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. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
88 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
90 PP/02 Make -n do something, by making it not do something.
91 When combined with -bP, the name of an option is not output.
93 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
96 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
97 $sender_host_name and config options to manage this, and basic check
100 PP/05 DSCP support for outbound connections and control modifier for inbound.
102 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
103 (Only plugin which currently uses this is kerberos4, which nobody should
104 be using, but we should make it available and other future plugins might
105 conceivably use it, even though it would break NAT; stuff *should* be
106 using channel bindings instead).
108 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
109 name; added for Sendmail compatibility; requires admin caller.
110 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
111 just ignore it); requires trusted caller.
112 Also parse but ignore: -Ac -Am -X<logfile>
115 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
117 TL/02 Add +smtp_confirmation as a default logging option.
119 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
120 Patch by Magnus Holmgren from 2007-02-20.
122 TL/04 Bugzilla 1281 - Spec typo.
123 Bugzilla 1283 - Spec typo.
124 Bugzilla 1290 - Spec grammar fixes.
126 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
128 TL/06 Add Experimental DMARC support using libopendmarc libraries.
130 TL/07 Fix an out of order global option causing a segfault. Reported to dev
131 mailing list by by Dmitry Isaikin.
133 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
135 JH/02 Support "G" suffix to numbers in ${if comparisons.
137 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
139 NM/01 Bugzilla 1197 - Spec typo
140 Bugzilla 1196 - Spec examples corrections
142 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
144 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
145 gnutls_enable_pkcs11, but renamed to more accurately indicate its
148 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
149 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
151 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
152 "acl {{name}{arg}...}", and optional args on acl condition
155 JH/05 Permit multiple router/transport headers_add/remove lines.
157 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
159 JH/07 Avoid using a waiting database for a single-message-only transport.
160 Performance patch from Paul Fisher. Bugzilla 1262.
162 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
165 JH/09 Add $headers_added variable, with content from use of ACL modifier
166 add_header (but not yet added to the message). Bugzilla 199.
168 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
169 Pulled from Bugzilla 817 by Wolfgang Breyha.
171 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
173 (nb: this is the same fix as in Exim 4.80.1)
175 JH/11 Add A= logging on delivery lines, and a client_set_id option on
178 JH/12 Add optional authenticated_sender logging to A= and a log_selector
181 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
183 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
184 advertise SMTP AUTH mechanism to us, instead of a generic
185 protocol violation error. Also, make Exim more robust to bad
186 data from the Dovecot auth socket.
188 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
190 When a queue runner is handling a message, Exim first routes the
191 recipient addresses, during which it prunes them based on the retry
192 hints database. After that it attempts to deliver the message to
193 any remaining recipients. It then updates the hints database using
196 So if a recipient address works intermittently, it can get repeatedly
197 deferred at routing time. The retry hints record remains fresh so the
198 address never reaches the final cutoff time.
200 This is a fairly common occurrence when a user is bumping up against
201 their storage quota. Exim had some logic in its local delivery code
202 to deal with this. However it did not apply to per-recipient defers
203 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
205 This change adds a proper retry rule check during routing so that the
206 final cutoff time is checked against the message's age. We only do
207 this check if there is an address retry record and there is not a
208 domain retry record; this implies that previous attempts to handle
209 the address had the retry_use_local_parts option turned on. We use
210 this as an approximation for the destination being like a local
211 delivery, as in LMTP.
213 I suspect this new check makes the old local delivery cutoff check
214 redundant, but I have not verified this so I left the code in place.
216 TF/02 Correct gecos expansion when From: is a prefix of the username.
218 Test 0254 submits a message to Exim with the header
222 When I ran the test suite under the user fanf2, Exim expanded
223 the header to contain my full name, whereas it should have added
224 a Resent-Sender: header. It erroneously treats any prefix of the
225 username as equal to the username.
227 This change corrects that bug.
229 GF/01 DCC debug and logging tidyup
230 Error conditions log to paniclog rather than rejectlog.
231 Debug lines prefixed by "DCC: " to remove any ambiguity.
233 TF/03 Avoid unnecessary rebuilds of lookup-related code.
235 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
236 Bug spotted by Jeremy Harris; was flawed since initial commit.
237 Would have resulted in OCSP responses post-SNI triggering an Exim
238 NULL dereference and crash.
240 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
242 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
243 Bug detection, analysis and fix by Samuel Thibault.
244 Bugzilla 1331, Debian bug #698092.
246 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
248 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
249 Server implementation by Todd Lyons, client by JH.
250 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
251 config variable "prdr_enable" controls whether the server
252 advertises the facility. If the client requests PRDR a new
253 acl_data_smtp_prdr ACL is called once for each recipient, after
254 the body content is received and before the acl_smtp_data ACL.
255 The client is controlled by bolth of: a hosts_try_prdr option
256 on the smtp transport, and the server advertisement.
257 Default client logging of deliveries and rejections involving
258 PRDR are flagged with the string "PRDR".
260 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
261 fclose(). Diagnosis by Todd Lyons.
263 PP/17 Update configure.default to handle IPv6 localhost better.
264 Patch by Alain Williams (plus minor tweaks).
267 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
268 This is now consistent with GnuTLS, and is now documented: the
269 previous undocumented portable approach to treating the option as
270 unset was to force an expansion failure. That still works, and
271 an empty string is now equivalent.
273 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
274 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
275 not performing validation itself.
277 PP/20 Added force_command boolean option to pipe transport.
278 Patch from Nick Koston, of cPanel Inc.
280 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
283 TF/04 Added udpsend ACL modifer and hexquote expansion operator
285 PP/21 Fix eximon continuous updating with timestamped log-files.
286 Broken in a format-string cleanup in 4.80, missed when I repaired the
287 other false fix of the same issue.
288 Report and fix from Heiko Schlichting.
291 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
292 Report from Prashanth Katuri.
294 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
295 It's SecureTransport, so affects any MacOS clients which use the
296 system-integrated TLS libraries, including email clients.
298 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
299 using a MIME ACL for non-SMTP local injection.
300 Report and assistance in diagnosis by Warren Baker.
302 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
304 JH/16 Fix comparisons for 64b. Bugzilla 1385.
306 TL/09 Add expansion variable $authenticated_fail_id to keep track of
307 last id that failed so it may be referenced in subsequent ACL's.
309 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
312 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
313 ldap library initialization, allowing self-signed CA's to be
314 used. Also properly sets require_cert option later in code by
315 using NULL (global ldap config) instead of ldap handle (per
316 session). Bug diagnosis and testing by alxgomz.
318 TL/12 Enhanced documentation in the ratelimit.pl script provided in
319 the src/util/ subdirectory.
321 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
322 renamed to Transport Post Delivery Action by Jeremy Harris, as
325 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
326 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
327 redis_servers = needs to be configured which will be used by the redis
328 lookup. Patch from Warren Baker, of The Packet Hub.
330 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
332 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
333 hostname or reverse DNS when processing a host list. Used suggestions
334 from multiple comments on this bug.
336 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
338 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
339 Missed a few lines, added it to make the runtest require no keyboard
342 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
343 contains upper case chars. Make router use caseful_local_part.
345 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
346 support when GnuTLS has been built with p11-kit.
352 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
354 This, or similar/improved, will also be change PP/11 of 4.82.
360 PP/01 Handle short writes when writing local log-files.
361 In practice, only affects FreeBSD (8 onwards).
362 Bugzilla 1053, with thanks to Dmitry Isaikin.
364 NM/01 Bugzilla 949 - Documentation tweak
366 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
369 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
371 PP/02 Implemented gsasl authenticator.
373 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
375 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
376 `pkg-config foo` for cflags/libs.
378 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
379 with rest of GSASL and with heimdal_gssapi.
381 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
382 `pkg-config foo` for cflags/libs for the TLS implementation.
384 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
385 properties get this fed in as external SSF. A number of robustness
386 and debugging improvements to the cyrus_sasl authenticator.
388 PP/08 cyrus_sasl server now expands the server_realm option.
390 PP/09 Bugzilla 1214 - Log authentication information in reject log.
391 Patch by Jeremy Harris.
393 PP/10 Added dbmjz lookup type.
395 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
397 PP/12 MAIL args handles TAB as well as SP, for better interop with
398 non-compliant senders.
399 Analysis and variant patch by Todd Lyons.
401 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
402 Bug report from Lars Müller <lars@samba.org> (via SUSE),
403 Patch from Dirk Mueller <dmueller@suse.com>
405 PP/13 tls_peerdn now print-escaped for spool files.
406 Observed some $tls_peerdn in wild which contained \n, which resulted
407 in spool file corruption.
409 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
410 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
411 or write after TLS renegotiation, which otherwise led to messages
414 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
415 as a tracking header (ie: a signed header comes before the signature).
416 Patch from Wolfgang Breyha.
418 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
419 comma-sep list; embedded commas doubled.
421 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
423 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
425 Report and patch from Dmitry Banschikov.
427 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
428 Removed SSL_clear() after SSL_new() which led to protocol negotiation
429 failures. We appear to now support TLS1.1+ with Exim.
431 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
432 lets Exim select keys and certificates based upon TLS SNI from client.
433 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
434 before an outbound SMTP session. New log_selector, +tls_sni.
436 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
437 NULL dereference. Report and patch from Alun Jones.
439 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
440 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
441 Not seeing resolver debug output on NetBSD, but suspect this is a
442 resolver implementation change.
444 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
445 Left warnings. Added "eximon gdb" invocation mode.
447 PP/21 Defaulting "accept_8bitmime" to true, not false.
449 PP/22 Added -bw for inetd wait mode support.
451 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
452 locate the relevant includes and libraries. Made this the default.
454 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
455 Bugzilla 1246, report and most of solution from Tomasz Kusy.
457 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
458 This may cause build issues on older platforms.
460 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
461 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
462 gnutls_require_mac & gnutls_require_protocols (no longer supported).
463 Added SNI support via GnuTLS too.
464 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
466 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
468 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
469 Applied second patch from Janne, implementing suggestion to default
470 multiple-strings-in-record handling to match SPF spec.
472 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
474 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
475 read-only, out of scope).
476 Patch from Wolfgang Breyha, report from Stuart Northfield.
478 PP/29 Fix three issues highlighted by clang analyser static analysis.
479 Only crash-plausible issue would require the Cambridge-specific
480 iplookup router and a misconfiguration.
481 Report from Marcin Mirosław.
483 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
485 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
486 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
487 As part of this, removing so much warning spew let me fix some minor
488 real issues in debug logging.
490 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
491 assignment on my part. Fixed.
493 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
494 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
495 Janne Snabb (who went above and beyond: thank you).
497 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
498 string otherwise requires a connection and a bunch more work and it's
499 relatively easy to get wrong. Should also expose TLS library linkage
502 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
503 64-bit ${eval} (JH/03).
505 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
506 GNU libc to support some of the 64-bit stuff, should not lead to
507 conflicts. Defined before os.h is pulled in, so if a given platform
508 needs to override this, it can.
510 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
511 protection layer was required, which is not implemented.
512 Bugzilla 1254, patch from Wolfgang Breyha.
514 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
515 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
516 tls_dhparam take prime identifiers. Also unbreak combination of
517 OpenSSL+DH_params+TLSSNI.
519 PP/39 Disable SSLv2 by default in OpenSSL support.
525 PP/01 Solaris build fix for Oracle's LDAP libraries.
526 Bugzilla 1109, patch from Stephen Usher.
528 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
530 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
533 TF/02 Fix a couple more cases where we did not log the error message
534 when unlink() failed. See also change 4.74-TF/03.
536 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
537 lock up or crash if it happened to be inside a call to libc when it
538 got a SIGUSR1 from exiwhat.
540 The SIGUSR1 handler appends the current process status to the process
541 log which is later printed by exiwhat. It used to use the general
542 purpose logging code to do this, but several functions it calls are
543 not safe for signals.
545 The new output code in the SIGUSR1 handler is specific to the process
546 log, and simple enough that it's easy to inspect for signal safety.
547 Removing some special cases also simplifies the general logging code.
548 Removing the spurious timestamps from the process log simplifies
551 TF/04 Improved ratelimit ACL condition.
553 The /noupdate option has been deprecated in favour of /readonly which
554 has clearer semantics. The /leaky, /strict, and /readonly update modes
555 are mutually exclusive. The update mode is no longer included in the
556 database key; it just determines when the database is updated. (This
557 means that when you upgrde Exim will forget old rate measurements.)
559 Exim now checks that the per_* options are used with an update mode that
560 makes sense for the current ACL. For example, when Exim is processing a
561 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
562 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
563 must specify per_mail/readonly. If you omit the update mode it defaults to
564 /leaky where that makes sense (as before) or /readonly where required.
566 The /noupdate option is now undocumented but still supported for
567 backwards compatibility. It is equivalent to /readonly except that in
568 ACLs where /readonly is required you may specify /leaky/noupdate or
569 /strict/noupdate which are treated the same as /readonly.
571 A useful new feature is the /count= option. This is a generalization
572 of the per_byte option, so that you can measure the throughput of other
573 aggregate values. For example, the per_byte option is now equivalent
574 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
576 The per_rcpt option has been generalized using the /count= mechanism
577 (though it's more complicated than the per_byte equivalence). When it is
578 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
579 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
580 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
581 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
582 Note that using per_rcpt with a non-readonly update mode in more than
583 one ACL will cause the recipients to be double-counted. (The per_mail
584 and per_byte options don't have this problem.)
586 The handling of very low rates has changed slightly. If the computed rate
587 is less than the event's count (usually one) then this event is the first
588 after a long gap. In this case the rate is set to the same as this event's
589 count, so that the first message of a spam run is counted properly.
591 The major new feature is a mechanism for counting the rate of unique
592 events. The new per_addr option counts the number of different
593 recipients that someone has sent messages to in the last time period. It
594 behaves like per_rcpt if all the recipient addresses are different, but
595 duplicate recipient addresses do not increase the measured rate. Like
596 the /count= option this is a general mechanism, so the per_addr option
597 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
598 example, measure the rate that a client uses different sender addresses
599 with the options per_mail/unique=$sender_address. There are further
600 details in the main documentation.
602 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
604 TF/06 Removed a few PCRE remnants.
606 TF/07 Automatically extract Exim's version number from tags in the git
607 repository when doing development or release builds.
609 PP/02 Raise smtp_cmd_buffer_size to 16kB.
610 Bugzilla 879. Patch from Paul Fisher.
612 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
613 Heavily based on revision 40f9a89a from Simon Arlott's tree.
616 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
618 PP/05 Variable $av_failed, true if the AV scanner deferred.
619 Bugzilla 1078. Patch from John Horne.
621 PP/06 Stop make process more reliably on build failure.
622 Bugzilla 1087. Patch from Heiko Schlittermann.
624 PP/07 Make maildir_use_size_file an _expandable_ boolean.
625 Bugzilla 1089. Patch from Heiko Schlittermann.
627 PP/08 Handle ${run} returning more data than OS pipe buffer size.
628 Bugzilla 1131. Patch from Holger Weiß.
630 PP/09 Handle IPv6 addresses with SPF.
631 Bugzilla 860. Patch from Wolfgang Breyha.
633 PP/10 GnuTLS: support TLS 1.2 & 1.1.
635 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
638 PP/11 match_* no longer expand right-hand-side by default.
639 New compile-time build option, EXPAND_LISTMATCH_RHS.
640 New expansion conditions, "inlist", "inlisti".
642 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
644 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
646 PP/14 fix log_write() format string regression from TF/03.
647 Bugzilla 1152. Patch from Dmitry Isaikin.
653 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
655 PP/02 Harmonised TLS library version reporting; only show if debugging.
656 Layout now matches that introduced for other libraries in 4.74 PP/03.
658 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
660 PP/04 New "dns_use_edns0" global option.
662 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
665 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
666 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
668 TK/01 Updated PolarSSL code to 0.14.2.
669 Bugzilla 1097. Patch from Andreas Metzler.
671 PP/07 Catch divide-by-zero in ${eval:...}.
674 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
677 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
678 format-string attack -- SECURITY: remote arbitrary code execution.
680 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
681 time unintentionally subject to list matching rules, letting the header
682 cause arbitrary Exim lookups (of items which can occur in lists, *not*
683 arbitrary string expansion). This allowed for information disclosure.
685 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
686 INT_MIN/-1 -- value coerced to INT_MAX.
692 NM/01 Workround for PCRE version dependency in version reporting
695 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
696 This fixes portability to compilers other than gcc, notably
697 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
699 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
700 makefiles for portability to HP-UX and POSIX correctness.
702 PP/01 Permit LOOKUP_foo enabling on the make command-line.
703 Also via indented variable definition in the Makefile.
704 (Debugging by Oliver Heesakkers).
706 PP/02 Restore caching of spamd results with expanded spamd_address.
707 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
709 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
710 Improves build reliability. Fix from: Frank Elsner
712 NM/02 Fix wide character breakage in the rfc2047 coding
713 Fixes bug 1064. Patch from Andrey N. Oktyabrski
715 NM/03 Allow underscore in dnslist lookups
716 Fixes bug 1026. Patch from Graeme Fowler
718 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
719 Code patches from Adam Ciarcinski of NetBSD.
721 NM/04 Fixed exiqgrep to cope with mailq missing size issue
724 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
725 is logged, to avoid truncation. Patch from John Horne.
727 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
728 Patch from Jakob Hirsch.
730 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
731 SQL string expansion failure details.
732 Patch from Andrey Oktyabrski.
734 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
735 Patch from Simon Arlott.
737 PP/09 New lookups functionality failed to compile on old gcc which rejects
738 extern declarations in function scope.
739 Patch from Oliver Fleischmann
741 PP/10 Use sig_atomic_t for flags set from signal handlers.
742 Check getgroups() return and improve debugging.
743 Fixed developed for diagnosis in bug 927 (which turned out to be
746 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
747 Patch from Mark Zealey.
749 PP/12 Bugzilla 1056: Improved spamd server selection.
750 Patch from Mark Zealey.
752 PP/13 Bugzilla 1086: Deal with maildir quota file races.
753 Based on patch from Heiko Schlittermann.
755 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
756 Patch from Uwe Doering, sign-off by Michael Haardt.
758 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
759 variable declaration deep within a block. Bug and patch from
762 PP/15 lookups-Makefile IRIX compatibilty coercion.
764 PP/16 Make DISABLE_DKIM build knob functional.
766 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
767 Patch by Simon Arlott
769 TF/03 Fix valgrind.h portability to C89 compilers that do not support
770 variable argument macros. Our copy now differs from upstream.
776 TF/01 Failure to get a lock on a hints database can have serious
777 consequences so log it to the panic log.
779 TF/02 Log LMTP confirmation messages in the same way as SMTP,
780 controlled using the smtp_confirmation log selector.
782 TF/03 Include the error message when we fail to unlink a spool file.
784 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
785 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
786 for maintaining out-of-tree patches for some time.
788 PP/01 Bugzilla 139: Documentation and portability issues.
789 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
790 Handle per-OS dynamic-module compilation flags.
792 PP/02 Let /dev/null have normal permissions.
793 The 4.73 fixes were a little too stringent and complained about the
794 permissions on /dev/null. Exempt it from some checks.
795 Reported by Andreas M. Kirchwitz.
797 PP/03 Report version information for many libraries, including
798 Exim version information for dynamically loaded libraries. Created
799 version.h, now support a version extension string for distributors
800 who patch heavily. Dynamic module ABI change.
802 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
803 privilege escalation vulnerability whereby the Exim run-time user
804 can cause root to append content of the attacker's choosing to
807 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
810 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
811 If dropping privileges for untrusted macros, we disabled normal logging
812 on the basis that it would fail; for the Exim run-time user, this is not
813 the case, and it resulted in successful deliveries going unlogged.
814 Fixed. Reported by Andreas Metzler.
820 PP/01 Date: & Message-Id: revert to normally being appended to a message,
821 only prepend for the Resent-* case. Fixes regression introduced in
822 Exim 4.70 by NM/22 for Bugzilla 607.
824 PP/02 Include check_rfc2047_length in configure.default because we're seeing
825 increasing numbers of administrators be bitten by this.
827 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
829 PP/03 Bugzilla 994: added openssl_options main configuration option.
831 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
833 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
835 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
837 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
838 without a peer certificate, leading to a segfault because of an
839 assumption that peers always have certificates. Be a little more
840 paranoid. Problem reported by Martin Tscholak.
842 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
843 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
844 NB: ClamAV planning to remove STREAM in "middle of 2010".
845 CL also introduces -bmalware, various -d+acl logging additions and
846 more caution in buffer sizes.
848 PP/09 Implemented reverse_ip expansion operator.
850 PP/10 Bugzilla 937: provide a "debug" ACL control.
852 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
854 PP/12 Bugzilla 973: Implement --version.
856 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
858 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
860 PP/15 Bugzilla 816: support multiple condition rules on Routers.
862 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
863 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
864 ignore trailing whitespace.
866 JJ/02 prevent non-panic DKIM error from being sent to paniclog
868 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
871 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
872 Notification from Dr Andrew Aitchison.
874 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
875 ExtendedDetectionInfo response format.
876 Notification from John Horne.
878 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
881 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
882 XSL and documented dependency on system catalogs, with examples of how
885 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
888 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
889 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
890 configuration file which is writeable by the Exim user or group.
892 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
893 of configuration files to cover files specified with the -C option if
894 they are going to be used with root privileges, not just the default
897 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
898 option (effectively making it always true).
900 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
901 files to be used while preserving root privileges.
903 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
904 that rogue child processes cannot use them.
906 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
907 run-time user, instead of root.
909 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
910 Exim run-time user without dropping privileges.
912 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
913 result string, instead of calling string_vformat() twice with the same
916 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
917 for other users. Others should always drop root privileges if they use
918 -C on the command line, even for a whitelisted configure file.
920 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
922 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
928 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
929 $data_path, and $header_path variables; fixed documentation bugs and
932 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
933 exipick to access non-standard spools, including the "frozen" queue
936 NM/01 Bugzilla 965: Support mysql stored procedures.
937 Patch from Alain Williams
939 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
941 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
942 Patch from Andreas Metzler
944 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
945 Patch from Kirill Miazine
947 NM/05 Bugzilla 671: Added umask to procmail example.
949 JJ/03 installed exipick 20100323.0, fixing doc bug
951 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
952 directory. Notification and patch from Dan Rosenberg.
954 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
956 TK/02 Improve log output when DKIM signing operation fails.
958 MH/01 Treat the transport option dkim_domain as a colon separated
959 list, not as a single string, and sign the message with each element,
960 omitting multiple occurences of the same signer.
962 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
963 Bugzilla 985, 986. Patch by Simon Arlott
965 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
966 Patch by Simon Arlott
968 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
969 MBX locking. Notification from Dan Rosenberg.
975 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
977 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
979 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
981 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
983 NM/04 Bugzilla 915: Flag broken perl installation during build.
989 TK/01 Added patch by Johannes Berg that expands the main option
990 "spamd_address" if it starts with a dollar sign.
992 TK/02 Write list of recipients to X-Envelope-Sender header when building
993 the mbox-format spool file for content scanning (suggested by Jakob
996 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
997 (http://www.dcc-servers.net/) support via dccifd. Activated by
998 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1000 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1001 by Mark Daniel Reidel <mr@df.eu>.
1003 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1004 When building exim an external PCRE library is now needed -
1005 PCRE is a system library on the majority of modern systems.
1006 See entry on PCRE_LIBS in EDITME file.
1008 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1009 conversation. Added nologin parameter to request.
1010 Patch contributed by Kirill Miazine.
1012 TF/01 Do not log submission mode rewrites if they do not change the address.
1014 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1016 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1017 log files in place. Contributed by Roberto Lima.
1019 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1021 TF/03 Bugzilla 615: When checking the local_parts router precondition
1022 after a local_part_suffix or local_part_prefix option, Exim now
1023 does not use the address's named list lookup cache, since this
1024 contains cached lookups for the whole local part.
1026 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1027 Robert Millan. Documentation is in experimental-spec.txt.
1029 TF/04 Bugzilla 668: Fix parallel build (make -j).
1031 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1033 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1034 Patch provided by Jan Srzednicki.
1036 TF/05 Leading white space used to be stripped from $spam_report which
1037 wrecked the formatting. Now it is preserved.
1039 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1040 that they are available at delivery time.
1042 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1044 TF/08 TLS error reporting now respects the incoming_interface and
1045 incoming_port log selectors.
1047 TF/09 Produce a more useful error message if an SMTP transport's hosts
1048 setting expands to an empty string.
1050 NM/06 Bugzilla 744: EXPN did not work under TLS.
1051 Patch provided by Phil Pennock.
1053 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1054 Patch provided by Richard Godbee.
1056 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1057 acl_smtp_notquit, added index entry.
1059 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1060 Patch provided by Eugene Bujak.
1062 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1063 accept(). Patch provided by Maxim Dounin.
1065 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1066 Patch provided by Phil Pennock.
1068 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1070 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1071 Patch provided by Brad "anomie" Jorsch.
1073 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1074 Patch provided by Dean Brooks.
1076 TK/05 Add native DKIM support (does not depend on external libraries).
1078 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1079 Patch provided by Graeme Fowler.
1081 NM/16 Bugzilla 851: Documentation example syntax fix.
1083 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1085 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1088 NM/19 Bugzilla 745: TLS version reporting.
1089 Patch provided by Phil Pennock.
1091 NM/20 Bugzilla 167: bool: condition support.
1092 Patch provided by Phil Pennock.
1094 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1095 clients. Patch provided by Phil Pennock.
1097 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1098 Patch provided by Brad "anomie" Jorsch.
1100 NM/23 Bugzilla 687: Fix misparses in eximstats.
1101 Patch provided by Heiko Schlittermann.
1103 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1104 Patch provided by Heiko Schlittermann.
1106 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1107 plus update to original patch.
1109 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1111 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1112 Patch provided by David Brownlee.
1114 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1116 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1118 NM/30 Bugzilla 888: TLS documentation bugfixes.
1120 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1122 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1123 Unlike the original bugzilla I have changed all shell scripts in src tree.
1125 NM/33 Bugzilla 898: Transport filter timeout fix.
1126 Patch by Todd Rinaldo.
1128 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1129 Patch by Serge Demonchaux.
1131 NM/35 Bugzilla 39: Base64 decode bug fixes.
1132 Patch by Jakob Hirsch.
1134 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1136 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1138 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1140 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1146 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1147 ALT-N's libdkim that I have put here:
1148 http://duncanthrax.net/exim-experimental/
1150 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1151 were called 'true' and it seems that C99 defines that as a reserved
1152 keyword to be used with 'bool' variable types. That means you could
1153 not include C99-style headers which use bools without triggering
1154 build errors in sieve.c.
1156 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1157 as mailq or other aliases. Changed the --help handling significantly
1158 to do whats expected. exim_usage() emits usage/help information.
1160 SC/01 Added the -bylocaldomain option to eximstats.
1162 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1164 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1166 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1172 PH/01 Another patch from the Sieve maintainer.
1174 PH/02 When an IPv6 address is converted to a string for single-key lookup
1175 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1176 dots are used instead of colons so that keys in lsearch files need not
1177 contain colons. This was done some time before quoting was made available
1178 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1179 (notated using the quote facility) so as to distinguish them from IPv4
1180 keys. This meant that lookups for IP addresses in host lists did not work
1181 for iplsearch lookups.
1183 This has been fixed by arranging for IPv6 addresses to be expressed with
1184 colons if the lookup type is iplsearch. This is not incompatible, because
1185 previously such lookups could never work.
1187 The situation is now rather anomolous, since one *can* have colons in
1188 ordinary lsearch keys. However, making the change in all cases is
1189 incompatible and would probably break a number of configurations.
1191 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1194 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1195 conversion specification without a maximum field width, thereby enabling
1196 a rogue spamd server to cause a buffer overflow. While nobody in their
1197 right mind would setup Exim to query an untrusted spamd server, an
1198 attacker that gains access to a server running spamd could potentially
1199 exploit this vulnerability to run arbitrary code as the Exim user.
1201 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1202 $primary_hostname instead of what libspf2 thinks the hosts name is.
1204 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1205 a directory entry by the name of the lookup key. Previously, if a
1206 symlink pointed to a non-existing file or a file in a directory that
1207 Exim lacked permissions to read, a lookup for a key matching that
1208 symlink would fail. Now it is enough that a matching directory entry
1209 exists, symlink or not. (Bugzilla 503.)
1211 PH/03 The body_linecount and body_zerocount variables are now exported in the
1214 PH/04 Added the $dnslist_matched variable.
1216 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1217 This means they are set thereafter only if the connection becomes
1220 PH/06 Added the client_condition to authenticators so that some can be skipped
1221 by clients under certain conditions.
1223 PH/07 The error message for a badly-placed control=no_multiline_responses left
1224 "_responses" off the end of the name.
1226 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1228 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1229 (without spaces) instead of just copying the configuration text.
1231 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1233 PH/11 Added $max_received_linelength.
1235 PH/12 Added +ignore_defer and +include_defer to host lists.
1237 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1238 way in which PCRE > 7.0 is built.
1240 PH/14 Implemented queue_only_load_latch.
1242 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1243 MAIL command. The effect was to mangle the value on 64-bit systems.
1245 PH/16 Another patch from the Sieve maintainer.
1247 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1249 PH/18 If a system quota error occurred while trying to create the file for
1250 a maildir delivery, the message "Mailbox is full" was not appended to the
1251 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1252 only to a quota excession during the actual writing of the file.
1254 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1255 characters?) which causes problems in log lines. The DN values are now
1256 passed through string_printing() before being added to log lines.
1258 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1259 and InterBase are left for another time.)
1261 PH/21 Added message_body_newlines option.
1263 PH/22 Guard against possible overflow in moan_check_errorcopy().
1265 PH/23 POSIX allows open() to be a macro; guard against that.
1267 PH/24 If the recipient of an error message contained an @ in the local part
1268 (suitably quoted, of course), incorrect values were put in $domain and
1269 $local_part during the evaluation of errors_copy.
1275 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1276 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1279 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1280 issue a MAIL command.
1282 PH/02 In an ACL statement such as
1284 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1286 if a client was not listed at all, or was listed with a value other than
1287 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1288 the condition was not true (as it should be), so access was not denied.
1289 The bug was that the ! inversion was incorrectly passed on to the second
1290 item. This has been fixed.
1292 PH/03 Added additional dnslists conditions == and =& which are different from
1293 = and & when the dns lookup returns more than one IP address.
1295 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1296 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1298 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1299 FSYNC, which compiles an option called disable_fsync that allows for
1300 bypassing fsync(). The documentation is heavily laced with warnings.
1302 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1304 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1305 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1306 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1307 including adding "make clean"; (3) Added -fPIC when compiling the test
1308 dynamically loaded module, to get rid of a warning.
1310 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1311 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1312 = 0s. The bug is otherwise harmless.
1314 PH/07 There was a bug in the dovecot authenticator such that the value of
1315 $auth1 could be overwritten, and so not correctly preserved, after a
1316 successful authentication. This usually meant that the value preserved by
1317 the server_setid option was incorrect.
1319 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1321 PH/09 Installed PCRE release 7.0.
1323 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1324 run for batched SMTP input. It is now run at the start of every message
1325 in the batch. While fixing this I discovered that the process information
1326 (output by running exiwhat) was not always getting set for -bs and -bS
1327 input. This is fixed, and it now also says "batched" for BSMTP.
1329 PH/11 Added control=no_pipelining.
1331 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1332 patch, slightly modified), and move the expansion of helo_data till after
1333 the connection is made in the smtp transport (so it can use these
1336 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1338 PH/14 Added log_selector = +pid.
1340 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1342 PH/16 Add ${if forany and ${if forall.
1344 PH/17 Added dsn_from option to vary the From: line in DSNs.
1346 PH/18 Flush SMTP output before performing a callout, unless control =
1347 no_callout_flush is set.
1349 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1350 was true (the default) a successful delivery failed to delete the retry
1351 item, thus causing premature timeout of the address. The bug is now
1354 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1356 PH/21 Long custom messages for fakedefer and fakereject are now split up
1357 into multiline reponses in the same way that messages for "deny" and
1358 other ACL rejections are.
1360 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1361 with slight modification.
1363 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1364 draft, changing the syntax and factoring some duplicate code".
1366 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1367 for deliveries of the second and subsequent messages over the same SMTP
1370 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1371 ${reduce, with only minor "tidies".
1373 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1375 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1376 expansion side effects.
1378 PH/27 When a message times out after an over-quota error from an Exim-imposed
1379 quota, the bounce message says "mailbox is full". This message was not
1380 being given when it was a system quota that was exceeded. It now should
1383 MH/03 Made $recipients available in local_scan(). local_scan() already has
1384 better access to the recipient list through recipients_list[], but
1385 $recipients can be useful in postmaster-provided expansion strings.
1387 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1388 in the case of a MAIL command with additional options following the
1389 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1390 were accidentally chopped off.
1392 PH/29 SMTP synchronization checks are implemented when a command is read -
1393 there is a check that no more input is waiting when there shouldn't be
1394 any. However, for some commands, a delay in an ACL can mean that it is
1395 some time before the response is written. In this time, more input might
1396 arrive, invalidly. So now there are extra checks after an ACL has run for
1397 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1398 pipelining has not been advertised.
1400 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1402 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1403 returned in the SMTP response when smtp_return_error_details was false.
1404 This has been fixed.
1406 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1407 instead of the C I/O that was originally supplied, because problems were
1408 reported on Solaris.
1410 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1411 Exim which did not show up earlier: it was assuming that a call to
1412 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1413 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1414 was a macro that became an assignment, so it seemed to work. This has
1415 changed to a proper function call with a void return, hence the compile
1416 error. Exim's code has been fixed.
1418 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1421 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1423 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1425 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1426 because it turns out to be too big or there is a timeout) there is no
1427 "Completed" line in the log. When some messages of this type were
1428 selected by exigrep, they were listed as "not completed". Others were
1429 picked up by some special patterns. I have improved the selection
1430 criteria to be more general.
1432 PH/38 The host_find_failed option in the manualroute router can now be set
1433 to "ignore", to completely ignore a host whose IP address cannot be
1434 found. If all hosts are ignored, the behaviour is controlled by the new
1435 host_all_ignored option.
1437 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1438 homing or because of multiple MX records with /mx) generated more than
1439 one IP address, and the following item turned out to be the local host,
1440 all the secondary addresses of the first item were incorrectly removed
1441 from the list, along with the local host and any following hosts (which
1442 is what is supposed to happen).
1444 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1445 whoever called Exim into the -H file. In the case of the daemon it was
1446 behaving confusingly. When first started, it used values for whoever
1447 started the daemon, but after a SIGHUP it used the Exim user (because it
1448 calls itself on a restart). I have changed the code so that it now always
1451 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1452 message are rejected with the same error (e.g. no authentication or bad
1453 sender address), and a DATA command is nevertheless sent (as can happen
1454 with PIPELINING or a stupid MUA), the error message that was given to the
1455 RCPT commands is included in the rejection of the DATA command. This is
1456 intended to be helpful for MUAs that show only the final error to their
1459 PH/42 Another patch from the Sieve maintainer.
1461 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1462 Eximstats - Fixed some broken HTML links and added missing column headers
1464 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1465 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1467 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1469 PH/43 Yet another patch from the Sieve maintainer.
1471 PH/44 I found a way to check for a TCP/IP connection going away before sending
1472 the response to the final '.' that terminates a message, but only in the
1473 case where the client has not sent further data following the '.'
1474 (unfortunately, this is allowed). However, in many cases there won't be
1475 any further data because there won't be any more messages to send. A call
1476 to select() can be used: if it shows that the input is "ready", there is
1477 either input waiting, or the socket has been closed. An attempt to read
1478 the next input character can distinguish the two cases. Previously, Exim
1479 would have sent an OK response which the client would never have see.
1480 This could lead to message repetition. This fix should cure that, at
1481 least in a lot of common cases.
1483 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1484 advertised in response to EHLO.
1490 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1491 fixed by 4.65/MH/01 (is this a record?) are fixed:
1493 (i) An empty string was always treated as zero by the numeric comparison
1494 operators. This behaviour has been restored.
1496 (ii) It is documented that the numeric comparison operators always treat
1497 their arguments as decimal numbers. This was broken in that numbers
1498 starting with 0 were being interpreted as octal.
1500 While fixing these problems I realized that there was another issue that
1501 hadn't been noticed. Values of message_size_limit (both the global option
1502 and the transport option) were treated as octal if they started with 0.
1503 The documentation was vague. These values are now always treated as
1504 decimal, and I will make that clear in the documentation.
1510 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1511 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1514 MH/01 Don't check that the operands of numeric comparison operators are
1515 integers when their expansion is in "skipping" mode (fixes bug
1516 introduced by 4.64-PH/07).
1518 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1519 child addresses, Exim now panics and dies. Previously, because the count
1520 is held in a short int, deliveries were likely to be lost. As such a
1521 large number of recipients for a single message is ridiculous
1522 (performance will be very, very poor), I have chosen to impose a limit
1523 rather than extend the field.
1529 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1530 leftover -K file (the existence of which was triggered by #402).
1531 While we were at it, introduced process PID as part of the -K
1532 filename. This should rule out race conditions when creating
1535 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1536 processing considerably. Previous code took too long for large mails,
1537 triggering a timeout which in turn triggers #401.
1539 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1540 in the DK code in transports.c. sendfile() is not really portable,
1541 hence the _LINUX specificness.
1543 TF/01 In the add_headers option to the mail command in an Exim filter,
1544 there was a bug that Exim would claim a syntax error in any
1545 header after the first one which had an odd number of characters
1548 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1549 callout verification, Exim cached a "reject" for the entire domain. This
1550 is correct for most verifications, but it is not correct for a recipient
1551 verification with use_sender or use_postmaster set, because in that case
1552 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1553 case of MAIL FROM:<> rejection from other early rejections (e.g.
1554 rejection of HELO). When verifying a recipient using a non-null MAIL
1555 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1556 Whatever the result of the callout, the value of the domain cache is
1557 left unchanged (for any other kind of callout, getting as far as trying
1558 RCPT means that the domain itself is ok).
1560 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1563 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1564 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1565 session, the daemon ignores ECONNECT errors and logs others; it now
1566 ignores EPIPE as well.
1568 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1569 (quoted-printable decoding).
1571 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1572 later the small subsequent patch to fix an introduced bug.
1574 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1576 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1578 PH/08 An error is now given if message_size_limit is specified negative.
1580 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1581 to be given (somewhat) arbitrary names.
1583 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1586 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1587 miscellaneous code fixes
1589 PH/10 Added the log_reject_target ACL modifier to specify where to log
1592 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1593 hostname. This is wrong, because it relates to the incoming message (and
1594 probably the interface on which it is arriving) and not to the outgoing
1595 callout (which could be using a different interface). This has been
1596 changed to use the value of the helo_data option from the smtp transport
1597 instead - this is what is used when a message is actually being sent. If
1598 there is no remote transport (possible with a router that sets up host
1599 addresses), $smtp_active_hostname is used.
1601 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1602 tweaks were necessary in order to get it to work (see also 21 below):
1603 (a) The code assumed that strncpy() returns a negative number on buffer
1604 overflow, which isn't the case. Replaced with Exim's string_format()
1606 (b) There were several signed/unsigned issues. I just did the minimum
1607 hacking in of casts. There is scope for a larger refactoring.
1608 (c) The code used strcasecmp() which is not a standard C function.
1609 Replaced with Exim's strcmpic() function.
1610 (d) The code set only $1; it now sets $auth1 as well.
1611 (e) A simple test gave the error "authentication client didn't specify
1612 service in request". It would seem that Dovecot has changed its
1613 interface. Fortunately there's a specification; I followed it and
1614 changed what the client sends and it appears to be working now.
1616 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1619 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1620 address A is aliased to B and C, where B exists and C does not. Without
1621 -v the output is "A verified" because verification stops after a
1622 successful redirection if more than one address is generated. However,
1623 with -v the child addresses are also verified. Exim was outputting "A
1624 failed to verify" and then showing the successful verification for C,
1625 with its parentage. It now outputs "B failed to verify", showing B's
1626 parentage before showing the successful verification of C.
1628 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1629 look up a TXT record in a specific list after matching in a combined
1632 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1633 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1634 they consult the DNS. I had assumed they would set it the way they
1635 wanted; and indeed my experiments on Linux seem to show that in some
1636 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1637 To be on the safe side, however, I have now made the interface to
1638 host_find_byname() similar to host_find_bydns(), with an argument
1639 containing the DNS resolver options. The host_find_byname() function now
1640 sets these options at its start, just as host_find_bydns() does. The smtp
1641 transport options dns_qualify_single and dns_search_parents are passed to
1642 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1643 of host_find_byname() use the default settings of RES_DEFNAMES
1644 (qualify_single) but not RES_DNSRCH (search_parents).
1646 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1647 spool_read_header() do less string testing, by means of a preliminary
1648 switch on the second character of optional "-foo" lines. (This is
1649 overdue, caused by the large number of possibilities that now exist.
1650 Originally there were few.) While I was there, I also converted the
1651 str(n)cmp tests so they don't re-test the leading "-" and the first
1652 character, in the hope this might squeeze out yet more improvement.
1654 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1655 flag allowing group syntax was set by the header_syntax check but not
1656 turned off, possible causing trouble later; (2) The flag was not being
1657 set at all for the header_verify test, causing "group"-style headers to
1658 be rejected. I have now set it in this case, and also caused header_
1659 verify to ignore an empty address taken from a group. While doing this, I
1660 came across some other cases where the code for allowing group syntax
1661 while scanning a header line wasn't quite right (mostly, not resetting
1662 the flag correctly in the right place). These bugs could have caused
1663 trouble for malformed header lines. I hope it is now all correct.
1665 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1666 with the "reply" argument non-NULL. The code, however (which originally
1667 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1668 but it didn't always do it. This confused somebody who was copying the
1669 code for some other use. I have removed all the tests.
1671 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1672 feature that was used to support insecure browsers during the U.S. crypto
1673 embargo. It requires special client support, and Exim is probably the
1674 only MTA that supported it -- and would never use it because real RSA is
1675 always available. This code has been removed, because it had the bad
1676 effect of slowing Exim down by computing (never used) parameters for the
1677 RSA_EXPORT functionality.
1679 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1680 authenticator to fail if there's a tab character in the incoming data
1681 (there should never be unless someone is messing about, as it's supposed
1682 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1683 if the connection is using TLS or if the remote IP is the same as the
1684 local IP, and the "valid-client-cert option" if a client certificate has
1687 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1688 authenticators. This can be used for authorization after authentication
1689 succeeds. (In the case of plaintext, it servers for both authentication
1692 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1693 if any retry times were supplied.
1695 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1696 connection, where there is no client IP address to check. In this
1697 situation, the verify now always succeeds.
1699 PH/25 Applied John Jetmore's -Mset patch.
1701 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1703 PH/27 In a string expansion for a processed (not raw) header when multiple
1704 headers of the same name were present, leading whitespace was being
1705 removed from all of them, but trailing whitespace was being removed only
1706 from the last one. Now trailing whitespace is removed from each header
1707 before concatenation. Completely empty headers in a concatenation (as
1708 before) are ignored.
1710 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1711 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1713 PH/29 [Removed. This was a change that I later backed out, and forgot to
1714 correct the ChangeLog entry (that I had efficiently created) before
1715 committing the later change.]
1717 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1718 address errors (4xx response to RCPT) over the same connection as other
1719 messages routed to the same hosts. Such deliveries are always "forced",
1720 so retry times are not inspected. This resulted in far too many retries
1721 for the affected addresses. The effect occurred only when there were more
1722 hosts than the hosts_max_try setting in the smtp transport when it had
1723 the 4xx errors. Those hosts that it had tried were not added to the list
1724 of hosts for which the message was waiting, so if all were tried, there
1725 was no problem. Two fixes have been applied:
1727 (i) If there are any address or message errors in an SMTP delivery, none
1728 of the hosts (tried or untried) are now added to the list of hosts
1729 for which the message is waiting, so the message should not be a
1730 candidate for sending over the same connection that was used for a
1731 successful delivery of some other message. This seems entirely
1732 reasonable: after all the message is NOT "waiting for some host".
1733 This is so "obvious" that I'm not sure why it wasn't done
1734 previously. Hope I haven't missed anything, but it can't do any
1735 harm, as the worst effect is to miss an optimization.
1737 (ii) If, despite (i), such a delivery is accidentally attempted, the
1738 routing retry time is respected, so at least it doesn't keep
1739 hammering the server.
1741 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1742 in ${readsocket because some servers need this prod.
1744 PH/32 Added some extra debug output when updating a wait-xxx database.
1746 PH/33 The hint "could be header name not terminated by colon", which has been
1747 given for certain expansion errors for a long time, was not being given
1748 for the ${if def:h_colon_omitted{... case.
1750 PH/34 The spec says: "With one important exception, whenever a domain list is
1751 being scanned, $domain contains the subject domain." There was at least
1752 one case where this was not true.
1754 PH/35 The error "getsockname() failed: connection reset by peer" was being
1755 written to the panic log as well as the main log, but it isn't really
1756 panic-worthy as it just means the connection died rather early on. I have
1757 removed the panic log writing for the ECONNRESET error when getsockname()
1760 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1761 runs only) independently of the message's sender address. This meant
1762 that, if the 4xx error was in fact related to the sender, a different
1763 message to the same recipient with a different sender could confuse
1764 things. In particualar, this can happen when sending to a greylisting
1765 server, but other circumstances could also provoke similar problems.
1766 I have changed the default so that the retry time for these errors is now
1767 based a combination of the sender and recipient addresses. This change
1768 can be overridden by setting address_retry_include_sender=false in the
1771 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1772 remote server are returned as part of bounce messages. This was not
1773 happening for LMTP over a pipe (the lmtp transport), but now it is the
1774 same for both kinds of LMTP.
1776 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1777 in header lines that were in fact CNAMEs. This is no longer the case.
1779 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1780 and queue runs started by the daemon processed all messages. This has
1781 been fixed so that -R and -S can now usefully be given with -q<time>.
1783 PH/40 Import PCRE release 6.7 (fixes some bugs).
1785 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1787 PH/42 Give an error if -q is specified more than once.
1789 PH/43 Renamed the variables $interface_address and $interface_port as
1790 $received_ip_address and $received_port, to make it clear that these
1791 values apply to message reception, and not to the outgoing interface when
1792 a message is delivered. (The old names remain recognized, of course.)
1794 PH/44 There was no timeout on the connect() call when using a Unix domain
1795 socket in the ${readsocket expansion. There now is.
1797 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1798 be meaningful with "accept".
1800 SC/01 Eximstats V1.43
1801 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1803 SC/02 Eximstats V1.44
1804 Use a glob alias rather than an array ref in the generated
1805 parser. This improves both readability and performance.
1807 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1808 Collect SpamAssassin and rejection statistics.
1809 Don't display local sender or destination tables unless
1810 there is data to show.
1811 Added average volumes into the top table text output.
1813 SC/04 Eximstats V1.46
1814 Collect data on the number of addresses (recipients)
1815 as well as the number of messages.
1817 SC/05 Eximstats V1.47
1818 Added 'Message too big' to the list of mail rejection
1819 reasons (thanks to Marco Gaiarin).
1821 SC/06 Eximstats V1.48
1822 Mainlog lines which have GMT offsets and are too short to
1823 have a flag are now skipped.
1825 SC/07 Eximstats V1.49 (Alain Williams)
1826 Added the -emptyok flag.
1828 SC/08 Eximstats V1.50
1829 Fixes for obtaining the IP address from reject messages.
1831 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1832 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1833 whitesspace changes from 4.64-PH/27
1835 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1838 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1839 are found, allow negative numbers in numeric criteria)
1841 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1843 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1844 to match changes made in 4.64-PH/43
1846 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1848 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1849 "domain" is actually an IP address, and force a failure. This locks out
1850 those revolvers/nameservers that support "A-for-A" lookups, in
1851 contravention of the specifications.
1853 PH/48 When a host name was looked up from an IP address, and the subsequent
1854 forward lookup of the name timed out, the host name was left in
1855 $sender_host_name, contrary to the specification.
1857 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1858 restricted to single-key lookups, Exim was not diagnosing an error if
1859 * or *@ was used with a query-style lookup.
1861 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1863 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1864 long ago, but noone interested enough thought of it. Let's just say that
1865 the "1.1" means that there are some new functions that weren't there at
1866 some point in the past.
1868 PH/51 Error processing for expansion failure of helo_data from an smtp
1869 transport during callout processing was broken.
1871 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1872 tested/used via the -bh/-bhc/-bs options.
1874 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1875 bug, fixed in subsequent PCRE releases).
1877 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1878 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1880 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1886 SC/01 Use a glob alias rather than an array ref in eximstats generated
1887 parser. This improves both readability and performance.
1889 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1890 Don't display local sender or destination tables in eximstats unless
1891 there is data to show.
1892 Added average volumes into the eximstats top table text output.
1894 SC/03 Collect data on the number of addresses (recipients) as well
1895 as the number of messages in eximstats.
1897 TF/01 Correct an error in the documentation for the redirect router. Exim
1898 does (usually) call initgroups() when daemonizing.
1900 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1901 with consistent privilege compared to when running as a daemon.
1903 TF/03 Note in the spec that $authenticated_id is not set for local
1904 submissions from trusted users.
1906 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1907 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1909 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1910 by adding some example configuration directives to the default
1911 configuration file. A little bit of work is required to uncomment the
1912 directives and define how usernames and passwords are checked, but
1913 there is now a framework to start from.
1915 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1916 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1917 without this. I don't know how relevant this is to other LDAP libraries.
1919 PH/02 Add the verb name to the "unknown ACL verb" error.
1921 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1923 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1925 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1926 directory not expanded when it should be if an expanded home directory
1927 was set for the address (which is overridden by the transport).
1929 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1932 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1933 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1934 because it is too late at that time, and has no effect.
1936 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1937 security issue with \' (bugzilla #107). I could not use the
1938 PQescapeStringConn() function, because it needs a PGconn value as one of
1941 PH/08 When testing addresses using -bt, indicate those final addresses that
1942 are duplicates that would not cause an additional delivery. At least one
1943 person was confused, thinking that -bt output corresponded to deliveries.
1944 (Suppressing duplicates isn't a good idea as you lose the information
1945 about possibly different redirections that led to the duplicates.)
1947 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1948 systems where poll() doesn't work, in particular OS X.
1950 PH/10 Added more information to debugging output for retry time not reached.
1952 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1953 operations in malware.c.
1955 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1958 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1959 syslog_duplication set false, log lines that would normally be written
1960 both the the main log and to the reject log were not written to syslog at
1963 PH/14 In the default configuration, change the use of "message" in ACL warn
1964 statements to "add_header".
1966 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1967 not followed by a command (e.g. "seen endif").
1969 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1970 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1973 PH/17 Added extra conditions to the default value of delay_warning_condition
1977 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1978 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1979 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1982 The Auto-Submitted: and various List- headers are standardised, whereas I
1983 don't think Precedence: ever was.
1985 PH/18 Refactored debugging code in route_finduser() to show more information,
1986 in particular, the error code if getpwnam() issues one.
1988 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1989 This is apparently needed in addition to the PH/07 change above to avoid
1990 any possible encoding problems.
1992 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1993 but not after initializing Perl.
1995 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1996 output them only if debugging. By default they are written stderr,
1997 apparently, which is not desirable.
1999 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2002 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2005 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2007 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2008 authenticated or an ident call has been made. Suppress the default
2009 values for $authenticated_id and $authenticated_sender (but permit -oMai
2010 and -oMas) when testing with -bh.
2012 PH/24 Re-jigged the order of the tests in the default configuration so that the
2013 tests for valid domains and recipients precede the DNS black list and CSA
2014 tests, on the grounds that those ones are more expensive.
2016 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2017 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2018 This bug exists in every version of Exim that I still have, right back to
2021 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2022 However, an attempt to turn on case-sensitivity in a regex key by
2023 including (?-i) didn't work because the subject string was already
2024 lowercased, and the effects were non-intuitive. It turns out that a
2025 one-line patch can be used to allow (?-i) to work as expected.
2031 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2032 other effects) broke the use of negated acl sub-conditions.
2034 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2037 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2038 "Deny" causes Exim to reject the incoming connection with a 554 error.
2039 Unfortunately, if there is a major crisis, such as a disk failure,
2040 tcp-wrappers gives "deny", whereas what one would like would be some
2041 kind of temporary error. A kludge has been added to help with this.
2042 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2043 554 error is used if errno is still zero or contains ENOENT (which occurs
2044 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2047 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2049 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2050 errors. Otherwise a message that provokes a temporary error (when other
2051 messages do not) can cause a whole host to time out.
2053 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2054 addresses were routed directly to files or pipes from a redirect router.
2055 File deliveries just didn't batch; pipe deliveries might have suffered
2058 PH/06 A failure to get a lock for a hints database would erroneously always say
2059 "Failed to get write lock", even when it was really a read lock.
2061 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2062 of 0600. This has been changed to use the value of the lockfile_mode
2063 option (which defaults to 0600).
2065 PH/08 Applied small patch from the Sieve maintainer.
2067 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2068 folder from quota calculations, a direct delivery into this folder messed
2069 up the contents of the maildirsize file. This was because the regex was
2070 used only to exclude .Trash (or whatever) when the size of the mailbox
2071 was calculated. There was no check that a delivery was happening into an
2072 excluded directory. This bug has been fixed by ignoring all quota
2073 processing for deliveries into excluded directories.
2075 PH/10 Added the maildirfolder_create_regex option to appendfile.
2081 PH/01 The code for finding all the local interface addresses on a FreeBSD
2082 system running IPv6 was broken. This may well have applied to all BSD
2083 systems, as well as to others that have similar system calls. The broken
2084 code found IPv4 interfaces correctly, but gave incorrect values for the
2085 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2086 that it would not match correctly against @[] and not recognize the IPv6
2089 PH/02 The ipliteral router was not recognizing addresses of the form user@
2090 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2092 PH/03 Added disable_ipv6.
2094 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2095 decoded form, because it is most often used to construct To: headers
2096 lines in autoreplies, and the decoded form may well be syntactically
2097 invalid. However, $reply_address has leading white space removed, and all
2098 newlines turned into spaces so that the autoreply transport does not
2101 PH/05 If group was specified without a user on a router, and no group or user
2102 was specified on a transport, the group from the router was ignored.
2104 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2105 for visible compile-time settings that can be used to change these
2106 numbers, for those that want even more. Backwards compatibility with old
2107 spool files has been maintained. However, going back to a previous Exim
2108 release will lost any variables that are in spool files.
2110 PH/07 Two small changes when running in the test harness: increase delay when
2111 passing a TCP/IP connection to a new process, in case the original
2112 process has to generate a bounce, and remove special handling of
2113 127.0.0.2 (sic), which is no longer necessary.
2115 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2116 be the same on different OS.
2118 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2121 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2122 whether --show-vars was specified or not
2124 JJ/02 exipick: Added support for new ACL variable spool format introduced
2127 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2128 syntactically invalid From: or Reply-to: line, and a filter used this to
2129 generate an autoreply, and therefore failed to obtain an address for the
2130 autoreply, Exim could try to deliver to a non-existent relative file
2131 name, causing unrelated and misleading errors. What now happens is that
2132 it logs this as a hard delivery error, but does not attempt to create a
2135 PH/11 The exinext utility has a -C option for testing purposes, but although
2136 the given file was scanned by exinext itself; it wasn't being passed on
2137 when Exim was called.
2139 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2140 an end-of-file indication when reading a command response.
2142 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2143 compiled. In many other places in Exim, IPv6 addresses are always
2144 recognized, so I have changed this. It also means that IPv4 domain
2145 literals of the form [IPV4:n.n.n.n] are now always recognized.
2147 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2148 used if the router is not running as root, for example, when verifying at
2149 ACL time, or when using -bh. The debugging output from this situation was
2150 non-existent - all you got was a failure to exec. I have made two
2153 (a) Failures to set uid/gid, the current directory, or a process leader
2154 in a subprocess such as that created by queryprogram now generate
2155 suitable debugging ouput when -d is set.
2157 (b) The queryprogram router detects when it is not running as root,
2158 outputs suitable debugging information if -d is set, and then runs
2159 the subprocess without attempting to change uid/gid.
2161 PH/15 Minor change to Makefile for building test_host (undocumented testing
2164 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2165 additional section of a DNS packet that returns MX or SRV records.
2166 Instead, it always explicitly searches for A/AAAA records. This avoids
2167 major problems that occur when a DNS server includes only records of one
2168 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2169 fixed another bug: if SRV records were looked up and the corresponding
2170 address records were *not* found in the additional section, the port
2171 values from the SRV records were lost.
2173 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2174 using the correct key (the original address) when searching the retry
2175 rules in order to find which one to use for generating the retry hint.
2177 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2178 adding the default one. Similarly, if it contains a Reply-To: header, the
2179 errors_reply_to option, if set, is not used.
2181 PH/19 When calculating a retry time, Exim used to measure the "time since
2182 failure" by looking at the "first failed" field in the retry record. Now
2183 it does not use this if it is later than than the arrival time of the
2184 message. Instead it uses the arrival time. This makes for better
2185 behaviour in cases where some deliveries succeed, thus re-setting the
2186 "first failed" field. An example is a quota failure for a huge message
2187 when small messages continue to be delivered. Without this change, the
2188 "time since failure" will always be short, possible causing more frequent
2189 delivery attempts for the huge message than are intended.
2190 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2192 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2193 $1, $2, $3) because the numerical variables can be reset during some
2194 expansion items (e.g. "match"), thereby losing the authentication data.
2196 PH/21 Make -bV show the size of off_t variables so that the test suite can
2197 decide whether to run tests for quotas > 2G.
2199 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2200 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2201 filecount value is greater than 2G or if a quota value is greater than 2G
2202 on a system where the size of off_t is not greater than 4, a panic error
2205 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2206 never match. The debug and -bh output now contains an explicit error
2207 message indicating a malformed IPv4 address or mask.
2209 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2210 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2211 PH/23 above applies.
2213 PH/25 Do not write to syslog when running in the test harness. The only
2214 occasion when this arises is a failure to open the main or panic logs
2215 (for which there is an explicit test).
2217 PH/26 Added the /no_tell option to "control=freeze".
2219 PH/27 If a host name lookup failed very early in a connection, for example, if
2220 the IP address matched host_lookup and the reverse lookup yielded a name
2221 that did not have a forward lookup, an error message of the form "no IP
2222 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2223 could be logged. Now it outputs the IP address instead of "NULL".
2225 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2226 allows the sender and the authenticated sender to be set when
2227 submitting a message from within Exim. Since child_open_exim() is
2228 documented for local_scan(), the new function should be too.
2230 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2231 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2232 results in an empty string is now treated as unset.
2234 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2236 PH/31 Added sender_verify_fail logging option.
2238 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2239 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2240 tidied the source and removed it altogether.
2242 PH/33 When a queue run was abandoned because the load average was too high, a
2243 log line was always written; now it is written only if the queue_run log
2244 selector is set. In addition, the log line for abandonment now contains
2245 information about the queue run such as the pid. This is always present
2246 in "start" and "stop" lines but was omitted from the "abandon" line.
2248 PH/34 Omit spaces between a header name and the colon in the error message that
2249 is given when verify = headers_syntax fails (if there are lots of them,
2250 the message gets confusing).
2252 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2253 names, as there are now some PTR records that contain slashes. This check
2254 is only to protect against broken name servers that fall over on strange
2255 characters, so the fact that it applies to all lookups doesn't matter.
2257 PH/36 Now that the new test suite is complete, we can remove some of the
2258 special code in Exim that was needed for the old test suite. For example,
2259 sorting DNS records because real resolvers return them in an arbitrary
2260 order. The new test suite's fake resolver always returns records in the
2263 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2264 bounces) except when queue_only is set, to avoid logging races between
2265 the different processes.
2267 PH/38 Panic-die if .include specifies a non-absolute path.
2269 PH/39 A tweak to the "H" retry rule from its user.
2271 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2272 a label. They prevented compilation on older perls.
2274 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2275 a warning to be raised on newish perls.
2277 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2278 on queue. Changes to match documented behaviour of showing count of
2279 messages matching specified criteria.
2281 PH/40 Changed the default ident timeout from 30s to 5s.
2283 PH/41 Added support for the use of login_cap features, on those BSD systems
2284 that have them, for controlling the resources used by pipe deliveries.
2286 PH/42 The content-scanning code uses fopen() to create files in which to put
2287 message data. Previously it was not paying any attention to the mode of
2288 the files. Exim runs with umask(0) because the rest of the code creates
2289 files with open(), and sets the required mode explicitly. Thus, these
2290 files were ending up world-writeable. This was not a big issue, because,
2291 being within the spool directory, they were not world-accessible. I have
2292 created a function called modefopen, which takes an additional mode
2293 argument. It sets umask(777), creates the file, chmods it to the required
2294 mode, then resets the umask. All the relevant calls to fopen() in the
2295 content scanning code have been changed to use this function.
2297 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2298 to 24 hours. This avoids potential overflow problems when processing G
2299 and H retry rules. I suspect nobody ever tinkers with this value.
2301 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2303 PH/45 When the plaintext authenticator is running as a client, the server's
2304 challenges are checked to ensure they are valid base64 strings. By
2305 default, the authentication attempt is cancelled if an invalid string is
2306 received. Setting client_ignore_invalid_base64 true ignores these errors.
2307 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2308 they are received. Thus, the responses can be made to depend on the
2309 challenges. If an invalid string is ignored, an empty string is placed in
2312 PH/46 Messages that are created by the autoreply transport now contains a
2313 References: header, in accordance with RFCs 2822 and 3834.
2315 PH/47 Added authenticated_sender_force to the smtp transport.
2317 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2319 PH/49 Installed latest patch from the Sieve maintainer.
2321 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2322 was also set, the appendfile transport was unnecessarily scanning a
2323 directory of message files (e.g. for maildir delivery) to find the count
2324 of files (along with the size), even though it did not need this
2325 information. It now does the scan only if it needs to find either the
2326 size of the count of files.
2328 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2330 PH/52 Two bugs concerned with error handling when the smtp transport is
2333 (i) Exim was not creating retry information for temporary errors given
2334 for individual recipients after the DATA command when the smtp transport
2335 was used in LMTP mode. This meant that they could be retried too
2336 frequently, and not timed out correctly.
2338 (ii) Exim was setting the flag that allows error details to be returned
2339 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2340 recipients that were returned after the DATA command.
2342 PH/53 This is related to PH/52, but is more general: for any failing address,
2343 when detailed error information was permitted to be returned to the
2344 sender, but the error was temporary, then after the final timeout, only
2345 "retry timeout exceeded" was returned. Now it returns the full error as
2346 well as "retry timeout exceeded".
2348 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2349 do this, and (what is worse) MTAs that accept it.
2351 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2352 will now be deprecated.
2354 PH/56 New os.c-cygwin from the Cygwin maintainer.
2356 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2357 formats (previously only available in exim formats via -bpr, -bpru,
2358 and -bpra. Now also available in native and exiqgrep formats)
2360 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2361 with very large, slow to parse queues
2363 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2365 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2367 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2368 responses to authentication challenges, though it was showing the
2369 challenges; (ii) I've removed the CR characters from the debug output for
2372 PH/58 Allow for the insertion of a newline as well as a space when a string
2373 is turned into more than one encoded-word during RFC 2047 encoding. The
2374 Sieve code now uses this.
2376 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2377 data_4xx, lost_connection, tls_required.
2379 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2380 message was being sent as an SMTP response.
2382 PH/61 Add -l and -k options to exicyclog.
2384 PH/62 When verifying, if an address was redirected to one new address, so that
2385 verification continued, and the new address failed or deferred after
2386 having set something in $address_data, the value of $address_data was not
2387 passed back to the ACL. This was different to the case when no
2388 redirection occurred. The value is now passed back in both cases.
2390 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2391 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2392 use login_cap.h, so on its own it isn't the distinguishing feature. The
2393 new name refers directly to the setclassresources() function.
2395 PH/65 Added configuration files for NetBSD3.
2397 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2399 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2400 is preferred over IPv4.
2402 PH/68 The bounce_return_message and bounce_return_body options were not being
2403 honoured for bounces generated during the reception of non-SMTP messages.
2404 In particular, this applied to messages rejected by the ACL. This bug has
2405 been fixed. However, if bounce_return_message is true and bounce_return_
2406 body is false, the headers that are returned for a non-SMTP message
2407 include only those that have been read before the error was detected.
2408 (In the case of an ACL rejection, they have all been read.)
2410 PH/69 The HTML version of the specification is now built in a directory called
2411 spec_html instead of spec.html, because the latter looks like a path with
2412 a MIME-type, and this confuses some software.
2414 PH/70 Catch two compiler warnings in sieve.c.
2416 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2417 function verify_get_ident() calls ip_connect() to connect a socket, but
2418 if the "connect()" function timed out, ip_connect() used to close the
2419 socket. However, verify_get_ident() also closes the socket later, and in
2420 between Exim writes to the log, which may get opened at this point. When
2421 the socket was closed in ip_connect(), the log could get the same file
2422 descriptor number as the socket. This naturally causes chaos. The fix is
2423 not to close the socket in ip_connect(); the socket should be closed by
2424 the function that creates it. There was only one place in the code where
2425 this was missing, in the iplookup router, which I don't think anybody now
2426 uses, but I've fixed it anyway.
2428 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2429 well as to direct DNS lookups. Otherwise the handling of names in host
2430 lists is inconsistent and therefore confusing.
2436 PH/01 Two changes to the default runtime configuration:
2438 (1) Move the checks for relay_from_hosts and authenticated clients from
2439 after to before the (commented out) DNS black list checks.
2441 (2) Add control=submission to the relay_from_hosts and authenticated
2442 clients checks, on the grounds that messages accepted by these
2443 statements are most likely to be submissions.
2445 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2447 (1) Generate an error if the third argument for the ${prvs expansion is
2450 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2453 (3) Reset the variables that are obtained from the first argument of
2454 ${prvscheck and used in the second argument before leaving the code,
2455 because their memory is reclaimed, so using them afterwards may do
2458 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2459 one (it's much easier than Tom thought :-).
2461 (5) Because of (4), we can now allow for the use of $prvscheck_result
2462 inside the third argument.
2464 PH/03 For some reason, the default setting of PATH when running a command from
2465 a pipe transport was just "/usr/bin". I have changed it to
2468 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2469 anything to be listed in the output from -bV.
2471 PH/05 When a filter generated an autoreply, the entire To: header line was
2472 quoted in the delivery log line, like this:
2474 => >A.N.Other <ano@some.domain> <original@ddress> ...
2476 This has been changed so that it extracts the operative address. There
2477 may be more than one such address. If so, they are comma-separated, like
2480 => >ano@some.domain,ona@other.domain <original@ddress> ...
2482 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2483 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2484 not being looked up in the rDNS to get a host name, Exim was showing the
2485 IP address twice in Received: lines, even though the IP addresses were
2486 identical. For example:
2488 Received: from [1.2.3.4] (helo=[1.2.3.4])
2490 However, if the real host name was known, it was omitting the HELO data
2491 if it matched the actual IP address. This has been tidied up so that it
2492 doesn't show the same IP address twice.
2494 PH/07 When both +timestamp and +memory debugging was on, the value given by
2495 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2496 called by the debug printing, thereby overwriting the timestamp buffer.
2497 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2499 PH/08 When the original message was included in an autoreply transport, it
2500 always said "this is a copy of the message, including all the headers",
2501 even if body_only or headers_only was set. It now gives an appropriate
2504 PH/09 Applied a patch from the Sieve maintainer which:
2506 o fixes some comments
2507 o adds the (disabled) notify extension core
2508 o adds some debug output for the result of if/elsif tests
2509 o points to the current vacation draft in the documentation
2510 and documents the missing references header update
2514 o fixes a bug in processing the envelope test (when testing
2515 multiple envelope elements, the last element determinted the
2518 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2519 Electronic Mail") by including:
2521 Auto-submitted: auto-generated
2523 in the messages that it generates (bounce messages and others, such as
2524 warnings). In the case of bounce messages for non-SMTP mesages, there was
2525 also a typo: it was using "Auto_submitted" (underscore instead of
2526 hyphen). Since every message generated by Exim is necessarily in response
2527 to another message, thes have all been changed to:
2529 Auto-Submitted: auto-replied
2531 in accordance with these statements in the RFC:
2533 The auto-replied keyword:
2535 - SHOULD be used on messages sent in direct response to another
2536 message by an automatic process,
2538 - MUST NOT be used on manually-generated messages,
2540 - MAY be used on Delivery Status Notifications (DSNs) and Message
2541 Disposition Notifications (MDNs),
2543 - MUST NOT be used on messages generated by automatic or periodic
2544 processes, except for messages which are automatic responses to
2547 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2548 to the default Received: header definition.
2550 PH/12 Added log selector acl_warn_skipped (default on).
2552 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2553 variables because (a) they are in the wrong storage pool and (b) even if
2554 they were copied, it wouldn't work properly because of the caching.
2556 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2557 checking when decoding. Apparently there are clients that generate
2558 overlong encoded strings. Why am I not surprised?
2560 PH/15 If the first argument of "${if match_address" was not empty, but did not
2561 contain an "@" character, Exim crashed. Now it writes a panic log message
2562 and treats the condition as false.
2564 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2566 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2567 extension "envelope-auth". The code is finished and in agreement with
2568 other implementations, but there is no documentation so far and in fact,
2569 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2570 not changing the active code.
2572 Print executed "if" and "elsif" statements when debugging is used. This
2573 helps a great deal to understand what a filter does.
2575 Document more things not specified clearly in RFC3028. I had all this
2576 sorted out, when out of a sudden new issues came to my mind. Oops."
2578 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2581 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2582 canonical form (as documented). However, after a host name lookup from
2583 the IP address, check_host() was doing a simple string comparison with
2584 addresses acquired from the DNS when checking that the found name did
2585 have the original IP as one of its addresses. Since any found IPv6
2586 addresses are likely to be in abbreviated form, the comparison could
2587 fail. Luckily, there already exists a function for doing the comparison
2588 by converting both addresses to binary, so now that is used instead of
2589 the text comparison.
2591 PH/20 There was another similar case to PH/19, when a complete host name was
2592 given in a host list; looking up its IP address could give an abbreviated
2593 form, whereas the current host's name might or might not be abbreviated.
2594 The same fix has been applied.
2600 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2601 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2604 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2605 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2607 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2609 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2610 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2611 Also, for "Auto-Submitted", treat anything other than "no" as
2612 non-personal, in accordance with RFC 3834. (Previously it treated
2613 anything starting "auto-" as non-personal.)
2615 TF/01 The control=submission/name=... option had a problem with syntax
2616 errors if the name included a slash character. The /name= option
2617 now slurps the rest of the string, so it can include any characters
2618 but it must come last in the list of options (after /sender_retain
2621 PH/05 Some modifications to the interface to the fake nameserver for the new
2629 TK/01 Added the "success_on_redirect" address verification option. See
2630 NewStuff for rationale and an example.
2632 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2634 PH/02 Patch to exigrep to allow it to work on syslog lines.
2636 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2637 fread() to skip over the body file's header line, because in Cygwin the
2638 header line is locked and is inaccessible.
2640 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2641 co-exist for some time) to make it clear that it is the Exim ID that is
2642 referenced, not the Message-ID: header line.
2644 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2645 string_format() function, because snprintf() does not exist on all
2648 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2649 ${stat: expansion item.
2651 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2652 protocol synchronization error", to keep the pedants happy.
2654 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2655 well as for IRIX systems, when gcc is being used. See the host.c source
2658 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2660 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2663 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2664 transport and to the smtp transport in LMTP mode.
2666 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2668 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2669 run a filter in a subprocess. This could lead to confusion in subsequent
2670 lookups in the parent process. There should also be a search_tidyup() at
2671 the end of the subprocess.
2673 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2674 only if the host matched helo_try_verify_hosts, which caused the
2675 verification to occur when the EHLO/HELO command was issued. The ACL just
2676 tested the remembered result. Now, if a previous verification attempt has
2677 not happened, "verify = helo" does it there and then.
2679 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2681 TK/03 Fix log output including CR from clamd.
2683 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2684 exist provoked a memory error which could cause a segfault.
2686 PH/15 Installed PCRE 6.2
2688 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2690 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2691 of the problem. Specifically, suggested +O2 rather than +O1 for the
2694 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2696 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2697 record, the port was not correctly logged when the outgoing_port log
2698 selector was set (it logged the transort's default port).
2700 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2701 fallback_hosts, and "hosts" in the smtp transport.
2703 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2704 host errors such as "Connection refused".
2706 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2707 authentication with radiusclient 0.4.9:
2709 - Error returned from rc_read_config was caught wrongly
2710 - Username/password not passed on to radius server due to wrong length.
2712 The presumption is that some radiusclient API changes for 4.51/PH/17
2713 were not taken care of correctly. The code is still untested by me (my
2714 Linux distribution still has 0.3.2 of radiusclient), but it was
2715 contributed by a Radius user.
2717 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2718 expanding the "port" option of the smtp transport.
2720 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2721 while reading a MIME header. Thanks to Tom Hughes for a patch.
2723 PH/24 Include config.h inside local_scan.h so that configuration settings are
2726 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2727 that in an ACL for RCPT (for example), you can examine exactly what was
2730 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2731 commands, but it was not correctly comparing the address with the actual
2732 client host address. Thus, it would show the EHLO address in Received:
2733 header lines when this was not necessary.
2735 PH/27 Added the % operator to ${eval:}.
2737 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2738 it should be ignoring failures (because with -C, for example, it has lost
2739 privilege). It wasn't ignoring creation failures other than "already
2742 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2745 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2746 because an input line was too long, either on its own, or by virtue of
2747 too many continuations, the temporary file was not being removed, and the
2748 return code was incorrect.
2750 PH/31 Missing "BOOL" in function definition in filtertest.c.
2752 PH/32 Applied Sieve patches from the maintainer.
2754 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2756 PH/33 Added "verify = not_blind".
2758 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2759 Local/Makefile (with some defaults set). These are used in built scripts
2760 such as exicyclog, but they have never been used in the exim_install
2761 script (though there are many overriding facilities there). I have
2762 arranged that the exim_install script now takes note of these two
2765 PH/35 Installed configuration files for Dragonfly.
2767 PH/36 When a locally submitted message by a trusted user did not contain a
2768 From: header, and the sender address was obtained from -f or from an SMTP
2769 MAIL command, and the trusted user did not use -F to supply a sender
2770 name, $originator_name was incorrectly used when constructing a From:
2771 header. Furthermore, $originator_name was used for submission mode
2772 messages from external hosts without From: headers in a similar way,
2773 which is clearly wrong.
2775 PH/37 Added control=suppress_local_fixups.
2777 PH/38 When log_selector = +received_sender was set, and the addition of the
2778 sender made the log line's construction buffer exactly full, or one byte
2779 less than full, an overflow happened when the terminating "\n" was
2782 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2783 when the result of a list match is failure because a DNS lookup failed.
2785 PH/40 RM_COMMAND is now used in the building process.
2787 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2788 the "build-* directories that it finds.
2790 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2791 address in a domain literal was a prefix of an interface address.
2793 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2794 when verifying a sender address, unless rewrite_headers is false.
2796 PH/44 Wrote a long comment about why errors_to addresses are verified as
2797 recipients, not senders.
2799 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2800 the ratelimit ACL was added.
2802 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2804 PH/46 Added extra information about PostgreSQL errors to the error string.
2806 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2807 avoiding the need to install special zones in a real server. This is
2808 backwards compatible; if it can't find the fake resolver, it drops back.
2809 Thus, both old and new test suites can be run.
2811 TF/02 Added util/ratelimit.pl
2813 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2814 clock is set back in time.
2816 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2817 Candler <B.Candler@pobox.com>.
2819 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2820 OK for addresses that are the result of redirections.
2822 PH/48 A number of further additions for the benefit of the new test suite,
2823 including a fake gethostbyname() that interfaces to the fake DNS resolver
2826 TF/06 The fix for widen_domains has also been applied to qualify_single and
2827 search_parents which are the other dnslookup options that can cause
2830 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2833 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2835 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2836 service" if it encountered an MX record with an empty target hostname.
2837 The message is now "an MX or SRV record indicated no SMTP service".
2839 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2840 if the DNS of the sending site is misconfigured. This is quite a
2841 common situation. This change restores the behaviour of treating a
2842 helo verification defer as a failure.
2844 PH/51 If self=fail was set on a router, the bounce message did not include the
2845 actual error message.
2851 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2853 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2854 command itself ended in error, the underlying message about the transport
2855 filter timeout was being overwritten with the pipe command error. Now the
2856 underlying error message should be appended to the second error message.
2858 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2860 PH/02 Reduce the amount of output that "make" produces by default. Full output
2861 can still be requested.
2863 PH/03 The warning log line about a condition test deferring for a "warn" verb
2864 was being output only once per connection, rather than after each
2865 occurrence (because it was using the same function as for successful
2866 "warn" verbs). This seems wrong, so I have changed it.
2868 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2869 it should not have, which might have caused a crash in the right
2870 circumstances, but probably never did.
2872 PH/04 Installed a modified version of Tony Finch's patch to make submission
2873 mode fix the return path as well as the Sender: header line, and to
2874 add a /name= option so that you can make the user's friendly name appear
2877 TF/03 Added the control = fakedefer ACL modifier.
2879 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2880 Mark Lowes for thorough testing.
2882 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2884 TK/03 Merged latest SRS patch from Miles Wilton.
2886 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2887 with the definition in sysexits.h (which is #included earlier).
2888 Fortunately, Exim does not actually use EX_OK. The code used to try to
2889 preserve the sysexits.h value, by assumimg that macro definitions were
2890 scanned for macro replacements. I have been disabused of this notion,
2891 so now the code just undefines EX_OK before #including unistd.h.
2893 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2894 in the smtp transport. When a block could not be written in a single
2895 write() function, the timeout was being re-applied to each part-write.
2896 This seems wrong - if the receiver was accepting one byte at a time it
2897 would take for ever. The timeout is now adjusted when this happens. It
2898 doesn't have to be particularly precise.
2900 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2901 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2903 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2904 without a domain if the check to <postmaster@domain> fails.
2906 SC/01 Eximstats: added -xls and the ability to specify output files
2907 (patch written by Frank Heydlauf).
2909 SC/02 Eximstats: use FileHandles for outputing results.
2911 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2913 SC/04 Eximstats: fixed display of large numbers with -nvr option
2915 SC/05 Eximstats: fixed merging of reports with empty tables.
2917 SC/06 Eximstats: added the -include_original_destination flag
2919 SC/07 Eximstats: removed tabs and trailing whitespace.
2921 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2923 TK/06 MBOX spool code: Add real "From " MBOX separator line
2924 so the .eml file is really in mbox format (even though
2925 most programs do not really care). Patch from Alex Miller.
2927 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2928 The latter is generated from $received_to and is only set if the
2929 message has one envelope recipient. SA can use these headers,
2930 obviously out-of-the-box. Patch from Alex Miller.
2932 PH/08 The ${def test on a variable was returning false if the variable's
2933 value was "0", contrary to what the specification has always said!
2934 The result should be true unless the variable is empty.
2936 PH/09 The syntax error of a character other than { following "${if
2937 def:variable_name" (after optional whitespace) was not being diagnosed.
2938 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2939 accidental colon was present, for example, could give incorrect results.
2941 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2942 result is used (not including appendfile, where other changes are about
2945 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2946 This involved changing a lot of size variables from int to off_t. It
2947 should work with maildirs and everything.
2949 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2950 spamd dying while we are connected to it.
2952 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2955 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2956 function for BDB 4.3.
2958 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2960 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2961 the sources for PCRE have been re-arranged and more files are now
2964 PH/15 The code I had for printing potentially long long variables in PH/11
2965 above was not the best (it lost precision). The length of off_t variables
2966 is now inspected at build time, and an appropriate printing format (%ld
2967 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2968 to be "long long int" or "long int". This is needed for the internal
2969 formatting function string_vformat().
2971 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2972 the configuration file to be ":syslog", then the script "guesses" where
2973 the logs files are, rather than using the compiled in default. In our
2974 case the guess is not the same as the compiled default, so the script
2975 suddenly stopped working when I started to use syslog. The patch checks
2976 to see if log_file_path is "". If so, it attempts to read it from exim
2977 with no configuration file to get the compiled in version, before it
2978 falls back to the previous guessing code."
2980 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2981 implementing BATV in an Exim configuration. See NewStuff for the gory
2984 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2985 Makefile that are specific to HP-UX.
2987 PH/18 If the "use_postmaster" option was set for a recipient callout together
2988 with the "random" option, the postmaster address was used as the MAIL
2989 FROM address for the random test, but not for the subsequent recipient
2990 test. It is now used for both.
2992 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2993 patch removes a few documentation additions to RFC 3028, because the
2994 latest draft now contains them. It adds the new en;ascii-case comparator
2995 and a new error check for 8bit text in MIME parts. Comparator and
2996 require names are now matched exactly. I enabled the subaddress
2997 extension, but it is not well tested yet (read: it works for me)."
2999 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3000 rework some of the code of TK/09 above to avoid the hardwired use of
3001 "%lld" and "long long". Replaced the call to snprintf() with a call to
3004 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3005 records point to non-existent hosts", "retry timeout exceeded", and
3006 "retry time not reached for any host after a long failure period".
3008 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3009 experimental DomainKeys support:
3011 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3012 (2) On an error such as an illegally used "control", the wrong name for
3013 the control was given.
3015 These problems did NOT occur unless DomainKeys support was compiled.
3017 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3019 PH/24 Added ${if match_ip condition.
3021 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3022 databases so that it will be absolutely obvious if a crash occurs in the
3023 DB library. This is a regular occurrence (often caused by mis-matched
3026 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3027 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3028 picked up on a user's system that detects such things. There doesn't seem
3029 to be a gcc warning option for this - only an attribute that has to be
3030 put on the function's prototype. It seems that in Fedora Core 4 they have
3031 set this on a number of new functions. No doubt there will be more in due
3034 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3035 specify a transport. However, if an address that was verified by such a
3036 router was the subject of a callout, Exim crashed because it tried to
3037 read the rcpt_include_affixes from the non-existent transport. Now it
3038 just assumes that the setting of that option is false. This bug was
3039 introduced by 4.51/PH/31.
3041 PH/28 Changed -d+all to exclude +memory, because that information is very
3042 rarely of interest, but it makes the output a lot bigger. People tend to
3043 do -d+all out of habit.
3045 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3046 code in os-type was giving problems when libc.so lives in lib64, like on
3049 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3050 aren't the modern standard, and it seems that some systems' include files
3051 don't always have them. Exim was already checking for some of the newer
3052 ones like T_AAAA, and defining it itself. I've added checks for all the
3053 record types that Exim uses.
3055 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3056 not automatically generating a new one, as it is supposed to. This
3057 prevented TLS from working. If the file did exist, but contained invalid
3058 data, a new version was generated, as expected. It was only the case of a
3059 non-existent file that was broken.
3061 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3062 with a change in libdomainkeys > 0.64.
3064 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3065 from DNS. If the selector record carries the flag, it now has
3066 precedence over the domain-wide flag.
3068 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3070 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3071 the use of a port name that isn't defined in /etc/services) occurred, the
3072 message was deferred as in a normal delivery, and thus remained on the
3073 spool, instead of being failed because of the mua_wrapper setting. This
3074 is now fixed, and I tidied up some of the mua_wrapper messages at the
3077 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3078 the messages in a hash of arrays rather than using individual hashes.
3079 This is a bit cleaner and results in dramatic memory savings, albeit
3080 at a slight CPU cost.
3082 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3083 as requested by Marc Sherman.
3085 SC/10 Eximstats: added histograms for user specified patterns as requested
3088 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3090 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3091 fopen() in the content-scanning modules that did not already have it.
3097 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3098 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3100 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3102 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3104 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3105 to test Sieve filters that use "vacation".
3107 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3108 that changes the way the GnuTLS parameters are stored in the cache file.
3109 The new format can be generated externally. For backward compatibility,
3110 if the data in the cache doesn't make sense, Exim assumes it has read an
3111 old-format file, and it generates new data and writes a new file. This
3112 means that you can't go back to an older release without removing the
3115 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3116 work if there are any delivery delays because "one_time" forces the
3117 parent to be marked "delivered", so its unseen clone is never tried
3118 again. For this reason, Exim now forbids the simultaneous setting of
3121 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3122 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3123 entry if you want to know the details. The fix, however, neglected to
3124 consider the case when local delivery batching is involved. The test for
3125 "previously delivered" was not happening when checking to see if an
3126 address could be batched with a previous (undelivered) one; under
3127 certain circumstances this could lead to multiple deliveries to the same
3130 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3131 in its include files, and this causes problems building Exim.
3133 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3134 header_syntax) but Exim was just ignoring anything given after a slash.
3135 In particular, this caused confusion with an attempt to use "verify =
3136 reverse_host_lookup/defer_ok". An error is now given when options are
3137 supplied for verify items that do not have them. (Maybe reverse_host_
3138 lookup should have a defer_ok option, but that's a different point.)
3140 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3141 defined by RFC 821) to 2048, because there were problems with some AUTH
3142 commands, and RFC 1869 says the size should be increased for extended
3143 SMTP commands that take arguments.
3145 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3148 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3149 "unknown" error; now it says that the functionality isn't in the binary.
3151 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3152 an address' error message when a string expansion fails (syntax or
3153 whatever). Otherwise the password may appear in the log. Following change
3154 PH/42 below, there is no longer a chance of it appearing in a bounce
3157 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3159 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3160 of its addresses was ever tried. (Bugzilla bug #2.)
3162 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3163 the result incorrectly in the debug output. (It correctly added a newline
3164 to what was transported.)
3166 TF/01 Added $received_time.
3168 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3169 commented out examples of how to interface to a virus scanner and to
3170 SpamAssassin. Also added commented examples of av_scanner and
3171 spamd_address settings.
3173 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3174 and controls are allowed in which ACLs. There were a couple of minor
3175 errors. Some of the entries in the conditions table (which is a table of
3176 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3177 negation of where the condition IS allowed.
3179 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3181 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3182 header file does not have a version number, so I've had to invent a new
3183 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3184 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3185 radiusclient), but it was contributed by a Radius user.
3187 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3188 files or directories, for OpenSSL.
3190 PH/19 When an Exim process that is running as root has to create an Exim log
3191 file, it does so in a subprocess that runs as exim:exim so as to get the
3192 ownership right at creation (otherwise, other Exim processes might see
3193 the file with the wrong ownership). There was no test for failure of this
3194 fork() call, which would lead to the process getting stuck as it waited
3195 for a non-existent subprocess. Forks do occasionally fail when resources
3196 run out. I reviewed all the other calls to fork(); they all seem to check
3199 PH/20 When checking for unexpected SMTP input at connect time (before writing
3200 the banner), Exim was not dealing correctly with a non-positive return
3201 from the read() function. If the client had disconnected by this time,
3202 the result was a log entry for a synchronization error with an empty
3203 string after "input=" when read() returned zero. If read() returned -1
3204 (an event I could not check), uninitialized data bytes were printed.
3205 There were reports of junk text (parts of files, etc) appearing after
3208 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3210 PH/22 Added support for macro redefinition, and (re)definition in between
3211 driver and ACL definitions.
3213 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3214 forgetting to use the resulting value; it was using the unexpanded value.
3216 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3217 hadn't been configured. The fix is from Juergen Kreileder, who
3218 understands it better than I do:
3220 "Here's what I see happening with three configured cyrus_sasl
3221 authenticators configured (plain, login, cram-md5):
3223 On startup auth_cyrus_sasl_init() gets called for each of these.
3224 This means three calls to sasl_listmech() without a specified mech_list.
3225 => SASL tests which mechs of all available mechs actually work
3226 => three warnings about OTP not working
3227 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3229 With the patch, sasl_listmech() also gets called three times. But now
3230 SASL's mech_list option is set to the server_mech specified in the the
3231 authenticator. Or in other words, the answer from sasl_listmech()
3232 gets limited to just the mech you're testing for (which is different
3234 => the return list contains just 'plain' or 'login', 'cram-md5' or
3235 nothing depending on the value of ob->server_mech.
3237 I've just tested the patch: Authentication still works fine,
3238 unavailable mechs specified in the exim configuration are still
3239 caught, and the auth.log warnings about OTP are gone."
3241 PH/25 When debugging is enabled, the contents of the command line are added
3242 to the debugging output, even when log_selector=+arguments is not
3245 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3246 answer is "GNU", and only if the return is "GNU/something" is the answer
3249 PH/27 $acl_verify_message is now set immediately after the failure of a
3250 verification in an ACL, and so is available in subsequent modifiers. In
3251 particular, the message can be preserved by coding like this:
3253 warn !verify = sender
3254 set acl_m0 = $acl_verify_message
3256 Previously, $acl_verify_message was set only while expanding "message"
3257 and "log_message" when a very denied access.
3259 PH/28 Modified OS/os.c-Linux with
3261 -#ifndef OS_LOAD_AVERAGE
3262 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3264 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3265 nomenclature these days.)
3267 PH/29 Installed patch from the Sieve maintainer that adds the options
3268 sieve_useraddress and sieve_subaddress to the redirect router.
3270 PH/30 In these circumstances:
3271 . Two addresses routed to the same list of hosts;
3272 . First host does not offer TLS;
3273 . First host accepts first address;
3274 . First host gives temporary error to second address;
3275 . Second host offers TLS and a TLS session is established;
3276 . Second host accepts second address.
3277 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3278 and peerdn, if requested) that were in fact used only for the second
3281 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3282 attention to any local part prefix or suffix that was matched by the
3283 router that accepted the address. It now behaves in the same way as it
3284 does for delivery: the affixes are removed from the local part unless
3285 rcpt_include_affixes is set on the transport.
3287 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3288 timeout during the DATA phase of an incoming message.
3290 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3291 applied a patch sanctioned by the Sieve maintainer.
3293 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3294 the uid or gid is negative. A case of a negative gid caused this to be
3295 noticed. The fix allows for either to be negative.
3297 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3298 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3301 PH/36 The change PH/12 above was broken. Fixed it.
3303 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3304 the grounds that routing the same address twice would always produce the
3305 same answer. This might have been true once, but it is certainly no
3306 longer true now. Routing a child address may depend on the previous
3307 routing that produced that child. Some complicated redirection strategies
3308 went wrong when messages had multiple recipients, and made Exim's
3309 behaviour dependent on the order in which the addresses were given.
3311 I have moved the duplicate checking until after the routing is complete.
3312 Exim scans the addresses that are assigned to local and remote
3313 transports, and removes any duplicates. This means that more work will be
3314 done, as duplicates will always all be routed, but duplicates are
3315 presumably rare, so I don't expect this is of any significance.
3317 For deliveries to pipes, files, and autoreplies, the duplicate checking
3318 still happens during the routing process, since they are not going to be
3321 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3322 It corrects a timeout issue with spamd. This is Ian's comment: "The
3323 background is that sometimes spamd either never reads data from a
3324 connection it has accepted, or it never writes response data. The exiscan
3325 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3326 blindly assumes that writes won't block so it may never time out."
3328 PH/39 Allow G after quota size as well as K and M.
3330 PH/40 The value set for $authenticated_id in an authenticator may not contain
3331 binary zeroes or newlines because the value is written to log lines and
3332 to spool files. There was no check on this. Now the value is run through
3333 the string_printing() function so that such characters are converted to
3334 printable escape sequences.
3336 PH/41 $message_linecount is a new variable that contains the total number of
3337 lines in the message. Compare $body_linecount, which is the count for the
3340 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3341 bounce and delay warning messages, except in certain special cases, which
3344 (a) An SMTP error message from a remote host;
3345 (b) A message specified in a :fail: redirection;
3346 (c) A message specified in a "fail" command in a system filter;
3347 (d) A message specified in a FAIL return from the queryprogram router;
3348 (e) A message specified by the cannot_route_message router option.
3350 In these cases only, Exim does include the error details in bounce and
3351 warning messages. There are also a few cases where bland messages such
3352 as "unrouteable address" or "local delivery error" are given.
3354 PH/43 $value is now also set for the "else" part of a ${run expansion.
3356 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3357 being worked on, but at least Exim now implements the latest version to
3360 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3361 process to complete was treated as a delivery failure, a timeout while
3362 writing the message to the pipe was logged, but erroneously treated as a
3363 successful delivery. Such timeouts include transport filter timeouts. For
3364 consistency with the overall process timeout, these timeouts are now
3365 treated as errors, giving rise to delivery failures by default. However,
3366 there is now a new Boolean option for the pipe transport called
3367 timeout_defer, which, if set TRUE, converts the failures into defers for
3368 both kinds of timeout. A transport filter timeout is now identified in
3371 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3372 systems where "make" and "gmake" are different, calling "gmake" at top
3373 level broke things. I've arranged for the value of $(MAKE) to be passed
3374 from the Makefile to this script so that it can call the same version of
3378 A note about Exim versions 4.44 and 4.50
3379 ----------------------------------------
3381 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3382 changes of various kinds. As a consequence, a big documentation update was
3383 needed. This delayed the release for rather longer than seemed good, especially
3384 in the light of a couple of (minor) security issues. Therefore, the changes
3385 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3386 release. So 4.44 and 4.50 are in effect two different branches that both start
3389 I have left the 4.50 change log unchanged; it contains all the changes since
3390 4.43. The change log for 4.44 is below; many of its items are identical to
3391 those for 4.50. This seems to be the most sensible way to preserve the
3392 historical information.
3398 1. Minor wording change to the doc/README.SIEVE file.
3400 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3401 computation of the current number of files was incorrect.
3403 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3404 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3407 4. Give more explanation in the error message when the command for a transport
3408 filter fails to execute.
3410 5. There are several places where Exim runs a non-Exim command in a
3411 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3412 was being done only for the command run by the queryprogram router. It is
3413 now done for all such subprocesses. The other cases are: ${run, transport
3414 filters, and the commands run by the lmtp and pipe transports.
3416 6. Added CONFIGURE_GROUP build-time option.
3418 7. Some older OS have a limit of 256 on the maximum number of file
3419 descriptors. Exim was using setrlimit() to set 1000 as a large value
3420 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3421 systems. I've change it so that if it can't get 1000, it tries for 256.
3423 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3424 was an oversight, and furthermore, ever since the addition of extra
3425 controls (e.g. 4.43/32), the checks on when to allow different forms of
3426 "control" were broken. There should now be diagnostics for all cases when a
3427 control that does not make sense is encountered.
3429 9. Added the /retain_sender option to "control=submission".
3431 10. $recipients is now available in the predata ACL (oversight).
3433 11. Tidy the search cache before the fork to do a delivery from a message
3434 received from the command line. Otherwise the child will trigger a lookup
3435 failure and thereby defer the delivery if it tries to use (for example) a
3436 cached ldap connection that the parent has called unbind on.
3438 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3439 of $address_data from the recipient verification was clobbered by the
3440 sender verification.
3442 13. The value of address_data from a sender verification is now available in
3443 $sender_address_data in subsequent conditions in the ACL statement.
3445 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3447 15. Added a new option "connect=<time>" to callout options, to set a different
3450 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3451 was its contents. (It was OK if the option was not defined at all.)
3453 17. A "Completed" log line is now written for messages that are removed from
3454 the spool by the -Mrm option.
3456 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3457 information about exactly what failed.
3459 19. Added -dd to debug only the daemon process.
3461 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3462 handles timeouts, both on the server side and network timeouts. Renamed the
3463 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3465 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3466 It is now set to "smtps".
3468 22. $host_address is now set to the target address during the checking of
3469 ignore_target_hosts.
3471 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3472 being passed; this would have caused $sender_host_name to have been used if
3473 matching the list had actually called for a host name (not very likely,
3474 since this list is usually IP addresses). A host name is now passed as
3477 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3478 code that specifies a non-restarting handler (typically sigaction() in
3479 modern systems) in an attempt to fix a rare and obscure crash bug.
3481 25. Narrowed the window for a race in the daemon that could cause it to ignore
3482 SIGCHLD signals. This is not a major problem, because they are used only to
3483 wake it up if nothing else does.
3485 26. A malformed maildirsize file could cause Exim to calculate negative values
3486 for the mailbox size or file count. Odd effects could occur as a result.
3487 The maildirsize information is now recalculated if the size or filecount
3490 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3491 support for a long time. Removed HAVE_SYS_VFS_H.
3493 28. Installed the latest version of exipick from John Jetmore.
3495 29. In an address list, if the pattern was not a regular expression, an empty
3496 subject address (from a bounce message) matched only if the pattern was an
3497 empty string. Non-empty patterns were not even tested. This was the wrong
3498 because it is perfectly reasonable to use an empty address as part of a
3499 database query. An empty address is now tested by patterns that are
3500 lookups. However, all the other forms of pattern expect the subject to
3501 contain a local part and a domain, and therefore, for them, an empty
3502 address still always fails if the pattern is not itself empty.
3504 30. Exim went into a mad DNS loop when attempting to do a callout where the
3505 host was specified on an smtp transport, and looking it up yielded more
3506 than one IP address.
3508 31. Re-factored the code for checking spool and log partition space into a
3509 function that finds that data and another that does the check. The former
3510 is then used to implement four new variables: $spool_space, $log_space,
3511 $spool_inodes, and $log_inodes.
3513 32. The RFC2047 encoding function was originally intended for short strings
3514 such as real names; it was not keeping to the 75-character limit for
3515 encoded words that the RFC imposes. It now respects the limit, and
3516 generates multiple encoded words if necessary. To be on the safe side, I
3517 have increased the buffer size for the ${rfc2047: expansion operator from
3520 33. It is now permitted to omit both strings after an "if" condition; if the
3521 condition is true, the result is "true". As before, when the second string
3522 is omitted, a false condition yields an empty string. This makes it less
3523 cumbersome to write custom ACL and router conditions.
3525 34. Failure to deliver a bounce message always caused it to be frozen, even if
3526 there was an errors_to setting on the router. The errors_to setting is now
3529 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3530 canonical form (fully expanded) before being placed in
3531 $sender_host_address.
3533 36. The table in the code that translates DNS record types into text (T_A to
3534 "A" for instance) was missing entries for NS and CNAME. It is just possible
3535 that this could have caused confusion if both these types were looked up
3536 for the same domain, because the text type is used as part of Exim's
3537 per-process caching. But the chance of anyone hitting this buglet seems
3540 37. The dnsdb lookup has been extended in a number of ways.
3542 (1) There is a new type, "zns", which walks up the domain tree until it
3543 finds some nameserver records. It should be used with care.
3545 (2) There is a new type, "mxh", which is like "mx" except that it returns
3546 just the host names, not the priorities.
3548 (3) It is now possible to give a list of domains (or IP addresses) to be
3549 looked up. The behaviour when one of the lookups defers can be
3550 controlled by a keyword.
3552 (4) It is now possible to specify the separator character for use when
3553 multiple records are returned.
3555 38. The dnslists ACL condition has been extended: it is now possible to supply
3556 a list of IP addresses and/or domains to be looked up in a particular DNS
3559 39. Added log_selector=+queue_time_overall.
3561 40. When running the queue in the test harness, wait just a tad after forking a
3562 delivery process, to get repeatability of debugging output.
3564 41. Include certificate and key file names in error message when GnuTLS fails
3565 to set them up, because the GnuTLS error message doesn't include the name
3566 of the failing file when there is a problem reading it.
3568 42. Allow both -bf and -bF in the same test run.
3570 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3572 44. The "Exiscan patch" is now merged into the mainline Exim source.
3574 45. Sometimes the final signoff response after QUIT could fail to get
3575 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3576 < 0 before doing a fflush(). This bug looks as though it goes back to the
3577 introduction of TLS in release 3.20, but "sometimes" must have been rare
3578 because the tests only now provoked it.
3580 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3581 (this can affect the format of dates).
3583 47. exim_tidydb, when checking for the continued existence of a message for
3584 which it has found a message-specific retry record, was not finding
3585 messages that were in split spool directories. Consequently, it was
3586 deleting retry records that should have stayed in existence.
3588 48. Steve fixed some bugs in eximstats.
3590 49. The SPA authentication driver was not abandoning authentication and moving
3591 on to the next authenticator when an expansion was forced to fail,
3592 contradicting the general specification for all authenticators. Instead it
3593 was generating a temporary error. It now behaves as specified.
3595 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3596 (the order specifies the preference for clients). The order is now AES256,
3597 AES128, 3DES, ARCFOUR128.
3599 51. Small patch to Sieve code - explicitly set From: when generating an
3602 52. Exim crashed if a remote delivery caused a very long error message to be
3603 recorded - for instance if somebody sent an entire SpamAssassin report back
3604 as a large number of 550 error lines. This bug was coincidentally fixed by
3605 increasing the size of one of Exim's internal buffers (big_buffer) that
3606 happened as part of the Exiscan merge. However, to be on the safe side, I
3607 have made the code more robust (and fixed the comments that describe what
3610 53. Now that there can be additional text after "Completed" in log lines (if
3611 the queue_time_overall log selector is set), a one-byte patch to exigrep
3612 was needed to allow it to recognize "Completed" as not the last thing in
3615 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3616 patch that reportedly fixes this has been added. I am not expert enough to
3617 create a test for it. This is what the patch creator wrote:
3619 "I found a little strange behaviour of ldap code when working with
3620 Windows 2003 AD Domain, where users was placed in more than one
3621 Organization Units. When I tried to give exim partial DN, the exit code
3622 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3623 But simultaneously result of request was absolutely normal ldap result,
3624 so I produce this patch..."
3626 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3627 REFERENCE, so I have modified the code to exclude the patch when that macro
3630 55. Some experimental protocols are using DNS PTR records for new purposes. The
3631 keys for these records are domain names, not reversed IP addresses. The
3632 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3633 leaves it alone. Component reversal etc. now happens only for IP addresses.
3636 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3638 57. Double the size of the debug message buffer (to 2048) so that more of very
3639 long debug lines gets shown.
3641 58. The exicyclog utility now does better if the number of log files to keep
3642 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3644 59. Two changes related to the smtp_active_hostname option:
3646 (1) $smtp_active_hostname is now available as a variable.
3647 (2) The default for smtp_banner uses $smtp_active_hostname instead
3648 of $primary_hostname.
3650 60. The host_aton() function is supposed to be passed a string that is known
3651 to be a valid IP address. However, in the case of IPv6 addresses, it was
3652 not checking this. This is a hostage to fortune. Exim now panics and dies
3653 if the condition is not met. A case was found where this could be provoked
3654 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3655 components; fortuitously, this particular loophole had already been fixed
3656 by change 4.50/55 above.
3658 If there are any other similar loopholes, the new check in host_aton()
3659 itself should stop them being exploited. The report I received stated that
3660 data on the command line could provoke the exploit when Exim was running as
3661 exim, but did not say which command line option was involved. All I could
3662 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3663 running as the user.
3666 61. There was a buffer overflow vulnerability in the SPA authentication code
3667 (which came originally from the Samba project). I have added a test to the
3668 spa_base64_to_bits() function which I hope fixes it.
3671 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3672 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3674 63. The daemon start-up calls getloadavg() while still root for those OS that
3675 need the first call to be done as root, but it missed one case: when
3676 deliver_queue_load_max is set with deliver_drop_privilege. This is
3677 necessary for the benefit of the queue runner, because there is no re-exec
3678 when deliver_drop_privilege is set.
3680 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3681 This has been fixed.
3683 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3684 was in use, was not putting the data itself into the right store pool;
3685 consequently, it could be overwritten for a subsequent message in the same
3686 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3689 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3691 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3692 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3693 the function were treating the return as a boolean value, which happened to
3694 work because 0=false and not-0=true, but is not correct code.
3696 68. The host_aton() function was not handling scoped IPv6 addresses (those
3697 with, for example, "%eth0" on the end) correctly.
3699 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3700 negated items (that is, ~something) in unsigned ints. Some compilers
3701 apparently mutter when there is no cast.
3703 70. If an address verification called from an ACL failed, and did not produce a
3704 user-specific message (i.e. there was only a "system" message), nothing was
3705 put in $acl_verify_message. In this situation, it now puts the system
3708 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3709 session; change 4.31/43 added the unwanted input to the log line - except
3710 that it did not do this in the start of session case. It now does.
3712 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3713 This is wrong and can cause the other end to generate a synchronization
3714 error if it is another Exim or anything else that does the synchronization
3715 check. A QUIT command is no longer sent after a timeout.
3717 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3718 during host lookups.
3720 74. The defer_ok option of callout verification was not working if it was used
3721 when verifying addresses in header lines, that is, for this case:
3723 verify = header_sender/callout=defer_ok
3725 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3726 those file descriptors could be used for SMTP connections. If anything
3727 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3728 could be sent to the SMTP client, causing chaos. The daemon now opens
3729 stdin, stdout, and stderr to /dev/null when it puts itself into the
3732 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3733 log by default. The user can override this with suitable Perl magic.
3735 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3736 the log message when discard triggers, was not working for the DATA ACL or
3737 for the non-SMTP ACL.
3739 78. Error message wording change in sieve.c.
3741 79. If smtp_accept_max_per_host was set, the number of connections could be
3742 restricted to fewer than expected, because the daemon was trying to set up
3743 a new connection before checking whether the processes handling previous
3744 connections had finished. The check for completed processes is now done
3745 earlier. On busy systems, this bug wouldn't be noticed because something
3746 else would have woken the daemon, and it would have reaped the completed
3749 80. If a message was submitted locally by a user whose login name contained one
3750 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3751 It caused a spool format error. I have fixed the spool reading code. A
3752 related problem was that the "from" clause in the Received: line became
3753 illegal because of the space(s). It is now covered by ${quote_local_part.
3755 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3758 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3760 83. Patch from Sieve maintainer to fix unterminated string problem in
3761 "vacation" handling.
3763 84. Some minor changes to the Linux configuration files to help with other
3764 OS variants using glibc.
3766 85. One more patch for Sieve to update vacation handling to latest spec.
3769 ----------------------------------------------------
3770 See the note above about the 4.44 and 4.50 releases.
3771 ----------------------------------------------------
3777 1. Change 4.43/35 introduced a bug that caused file counts to be
3778 incorrectly computed when quota_filecount was set in an appendfile
3781 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3782 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3785 3. Give more explanation in the error message when the command for a transport
3786 filter fails to execute.
3788 4. There are several places where Exim runs a non-Exim command in a
3789 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3790 was being done only for the command run by the queryprogram router. It is
3791 now done for all such subprocesses. The other cases are: ${run, transport
3792 filters, and the commands run by the lmtp and pipe transports.
3794 5. Some older OS have a limit of 256 on the maximum number of file
3795 descriptors. Exim was using setrlimit() to set 1000 as a large value
3796 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3797 systems. I've change it so that if it can't get 1000, it tries for 256.
3799 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3800 was an oversight, and furthermore, ever since the addition of extra
3801 controls (e.g. 4.43/32), the checks on when to allow different forms of
3802 "control" were broken. There should now be diagnostics for all cases when a
3803 control that does not make sense is encountered.
3805 7. $recipients is now available in the predata ACL (oversight).
3807 8. Tidy the search cache before the fork to do a delivery from a message
3808 received from the command line. Otherwise the child will trigger a lookup
3809 failure and thereby defer the delivery if it tries to use (for example) a
3810 cached ldap connection that the parent has called unbind on.
3812 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3813 of $address_data from the recipient verification was clobbered by the
3814 sender verification.
3816 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3817 was its contents. (It was OK if the option was not defined at all.)
3819 11. A "Completed" log line is now written for messages that are removed from
3820 the spool by the -Mrm option.
3822 12. $host_address is now set to the target address during the checking of
3823 ignore_target_hosts.
3825 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3826 being passed; this would have caused $sender_host_name to have been used if
3827 matching the list had actually called for a host name (not very likely,
3828 since this list is usually IP addresses). A host name is now passed as
3831 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3832 code that specifies a non-restarting handler (typically sigaction() in
3833 modern systems) in an attempt to fix a rare and obscure crash bug.
3835 15. Narrowed the window for a race in the daemon that could cause it to ignore
3836 SIGCHLD signals. This is not a major problem, because they are used only to
3837 wake it up if nothing else does.
3839 16. A malformed maildirsize file could cause Exim to calculate negative values
3840 for the mailbox size or file count. Odd effects could occur as a result.
3841 The maildirsize information is now recalculated if the size or filecount
3844 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3845 support for a long time. Removed HAVE_SYS_VFS_H.
3847 18. Updated exipick to current release from John Jetmore.
3849 19. Allow an empty sender to be matched against a lookup in an address list.
3850 Previously the only cases considered were a regular expression, or an
3853 20. Exim went into a mad DNS lookup loop when doing a callout where the
3854 host was specified on the transport, if the DNS lookup yielded more than
3857 21. The RFC2047 encoding function was originally intended for short strings
3858 such as real names; it was not keeping to the 75-character limit for
3859 encoded words that the RFC imposes. It now respects the limit, and
3860 generates multiple encoded words if necessary. To be on the safe side, I
3861 have increased the buffer size for the ${rfc2047: expansion operator from
3864 22. Failure to deliver a bounce message always caused it to be frozen, even if
3865 there was an errors_to setting on the router. The errors_to setting is now
3868 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3869 canonical form (fully expanded) before being placed in
3870 $sender_host_address.
3872 24. Updated eximstats to version 1.33
3874 25. Include certificate and key file names in error message when GnuTLS fails
3875 to set them up, because the GnuTLS error message doesn't include the name
3876 of the failing file when there is a problem reading it.
3878 26. Expand error message when OpenSSL has problems setting up cert/key files.
3881 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3882 (this can affect the format of dates).
3884 28. exim_tidydb, when checking for the continued existence of a message for
3885 which it has found a message-specific retry record, was not finding
3886 messages that were in split spool directories. Consequently, it was
3887 deleting retry records that should have stayed in existence.
3889 29. eximstats updated to version 1.35
3890 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3891 1.35 - bugfix such that pie charts by volume are generated correctly
3893 30. The SPA authentication driver was not abandoning authentication and moving
3894 on to the next authenticator when an expansion was forced to fail,
3895 contradicting the general specification for all authenticators. Instead it
3896 was generating a temporary error. It now behaves as specified.
3898 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3899 (the order specifies the preference for clients). The order is now AES256,
3900 AES128, 3DES, ARCFOUR128.
3902 31. Small patch to Sieve code - explicitly set From: when generating an
3905 32. Exim crashed if a remote delivery caused a very long error message to be
3906 recorded - for instance if somebody sent an entire SpamAssassin report back
3907 as a large number of 550 error lines. This bug was coincidentally fixed by
3908 increasing the size of one of Exim's internal buffers (big_buffer) that
3909 happened as part of the Exiscan merge. However, to be on the safe side, I
3910 have made the code more robust (and fixed the comments that describe what
3913 33. Some experimental protocols are using DNS PTR records for new purposes. The
3914 keys for these records are domain names, not reversed IP addresses. The
3915 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3916 leaves it alone. Component reversal etc. now happens only for IP addresses.
3919 34. The host_aton() function is supposed to be passed a string that is known
3920 to be a valid IP address. However, in the case of IPv6 addresses, it was
3921 not checking this. This is a hostage to fortune. Exim now panics and dies
3922 if the condition is not met. A case was found where this could be provoked
3923 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3924 components; fortuitously, this particular loophole had already been fixed
3925 by change 4.50/55 or 4.44/33 above.
3927 If there are any other similar loopholes, the new check in host_aton()
3928 itself should stop them being exploited. The report I received stated that
3929 data on the command line could provoke the exploit when Exim was running as
3930 exim, but did not say which command line option was involved. All I could
3931 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3932 running as the user.
3935 35. There was a buffer overflow vulnerability in the SPA authentication code
3936 (which came originally from the Samba project). I have added a test to the
3937 spa_base64_to_bits() function which I hope fixes it.
3940 36. The daemon start-up calls getloadavg() while still root for those OS that
3941 need the first call to be done as root, but it missed one case: when
3942 deliver_queue_load_max is set with deliver_drop_privilege. This is
3943 necessary for the benefit of the queue runner, because there is no re-exec
3944 when deliver_drop_privilege is set.
3946 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3947 was in use, was not putting the data itself into the right store pool;
3948 consequently, it could be overwritten for a subsequent message in the same
3949 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3952 38. Sometimes the final signoff response after QUIT could fail to get
3953 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3954 < 0 before doing a fflush(). This bug looks as though it goes back to the
3955 introduction of TLS in release 3.20, but "sometimes" must have been rare
3956 because the tests only now provoked it.
3962 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3963 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3964 Now it returns an integer. A number of calls were still expecting a T/F
3965 return. Fortuitously, in all cases, the tests worked in OK situations,
3966 which is the norm. However, things would have gone wrong on any write
3967 failures on the smtp file descriptor. This function is used when sending
3968 messages over SMTP and also when doing verify callouts.
3970 2. When Exim is called to do synchronous delivery of a locally submitted
3971 message (the -odf or -odi options), it no longer closes stderr before doing
3974 3. Implemented the mua_wrapper option.
3976 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3978 5. Implemented the functions header_remove(), header_testname(),
3979 header_add_at_position(), and receive_remove_recipient(), and exported them
3982 6. If an ACL "warn" statement specified the addition of headers, Exim already
3983 inserted X-ACL-Warn: at the start if there was no header name. However, it
3984 was not making this test for the second and subsequent header lines if
3985 there were newlines in the string. This meant that an invalid header could
3986 be inserted if Exim was badly configured.
3988 7. Allow an ACL "warn" statement to add header lines at the start or after all
3989 the Received: headers, as well as at the end.
3991 8. Added the rcpt_4xx retry error code.
3993 9. Added postmaster_mailfrom=xxx to callout verification option.
3995 10. Added mailfrom=xxxx to the callout verification option, for verify=
3998 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3999 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4001 12. Inserted some casts to stop certain compilers warning when using pointer
4002 differences as field lengths or precisions in printf-type calls (mostly
4003 affecting debugging statements).
4005 13. Added optional readline() support for -be (dynamically loaded).
4007 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4008 same clock tick as a message's arrival, so that its received time was the
4009 same as the "first fail" time on the retry record, and that message
4010 remained on the queue past the ultimate address timeout, every queue runner
4011 would try a delivery (because it was past the ultimate address timeout) but
4012 after another failure, the ultimate address timeout, which should have then
4013 bounced the address, did not kick in. This was a "< instead of <=" error;
4014 in most cases the first failure would have been in the next clock tick
4015 after the received time, and all would be well.
4017 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4018 being recognized when the domain list was tested by the match_domain
4019 condition in an expansion string.
4021 16. Added the ${str2b64: operator.
4023 17. Exim was always calling setrlimit() to set a large limit for the number of
4024 processes, without checking whether the existing limit was already
4025 adequate. (It did check for the limit on file descriptors.) Furthermore,
4026 errors from getrlimit() and setrlimit() were being ignored. Now they are
4027 logged to the main and panic logs, but Exim does carry on, to try to do its
4028 job under whatever limits there are.
4030 18. Imported PCRE 5.0.
4032 19. Trivial typo in log message " temporarily refused connection" (the leading
4035 20. If the log selector return_path_on_delivery was set and an address was
4036 redirected to /dev/null, the delivery process crashed because it assumed
4037 that a return path would always be set for a "successful" delivery. In this
4038 case, the whole delivery is bypassed as an optimization, and therefore no
4041 21. Internal re-arrangement: the function for sending a challenge and reading
4042 a response while authentication was assuming a zero-terminated challenge
4043 string. It's now changed to take a pointer and a length, to allow for
4044 binary data in such strings.
4046 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4048 23. Exim was not respecting finduser_retries when seeking the login of the
4049 uid under which it was called; it was always trying 10 times. (The default
4050 setting of finduser_retries is zero.) Also, it was sleeping after the final
4051 failure, which is pointless.
4053 24. Implemented tls_on_connect_ports.
4055 25. Implemented acl_smtp_predata.
4057 26. If the domain in control=submission is set empty, Exim assumes that the
4058 authenticated id is a complete email address when it generates From: or
4059 Sender: header lines.
4061 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4062 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4063 chown and chgrp in /bin and hostname in /usr/bin.
4065 28. Exim was keeping the "process log" file open after each use, just as it
4066 does for the main log. This opens the possibility of it remaining open for
4067 long periods when the USR1 signal hits a daemon. Occasional processlog
4068 errors were reported, that could have been caused by this. Anyway, it seems
4069 much more sensible not to leave this file open at all, so that is what now
4072 29. The long-running daemon process does not normally write to the log once it
4073 has entered its main loop, and it closes the log before doing so. This is
4074 so that log files can straightforwardly be renamed and moved. However,
4075 there are a couple of unusual error situations where the daemon does write
4076 log entries, and I had neglected to close the log afterwards.
4078 30. The text of an SMTP error response that was received during a remote
4079 delivery was being truncated at 512 bytes. This is too short for some of
4080 the long messages that one sometimes sees. I've increased the limit to
4083 31. It is now possible to make retry rules that apply only when a message has a
4084 specific sender, in particular, an empty sender.
4086 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4087 it possible to be selective about when SMTP synchronization is enforced.
4089 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4091 32. Implemented hosts_connection_nolog.
4093 33. Added an ACL for QUIT.
4095 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4098 35. Added mailbox_size and mailbox_filecount to appendfile.
4100 36. Added control = no_multiline_responses to ACLs.
4102 37. There was a bug in the logic of the code that waits for the clock to tick
4103 in the case where the clock went backwards by a substantial amount such
4104 that the microsecond fraction of "now" was more than the microsecond
4105 fraction of "then" (but the whole seconds number was less).
4107 38. Added support for the libradius Radius client library this is found on
4108 FreeBSD (previously only the radiusclient library was supported).
4114 1. When certain lookups returned multiple values in the form name=value, the
4115 quoting of the values was not always being done properly. Specifically:
4116 (a) If the value started with a double quote, but contained no whitespace,
4118 (b) If the value contained whitespace other than a space character (i.e.
4119 tabs or newlines or carriage returns) it was not quoted.
4120 This fix has been applied to the mysql and pgsql lookups by writing a
4121 separate quoting function and calling it from the lookup code. The fix
4122 should probably also be applied to nisplus, ibase and oracle lookups, but
4123 since I cannot test any of those, I have not disturbed their existing code.
4125 2. A hit in the callout cache for a specific address caused a log line with no
4126 reason for rejecting RCPT. Now it says "Previous (cached) callout
4127 verification failure".
4129 3. There was an off-by-one bug in the queryprogram router. An over-long
4130 return line was truncated at 256 instead of 255 characters, thereby
4131 overflowing its buffer with the terminating zero. As well as fixing this, I
4132 have increased the buffer size to 1024 (and made a note to document this).
4134 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4135 when Exim is waiting for an SMTP response from a remote server, Exim
4136 restarts its select() call on the socket, thereby resetting its timeout.
4137 This is not a problem when such interrupts are rare. Somebody set up a cron
4138 job to run exiwhat every 2 minutes, which is less than the normal select()
4139 timeout (5 or 10 minutes). This meant that the select() timeout never
4140 kicked in because it was always reset. I have fixed this by comparing the
4141 time when an interrupt arrives with the time at the start of the first call
4142 to select(). If more time than the timeout has elapsed, the interrupt is
4143 treated as a timeout.
4145 5. Some internal re-factoring in preparation for the addition of Sieve
4146 extensions (by MH). In particular, the "personal" test is moved to a
4147 separate function, and given an option for scanning Cc: and Bcc: (which is
4148 not set for Exim filters).
4150 6. When Exim created an email address using the login of the caller as the
4151 local part (e.g. when creating a From: or Sender: header line), it was not
4152 quoting the local part when it contained special characters such as @.
4154 7. Installed new OpenBSD configuration files.
4156 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4157 try to make them clearer.
4159 9. Callout options, other than the timeout value, were being ignored when
4160 verifying sender addresses in header lines. For example, when using
4162 verify = header_sender/callout=no_cache
4164 the cache was (incorrectly) being used.
4166 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4167 only the Cygwin environment.
4169 11. When return_path_on_delivery was set as a log selector, if different remote
4170 addresses in the same message used different return paths and parallel
4171 remote delivery occurred, the wrong values would sometimes be logged.
4172 (Whenever a remote delivery process finished, the return path value from
4173 the most recently started remote delivery process was logged.)
4175 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4176 lines when AUTH and/or TLS are in use. This is the "received protocol"
4177 field. Exim used to use "asmtp" for authenticated SMTP, without any
4178 indication (in the protocol name) for TLS use. Now it follows the RFC and
4179 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4180 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4181 names appear in log lines as well as in Received: header lines.
4183 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4184 extensions, and comparison tests, and to fix some bugs.
4186 14. Changes to the "personal" filter test:
4188 (1) The test was buggy in that it was just doing the equivalent of
4189 "contains" tests on header lines. For example, if a user's address was
4190 anne@some.where, the "personal" test would incorrectly be true for
4192 To: susanne@some.where
4194 This test is now done by extracting each address from the header in turn,
4195 and checking the entire address. Other tests that are part of "personal"
4196 are now done using regular expressions (for example, to check local parts
4197 of addresses in From: header lines).
4199 (2) The list of non-personal local parts in From: addresses has been
4200 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4201 taken from the Sieve specification recommendations.
4203 (3) If the message contains any header line starting with "List-" it is
4204 treated as non-personal.
4206 (4) The test for "circular" in the Subject: header line has been removed
4207 because it now seems ill-conceived.
4209 15. Minor typos in src/EDITME comments corrected.
4211 16. Installed latest exipick from John Jetmore.
4213 17. If headers_add on a router specified a text string that was too long for
4214 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4215 of string_sprintf() is now avoided.
4217 18. $message_body_size was not set (it was always zero) when running the DATA
4218 ACL and the local_scan() function.
4220 19. For the "mail" command in an Exim filter, no default was being set for
4221 the once_repeat time, causing a random time value to be used if "once" was
4222 specified. (If the value happened to be <= 0, no repeat happened.) The
4223 default is now 0s, meaning "never repeat". The "vacation" command was OK
4224 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4225 (I found it when inspecting the code).
4227 20. There is now an overall timeout for performing a callout verification. It
4228 defaults to 4 times the callout timeout, which applies to individual SMTP
4229 commands during the callout. The overall timeout applies when there is more
4230 than one host that can be tried. The timeout is checked before trying the
4231 next host. This prevents very long delays if there are a large number of
4232 hosts and all are timing out (e.g. when the network connections are timing
4233 out). The value of the overall timeout can be changed by specifying an
4234 additional sub-option for "callout", called "maxwait". For example:
4236 verify = sender/callout=5s,maxwait=20s
4238 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4239 to the end before writing, but this should make it even safer).
4241 22. Exim was forgetting that it had advertised PIPELINING for the second and
4242 subsequent messages on an SMTP connection. It was also not resetting its
4243 memory on STARTTLS and an internal HELO.
4245 23. When Exim logs an SMTP synchronization error within a session, it now
4246 records whether PIPELINING has been advertised or not.
4248 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4249 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4250 rather than long int.
4252 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4254 26. Added the never_mail option to autoreply.
4260 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4261 crash if the getsockname() call failed; this can happen if a connection is
4262 closed very soon after it is established. The problem was simply in the
4263 order in which certain operations were done, causing Exim to try to write
4264 to the SMTP stream before it had set up the file descriptor. The bug has
4265 been fixed by making things happen in the correct order.
4271 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4272 before the connection was closed, thus losing the rejection response.
4274 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4275 some early Solaris releases, but causes trouble in current releases where
4276 socklen_t is defined.
4278 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4281 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4284 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4285 that releasing the top of it at the end releases what was used for sub-
4286 expansions (unless the block got too big). However, discard this block if
4287 the first thing is a variable or header, so that we can use its block when
4288 it is dynamic (useful for very large $message_headers, for example).
4290 6. Lookups now cache *every* query, not just the most recent. A new, separate
4291 store pool is used for this. It can be recovered when all lookup caches are
4292 flushed. Lookups now release memory at the end of their result strings.
4293 This has involved some general refactoring of the lookup sources.
4295 7. Some code has been added to the store_xxx() functions to reduce the amount
4296 of flapping under certain conditions.
4298 8. log_incoming_interface used to affect only the <= reception log lines. Now
4299 it causes the local interface and port to be added to several more SMTP log
4300 lines, for example "SMTP connection from", and rejection lines.
4302 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4304 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4306 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4307 could be overwritten at the end of the current message (or the start of a
4308 new message if it was set in a HELO ACL). The value is now preserved for
4309 the duration of the SMTP connection.
4311 12. If a transport had a headers_rewrite setting, and a matching header line
4312 contained an unqualified address, that address was qualified, even if it
4313 did not match any rewriting rules. The underlying bug was that the values
4314 of the flags that permit the existence of unqualified sender and recipient
4315 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4316 non-local messages, and by -bnq for local messages) were not being
4317 preserved with the message after it was received.
4319 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4320 "next input=" as part of the text comprising the host identity instead of
4321 the correct text. The code was using the same buffer for two different
4322 strings. However, depending on which order the printing function evaluated
4323 its arguments, the bug did not always show up. Under Linux, for example, my
4324 test suite worked just fine.
4326 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4327 doesn't work with some older versions of Perl. It has been changed to "my",
4328 which in any case is probably the better facility to use.
4330 15. A really picky compiler found some instances of statements for creating
4331 error messages that either had too many or two few arguments for the format
4334 16. The size of the buffer for calls to the DNS resolver has been increased
4335 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4336 for addresses that have a lot of PTR records. This alleviates a problem; it
4337 does not fully solve it.
4339 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4340 buffer now truncates the list and logs the incident, which is the same
4341 action as happens when Exim is looking up a host name and its aliases.
4342 Previously in this situation something unpredictable would happen;
4343 sometimes it was "internal error: store_reset failed".
4345 18. If a server dropped the connection unexpectedly when an Exim client was
4346 using GnuTLS and trying to read a response, the client delivery process
4347 crashed while trying to generate an error log message.
4349 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4350 string, for example:
4352 warn message = H1: something\nH2: something
4354 the text was added as a single header line from Exim's point of view
4355 though it ended up OK in the delivered message. However, searching for the
4356 second and subsequent header lines using $h_h2: did not work. This has been
4357 fixed. Similarly, if a system filter added multiple headers in this way,
4358 the routers could not see them.
4360 20. Expanded the error message when iplsearch is called with an invalid key to
4361 suggest using net-iplsearch in a host list.
4363 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4364 ACLs are no longer actually imposed (and a message to that effect is
4367 22. If a "gecos" field in a passwd entry contained escaped characters, in
4368 particular, if it contained a \" sequence, Exim got it wrong when building
4369 a From: or a Sender: header from that name. A second bug also caused
4370 incorrect handling when an unquoted " was present following a character
4371 that needed quoting.
4373 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4374 was not being matched caselessly.
4376 24. Arranged for all hyphens in the exim.8 source to be escaped with
4379 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4380 properly. Recipient callout cache records were still being keyed to include
4381 the sender, even when use_sender was set false. This led to far more
4382 callouts that were necessary. The sender is no longer included in the key
4383 when use_sender is false.
4385 26. Added "control = submission" modifier to ACLs.
4387 27. Added the ${base62d: operator to decode base 62 numbers.
4389 28. dnsdb lookups can now access SRV records.
4391 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4392 the configuration file.
4394 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4395 (-v) mode. This makes the output for a verbose queue run more intelligible.
4397 31. Added a use_postmaster feature to recipient callouts.
4399 32. Added the $body_zerocount variable, containing the number of binary zero
4400 bytes in the message body.
4402 33. The time of last modification of the "new" subdirectory is now used as the
4403 "mailbox time last read" when there is a quota error for a maildir
4406 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4408 35. Added +ignore_unknown as a special item in host lists.
4410 36. Code for decoding IPv6 addresses in host lists is now included, even if
4411 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4412 address was recognized as an IP address, but was then not correctly decoded
4413 into binary, causing unexpected and incorrect effects when compared with
4420 1. Very minor rewording of debugging text in manualroute to say "list of
4421 hosts" instead of "hostlist".
4423 2. If verify=header_syntax was set, and a header line with an unqualified
4424 address (no domain) and a large number of spaces between the end of the
4425 name and the colon was received, the reception process suffered a buffer
4426 overflow, and (when I tested it) crashed. This was caused by some obsolete
4427 code that should have been removed. The fix is to remove it!
4429 3. When running in the test harness, delay a bit after writing a bounce
4430 message to get a bit more predictability in the log output.
4432 4. Added a call to search_tidyup() just before forking a reception process. In
4433 theory, someone could use a lookup in the expansion of smtp_accept_max_
4434 per_host which, without the tidyup, could leave open a database connection.
4436 5. Added the variables $recipient_data and $sender_data which get set from a
4437 lookup success in an ACL "recipients" or "senders" condition, or a router
4438 "senders" option, similar to $domain_data and $local_part_data.
4440 6. Moved the writing of debug_print from before to after the "senders" test
4443 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4444 problems for message scanning, either using a data ACL, or using
4445 local_scan() because the Received: header was not generated till after they
4446 were called (in order to set the time as the time of reception completion).
4447 I have revised the way this works. The header is now generated after the
4448 body is received, but before the ACL or local_scan() are called. After they
4449 are run, the timestamp in the header is updated.
4455 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4456 before starting a queue runner without re-exec. This happened only when
4457 deliver_drop_privilege was set or when the Exim user was set to root. The
4458 effect of the bug was that timeouts during subsequent deliveries caused
4459 crashes instead of being properly handled. The handler is now left at its
4460 default (and expected) setting.
4462 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4463 message, again when deliver_drop_privilege is set or Exim is run as root.
4464 The bug described in (1) was not present in this case, but the tidying up
4465 of the other signals was missing. I have made the two cases consistent.
4467 3. The ignore_target_hosts setting on a manualroute router was being ignored
4468 for hosts that were looked up using the /MX notation.
4470 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4473 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4474 operated on the sender address. After changing the $sender_address to <>
4475 for the sender address verify, Exim was re-instated it as the original
4476 (before rewriting) address, but remembering that it had rewritten it, so it
4477 wasn't rewriting it again. This bug also had the effect of breaking the
4478 sender address verification caching when the sender address was rewritten.
4480 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4481 This has been changed so that if the ip literal address matches
4482 ignore_target_hosts, the router declines.
4484 7. Added expansion conditions match_domain, match_address, and match_local_
4485 part (NOT match_host).
4487 8. The placeholder for the Received: header didn't have a length field set.
4489 9. Added code to Exim itself and to exim_lock to test for a specific race
4490 condition that could lead to file corruption when using MBX delivery. The
4491 issue is with the lockfile that is created in /tmp. If this file is removed
4492 after a process has opened it but before that process has acquired a lock,
4493 there is the potential for a second process to recreate the file and also
4494 acquire a lock. This could lead to two Exim processes writing to the file
4495 at the same time. The added code performs the same test as UW imapd; it
4496 checks after acquiring the lock that its file descriptor still refers to
4497 the same named file.
4499 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4500 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4503 11. Added the smtp_active_hostname option. If used, this will typically be made
4504 to depend on the incoming interface address. Because $interface_address is
4505 not set up until the daemon has forked a reception process, error responses
4506 that can happen earlier (such as "too many connections") no longer contain
4509 12. If an expansion in a condition on a "warn" statement fails because a lookup
4510 defers, the "warn" statement is abandoned, and the next ACL statement is
4511 processed. Previously this caused the whole ACL to be aborted.
4513 13. Added the iplsearch lookup type.
4515 14. Added ident_timeout as a log selector.
4517 15. Added tls_certificate_verified as a log selector.
4519 16. Added a global option tls_require_ciphers (compare the smtp transport
4520 option of the same name). This controls incoming TLS connections.
4522 17. I finally figured out how to make tls_require_ciphers do a similar thing
4523 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4524 before starting the TLS session.
4526 18. Tabs are now shown as \t in -bP output.
4528 19. If the log selector return_path_on_delivery was set, Exim crashed when
4529 bouncing a message because it had too many Received: header lines.
4531 20. If two routers both had headers_remove settings, and the first one included
4532 a superfluous trailing colon, the final name in the first list and the
4533 first name in the second list were incorrectly joined into one item (with a
4534 colon in the middle).
4540 1. Added -C and -D options to the exinext utility, mainly to make it easier
4541 to include in the automated testing, but these could be helpful when
4542 multiple configurations are in use.
4544 2. The exinext utility was not formatting the output nicely when there was
4545 an alternate port involved in the retry record key, nor when there was a
4546 message id as well (for retries that were specific to a specific message
4547 and a specific host). It was also confused by IPv6 addresses, because of
4548 the additional colons they contain. I have fixed the IPv4 problem, and
4549 patched it up to do a reasonable job for IPv6.
4551 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4552 delivery, the log line now contains "pipelined" if PIPELINING was used.
4554 4. An SMTP transport process used to panic and die if the bind() call to set
4555 an explicit outgoing interface failed. This has been changed; it is now
4556 treated in the same way as a connect() failure.
4558 5. A reference to $sender_host_name in the part of a conditional expansion
4559 that was being skipped was still causing a DNS lookup. This no longer
4562 6. The def: expansion condition was not recognizing references to header lines
4563 that used bh_ and bheader_.
4565 7. Added the _cache feature to named lists.
4567 8. The code for checking quota_filecount in the appendfile transport was
4568 allowing one more file than it should have been.
4570 9. For compatibility with Sendmail, the command line option
4578 and sets the incoming protocol and host name (for trusted callers). The
4579 host name and its colon can be omitted when only the protocol is to be set.
4580 Note the Exim already has two private options, -pd and -ps, that refer to
4581 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4582 "s", but I don't think that's a major issue.
4584 10. A number of refactoring changes to the code, none of which should affect
4587 (a) The number of logging options was getting close to filling up the
4588 32-bit word that was used as a bit map. I have split them into two classes:
4589 those that are passed in the argument to log_write(), and those that are
4590 only ever tested independently outside of that function. These are now in
4591 separate 32-bit words, so there is plenty of room for expansion again.
4592 There is no change in the user interface or the logging behaviour.
4594 (b) When building, for example, log lines, the code previously used a
4595 macro that called string_cat() twice, in order to add two strings. This is
4596 not really sufficiently general. Furthermore, there was one instance where
4597 it was actually wrong because one of the argument was used twice, and in
4598 one call a function was used. (As it happened, calling the function twice
4599 did not affect the overall behaviour.) The macro has been replaced by a
4600 function that can join an arbitrary number of extra strings onto a growing
4603 (c) The code for expansion conditions now uses a table and a binary chop
4604 instead of a serial search (which was left over from when there were very
4605 few conditions). Also, it now recognizes conditions like "pam" even when
4606 the relevant support is not compiled in: a suitably worded error message is
4607 given if an attempt is made to use such a condition.
4609 11. Added ${time_interval:xxxxx}.
4611 12. A bug was causing one of the ddress fields not to be passed back correctly
4612 from remote delivery subprocesses. The field in question was not being
4613 subsequently used, so this caused to problems in practice.
4615 13. Added new log selectors queue_time and deliver_time.
4617 14. Might have fixed a bug in maildirsizefile handling that threw up
4618 "unexpected character" debug warnings, and recalculated the data
4619 unnecessarily. In any case, I expanded the warning message to give more
4622 15. Added the message "Restricted characters in address" to the statements in
4623 the default ACL that block characters like @ and % in local parts.
4625 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4626 Three changes have been made:
4628 (a) There was a serious bug; a negative response to MAIL caused the whole
4629 recipient domain to be cached as invalid, thereby blocking all messages
4630 to all local parts at the same domain, from all senders. This bug has
4631 been fixed. The domain is no longer cached after a negative response to
4632 MAIL if the sender used is not empty.
4634 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4637 (c) A new callout option, "use_sender" has been added for people who want
4638 the modified behaviour.
4644 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4647 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4648 indeed breaks things for older releases.
4650 3. Added additional logging to the case where there is a problem reading data
4651 from a filter that is running in a subprocess using a pipe, in order to
4652 try to track down a specific problem.
4654 4. Testing facility fudge: when running in the test harness and attempting
4655 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4656 getting "No route to host". Convert this to a timeout.
4658 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4661 6. Some OS don't have socklen_t but use size_t instead. This affects the
4662 fifth argument of getsockopt() amongst other things. This is now
4663 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4664 can be set for individual OS. I have set it for SunOS5, OSF1, and
4665 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4666 some earlier ones do not.
4668 7. Change 4.30/15 was not doing the test caselessly.
4670 8. The standard form for an IPv6 address literal was being rejected by address
4671 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4672 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4673 this, as well as the form without the "IPv6" on the front (but only when
4674 address literals are enabled, of course).
4676 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4678 10. Exim crashed if a message with an empty sender address specified by -f
4679 encountered a router with an errors_to setting. This could be provoked only
4680 by a command such as
4684 where an empty string was supplied; "<>" did not hit this bug.
4686 11. Installed PCRE release 4.5.
4688 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4689 remained set. It is now erased.
4691 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4692 times from message ids (which are base 36 rather than the normal 62).
4694 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4695 were being counted as actual protocol errors, and logged if the log
4696 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4697 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4698 and DATA following a set of rejected RCPTs do not count as protocol errors.
4699 In other words, Exim assumes they were pipelined, though this may not
4700 actually be the case. Of course, in all cases the client gets an
4701 appropriate error code.
4703 15. If a lookup fails in an ACL condition, a message about the failure may
4704 be available; it is used if testing the ACL cannot continue, because most
4705 such messages specify what the cause of the deferral is. However, some
4706 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4707 that caused an old message to be retained and used if a later statement
4708 caused a defer, replacing the real cause of the deferral.
4710 16. If an IP address had so many PTR records that the DNS lookup buffer
4711 was not large enough to hold them, Exim could crash while trying to process
4712 the truncated data. It now detects and logs this case.
4714 17. Further to 4.21/58, another change has been made: if (and only if) the
4715 first line of a message (the first header line) ends with CRLF, a bare LF
4716 in a subsequent header line has a space inserted after it, so as not to
4717 terminate the header.
4719 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4720 unnecessarily, used atoi() instead of strtol(), and didn't check the
4721 termination when getting file sizes from file names by regex.
4723 19. Completely re-implemented the support for maildirsize files, in the light
4724 of a number of problems with the previous contributed implementation
4725 (4.30/29). In particular:
4727 . If the quota is zero, the maildirsize file is maintained, but no quota is
4730 . If the maildir directory does not exist, it is created before any attempt
4731 to write a maildirsize file.
4733 . The quota value in the file is just a cache; if the quota is changed in
4734 the transport, the new value overrides.
4736 . A regular expression is available for excluding directories from the
4739 20. The autoreply transport checks the characters in options that define the
4740 message's headers; it allows continued headers, but it was checking with
4741 isspace() after an embedded newline instead of explicitly looking for a
4744 21. If all the "regular" hosts to which an address was routed had passed their
4745 expiry times, and had not reached their retry times, the address was
4746 bounced, even if fallback hosts were defined. Now Exim should go on to try
4749 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4750 equivalent code in the SMTP transport. Some hosts send humungous responses
4751 to HELO/EHLO, more than 1024 it seems.
4753 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4754 gives compiler warnings in some environments. I've now done it "properly",
4757 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4758 (because of problems with the built-in one) was declared to return uschar *
4759 instead of char *, causing compiler failure.
4761 25. Fixed a file descriptor leak when processing alias/forward files.
4763 26. Fixed a minor format string issue in dbfn.c.
4765 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4767 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4768 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4771 29. When a sender address is verified, it is cached, to save repeating the test
4772 when there is more than one recipient in a message. However, when the
4773 verification involves a callout, it is possible for different callout
4774 options to be set for different recipients. It is too complicated to keep
4775 track of this in the cache, so now Exim always runs a verification when a
4776 callout is required, relying on the callout cache for the optimization.
4777 The overhead is duplication of the address routing, but this should not be
4780 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4781 to be verified with callout=postmaster, and the main callout worked but the
4782 postmaster check failed, the verification correctly failed. However, if a
4783 subsequent RCPT command asked for sender verification *without* the
4784 postmaster check, incorrect caching caused this verification also to fail,
4787 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4788 it was not caching the DNS options (qualify_single, search_parents) that
4789 were used when the lookup failed. A subsequent lookup with different
4790 options therefore always gave the same answer, though there were cases
4791 where it should not have. (Example: a "domains = !$mx_any" option on a
4792 dnslookup router: the "domains" option is always processed without any
4793 widening, but the router might have qualify_single set.) Now Exim uses the
4794 cached value only when the same options are set.
4796 32. Added John Jetmore's "exipick" utility to the distribution.
4798 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4799 than a timeout (e.g. a certificate is required, and is not provided), an
4800 Exim server now closes the connection immediately. Previously it waited for
4801 the client to close - but if the client is SSL, it seems that they each
4802 wait for each other, leading to a delay before one of them times out.
4804 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4805 maintained 0.8.x compatibility because I don't think many are using it, and
4806 it is clearly obsolete.
4808 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4811 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4812 client certificate was expired. A simple patch fixes this, though I don't
4813 understand the full logic of why the verify callback is called multiple
4816 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4817 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4818 which causes problems with some clients (such as the Certicom SSL Plus
4819 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4820 disables the coutermeasure allowing Eudora to connect."
4822 38. Exim was not checking that a write() to a log file succeeded. This could
4823 lead to Bad Things if a log got too big, in particular if it hit a file
4824 size limit. Exim now panics and dies if it cannot write to a log file, just
4825 as it does if it cannot open a log file.
4827 39. Modified OS/Makefile-Linux so that it now contains
4829 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4831 The two -D definitions ensure that Exim is compiled with large file
4832 support, which makes it possible to handle log files that are bigger than
4835 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4836 instance) a domain was checked against a named list that involved a lookup,
4837 causing $domain_data to be set, then another domain was checked against the
4838 same list, then the first domain was re-checked, the value of $domain_data
4839 after the final check could be wrong. In particular, if the second check
4840 failed, it could be set empty. This bug probably also applied to
4843 41. The strip_trailing_dot option was not being applied to the address given
4844 with the -f command-line option.
4846 42. The code for reading a message's header from the spool was incrementing
4847 $received_count, but never initializing it. This meant that the value was
4848 incorrect (doubled) while delivering a message in the same process in which
4849 it was received. In the most common configuration of Exim, this never
4850 happens - a fresh exec is done - but it can happen when
4851 deliver_drop_privilege is set.
4853 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4854 it now includes up to 150 characters of the unexpected data in the log
4857 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4858 and building data strings. The size of both of these buffers was 10 000
4859 bytes - far larger than anybody would *ever* want, thought I. Needless to
4860 say, somebody hit the limit. I have increased the maximum line length to
4861 20 000 and the maximum data length of concatenated lines to 100 000. I have
4862 also fixed two bugs, because there was no checking on these buffers. Tsk,
4863 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4864 buffer is too small.
4866 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4867 lsearch lookups. Now it does.
4869 46. When parsing a route_list item in a manualroute router, a fixed-length
4870 buffer was used for the list of hosts. I made this 1024 bytes long,
4871 thinking that nobody would ever have a list of hosts that long. Wrong.
4872 Somebody had a whole pile of complicated expansion conditions, and the
4873 string was silently truncated, leading to an expansion error. It turns out
4874 that it is easier to change to an unlimited length (owing to other changes
4875 that have happened since this code was originally written) than to build
4876 structure for giving a limitation error. The length of the item that
4877 expands into the list of hosts is now unlimited.
4879 47. The lsearch lookup could not handle data where the length of text line was
4880 more than 4095 characters. Such lines were truncated, leading to shortened
4881 data being returned. It should now handle lines of any length.
4883 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4884 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4887 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4888 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4889 no longer happens in comment lines. A list of replacements is now placed
4890 at the head of all of the source files, except those whose only change is
4891 to replace PERL_COMMAND in the very first #! line.
4893 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4894 messages on the queue, with a bottom-up merge sort, using code contributed
4895 by Michael Haardt. This should make operations like -bp somewhat faster on
4896 large queues. It won't affect queue runners, except when queue_run_in_order
4899 51. Installed eximstats 1.31 in the distribution.
4901 52. Added support for SRV lookups to the dnslookup router.
4903 53. If an ACL referred to $message_body or $message_body_end, the value was not
4904 reset for any messages that followed in the same SMTP session.
4906 54. The store-handling optimization for building very long strings was not
4907 differentiating between the different store pools. I don't think this
4908 actually made any difference in practice, but I've tidied it.
4910 55. While running the routers to verify a sender address, $sender_address
4911 was still set to the sender address. This is wrong, because when routing to
4912 send a bounce to the sender, it would be empty. Therefore, I have changed
4913 it so that, while verifying a sender address, $sender_address is set to <>.
4914 (There is no change to what happens when verifying a recipient address.)
4916 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4917 target A or AAAA records (if not already returned) without resetting the
4918 qualify_single or search_parents options of the DNS resolver. These are
4919 inappropriate in this case because the targets of MX and SRV records must
4920 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4921 target that, when qualified, matched something in the local domain. These
4922 two options are now turned off when doing these lookups.
4924 57. It seems that at least some releases of Reiserfs (which does not have the
4925 concept of a fixed number of inodes) returns zero and not -1 for the
4926 number of available inodes. This interacted badly with check_spool_inodes,
4927 which assumed that -1 was the "no such thing" setting. What I have done is
4928 to check that the total number of inodes is greater than zero before doing
4929 the test of how many are available.
4931 58. When a "warn" ACL statement has a log_message modifier, the message is
4932 remembered, and not repeated. This is to avoid a lot of repetition when a
4933 message has many recipients that cause the same warning to be written.
4934 Howewer, Exim was preserving the list of already written lines for an
4935 entire SMTP session, which doesn't seem right. The memory is now reset if a
4936 new message is started.
4938 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4939 debugging output unless log_rewrite was also set.
4941 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4942 of (int)(handle) when we know that handle contains (void *)(-1).
4944 61. The Exim daemon panic-logs an error return when it closes the incoming
4945 connection. However "connection reset by peer" seems to be common, and
4946 isn't really an error worthy of noting specially, so that particular error
4949 62. When Exim is trying to find all the local interfaces, it used to panic and
4950 die if the ioctl to get the interface flags failed. However, it seems that
4951 on at least one OS (Solaris 9) it is possible to have an interface that is
4952 included in the list of interfaces, but for which you get a failure error
4953 for this call. This happens when the interface is not "plumbed" into a
4954 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4955 failure of the "get flags" call assumes that the interface is down.
4957 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4958 makes life easier for people who are doing arithmetic on fields extracted
4959 from dates, where you often get leading zeros that should not be
4960 interpreted as octal.
4962 64. Added qualify_domain to the redirect router, to override the global
4965 65. If a pathologically long header line contained very many addresses (the
4966 report of this problem mentioned 10 000) and each of them was rewritten,
4967 Exim could use up a very large amount of memory. (It kept on making new
4968 copies of the header line as it rewrote, and never released the old ones.)
4969 At the expense of a bit more processing, the header rewriting function has
4970 been changed so that it no longer eats memory in this way.
4972 66. The generation of the Received: header has been moved from the time that a
4973 message starts to be received, to the time that it finishes. The timestamp
4974 in the Received: header should now be very close to that of the <= log
4975 line. There are two side-effects of this change:
4977 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4978 logged header lines no longer include the local Received: line, because
4979 it has not yet been created. The same applies to a copy of the message
4980 that is returned to a non-SMTP sender when a message is rejected.
4982 (b) When a filter file is tested using -bf, no additional Received: header
4983 is added to the test message. After some thought, I decided that this
4986 This change does not affect the value of $received_for. It is still set
4987 after address rewriting, but before local_scan() is called.
4989 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4991 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4992 gave an unhelpful panic error message, and a defer error. I have managed to
4993 change this behaviour so that it now rejects any supplied certificate,
4994 which seems right, as the list of acceptable certificates is empty.
4996 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4997 gave an unhelpful defer error. I have not managed to make this reject any
4998 supplied certificates, but the error message it gives is "no certificate
4999 supplied", which is not helpful.
5001 70. exigrep's output now also includes lines that are not associated with any
5002 message, but which match the given pattern. Implemented by a patch from
5003 Martin Sluka, which also tidied up the Perl a bit.
5005 71. Recipient callout verification, like sender verification, was using <> in
5006 the MAIL FROM command. This isn't really the right thing, since the actual
5007 sender may affect whether the remote host accepts the recipient or not. I
5008 have changed it to use the actual sender in the callout; this means that
5009 the cache record is now keyed on a recipient/sender pair, not just the
5010 recipient address. There doesn't seem to be a real danger of callout loops,
5011 since a callout by the remote host to check the sender would use <>.
5012 [SEE ABOVE: changed after hitting problems.]
5014 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5015 temporary errors. However, in the case of such a code being given after
5016 the end of a data transmission (i.e. after ".") Exim was failing to write
5017 a retry record for the message. (Yes, there was some broken host that was
5018 actually sending 8xx at this point.)
5020 73. An unknown lookup type in a host list could cause Exim to panic-die when
5021 the list was checked. (An example that provoked this was putting <; in the
5022 middle of a list instead of at the start.) If this happened during a DATA
5023 ACL check, a -D file could be left lying around. This kind of configuration
5024 error no longer causes Exim to die; instead it causes a defer errror. The
5025 incident is still logged to the main and panic logs.
5027 74. Buglet left over from Exim 3 conversion. The message "too many messages
5028 in one connection" was written to the rejectlog but not the mainlog, except
5029 when address rewriting (yes!) was being logged.
5031 75. Added write_rejectlog option.
5033 76. When a system filter was run not as root (that is, when system_filter_user
5034 was set), the values of the $n variables were not being returned to the
5035 main process; thus, they were not subsequently available in the $sn
5038 77. Added +return_path_on_delivery log selector.
5040 78. A connection timeout was being treated differently from recipients deferred
5041 when testing hosts_max_try with a message that was older than the host's
5042 retry timeout. (The host should not be counted, thus allowing all hosts to
5043 be tried at least once before bouncing.) This may have been the cause of an
5044 occasionally reported bug whereby a message would remain on the queue
5045 longer than the retry timeout, but would be bounced if a delivery was
5046 forced. I say "may" because I never totally pinned down the problem;
5047 setting up timeout/retry tests is difficult. See also the next item.
5049 79. The ultimate address timeout was not being applied to errors that involved
5050 a combination of host plus message (for example, a timeout on a MAIL
5051 command). When an address resolved to a number of possible hosts, and they
5052 were not all tried for each delivery (e.g. because of hosts_max_try), a
5053 message could remain on the queue longer than the retry timeout.
5055 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5058 81. Fixed an obscure SMTP outgoing bug which required at least the following
5059 conditions: (a) there was another message waiting for the same server;
5060 (b) the server returned 5xx to all RCPT commands in the first message so
5061 that the message was not completed; (c) the server dropped the connection
5062 or gave a negative response to the RSET that Exim sends to abort the
5063 transaction. The observed case was a dropped connection after DATA that had
5064 been sent in pipelining mode. That is, the server had advertised PIPELINING
5065 but was not implementing it correctly. The effect of the bug was incorrect
5066 behaviour, such as trying another host, and this could lead to a crash.
5072 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5073 and daemon.c were passed as pointers to ints; they should have been
5074 pointers to socklen_t variables (which are typically unsigned ints).
5076 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5079 3. Fixed a really odd bug that affected only the testing scheme; patching a
5080 certain fixed string in the binary changed the value of another string that
5081 happened to be identical to the end of the original first string.
5083 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5084 name", it returns that address as the IP address. On some operating
5085 systems (e.g. Solaris), it also passes back the IP address string as the
5086 "host name". However, on others (e.g. Linux), it passes back an empty
5087 string. Exim wasn't checking for this, and was changing the host name to an
5088 empty string, assuming it had been canonicized.
5090 5. Although rare, it is permitted to have more than one PTR record for a given
5091 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5092 all the names associated with an address, because they do in Solaris.
5093 However, it seems that they do not in Linux for data that comes from the
5094 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5095 I found this out when I moved to a new Linux workstation and tried to run
5096 the Exim test suite.
5098 To get round this problem I have changed the code so that it now does its
5099 own call to the DNS to look up PTR records when searching for a host name.
5100 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5101 addresses that are only in /etc/hosts are still found.
5103 This behaviour is, however, controlled by an option called host_lookup_
5104 order, which defaults to "bydns:byaddr". If people want to use the other
5105 order, or indeed, just use one or the other means of lookup, they can
5106 specify it in this variable.
5108 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5109 some operating systems, this comes back from gethostbyaddr() as an empty
5110 string, and this is what Exim used to test for. However, it seems that in
5111 other systems, "." is yielded. Exim now tests for this case too.
5113 7. The values of check_spool_space and check_log_space are now held internally
5114 as a number of kilobytes instead of an absolute number of bytes. If a
5115 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5116 kilobyte. This means that much larger values can be stored.
5118 8. Exim monitor: an attempt to get the action menu when not actually pointing
5119 at a message produces an empty menu entitled "No message selected". This
5120 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5121 no entries in it ("Shell widget menu has zero width and/or height"). So I
5122 have added a single, blank menu entry in this case.
5124 9. Added ${quote_local_part.
5126 10. MIME decoding is now applied to the contents of Subject: header lines when
5129 11. Now that a reference to $sender_host_address automatically causes a reverse
5130 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5131 host lookup before query-style lookups in lists that might use this
5132 variable. This has therefore been abolished, and the "net-" prefix is no
5133 longer necessary for query-style lookups.
5135 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5136 have been a typo for LFLAGS, so it has been changed.
5138 13. The install script calls Exim with "-C /dev/null" in order to find the
5139 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5140 to be output. Howeve, since Exim outputs its version number before the
5141 error, it didn't break the script. It just looked ugly. I fixed this by
5142 always allowing "-C /dev/null" if the caller is root.
5144 14. Ignore overlarge ACL variable number when reading spool file - insurance
5145 against a later release with more variables having written the file.
5147 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5148 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5149 this, as well as the form without the "IPv6" on the front.
5151 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5152 OS/Makefile-Darwin file.
5154 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5155 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5157 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5158 variable. However, some OS use ints for inodes. I've added cast to long int
5159 to get rid of the compiler warning.
5161 19. I had forgotten to lock out "/../" in configuration file names when
5162 ALT_CONFIG_PREFIX was set.
5164 20. Routers used for verification do not need to specify transports. However,
5165 if such a router generated a host list, and callout was configured, Exim
5166 crashed, because it could not find a port number from the (non-existent)
5167 transport. It now assumes port 25 in this circumstance.
5169 21. Added the -t option to exigrep.
5171 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5172 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5173 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5174 them is set, LOOKUP_LSEARCH is forced.
5176 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5178 24. Added sender and host information to the "rejected by local_scan()" log
5179 line; previously there was no indication of these.
5181 25. Added .include_if_exists.
5183 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5184 Linux. It turns out that not all file systems support this. Apparently some
5185 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5186 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5187 not supported on the file descriptor, is now ignored when Exim is trying to
5188 sync a directory. This applies only to Linux.
5190 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5192 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5193 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5196 29. Added support for maildirsize files from supplied patch (modified a bit).
5198 30. The use of :fail: followed by an empty string could lead Exim to respond to
5199 sender verification failures with (e.g.):
5201 550 Verification failed for <xxx>
5202 550 Sender verify failed
5204 where the first response line was missing the '-' that indicates it is not
5205 the final line of the response.
5207 31. The loop for finding the name of the user that called Exim had a hardwired
5208 limit of 10; it now uses the value of finduser_retries, which is used for
5209 all other user lookups.
5211 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5214 33. Exim was neglecting to zero errno before one call of strtol() when
5215 expanding a string and expecting an integer value. On some systems this
5216 resulted in spurious "integer overflow" errors. Also, it was casting the
5217 result into an int without checking.
5219 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5220 did not work. The code looks as if it has *never* worked, though it appears
5221 to have been documented since at least releast 1.62. I have made it work.
5223 35. The "timeout_DNS" error in retry rules, also documented since at least
5224 1.62, also never worked. As it isn't clear exactly what this means, and
5225 clearly it isn't a major issue, I have abolished the feature by treating it
5226 as "timeout", and writing a warning to the main and panic logs.
5228 36. The display of retry rules for -brt wasn't always showing the error code
5231 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5232 timeout_connect_A, timeout_connect_MX.
5234 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5235 to the empty sender.
5237 39. The daemon was not analysing the content of -oX till after it had closed
5238 stderr and disconnected from the controlling terminal. This meant that any
5239 syntax errors were only noted on the panic log, and the return code from
5240 the command was 0. By re-arranging the code a little, I've made the
5241 decoding happen first, so such errors now appear on stderr, and the return
5242 code is 1. However, the actual setting up of the sockets still happens in
5243 the disconnected process, so errors there are still only recorded on the
5246 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5247 connections (as happens on some IP stacks) was logged at start up time as
5248 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5249 it from "IPv6 and IPv4", which means that two separate sockets are being
5252 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5253 says whether AF_INET or AF_INET6 was passed as an argument.
5255 42. Exiwhat output was messed up when time zones were included in log
5258 43. Exiwhat now gives more information about the daemon's listening ports,
5259 and whether -tls-on-connect was used.
5261 44. The "port" option of the smtp transport is now expanded.
5263 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5264 silently ignored. Now an error message is written to the main and panic
5267 46. There's a new ACL modifier called "logwrite" which writes to a log file
5268 as soon as it is encountered.
5270 47. Added $local_user_uid and $local_user_gid at routing time.
5272 48. Exim crashed when trying to verify a sender address that was being
5275 49. Exim was recognizing only a space character after ".include". It now also
5276 recognizes a tab character.
5278 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5279 extracting the relevant information from the specification. The man page no
5280 longer contains scrambled data for the -d option, and I've added a section
5281 at the front about calling Exim under different names.
5283 51. Added "extra_headers" argument to the "mail" command in filter files.
5285 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5288 53. Installed eximstats 1.29.
5290 54. Added transport_filter_timeout as a generic transport option.
5292 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5293 Cc: header lines. This was required by RFC 822, but it not required by RFC
5296 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5297 incoming messages that did not have them. Now it does so only if the
5298 message originates locally, that is, if there is no associated remote host
5299 address. When Resent- header lines are present, this applies to the Resent-
5300 lines rather than the non-Resent- lines.
5302 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5303 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5305 58. Messages for configuration errors now include the name of the main
5306 configuration file - useful now that there may be more than one file in a
5307 list (.included file names were always shown).
5309 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5310 for those rare installations that do not start the daemon as root or run it
5311 setuid root. I've cut out the call to initgroups() if the daemon is not
5314 60. The Exim user and group can now be bound into the binary as text strings
5315 that are looked up at the start of Exim's processing.
5317 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5319 62. Added $mailstore_basename variable.
5321 63. Installed patch to sieve.c from Michael Haardt.
5323 64. When Exim failed to open the panic log after failing to open the main log,
5324 the original message it was trying to log was written to stderr and debug
5325 output, but if they were not available (the usual case in production), it
5326 was lost. Now it is written to syslog before the two lines that record the
5327 failures to open the logs.
5329 65. Users' Exim filters run in subprocesses under the user's uid. It is
5330 possible for a "deliver" command or an alias in a "personal" command to
5331 provoke an address rewrite. If logging of address rewriting is configured,
5332 this fails because the process is not running as root or exim. There may be
5333 a better way of dealing with this, but for the moment (because 4.30 needs
5334 to be released), I have disabled address rewrite logging when running a
5335 filter in a non-root, non-exim process.
5341 1. The buildconfig auxiliary program wasn't quoting the value set for
5342 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5343 not defined. This bug was masked in 4.22 by the effect that was fixed in
5346 2. Some messages that were rejected after a message id was allocated were
5347 shown as "incomplete" by exigrep. It no longer does this for messages that
5348 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5350 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5351 have allow_domain_literals set, the ID did not get logged in the <= line.
5352 Domain literals are now always recognized in Message-ID: header lines.
5354 4. The first argument for a ${extract expansion item is the key name or field
5355 number. Leading and trailing spaces in this item were not being ignored,
5356 causing some misleading effects.
5358 5. When deliver_drop_privilege was set, single queue runner processes started
5359 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5360 same command in the process it spins off) were not dropping privilege.
5362 6. When the daemon running as "exim" started a queue runner, it always
5363 re-executed Exim in the spun-off process. This is a waste of effort when
5364 deliver_drop_privilege is set. The new process now just calls the
5365 queue-runner function directly.
5371 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5374 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5375 ignored. Though the use of -oP was forcing the writing of a pid file, it
5376 was always written to the default place.
5378 3. If the message "no IP address found for host xxxx" is generated during
5379 incoming verification, it is now followed by identification of the incoming
5380 connection (so you can more easily find what provoked it).
5382 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5384 5. Added some features to "harden" Exim a bit more against certain attacks:
5386 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5387 be put in Local/Makefile. This is like the never_users runtime option,
5388 but it cannot be overridden. The default setting is "root".
5390 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5391 prefix string with which any file named in a -C command line option
5394 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5395 is retained for -C and -D only if the caller of Exim is root. Without
5396 it, the exim user may also use -C and -D and retain privilege.
5398 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5399 command line option is disabled.
5401 6. Macro names set by the -D option must start with an upper case letter, just
5402 like macro names defined in the configuration file.
5404 7. Added "dereference=" facility to LDAP.
5406 8. Two instances of the typo "uknown" in the source files are fixed.
5408 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5409 the Configure-Makefile script screwed up while processing it.
5411 10. Incorporated PCRE 4.4.
5413 11. The SMTP synchronization check was not operating right at the start of an
5414 SMTP session. For example, it could not catch a HELO sent before the client
5415 waited for the greeting. There is now a check for outstanding input at the
5416 point when the greeting is written. Because of the duplex, asynchronous
5417 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5418 way, but not yet received, when the check is performed.
5420 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5421 on TCP/IP sockets, because this apparently causes some broken clients to
5424 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5425 unchanged) from the Cygwin maintainer.
5427 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5428 was supported instead of testing for mbx. Effectively a typo.
5430 15. The spa authenticator server code was not checking that the input it
5431 received was valid base64.
5433 16. The debug output line for the "set" modifier in ACLs was not showing the
5434 name of the variable that was being set.
5436 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5438 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5439 lookup on its own. Something else was needed to trigger the lookup. For
5440 example, a match in host_lookup or the need for a host name in a host list.
5441 Now, if $sender_host_name is referenced and the host name has not yet been
5442 looked up, a lookup is performed. If the lookup fails, the variable remains
5443 empty, and $host_lookup_failed is set to "1".
5445 19. Added "eqi" as a case-independent comparison operator.
5447 20. The saslauthd authentication condition could segfault if neither service
5448 nor realm was specified.
5450 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5451 error message that was logged was misleading, and incoming SMTP
5452 connections were dropped. The message is now more accurate, and temporary
5453 errors are given to SMTP connections.
5455 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5456 (except RSET) with a 421 error, until QUIT is received. However, it was
5457 failing to send a response to QUIT.
5459 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5460 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5461 ACL is now run after the helo_try_verify_hosts code.
5463 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5464 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5465 case-independent, so other case variants are also recognized). Apparently
5466 some systems use these upper case variants.
5468 25. If more than two messages were waiting for the same host, and a transport
5469 filter was specified for the transport, Exim sent two messages over the
5470 same TCP/IP connection, and then failed with "socket operation on non-
5471 socket" when it tried to send the third.
5473 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5475 27. The extern definition of crypt16() in expand.c was not being excluded when
5476 the OS had its own crypt16() function.
5478 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5479 as a preferred synonym for return_size_limit, both as an option and as an
5482 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5484 30. Changed the default configuration ACL to relax the local part checking rule
5485 for addresses that are not in any local domains. For these addresses,
5486 slashes and pipe symbols are allowed within local parts, but the sequence
5487 /../ is explicitly forbidden.
5489 31. SPA server authentication was not clearing the challenge buffer before
5492 32. log_message in a "warn" ACL statement was writing to the reject log as
5493 well as to the main log, which contradicts the documentation and doesn't
5494 seem right (because no rejection is happening). So I have stopped it.
5496 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5497 However, I am unable to do any testing of this.
5499 34. Fixed an infelicity in the appendfile transport. When checking directories
5500 for a mailbox, to see if any needed to be created, it was accidentally
5501 using path names with one or more superfluous leading slashes; tracing
5502 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5504 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5505 contents are added to the log line that is written for every discarded
5506 recipient. (Previously a log_message setting was ignored.)
5508 36. The ${quote: operator now quotes the string if it is empty.
5510 37. The install script runs exim in order to find its version number. If for
5511 some reason other than non-existence or emptiness, which it checks, it
5512 could not run './exim', it was installing it with an empty version number,
5513 i.e. as "exim-". This error state is now caught, and the installation is
5516 38. An argument was missing from the function that creates an error message
5517 when Exim fails to connect to the socket for saslauthd authentication.
5518 This could cause Exim to crash, or give a corrupted message.
5520 39. Added isip, isip4, and isip6 to ${if conditions.
5522 40. The ACL variables $acl_xx are now saved with the message, and can be
5523 accessed later in routers, transports, and filters.
5525 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5526 strings in the file are not string-expanded.
5528 42. If a MAIL command specified a SIZE value that was too large to fit into an
5529 int variable, the check against message_size_limit failed. Such values are
5530 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5531 day this will have to be increased, but I don't think I want to be around
5532 when emails are that large.
5539 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5540 iconv() is not standard in FreeBSD.
5542 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5543 IPv6 enabled. The observed symptom was a segmentation fault on return from
5544 the function os_common_find_running_interfaces() in src/os.c.
5546 3. In the check_special_case() function in daemon.c I had used "errno" as an
5547 argument name, which causes warnings on some systems. This was basically a
5548 typo, since it was named "eno" in the comments!
5550 4. The code that waits for the clock to tick (at a resolution of some fraction
5551 of a second) so as to ensure message-id uniqueness was always waiting for
5552 at least one whole tick, when it could have waited for less. [This is
5553 almost certainly not relevant at current processor speeds, where it is
5554 unlikely to ever wait at all. But we try to future-proof.]
5556 5. The function that sleeps for a time interval that includes fractions of a
5557 second contained a race. It did not block SIGALRM between setting the
5558 timer, and suspending (a couple of lines later). If the interval was short
5559 and the sigsuspend() was delayed until after it had expired, the suspension
5560 never ended. On busy systems this could lead to processes getting stuck for
5563 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5564 process, before it forks any delivery processes. The open lookup caching
5565 mechanism meant that the open file or database connection was passed into
5566 the delivery process. The problem was that delivery processes always tidy
5567 up cached lookup data. This could cause a problem for the next delivery
5568 process started by the queue runner, because the external queue runner
5569 process does not know about the closure. So the next delivery process
5570 still has data in the lookup cache. In the case of a file lookup, there was
5571 no problem because closing a file descriptor in a subprocess doesn't affect
5572 the parent. However, if the lookup was caching a connection to a database,
5573 the connection was closed, and the second delivery process was likely to
5574 see errors such as "PGSQL: query failed: server closed the connection
5575 unexpectedly". The problem has been fixed by closing all cached lookups
5576 in a queue runner before running a delivery process.
5578 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5579 seem to have the "const" qualifier which it has on other OS. I've
5582 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5583 *of the same type* (client or server) with the same public name that an
5584 error should be diagnosed.
5586 9. When Exim looked up a host name for an IP address, but failed to find the
5587 original IP address when looking up the host name (a safety check), it
5588 output the message "<ip address> does not match any IP for NULL", which was
5589 confusing, to say the least. The bug was that the host name should have
5590 appeared instead of "NULL".
5592 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5593 user that is built into the binary, and the -C or -D options is used, root
5594 privilege is dropped before the configuration file is read. In addition,
5595 logging is switched to stderr instead of the normal log files. If the
5596 configuration then re-defines the Exim user, the unprivileged environment
5597 is probably not what is expected, so Exim logs a panic warning message (but
5600 However, if deliver_drop_privilege is set, the unprivileged state may well
5601 be exactly what is intended, so the warning has been cut out in that case,
5602 and Exim is allowed to try to write to its normal log files.
5608 1. smtp_return_error_details was not giving details for temporary sender
5609 or receiver verification errors.
5611 2. Diagnose a configuration error if two authenticators have the same public
5614 3. Exim used not to create the message log file for a message until the first
5615 delivery attempt. This could be confusing when incoming messages were held
5616 for policy or load reasons. The message log file is now created at the time
5617 the message is received, and an initial "Received" line is written to it.
5619 4. The automatically generated man page for command line options had a minor
5620 bug that caused no ill effects; however, a more serious problem was that
5621 the procedure for building the man page automatically didn't always
5622 operate. Consequently, release 4.20 contains an out-of-date version. This
5623 shouldn't happen again.
5625 5. When building Exim with embedded Perl support, the script that builds the
5626 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5627 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5629 6. The freeze_tell option was not being used for messages that were frozen on
5630 arrival, either by an ACL or by local_scan().
5632 7. Added the smtp_incomplete_transaction log selector.
5634 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5635 was accepting AUTH without a new EHLO.
5637 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5638 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5641 10. Logging of TCP/IP connections (when configured) now happens in the main
5642 daemon process instead of the child process, so that the TCP/IP connection
5643 count is more accurate (but it can never be perfect).
5645 11. The use of "drop" in a nested ACL was not being handled correctly in the
5646 outer ACL. Now, if condition failure induced by the nested "drop" causes
5647 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5648 or "require"), the connection is dropped.
5650 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5651 that yield "discard" can now be used with an "accept" or a "discard" verb,
5652 but an error is generated for any others (because I can't see a useful way
5653 to define what should happen).
5655 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5656 are now processed in the same way as lines in the Exim configuration file.
5657 In particular, continuation lines are supported.
5659 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5661 15. Added -ti meaning -t -i.
5663 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5664 domains, and warn by logging if others are found.
5666 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5667 returned from GIFCONF when Exim is trying to find the list of interfaces on
5668 a host. The code in os.c has been modified to copy each ifreq to an aligned
5669 structure in all cases.
5671 Also, in some cases, the returned ifreq's were being copied to a 'struct
5672 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5673 means the last couple of bytes of an IPv6 address could be chopped if the
5674 ifreq contained only a normal sockaddr (14 bytes storage).
5676 18. Named domain lists were not supported in the hosts_treat_as_local option.
5677 An entry such as +xxxx was not recognized, and was treated as a literal
5680 19. Ensure that header lines added by a DATA ACL are included in the reject log
5681 if the ACL subsequently rejects the message.
5683 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5684 MD5 (which is deprecated).
5686 21. When testing a filter file using -bf, Exim was writing a message when it
5687 took the sender from a "From " line in the message, but it was not doing so
5688 when it took $return_path from a Return-Path: header line. It now does.
5690 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5691 with a valid header line field name (a series of printing characters
5692 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5694 23. Changed "disc" in the source to "disk" to conform to the documentation and
5695 the book and for uniformity.
5697 24. Ignore Sendmail's -Ooption=value command line item.
5699 25. When execve() failed while trying to run a command in a pipe transport,
5700 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5701 could be confused with a return value of 69 from the command itself. This
5702 has been changed to 127, the value the shell returns if it is asked to run
5703 a non-existent command. The wording for the related log line suggests a
5704 non-existent command as the problem.
5706 26. If received_header_text expands to an empty string, do not add a Received:
5707 header line to the message. (Well, it adds a token one on the spool, but
5708 marks it "old" so that it doesn't get used or transmitted.)
5710 27. Installed eximstats 1.28 (addition of -nt option).
5712 28. There was no check for failure on the call to getsockname() in the daemon
5713 code. This can fail if there is a shortage of resources on the system, with
5714 ENOMEM, for example. A temporary error is now given on failure.
5716 29. Contrary to the C standard, it seems that in some environments, the
5717 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5718 program. Exim now does this explicitly; it affects the formatting of
5719 timestamps using strftime().
5721 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5722 complaints. I've now initialized all the variables, to avoid this.
5724 32. Header lines added by a system filter were not being "seen" during
5725 transport-time rewrites.
5727 33. The info_callback() function passed to OpenSSL is set up with type void
5728 (*)(SSL *, int, int), as described somewhere. However, when calling the
5729 function (actually a macro) that sets it up, the type void(*)() is
5730 expected. I've put in a cast to prevent warnings from picky compilers.
5732 34. If a DNS black list lookup found a CNAME record, but there were no A
5733 records associated with the domain it pointed at, Exim crashed.
5735 35. If a DNS black list lookup returned more than one A record, Exim ignored
5736 all but the first. It now scans all returned addresses if a particular IP
5737 value is being sought. In this situation, the contents of the
5738 $dnslist_value variable are a list of all the addresses, separated by a
5741 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5742 to accept a host name and all its aliases if the forward lookup for any of
5743 them yielded the IP address of the incoming connection. Now it accepts only
5744 those names whose forward lookup yields the correct IP address. Any other
5745 names are discarded. This closes a loophole whereby a rogue DNS
5746 administrator could create reverse DNS records to break through a
5747 wildcarded host restriction in an ACL.
5749 37. If a user filter or a system filter that ran in a subprocess used any of
5750 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5751 the wrong values were passed to the pipe command ($thisaddress had the
5752 value of $0, $0 had the value of $1, etc). This bug was introduced by
5753 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5755 38. Improved the line breaking for long SMTP error messages from ACLs.
5756 Previously, if there was no break point between 40 and 75 characters, Exim
5757 left the rest of the message alone. Two changes have been made: (a) I've
5758 reduced the minimum length to 35 characters; (b) if it can't find a break
5759 point between 35 and 75 characters, it looks ahead and uses the first one
5760 that it finds. This may give the occasional overlong line, but at least the
5761 remaining text gets split now.
5763 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5764 file descriptors might be low, and that setting 1000 would always raise it.
5765 It turns out that in some environments, the limit is already over 1000 and
5766 that lowering it causes trouble. So now Exim takes care not to decrease it.
5768 40. When delivering a message, the value of $return_path is set to $sender_
5769 address at the start of routing (routers may change the value). By an
5770 oversight, this default was not being set up when an address was tested by
5771 -bt or -bv, which affected the outcome if any router or filter referred to
5774 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5775 log only when "message" or "log_message" are set. However, if one of the
5776 conditions was an address verification, or a call to a nested ACL, the
5777 messages generated by the underlying test were being passed through. This
5778 no longer happens. The underlying message is available in $acl_verify_
5779 message for both "message" and "log_message" expansions, so it can be
5780 passed through if needed.
5782 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5783 new expansion $bh_ to give the encoded byte string without charset
5784 translation. Translation happens only if iconv() is available; HAVE_ICONV
5785 indicates this at build time. HEADERS_CHARSET gives the charset to
5786 translate to; headers_charset can change it in the configuration, and
5787 "headers charset" can change it in an individual filter file.
5789 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5790 that creates RFC 2047 encoded "words" labels them as that charset instead
5791 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5792 expansion operator; (ii) when Exim creates a From: line for a local
5793 message; (iii) when a header line is rewritten to include a "phrase" part.
5795 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5796 buggy, causing it to skip the first lines of messages whose message ID
5797 ended in 'D'. This would not have bitten before Exim release 4.14, because
5798 message IDs were unlikely to end in 'D' before then. The effect was to have
5799 incorrect size information for certain domains.
5801 45. #include "config.h" was missing at the start of the crypt16.c module. This
5802 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5805 46. If there was a timeout during a "random" callout check, Exim treated it as
5806 a failure of the random address, and carried on sending RSET and the real
5807 address. If the delay was just some slowness somewhere, the response to the
5808 original RCPT would be taken as a response to RSET and so on, causing
5809 mayhem of various kinds.
5811 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5812 when I implemented it. It didn't allow for the fact that some option values
5813 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5814 the right test for positive values.
5816 48. Domain names in DNS records are case-independent. Exim always looks them up
5817 in lower case. Some resolvers return domain names in exactly the case they
5818 appear in the zone file, that is, they may contain uppercase letters. Not
5819 all resolvers do this - some return always lower case. Exim was treating a
5820 change of case by a resolver as a change of domain, similar to a widening
5821 of a domain abbreviation. This triggered its re-routing code and so it was
5822 trying to route what was effectively the same domain again. This normally
5823 caused routing to fail (because the router wouldn't handle the domain
5824 twice). Now Exim checks for this case specially, and just changes the
5825 casing of the domain that it ultimately uses when it transmits the message
5828 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5831 50. If a filter generated a file delivery with a non-absolute name (possible if
5832 no home directory exists for the router), the forbid_file option was not
5835 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5836 the existing equality matching.
5838 52. Exim was using ints instead of ino_t variables in some places where it was
5839 dealing with inode numbers.
5841 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5842 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5843 called TMPDIR, and if it finds it is different, it changes its value.
5845 54. The smtp_printf() function is now made available to local_scan() so
5846 additional output lines can be written before returning. There is also an
5847 smtp_fflush() function to enable the detection of a dropped connection.
5848 The variables smtp_input and smtp_batched_input are exported to
5851 55. Changed the default runtime configuration: the message "Unknown user"
5852 has been removed from the ACL, and instead placed on the localuser router,
5853 using the cannot_route_message feature. This means that any verification
5854 failures that generate their own messages won't get overridden. Similarly,
5855 the "Unrouteable address" message that was in the ACL for unverifiable
5856 relay addresses has also been removed.
5858 56. Added hosts_avoid_esmtp to the smtp transport.
5860 57. The exicyclog script was not checking for the esoteric option
5861 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5862 will work only if exicyclog is run under the appropriate euid.
5864 58. Following a discussion on the list, the rules by which Exim recognises line
5865 endings on incoming messages have been changed. The -dropcr and drop_cr
5866 options are now no-ops, retained only for backwards compatibility. The
5867 following line terminators are recognized: LF CRLF CR. However, special
5868 processing applies to CR:
5870 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5871 nor a local message in the state where . is a terminator.
5873 (ii) If a bare CR is encountered in a header line, an extra space is added
5874 after the line terminator so as not to end the header. The reasoning
5875 behind this is that bare CRs in header lines are most likely either
5876 to be mistakes, or people trying to play silly games.
5878 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5879 was being incorrectly given as 18 bytes larger than it should have been.
5880 This is a VOB (very old bug).
5882 60. This may never have affected anything current, but just in case it has:
5883 When the local host is found other than at the start of a list of hosts,
5884 the local host, those with the same MX, and any that follow, are discarded.
5885 When the list in question was part of a longer list of hosts, the following
5886 hosts (not currently being processed) were also being discarded. This no
5887 longer happens. I'm not sure if this situation could ever has previously
5890 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5893 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5894 header. This is something that is recommended in a new Internet Draft, and
5895 is something that is documented as being done by Sendmail. There are two
5896 possible values. For messages generated by the autoreply transport, Exim
5899 Auto-Submitted: auto-replied
5901 whereas for all other generated messages (e.g. bounces) it adds
5903 Auto-Submitted: auto-generated
5905 63. The "personal" condition in filters now includes a test for the
5906 Auto-Submitted: header. If it contains the string "auto-" the message it
5907 not considered personal.
5909 64. Added rcpt_include_affixes as a generic transport option.
5911 65. Added queue_only_override (default true).
5913 66. Added the syslog_duplication option.
5915 67. If what should have been the first header line of a message consisted of
5916 a space followed by a colon, Exim was mis-interpreting it as a header line.
5917 It isn't of course - it is syntactically invalid and should therefore be
5918 treated as the start of the message body. The misbehaviour could have
5919 caused a number of strange effects, including loss of data in subsequent
5920 header lines, and spool format errors.
5922 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5923 client host had authenticated. This control can now be exercised by an ACL
5924 for more flexibility.
5926 69. By default, callouts do not happen when testing with -bh. There is now a
5927 variant, -bhc, which does actually run the callout code, including
5928 consulting and updating the callout cache.
5930 70. Added support for saslauthd authentication, courtesy of Alexander
5933 71. If statvfs() failed on the spool or log directories while checking their
5934 size for availability, Exim confusingly gave the error "space shortage".
5935 Furthermore, in debugging mode it crashed with a floating point exception.
5936 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5937 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5938 serious problem, Exim now writes to the main and panic logs when this
5939 happens, with details of the failure. It then refuses to accept the
5940 incoming message, giving the message "spool directory problem" or "log
5941 directory problem" with a 421 code for SMTP messages.
5943 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5944 0, 1, and 2 exist, because some OS complain for execs without them (see
5945 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5946 descriptors. However, the code omitted to check that the open succeeded,
5947 causing mysterious errors if for some reason the permissions on /dev/null
5948 got screwed. Now Exim writes a message to the main and panic logs, and
5949 bombs out if it can't open /dev/null.
5951 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5952 interact so that it is all more flexible. It is supposed to remain
5953 backwards compatible. Also added extra_local_interfaces.
5955 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5956 to bomb out with an assertion failure - to the client this appears as a
5957 connection drop. This problem occurs in the part of the code that was taken
5958 from the Samba project. Fortunately, the assertion is in a very simple
5959 function, so I have fixed this by reproducing the function inline in the
5960 one place where it is called, and arranging for authentication to fail
5961 instead of killing the process with assert().
5963 75. The SPA client code was not working when the server requested OEM rather
5964 than Unicode encoding.
5966 76. Added code to make require_files with a specific uid setting more usable in
5967 the case where statting the file as root fails - usually a non-root-mounted
5968 NFS file system. When this happens and the failure is EACCES, Exim now
5969 forks a subprocess and does the per-uid checking as the relevant uid.
5971 77. Added process_log_path.
5973 78. If log_file_path was not explicitly set, a setting of check_log_space or
5974 check_log_inodes was ignored.
5976 79. If a space check for the spool or log partitions fails, the incident is now
5977 logged. Of course, in the latter case the data may get lost...
5979 80. Added the %p formatting code to string_format() so that it can be used to
5980 print addresses in debug_print(). Adjusted all the address printing in the
5981 debugging in store.c to use %p rather than %d.
5983 81. There was a concern that a line of code in smtp_in.c could overflow a
5984 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5985 initially expressed, the concern was not well-founded, because trailing
5986 spaces are removed early. However, if the trailing spaces were followed by
5987 a NULL, they did not get removed, so the overflow was possible. Two fixes
5990 (a) I re-wrote the offending code in a cleaner fashion.
5991 (b) If an incoming SMTP command contains a NULL character, it is rejected
5994 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5995 runs initgroups(), so that if the Exim user is in any additional groups,
5996 they will be used during message reception.
6002 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob