actually run the callout code, including
5911 consulting and updating the callout cache.
5913 70. Added support for saslauthd authentication, courtesy of Alexander
5916 71. If statvfs() failed on the spool or log directories while checking their
5917 size for availability, Exim confusingly gave the error "space shortage".
5918 Furthermore, in debugging mode it crashed with a floating point exception.
5919 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5920 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5921 serious problem, Exim now writes to the main and panic logs when this
5922 happens, with details of the failure. It then refuses to accept the
5923 incoming message, giving the message "spool directory problem" or "log
5924 directory problem" with a 421 code for SMTP messages.
5926 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5927 0, 1, and 2 exist, because some OS complain for execs without them (see
5928 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5929 descriptors. However, the code omitted to check that the open succeeded,
5930 causing mysterious errors if for some reason the permissions on /dev/null
5931 got screwed. Now Exim writes a message to the main and panic logs, and
5932 bombs out if it can't open /dev/null.
5934 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5935 interact so that it is all more flexible. It is supposed to remain
5936 backwards compatible. Also added extra_local_interfaces.
5938 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5939 to bomb out with an assertion failure - to the client this appears as a
5940 connection drop. This problem occurs in the part of the code that was taken
5941 from the Samba project. Fortunately, the assertion is in a very simple
5942 function, so I have fixed this by reproducing the function inline in the
5943 one place where it is called, and arranging for authentication to fail
5944 instead of killing the process with assert().
5946 75. The SPA client code was not working when the server requested OEM rather
5947 than Unicode encoding.
5949 76. Added code to make require_files with a specific uid setting more usable in
5950 the case where statting the file as root fails - usually a non-root-mounted
5951 NFS file system. When this happens and the failure is EACCES, Exim now
5952 forks a subprocess and does the per-uid checking as the relevant uid.
5954 77. Added process_log_path.
5956 78. If log_file_path was not explicitly set, a setting of check_log_space or
5957 check_log_inodes was ignored.
5959 79. If a space check for the spool or log partitions fails, the incident is now
5960 logged. Of course, in the latter case the data may get lost...
5962 80. Added the %p formatting code to string_format() so that it can be used to
5963 print addresses in debug_print(). Adjusted all the address printing in the
5964 debugging in store.c to use %p rather than %d.
5966 81. There was a concern that a line of code in smtp_in.c could overflow a
5967 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5968 initially expressed, the concern was not well-founded, because trailing
5969 spaces are removed early. However, if the trailing spaces were followed by
5970 a NULL, they did not get removed, so the overflow was possible. Two fixes
5973 (a) I re-wrote the offending code in a cleaner fashion.
5974 (b) If an incoming SMTP command contains a NULL character, it is rejected
5977 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5978 runs initgroups(), so that if the Exim user is in any additional groups,
5979 they will be used during message reception.
5985 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob