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
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router (applying to the forward lookup).
90 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
91 of ldap servers used for a specific lookup. Patch provided by Heiko
94 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
100 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
102 PP/02 Make -n do something, by making it not do something.
103 When combined with -bP, the name of an option is not output.
105 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
108 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
109 $sender_host_name and config options to manage this, and basic check
112 PP/05 DSCP support for outbound connections and control modifier for inbound.
114 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
115 (Only plugin which currently uses this is kerberos4, which nobody should
116 be using, but we should make it available and other future plugins might
117 conceivably use it, even though it would break NAT; stuff *should* be
118 using channel bindings instead).
120 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
121 name; added for Sendmail compatibility; requires admin caller.
122 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
123 just ignore it); requires trusted caller.
124 Also parse but ignore: -Ac -Am -X<logfile>
127 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
129 TL/02 Add +smtp_confirmation as a default logging option.
131 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
132 Patch by Magnus Holmgren from 2007-02-20.
134 TL/04 Bugzilla 1281 - Spec typo.
135 Bugzilla 1283 - Spec typo.
136 Bugzilla 1290 - Spec grammar fixes.
138 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
140 TL/06 Add Experimental DMARC support using libopendmarc libraries.
142 TL/07 Fix an out of order global option causing a segfault. Reported to dev
143 mailing list by by Dmitry Isaikin.
145 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
147 JH/02 Support "G" suffix to numbers in ${if comparisons.
149 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
151 NM/01 Bugzilla 1197 - Spec typo
152 Bugzilla 1196 - Spec examples corrections
154 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
156 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
157 gnutls_enable_pkcs11, but renamed to more accurately indicate its
160 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
161 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
163 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
164 "acl {{name}{arg}...}", and optional args on acl condition
167 JH/05 Permit multiple router/transport headers_add/remove lines.
169 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
171 JH/07 Avoid using a waiting database for a single-message-only transport.
172 Performance patch from Paul Fisher. Bugzilla 1262.
174 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
177 JH/09 Add $headers_added variable, with content from use of ACL modifier
178 add_header (but not yet added to the message). Bugzilla 199.
180 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
181 Pulled from Bugzilla 817 by Wolfgang Breyha.
183 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
185 (nb: this is the same fix as in Exim 4.80.1)
187 JH/11 Add A= logging on delivery lines, and a client_set_id option on
190 JH/12 Add optional authenticated_sender logging to A= and a log_selector
193 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
195 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
196 advertise SMTP AUTH mechanism to us, instead of a generic
197 protocol violation error. Also, make Exim more robust to bad
198 data from the Dovecot auth socket.
200 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
202 When a queue runner is handling a message, Exim first routes the
203 recipient addresses, during which it prunes them based on the retry
204 hints database. After that it attempts to deliver the message to
205 any remaining recipients. It then updates the hints database using
208 So if a recipient address works intermittently, it can get repeatedly
209 deferred at routing time. The retry hints record remains fresh so the
210 address never reaches the final cutoff time.
212 This is a fairly common occurrence when a user is bumping up against
213 their storage quota. Exim had some logic in its local delivery code
214 to deal with this. However it did not apply to per-recipient defers
215 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
217 This change adds a proper retry rule check during routing so that the
218 final cutoff time is checked against the message's age. We only do
219 this check if there is an address retry record and there is not a
220 domain retry record; this implies that previous attempts to handle
221 the address had the retry_use_local_parts option turned on. We use
222 this as an approximation for the destination being like a local
223 delivery, as in LMTP.
225 I suspect this new check makes the old local delivery cutoff check
226 redundant, but I have not verified this so I left the code in place.
228 TF/02 Correct gecos expansion when From: is a prefix of the username.
230 Test 0254 submits a message to Exim with the header
234 When I ran the test suite under the user fanf2, Exim expanded
235 the header to contain my full name, whereas it should have added
236 a Resent-Sender: header. It erroneously treats any prefix of the
237 username as equal to the username.
239 This change corrects that bug.
241 GF/01 DCC debug and logging tidyup
242 Error conditions log to paniclog rather than rejectlog.
243 Debug lines prefixed by "DCC: " to remove any ambiguity.
245 TF/03 Avoid unnecessary rebuilds of lookup-related code.
247 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
248 Bug spotted by Jeremy Harris; was flawed since initial commit.
249 Would have resulted in OCSP responses post-SNI triggering an Exim
250 NULL dereference and crash.
252 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
254 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
255 Bug detection, analysis and fix by Samuel Thibault.
256 Bugzilla 1331, Debian bug #698092.
258 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
260 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
261 Server implementation by Todd Lyons, client by JH.
262 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
263 config variable "prdr_enable" controls whether the server
264 advertises the facility. If the client requests PRDR a new
265 acl_data_smtp_prdr ACL is called once for each recipient, after
266 the body content is received and before the acl_smtp_data ACL.
267 The client is controlled by bolth of: a hosts_try_prdr option
268 on the smtp transport, and the server advertisement.
269 Default client logging of deliveries and rejections involving
270 PRDR are flagged with the string "PRDR".
272 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
273 fclose(). Diagnosis by Todd Lyons.
275 PP/17 Update configure.default to handle IPv6 localhost better.
276 Patch by Alain Williams (plus minor tweaks).
279 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
280 This is now consistent with GnuTLS, and is now documented: the
281 previous undocumented portable approach to treating the option as
282 unset was to force an expansion failure. That still works, and
283 an empty string is now equivalent.
285 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
286 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
287 not performing validation itself.
289 PP/20 Added force_command boolean option to pipe transport.
290 Patch from Nick Koston, of cPanel Inc.
292 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
295 TF/04 Added udpsend ACL modifer and hexquote expansion operator
297 PP/21 Fix eximon continuous updating with timestamped log-files.
298 Broken in a format-string cleanup in 4.80, missed when I repaired the
299 other false fix of the same issue.
300 Report and fix from Heiko Schlichting.
303 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
304 Report from Prashanth Katuri.
306 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
307 It's SecureTransport, so affects any MacOS clients which use the
308 system-integrated TLS libraries, including email clients.
310 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
311 using a MIME ACL for non-SMTP local injection.
312 Report and assistance in diagnosis by Warren Baker.
314 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
316 JH/16 Fix comparisons for 64b. Bugzilla 1385.
318 TL/09 Add expansion variable $authenticated_fail_id to keep track of
319 last id that failed so it may be referenced in subsequent ACL's.
321 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
324 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
325 ldap library initialization, allowing self-signed CA's to be
326 used. Also properly sets require_cert option later in code by
327 using NULL (global ldap config) instead of ldap handle (per
328 session). Bug diagnosis and testing by alxgomz.
330 TL/12 Enhanced documentation in the ratelimit.pl script provided in
331 the src/util/ subdirectory.
333 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
334 renamed to Transport Post Delivery Action by Jeremy Harris, as
337 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
338 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
339 redis_servers = needs to be configured which will be used by the redis
340 lookup. Patch from Warren Baker, of The Packet Hub.
342 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
344 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
345 hostname or reverse DNS when processing a host list. Used suggestions
346 from multiple comments on this bug.
348 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
350 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
351 Missed a few lines, added it to make the runtest require no keyboard
354 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
355 contains upper case chars. Make router use caseful_local_part.
357 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
358 support when GnuTLS has been built with p11-kit.
364 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
366 This, or similar/improved, will also be change PP/11 of 4.82.
372 PP/01 Handle short writes when writing local log-files.
373 In practice, only affects FreeBSD (8 onwards).
374 Bugzilla 1053, with thanks to Dmitry Isaikin.
376 NM/01 Bugzilla 949 - Documentation tweak
378 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
381 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
383 PP/02 Implemented gsasl authenticator.
385 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
387 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
388 `pkg-config foo` for cflags/libs.
390 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
391 with rest of GSASL and with heimdal_gssapi.
393 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
394 `pkg-config foo` for cflags/libs for the TLS implementation.
396 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
397 properties get this fed in as external SSF. A number of robustness
398 and debugging improvements to the cyrus_sasl authenticator.
400 PP/08 cyrus_sasl server now expands the server_realm option.
402 PP/09 Bugzilla 1214 - Log authentication information in reject log.
403 Patch by Jeremy Harris.
405 PP/10 Added dbmjz lookup type.
407 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
409 PP/12 MAIL args handles TAB as well as SP, for better interop with
410 non-compliant senders.
411 Analysis and variant patch by Todd Lyons.
413 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
414 Bug report from Lars Müller <lars@samba.org> (via SUSE),
415 Patch from Dirk Mueller <dmueller@suse.com>
417 PP/13 tls_peerdn now print-escaped for spool files.
418 Observed some $tls_peerdn in wild which contained \n, which resulted
419 in spool file corruption.
421 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
422 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
423 or write after TLS renegotiation, which otherwise led to messages
426 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
427 as a tracking header (ie: a signed header comes before the signature).
428 Patch from Wolfgang Breyha.
430 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
431 comma-sep list; embedded commas doubled.
433 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
435 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
437 Report and patch from Dmitry Banschikov.
439 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
440 Removed SSL_clear() after SSL_new() which led to protocol negotiation
441 failures. We appear to now support TLS1.1+ with Exim.
443 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
444 lets Exim select keys and certificates based upon TLS SNI from client.
445 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
446 before an outbound SMTP session. New log_selector, +tls_sni.
448 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
449 NULL dereference. Report and patch from Alun Jones.
451 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
452 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
453 Not seeing resolver debug output on NetBSD, but suspect this is a
454 resolver implementation change.
456 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
457 Left warnings. Added "eximon gdb" invocation mode.
459 PP/21 Defaulting "accept_8bitmime" to true, not false.
461 PP/22 Added -bw for inetd wait mode support.
463 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
464 locate the relevant includes and libraries. Made this the default.
466 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
467 Bugzilla 1246, report and most of solution from Tomasz Kusy.
469 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
470 This may cause build issues on older platforms.
472 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
473 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
474 gnutls_require_mac & gnutls_require_protocols (no longer supported).
475 Added SNI support via GnuTLS too.
476 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
478 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
480 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
481 Applied second patch from Janne, implementing suggestion to default
482 multiple-strings-in-record handling to match SPF spec.
484 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
486 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
487 read-only, out of scope).
488 Patch from Wolfgang Breyha, report from Stuart Northfield.
490 PP/29 Fix three issues highlighted by clang analyser static analysis.
491 Only crash-plausible issue would require the Cambridge-specific
492 iplookup router and a misconfiguration.
493 Report from Marcin Mirosław.
495 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
497 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
498 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
499 As part of this, removing so much warning spew let me fix some minor
500 real issues in debug logging.
502 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
503 assignment on my part. Fixed.
505 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
506 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
507 Janne Snabb (who went above and beyond: thank you).
509 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
510 string otherwise requires a connection and a bunch more work and it's
511 relatively easy to get wrong. Should also expose TLS library linkage
514 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
515 64-bit ${eval} (JH/03).
517 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
518 GNU libc to support some of the 64-bit stuff, should not lead to
519 conflicts. Defined before os.h is pulled in, so if a given platform
520 needs to override this, it can.
522 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
523 protection layer was required, which is not implemented.
524 Bugzilla 1254, patch from Wolfgang Breyha.
526 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
527 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
528 tls_dhparam take prime identifiers. Also unbreak combination of
529 OpenSSL+DH_params+TLSSNI.
531 PP/39 Disable SSLv2 by default in OpenSSL support.
537 PP/01 Solaris build fix for Oracle's LDAP libraries.
538 Bugzilla 1109, patch from Stephen Usher.
540 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
542 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
545 TF/02 Fix a couple more cases where we did not log the error message
546 when unlink() failed. See also change 4.74-TF/03.
548 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
549 lock up or crash if it happened to be inside a call to libc when it
550 got a SIGUSR1 from exiwhat.
552 The SIGUSR1 handler appends the current process status to the process
553 log which is later printed by exiwhat. It used to use the general
554 purpose logging code to do this, but several functions it calls are
555 not safe for signals.
557 The new output code in the SIGUSR1 handler is specific to the process
558 log, and simple enough that it's easy to inspect for signal safety.
559 Removing some special cases also simplifies the general logging code.
560 Removing the spurious timestamps from the process log simplifies
563 TF/04 Improved ratelimit ACL condition.
565 The /noupdate option has been deprecated in favour of /readonly which
566 has clearer semantics. The /leaky, /strict, and /readonly update modes
567 are mutually exclusive. The update mode is no longer included in the
568 database key; it just determines when the database is updated. (This
569 means that when you upgrde Exim will forget old rate measurements.)
571 Exim now checks that the per_* options are used with an update mode that
572 makes sense for the current ACL. For example, when Exim is processing a
573 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
574 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
575 must specify per_mail/readonly. If you omit the update mode it defaults to
576 /leaky where that makes sense (as before) or /readonly where required.
578 The /noupdate option is now undocumented but still supported for
579 backwards compatibility. It is equivalent to /readonly except that in
580 ACLs where /readonly is required you may specify /leaky/noupdate or
581 /strict/noupdate which are treated the same as /readonly.
583 A useful new feature is the /count= option. This is a generalization
584 of the per_byte option, so that you can measure the throughput of other
585 aggregate values. For example, the per_byte option is now equivalent
586 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
588 The per_rcpt option has been generalized using the /count= mechanism
589 (though it's more complicated than the per_byte equivalence). When it is
590 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
591 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
592 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
593 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
594 Note that using per_rcpt with a non-readonly update mode in more than
595 one ACL will cause the recipients to be double-counted. (The per_mail
596 and per_byte options don't have this problem.)
598 The handling of very low rates has changed slightly. If the computed rate
599 is less than the event's count (usually one) then this event is the first
600 after a long gap. In this case the rate is set to the same as this event's
601 count, so that the first message of a spam run is counted properly.
603 The major new feature is a mechanism for counting the rate of unique
604 events. The new per_addr option counts the number of different
605 recipients that someone has sent messages to in the last time period. It
606 behaves like per_rcpt if all the recipient addresses are different, but
607 duplicate recipient addresses do not increase the measured rate. Like
608 the /count= option this is a general mechanism, so the per_addr option
609 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
610 example, measure the rate that a client uses different sender addresses
611 with the options per_mail/unique=$sender_address. There are further
612 details in the main documentation.
614 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
616 TF/06 Removed a few PCRE remnants.
618 TF/07 Automatically extract Exim's version number from tags in the git
619 repository when doing development or release builds.
621 PP/02 Raise smtp_cmd_buffer_size to 16kB.
622 Bugzilla 879. Patch from Paul Fisher.
624 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
625 Heavily based on revision 40f9a89a from Simon Arlott's tree.
628 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
630 PP/05 Variable $av_failed, true if the AV scanner deferred.
631 Bugzilla 1078. Patch from John Horne.
633 PP/06 Stop make process more reliably on build failure.
634 Bugzilla 1087. Patch from Heiko Schlittermann.
636 PP/07 Make maildir_use_size_file an _expandable_ boolean.
637 Bugzilla 1089. Patch from Heiko Schlittermann.
639 PP/08 Handle ${run} returning more data than OS pipe buffer size.
640 Bugzilla 1131. Patch from Holger Weiß.
642 PP/09 Handle IPv6 addresses with SPF.
643 Bugzilla 860. Patch from Wolfgang Breyha.
645 PP/10 GnuTLS: support TLS 1.2 & 1.1.
647 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
650 PP/11 match_* no longer expand right-hand-side by default.
651 New compile-time build option, EXPAND_LISTMATCH_RHS.
652 New expansion conditions, "inlist", "inlisti".
654 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
656 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
658 PP/14 fix log_write() format string regression from TF/03.
659 Bugzilla 1152. Patch from Dmitry Isaikin.
665 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
667 PP/02 Harmonised TLS library version reporting; only show if debugging.
668 Layout now matches that introduced for other libraries in 4.74 PP/03.
670 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
672 PP/04 New "dns_use_edns0" global option.
674 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
677 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
678 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
680 TK/01 Updated PolarSSL code to 0.14.2.
681 Bugzilla 1097. Patch from Andreas Metzler.
683 PP/07 Catch divide-by-zero in ${eval:...}.
686 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
689 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
690 format-string attack -- SECURITY: remote arbitrary code execution.
692 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
693 time unintentionally subject to list matching rules, letting the header
694 cause arbitrary Exim lookups (of items which can occur in lists, *not*
695 arbitrary string expansion). This allowed for information disclosure.
697 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
698 INT_MIN/-1 -- value coerced to INT_MAX.
704 NM/01 Workround for PCRE version dependency in version reporting
707 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
708 This fixes portability to compilers other than gcc, notably
709 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
711 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
712 makefiles for portability to HP-UX and POSIX correctness.
714 PP/01 Permit LOOKUP_foo enabling on the make command-line.
715 Also via indented variable definition in the Makefile.
716 (Debugging by Oliver Heesakkers).
718 PP/02 Restore caching of spamd results with expanded spamd_address.
719 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
721 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
722 Improves build reliability. Fix from: Frank Elsner
724 NM/02 Fix wide character breakage in the rfc2047 coding
725 Fixes bug 1064. Patch from Andrey N. Oktyabrski
727 NM/03 Allow underscore in dnslist lookups
728 Fixes bug 1026. Patch from Graeme Fowler
730 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
731 Code patches from Adam Ciarcinski of NetBSD.
733 NM/04 Fixed exiqgrep to cope with mailq missing size issue
736 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
737 is logged, to avoid truncation. Patch from John Horne.
739 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
740 Patch from Jakob Hirsch.
742 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
743 SQL string expansion failure details.
744 Patch from Andrey Oktyabrski.
746 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
747 Patch from Simon Arlott.
749 PP/09 New lookups functionality failed to compile on old gcc which rejects
750 extern declarations in function scope.
751 Patch from Oliver Fleischmann
753 PP/10 Use sig_atomic_t for flags set from signal handlers.
754 Check getgroups() return and improve debugging.
755 Fixed developed for diagnosis in bug 927 (which turned out to be
758 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
759 Patch from Mark Zealey.
761 PP/12 Bugzilla 1056: Improved spamd server selection.
762 Patch from Mark Zealey.
764 PP/13 Bugzilla 1086: Deal with maildir quota file races.
765 Based on patch from Heiko Schlittermann.
767 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
768 Patch from Uwe Doering, sign-off by Michael Haardt.
770 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
771 variable declaration deep within a block. Bug and patch from
774 PP/15 lookups-Makefile IRIX compatibilty coercion.
776 PP/16 Make DISABLE_DKIM build knob functional.
778 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
779 Patch by Simon Arlott
781 TF/03 Fix valgrind.h portability to C89 compilers that do not support
782 variable argument macros. Our copy now differs from upstream.
788 TF/01 Failure to get a lock on a hints database can have serious
789 consequences so log it to the panic log.
791 TF/02 Log LMTP confirmation messages in the same way as SMTP,
792 controlled using the smtp_confirmation log selector.
794 TF/03 Include the error message when we fail to unlink a spool file.
796 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
797 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
798 for maintaining out-of-tree patches for some time.
800 PP/01 Bugzilla 139: Documentation and portability issues.
801 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
802 Handle per-OS dynamic-module compilation flags.
804 PP/02 Let /dev/null have normal permissions.
805 The 4.73 fixes were a little too stringent and complained about the
806 permissions on /dev/null. Exempt it from some checks.
807 Reported by Andreas M. Kirchwitz.
809 PP/03 Report version information for many libraries, including
810 Exim version information for dynamically loaded libraries. Created
811 version.h, now support a version extension string for distributors
812 who patch heavily. Dynamic module ABI change.
814 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
815 privilege escalation vulnerability whereby the Exim run-time user
816 can cause root to append content of the attacker's choosing to
819 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
822 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
823 If dropping privileges for untrusted macros, we disabled normal logging
824 on the basis that it would fail; for the Exim run-time user, this is not
825 the case, and it resulted in successful deliveries going unlogged.
826 Fixed. Reported by Andreas Metzler.
832 PP/01 Date: & Message-Id: revert to normally being appended to a message,
833 only prepend for the Resent-* case. Fixes regression introduced in
834 Exim 4.70 by NM/22 for Bugzilla 607.
836 PP/02 Include check_rfc2047_length in configure.default because we're seeing
837 increasing numbers of administrators be bitten by this.
839 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
841 PP/03 Bugzilla 994: added openssl_options main configuration option.
843 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
845 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
847 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
849 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
850 without a peer certificate, leading to a segfault because of an
851 assumption that peers always have certificates. Be a little more
852 paranoid. Problem reported by Martin Tscholak.
854 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
855 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
856 NB: ClamAV planning to remove STREAM in "middle of 2010".
857 CL also introduces -bmalware, various -d+acl logging additions and
858 more caution in buffer sizes.
860 PP/09 Implemented reverse_ip expansion operator.
862 PP/10 Bugzilla 937: provide a "debug" ACL control.
864 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
866 PP/12 Bugzilla 973: Implement --version.
868 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
870 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
872 PP/15 Bugzilla 816: support multiple condition rules on Routers.
874 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
875 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
876 ignore trailing whitespace.
878 JJ/02 prevent non-panic DKIM error from being sent to paniclog
880 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
883 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
884 Notification from Dr Andrew Aitchison.
886 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
887 ExtendedDetectionInfo response format.
888 Notification from John Horne.
890 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
893 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
894 XSL and documented dependency on system catalogs, with examples of how
897 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
900 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
901 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
902 configuration file which is writeable by the Exim user or group.
904 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
905 of configuration files to cover files specified with the -C option if
906 they are going to be used with root privileges, not just the default
909 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
910 option (effectively making it always true).
912 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
913 files to be used while preserving root privileges.
915 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
916 that rogue child processes cannot use them.
918 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
919 run-time user, instead of root.
921 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
922 Exim run-time user without dropping privileges.
924 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
925 result string, instead of calling string_vformat() twice with the same
928 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
929 for other users. Others should always drop root privileges if they use
930 -C on the command line, even for a whitelisted configure file.
932 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
934 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
940 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
941 $data_path, and $header_path variables; fixed documentation bugs and
944 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
945 exipick to access non-standard spools, including the "frozen" queue
948 NM/01 Bugzilla 965: Support mysql stored procedures.
949 Patch from Alain Williams
951 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
953 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
954 Patch from Andreas Metzler
956 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
957 Patch from Kirill Miazine
959 NM/05 Bugzilla 671: Added umask to procmail example.
961 JJ/03 installed exipick 20100323.0, fixing doc bug
963 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
964 directory. Notification and patch from Dan Rosenberg.
966 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
968 TK/02 Improve log output when DKIM signing operation fails.
970 MH/01 Treat the transport option dkim_domain as a colon separated
971 list, not as a single string, and sign the message with each element,
972 omitting multiple occurences of the same signer.
974 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
975 Bugzilla 985, 986. Patch by Simon Arlott
977 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
978 Patch by Simon Arlott
980 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
981 MBX locking. Notification from Dan Rosenberg.
987 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
989 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
991 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
993 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
995 NM/04 Bugzilla 915: Flag broken perl installation during build.
1001 TK/01 Added patch by Johannes Berg that expands the main option
1002 "spamd_address" if it starts with a dollar sign.
1004 TK/02 Write list of recipients to X-Envelope-Sender header when building
1005 the mbox-format spool file for content scanning (suggested by Jakob
1008 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1009 (http://www.dcc-servers.net/) support via dccifd. Activated by
1010 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1012 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1013 by Mark Daniel Reidel <mr@df.eu>.
1015 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1016 When building exim an external PCRE library is now needed -
1017 PCRE is a system library on the majority of modern systems.
1018 See entry on PCRE_LIBS in EDITME file.
1020 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1021 conversation. Added nologin parameter to request.
1022 Patch contributed by Kirill Miazine.
1024 TF/01 Do not log submission mode rewrites if they do not change the address.
1026 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1028 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1029 log files in place. Contributed by Roberto Lima.
1031 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1033 TF/03 Bugzilla 615: When checking the local_parts router precondition
1034 after a local_part_suffix or local_part_prefix option, Exim now
1035 does not use the address's named list lookup cache, since this
1036 contains cached lookups for the whole local part.
1038 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1039 Robert Millan. Documentation is in experimental-spec.txt.
1041 TF/04 Bugzilla 668: Fix parallel build (make -j).
1043 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1045 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1046 Patch provided by Jan Srzednicki.
1048 TF/05 Leading white space used to be stripped from $spam_report which
1049 wrecked the formatting. Now it is preserved.
1051 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1052 that they are available at delivery time.
1054 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1056 TF/08 TLS error reporting now respects the incoming_interface and
1057 incoming_port log selectors.
1059 TF/09 Produce a more useful error message if an SMTP transport's hosts
1060 setting expands to an empty string.
1062 NM/06 Bugzilla 744: EXPN did not work under TLS.
1063 Patch provided by Phil Pennock.
1065 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1066 Patch provided by Richard Godbee.
1068 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1069 acl_smtp_notquit, added index entry.
1071 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1072 Patch provided by Eugene Bujak.
1074 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1075 accept(). Patch provided by Maxim Dounin.
1077 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1078 Patch provided by Phil Pennock.
1080 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1082 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1083 Patch provided by Brad "anomie" Jorsch.
1085 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1086 Patch provided by Dean Brooks.
1088 TK/05 Add native DKIM support (does not depend on external libraries).
1090 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1091 Patch provided by Graeme Fowler.
1093 NM/16 Bugzilla 851: Documentation example syntax fix.
1095 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1097 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1100 NM/19 Bugzilla 745: TLS version reporting.
1101 Patch provided by Phil Pennock.
1103 NM/20 Bugzilla 167: bool: condition support.
1104 Patch provided by Phil Pennock.
1106 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1107 clients. Patch provided by Phil Pennock.
1109 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1110 Patch provided by Brad "anomie" Jorsch.
1112 NM/23 Bugzilla 687: Fix misparses in eximstats.
1113 Patch provided by Heiko Schlittermann.
1115 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1116 Patch provided by Heiko Schlittermann.
1118 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1119 plus update to original patch.
1121 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1123 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1124 Patch provided by David Brownlee.
1126 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1128 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1130 NM/30 Bugzilla 888: TLS documentation bugfixes.
1132 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1134 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1135 Unlike the original bugzilla I have changed all shell scripts in src tree.
1137 NM/33 Bugzilla 898: Transport filter timeout fix.
1138 Patch by Todd Rinaldo.
1140 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1141 Patch by Serge Demonchaux.
1143 NM/35 Bugzilla 39: Base64 decode bug fixes.
1144 Patch by Jakob Hirsch.
1146 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1148 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1150 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1152 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1158 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1159 ALT-N's libdkim that I have put here:
1160 http://duncanthrax.net/exim-experimental/
1162 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1163 were called 'true' and it seems that C99 defines that as a reserved
1164 keyword to be used with 'bool' variable types. That means you could
1165 not include C99-style headers which use bools without triggering
1166 build errors in sieve.c.
1168 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1169 as mailq or other aliases. Changed the --help handling significantly
1170 to do whats expected. exim_usage() emits usage/help information.
1172 SC/01 Added the -bylocaldomain option to eximstats.
1174 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1176 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1178 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1184 PH/01 Another patch from the Sieve maintainer.
1186 PH/02 When an IPv6 address is converted to a string for single-key lookup
1187 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1188 dots are used instead of colons so that keys in lsearch files need not
1189 contain colons. This was done some time before quoting was made available
1190 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1191 (notated using the quote facility) so as to distinguish them from IPv4
1192 keys. This meant that lookups for IP addresses in host lists did not work
1193 for iplsearch lookups.
1195 This has been fixed by arranging for IPv6 addresses to be expressed with
1196 colons if the lookup type is iplsearch. This is not incompatible, because
1197 previously such lookups could never work.
1199 The situation is now rather anomolous, since one *can* have colons in
1200 ordinary lsearch keys. However, making the change in all cases is
1201 incompatible and would probably break a number of configurations.
1203 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1206 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1207 conversion specification without a maximum field width, thereby enabling
1208 a rogue spamd server to cause a buffer overflow. While nobody in their
1209 right mind would setup Exim to query an untrusted spamd server, an
1210 attacker that gains access to a server running spamd could potentially
1211 exploit this vulnerability to run arbitrary code as the Exim user.
1213 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1214 $primary_hostname instead of what libspf2 thinks the hosts name is.
1216 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1217 a directory entry by the name of the lookup key. Previously, if a
1218 symlink pointed to a non-existing file or a file in a directory that
1219 Exim lacked permissions to read, a lookup for a key matching that
1220 symlink would fail. Now it is enough that a matching directory entry
1221 exists, symlink or not. (Bugzilla 503.)
1223 PH/03 The body_linecount and body_zerocount variables are now exported in the
1226 PH/04 Added the $dnslist_matched variable.
1228 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1229 This means they are set thereafter only if the connection becomes
1232 PH/06 Added the client_condition to authenticators so that some can be skipped
1233 by clients under certain conditions.
1235 PH/07 The error message for a badly-placed control=no_multiline_responses left
1236 "_responses" off the end of the name.
1238 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1240 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1241 (without spaces) instead of just copying the configuration text.
1243 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1245 PH/11 Added $max_received_linelength.
1247 PH/12 Added +ignore_defer and +include_defer to host lists.
1249 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1250 way in which PCRE > 7.0 is built.
1252 PH/14 Implemented queue_only_load_latch.
1254 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1255 MAIL command. The effect was to mangle the value on 64-bit systems.
1257 PH/16 Another patch from the Sieve maintainer.
1259 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1261 PH/18 If a system quota error occurred while trying to create the file for
1262 a maildir delivery, the message "Mailbox is full" was not appended to the
1263 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1264 only to a quota excession during the actual writing of the file.
1266 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1267 characters?) which causes problems in log lines. The DN values are now
1268 passed through string_printing() before being added to log lines.
1270 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1271 and InterBase are left for another time.)
1273 PH/21 Added message_body_newlines option.
1275 PH/22 Guard against possible overflow in moan_check_errorcopy().
1277 PH/23 POSIX allows open() to be a macro; guard against that.
1279 PH/24 If the recipient of an error message contained an @ in the local part
1280 (suitably quoted, of course), incorrect values were put in $domain and
1281 $local_part during the evaluation of errors_copy.
1287 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1288 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1291 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1292 issue a MAIL command.
1294 PH/02 In an ACL statement such as
1296 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1298 if a client was not listed at all, or was listed with a value other than
1299 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1300 the condition was not true (as it should be), so access was not denied.
1301 The bug was that the ! inversion was incorrectly passed on to the second
1302 item. This has been fixed.
1304 PH/03 Added additional dnslists conditions == and =& which are different from
1305 = and & when the dns lookup returns more than one IP address.
1307 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1308 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1310 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1311 FSYNC, which compiles an option called disable_fsync that allows for
1312 bypassing fsync(). The documentation is heavily laced with warnings.
1314 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1316 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1317 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1318 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1319 including adding "make clean"; (3) Added -fPIC when compiling the test
1320 dynamically loaded module, to get rid of a warning.
1322 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1323 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1324 = 0s. The bug is otherwise harmless.
1326 PH/07 There was a bug in the dovecot authenticator such that the value of
1327 $auth1 could be overwritten, and so not correctly preserved, after a
1328 successful authentication. This usually meant that the value preserved by
1329 the server_setid option was incorrect.
1331 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1333 PH/09 Installed PCRE release 7.0.
1335 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1336 run for batched SMTP input. It is now run at the start of every message
1337 in the batch. While fixing this I discovered that the process information
1338 (output by running exiwhat) was not always getting set for -bs and -bS
1339 input. This is fixed, and it now also says "batched" for BSMTP.
1341 PH/11 Added control=no_pipelining.
1343 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1344 patch, slightly modified), and move the expansion of helo_data till after
1345 the connection is made in the smtp transport (so it can use these
1348 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1350 PH/14 Added log_selector = +pid.
1352 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1354 PH/16 Add ${if forany and ${if forall.
1356 PH/17 Added dsn_from option to vary the From: line in DSNs.
1358 PH/18 Flush SMTP output before performing a callout, unless control =
1359 no_callout_flush is set.
1361 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1362 was true (the default) a successful delivery failed to delete the retry
1363 item, thus causing premature timeout of the address. The bug is now
1366 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1368 PH/21 Long custom messages for fakedefer and fakereject are now split up
1369 into multiline reponses in the same way that messages for "deny" and
1370 other ACL rejections are.
1372 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1373 with slight modification.
1375 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1376 draft, changing the syntax and factoring some duplicate code".
1378 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1379 for deliveries of the second and subsequent messages over the same SMTP
1382 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1383 ${reduce, with only minor "tidies".
1385 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1387 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1388 expansion side effects.
1390 PH/27 When a message times out after an over-quota error from an Exim-imposed
1391 quota, the bounce message says "mailbox is full". This message was not
1392 being given when it was a system quota that was exceeded. It now should
1395 MH/03 Made $recipients available in local_scan(). local_scan() already has
1396 better access to the recipient list through recipients_list[], but
1397 $recipients can be useful in postmaster-provided expansion strings.
1399 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1400 in the case of a MAIL command with additional options following the
1401 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1402 were accidentally chopped off.
1404 PH/29 SMTP synchronization checks are implemented when a command is read -
1405 there is a check that no more input is waiting when there shouldn't be
1406 any. However, for some commands, a delay in an ACL can mean that it is
1407 some time before the response is written. In this time, more input might
1408 arrive, invalidly. So now there are extra checks after an ACL has run for
1409 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1410 pipelining has not been advertised.
1412 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1414 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1415 returned in the SMTP response when smtp_return_error_details was false.
1416 This has been fixed.
1418 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1419 instead of the C I/O that was originally supplied, because problems were
1420 reported on Solaris.
1422 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1423 Exim which did not show up earlier: it was assuming that a call to
1424 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1425 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1426 was a macro that became an assignment, so it seemed to work. This has
1427 changed to a proper function call with a void return, hence the compile
1428 error. Exim's code has been fixed.
1430 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1433 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1435 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1437 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1438 because it turns out to be too big or there is a timeout) there is no
1439 "Completed" line in the log. When some messages of this type were
1440 selected by exigrep, they were listed as "not completed". Others were
1441 picked up by some special patterns. I have improved the selection
1442 criteria to be more general.
1444 PH/38 The host_find_failed option in the manualroute router can now be set
1445 to "ignore", to completely ignore a host whose IP address cannot be
1446 found. If all hosts are ignored, the behaviour is controlled by the new
1447 host_all_ignored option.
1449 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1450 homing or because of multiple MX records with /mx) generated more than
1451 one IP address, and the following item turned out to be the local host,
1452 all the secondary addresses of the first item were incorrectly removed
1453 from the list, along with the local host and any following hosts (which
1454 is what is supposed to happen).
1456 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1457 whoever called Exim into the -H file. In the case of the daemon it was
1458 behaving confusingly. When first started, it used values for whoever
1459 started the daemon, but after a SIGHUP it used the Exim user (because it
1460 calls itself on a restart). I have changed the code so that it now always
1463 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1464 message are rejected with the same error (e.g. no authentication or bad
1465 sender address), and a DATA command is nevertheless sent (as can happen
1466 with PIPELINING or a stupid MUA), the error message that was given to the
1467 RCPT commands is included in the rejection of the DATA command. This is
1468 intended to be helpful for MUAs that show only the final error to their
1471 PH/42 Another patch from the Sieve maintainer.
1473 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1474 Eximstats - Fixed some broken HTML links and added missing column headers
1476 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1477 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1479 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1481 PH/43 Yet another patch from the Sieve maintainer.
1483 PH/44 I found a way to check for a TCP/IP connection going away before sending
1484 the response to the final '.' that terminates a message, but only in the
1485 case where the client has not sent further data following the '.'
1486 (unfortunately, this is allowed). However, in many cases there won't be
1487 any further data because there won't be any more messages to send. A call
1488 to select() can be used: if it shows that the input is "ready", there is
1489 either input waiting, or the socket has been closed. An attempt to read
1490 the next input character can distinguish the two cases. Previously, Exim
1491 would have sent an OK response which the client would never have see.
1492 This could lead to message repetition. This fix should cure that, at
1493 least in a lot of common cases.
1495 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1496 advertised in response to EHLO.
1502 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1503 fixed by 4.65/MH/01 (is this a record?) are fixed:
1505 (i) An empty string was always treated as zero by the numeric comparison
1506 operators. This behaviour has been restored.
1508 (ii) It is documented that the numeric comparison operators always treat
1509 their arguments as decimal numbers. This was broken in that numbers
1510 starting with 0 were being interpreted as octal.
1512 While fixing these problems I realized that there was another issue that
1513 hadn't been noticed. Values of message_size_limit (both the global option
1514 and the transport option) were treated as octal if they started with 0.
1515 The documentation was vague. These values are now always treated as
1516 decimal, and I will make that clear in the documentation.
1522 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1523 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1526 MH/01 Don't check that the operands of numeric comparison operators are
1527 integers when their expansion is in "skipping" mode (fixes bug
1528 introduced by 4.64-PH/07).
1530 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1531 child addresses, Exim now panics and dies. Previously, because the count
1532 is held in a short int, deliveries were likely to be lost. As such a
1533 large number of recipients for a single message is ridiculous
1534 (performance will be very, very poor), I have chosen to impose a limit
1535 rather than extend the field.
1541 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1542 leftover -K file (the existence of which was triggered by #402).
1543 While we were at it, introduced process PID as part of the -K
1544 filename. This should rule out race conditions when creating
1547 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1548 processing considerably. Previous code took too long for large mails,
1549 triggering a timeout which in turn triggers #401.
1551 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1552 in the DK code in transports.c. sendfile() is not really portable,
1553 hence the _LINUX specificness.
1555 TF/01 In the add_headers option to the mail command in an Exim filter,
1556 there was a bug that Exim would claim a syntax error in any
1557 header after the first one which had an odd number of characters
1560 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1561 callout verification, Exim cached a "reject" for the entire domain. This
1562 is correct for most verifications, but it is not correct for a recipient
1563 verification with use_sender or use_postmaster set, because in that case
1564 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1565 case of MAIL FROM:<> rejection from other early rejections (e.g.
1566 rejection of HELO). When verifying a recipient using a non-null MAIL
1567 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1568 Whatever the result of the callout, the value of the domain cache is
1569 left unchanged (for any other kind of callout, getting as far as trying
1570 RCPT means that the domain itself is ok).
1572 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1575 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1576 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1577 session, the daemon ignores ECONNECT errors and logs others; it now
1578 ignores EPIPE as well.
1580 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1581 (quoted-printable decoding).
1583 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1584 later the small subsequent patch to fix an introduced bug.
1586 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1588 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1590 PH/08 An error is now given if message_size_limit is specified negative.
1592 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1593 to be given (somewhat) arbitrary names.
1595 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1598 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1599 miscellaneous code fixes
1601 PH/10 Added the log_reject_target ACL modifier to specify where to log
1604 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1605 hostname. This is wrong, because it relates to the incoming message (and
1606 probably the interface on which it is arriving) and not to the outgoing
1607 callout (which could be using a different interface). This has been
1608 changed to use the value of the helo_data option from the smtp transport
1609 instead - this is what is used when a message is actually being sent. If
1610 there is no remote transport (possible with a router that sets up host
1611 addresses), $smtp_active_hostname is used.
1613 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1614 tweaks were necessary in order to get it to work (see also 21 below):
1615 (a) The code assumed that strncpy() returns a negative number on buffer
1616 overflow, which isn't the case. Replaced with Exim's string_format()
1618 (b) There were several signed/unsigned issues. I just did the minimum
1619 hacking in of casts. There is scope for a larger refactoring.
1620 (c) The code used strcasecmp() which is not a standard C function.
1621 Replaced with Exim's strcmpic() function.
1622 (d) The code set only $1; it now sets $auth1 as well.
1623 (e) A simple test gave the error "authentication client didn't specify
1624 service in request". It would seem that Dovecot has changed its
1625 interface. Fortunately there's a specification; I followed it and
1626 changed what the client sends and it appears to be working now.
1628 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1631 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1632 address A is aliased to B and C, where B exists and C does not. Without
1633 -v the output is "A verified" because verification stops after a
1634 successful redirection if more than one address is generated. However,
1635 with -v the child addresses are also verified. Exim was outputting "A
1636 failed to verify" and then showing the successful verification for C,
1637 with its parentage. It now outputs "B failed to verify", showing B's
1638 parentage before showing the successful verification of C.
1640 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1641 look up a TXT record in a specific list after matching in a combined
1644 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1645 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1646 they consult the DNS. I had assumed they would set it the way they
1647 wanted; and indeed my experiments on Linux seem to show that in some
1648 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1649 To be on the safe side, however, I have now made the interface to
1650 host_find_byname() similar to host_find_bydns(), with an argument
1651 containing the DNS resolver options. The host_find_byname() function now
1652 sets these options at its start, just as host_find_bydns() does. The smtp
1653 transport options dns_qualify_single and dns_search_parents are passed to
1654 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1655 of host_find_byname() use the default settings of RES_DEFNAMES
1656 (qualify_single) but not RES_DNSRCH (search_parents).
1658 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1659 spool_read_header() do less string testing, by means of a preliminary
1660 switch on the second character of optional "-foo" lines. (This is
1661 overdue, caused by the large number of possibilities that now exist.
1662 Originally there were few.) While I was there, I also converted the
1663 str(n)cmp tests so they don't re-test the leading "-" and the first
1664 character, in the hope this might squeeze out yet more improvement.
1666 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1667 flag allowing group syntax was set by the header_syntax check but not
1668 turned off, possible causing trouble later; (2) The flag was not being
1669 set at all for the header_verify test, causing "group"-style headers to
1670 be rejected. I have now set it in this case, and also caused header_
1671 verify to ignore an empty address taken from a group. While doing this, I
1672 came across some other cases where the code for allowing group syntax
1673 while scanning a header line wasn't quite right (mostly, not resetting
1674 the flag correctly in the right place). These bugs could have caused
1675 trouble for malformed header lines. I hope it is now all correct.
1677 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1678 with the "reply" argument non-NULL. The code, however (which originally
1679 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1680 but it didn't always do it. This confused somebody who was copying the
1681 code for some other use. I have removed all the tests.
1683 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1684 feature that was used to support insecure browsers during the U.S. crypto
1685 embargo. It requires special client support, and Exim is probably the
1686 only MTA that supported it -- and would never use it because real RSA is
1687 always available. This code has been removed, because it had the bad
1688 effect of slowing Exim down by computing (never used) parameters for the
1689 RSA_EXPORT functionality.
1691 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1692 authenticator to fail if there's a tab character in the incoming data
1693 (there should never be unless someone is messing about, as it's supposed
1694 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1695 if the connection is using TLS or if the remote IP is the same as the
1696 local IP, and the "valid-client-cert option" if a client certificate has
1699 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1700 authenticators. This can be used for authorization after authentication
1701 succeeds. (In the case of plaintext, it servers for both authentication
1704 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1705 if any retry times were supplied.
1707 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1708 connection, where there is no client IP address to check. In this
1709 situation, the verify now always succeeds.
1711 PH/25 Applied John Jetmore's -Mset patch.
1713 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1715 PH/27 In a string expansion for a processed (not raw) header when multiple
1716 headers of the same name were present, leading whitespace was being
1717 removed from all of them, but trailing whitespace was being removed only
1718 from the last one. Now trailing whitespace is removed from each header
1719 before concatenation. Completely empty headers in a concatenation (as
1720 before) are ignored.
1722 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1723 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1725 PH/29 [Removed. This was a change that I later backed out, and forgot to
1726 correct the ChangeLog entry (that I had efficiently created) before
1727 committing the later change.]
1729 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1730 address errors (4xx response to RCPT) over the same connection as other
1731 messages routed to the same hosts. Such deliveries are always "forced",
1732 so retry times are not inspected. This resulted in far too many retries
1733 for the affected addresses. The effect occurred only when there were more
1734 hosts than the hosts_max_try setting in the smtp transport when it had
1735 the 4xx errors. Those hosts that it had tried were not added to the list
1736 of hosts for which the message was waiting, so if all were tried, there
1737 was no problem. Two fixes have been applied:
1739 (i) If there are any address or message errors in an SMTP delivery, none
1740 of the hosts (tried or untried) are now added to the list of hosts
1741 for which the message is waiting, so the message should not be a
1742 candidate for sending over the same connection that was used for a
1743 successful delivery of some other message. This seems entirely
1744 reasonable: after all the message is NOT "waiting for some host".
1745 This is so "obvious" that I'm not sure why it wasn't done
1746 previously. Hope I haven't missed anything, but it can't do any
1747 harm, as the worst effect is to miss an optimization.
1749 (ii) If, despite (i), such a delivery is accidentally attempted, the
1750 routing retry time is respected, so at least it doesn't keep
1751 hammering the server.
1753 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1754 in ${readsocket because some servers need this prod.
1756 PH/32 Added some extra debug output when updating a wait-xxx database.
1758 PH/33 The hint "could be header name not terminated by colon", which has been
1759 given for certain expansion errors for a long time, was not being given
1760 for the ${if def:h_colon_omitted{... case.
1762 PH/34 The spec says: "With one important exception, whenever a domain list is
1763 being scanned, $domain contains the subject domain." There was at least
1764 one case where this was not true.
1766 PH/35 The error "getsockname() failed: connection reset by peer" was being
1767 written to the panic log as well as the main log, but it isn't really
1768 panic-worthy as it just means the connection died rather early on. I have
1769 removed the panic log writing for the ECONNRESET error when getsockname()
1772 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1773 runs only) independently of the message's sender address. This meant
1774 that, if the 4xx error was in fact related to the sender, a different
1775 message to the same recipient with a different sender could confuse
1776 things. In particualar, this can happen when sending to a greylisting
1777 server, but other circumstances could also provoke similar problems.
1778 I have changed the default so that the retry time for these errors is now
1779 based a combination of the sender and recipient addresses. This change
1780 can be overridden by setting address_retry_include_sender=false in the
1783 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1784 remote server are returned as part of bounce messages. This was not
1785 happening for LMTP over a pipe (the lmtp transport), but now it is the
1786 same for both kinds of LMTP.
1788 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1789 in header lines that were in fact CNAMEs. This is no longer the case.
1791 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1792 and queue runs started by the daemon processed all messages. This has
1793 been fixed so that -R and -S can now usefully be given with -q<time>.
1795 PH/40 Import PCRE release 6.7 (fixes some bugs).
1797 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1799 PH/42 Give an error if -q is specified more than once.
1801 PH/43 Renamed the variables $interface_address and $interface_port as
1802 $received_ip_address and $received_port, to make it clear that these
1803 values apply to message reception, and not to the outgoing interface when
1804 a message is delivered. (The old names remain recognized, of course.)
1806 PH/44 There was no timeout on the connect() call when using a Unix domain
1807 socket in the ${readsocket expansion. There now is.
1809 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1810 be meaningful with "accept".
1812 SC/01 Eximstats V1.43
1813 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1815 SC/02 Eximstats V1.44
1816 Use a glob alias rather than an array ref in the generated
1817 parser. This improves both readability and performance.
1819 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1820 Collect SpamAssassin and rejection statistics.
1821 Don't display local sender or destination tables unless
1822 there is data to show.
1823 Added average volumes into the top table text output.
1825 SC/04 Eximstats V1.46
1826 Collect data on the number of addresses (recipients)
1827 as well as the number of messages.
1829 SC/05 Eximstats V1.47
1830 Added 'Message too big' to the list of mail rejection
1831 reasons (thanks to Marco Gaiarin).
1833 SC/06 Eximstats V1.48
1834 Mainlog lines which have GMT offsets and are too short to
1835 have a flag are now skipped.
1837 SC/07 Eximstats V1.49 (Alain Williams)
1838 Added the -emptyok flag.
1840 SC/08 Eximstats V1.50
1841 Fixes for obtaining the IP address from reject messages.
1843 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1844 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1845 whitesspace changes from 4.64-PH/27
1847 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1850 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1851 are found, allow negative numbers in numeric criteria)
1853 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1855 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1856 to match changes made in 4.64-PH/43
1858 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1860 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1861 "domain" is actually an IP address, and force a failure. This locks out
1862 those revolvers/nameservers that support "A-for-A" lookups, in
1863 contravention of the specifications.
1865 PH/48 When a host name was looked up from an IP address, and the subsequent
1866 forward lookup of the name timed out, the host name was left in
1867 $sender_host_name, contrary to the specification.
1869 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1870 restricted to single-key lookups, Exim was not diagnosing an error if
1871 * or *@ was used with a query-style lookup.
1873 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1875 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1876 long ago, but noone interested enough thought of it. Let's just say that
1877 the "1.1" means that there are some new functions that weren't there at
1878 some point in the past.
1880 PH/51 Error processing for expansion failure of helo_data from an smtp
1881 transport during callout processing was broken.
1883 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1884 tested/used via the -bh/-bhc/-bs options.
1886 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1887 bug, fixed in subsequent PCRE releases).
1889 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1890 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1892 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1898 SC/01 Use a glob alias rather than an array ref in eximstats generated
1899 parser. This improves both readability and performance.
1901 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1902 Don't display local sender or destination tables in eximstats unless
1903 there is data to show.
1904 Added average volumes into the eximstats top table text output.
1906 SC/03 Collect data on the number of addresses (recipients) as well
1907 as the number of messages in eximstats.
1909 TF/01 Correct an error in the documentation for the redirect router. Exim
1910 does (usually) call initgroups() when daemonizing.
1912 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1913 with consistent privilege compared to when running as a daemon.
1915 TF/03 Note in the spec that $authenticated_id is not set for local
1916 submissions from trusted users.
1918 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1919 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1921 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1922 by adding some example configuration directives to the default
1923 configuration file. A little bit of work is required to uncomment the
1924 directives and define how usernames and passwords are checked, but
1925 there is now a framework to start from.
1927 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1928 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1929 without this. I don't know how relevant this is to other LDAP libraries.
1931 PH/02 Add the verb name to the "unknown ACL verb" error.
1933 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1935 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1937 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1938 directory not expanded when it should be if an expanded home directory
1939 was set for the address (which is overridden by the transport).
1941 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1944 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1945 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1946 because it is too late at that time, and has no effect.
1948 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1949 security issue with \' (bugzilla #107). I could not use the
1950 PQescapeStringConn() function, because it needs a PGconn value as one of
1953 PH/08 When testing addresses using -bt, indicate those final addresses that
1954 are duplicates that would not cause an additional delivery. At least one
1955 person was confused, thinking that -bt output corresponded to deliveries.
1956 (Suppressing duplicates isn't a good idea as you lose the information
1957 about possibly different redirections that led to the duplicates.)
1959 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1960 systems where poll() doesn't work, in particular OS X.
1962 PH/10 Added more information to debugging output for retry time not reached.
1964 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1965 operations in malware.c.
1967 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1970 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1971 syslog_duplication set false, log lines that would normally be written
1972 both the the main log and to the reject log were not written to syslog at
1975 PH/14 In the default configuration, change the use of "message" in ACL warn
1976 statements to "add_header".
1978 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1979 not followed by a command (e.g. "seen endif").
1981 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1982 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1985 PH/17 Added extra conditions to the default value of delay_warning_condition
1989 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1990 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1991 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1994 The Auto-Submitted: and various List- headers are standardised, whereas I
1995 don't think Precedence: ever was.
1997 PH/18 Refactored debugging code in route_finduser() to show more information,
1998 in particular, the error code if getpwnam() issues one.
2000 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2001 This is apparently needed in addition to the PH/07 change above to avoid
2002 any possible encoding problems.
2004 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2005 but not after initializing Perl.
2007 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2008 output them only if debugging. By default they are written stderr,
2009 apparently, which is not desirable.
2011 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2014 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2017 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2019 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2020 authenticated or an ident call has been made. Suppress the default
2021 values for $authenticated_id and $authenticated_sender (but permit -oMai
2022 and -oMas) when testing with -bh.
2024 PH/24 Re-jigged the order of the tests in the default configuration so that the
2025 tests for valid domains and recipients precede the DNS black list and CSA
2026 tests, on the grounds that those ones are more expensive.
2028 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2029 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2030 This bug exists in every version of Exim that I still have, right back to
2033 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2034 However, an attempt to turn on case-sensitivity in a regex key by
2035 including (?-i) didn't work because the subject string was already
2036 lowercased, and the effects were non-intuitive. It turns out that a
2037 one-line patch can be used to allow (?-i) to work as expected.
2043 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2044 other effects) broke the use of negated acl sub-conditions.
2046 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2049 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2050 "Deny" causes Exim to reject the incoming connection with a 554 error.
2051 Unfortunately, if there is a major crisis, such as a disk failure,
2052 tcp-wrappers gives "deny", whereas what one would like would be some
2053 kind of temporary error. A kludge has been added to help with this.
2054 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2055 554 error is used if errno is still zero or contains ENOENT (which occurs
2056 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2059 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2061 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2062 errors. Otherwise a message that provokes a temporary error (when other
2063 messages do not) can cause a whole host to time out.
2065 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2066 addresses were routed directly to files or pipes from a redirect router.
2067 File deliveries just didn't batch; pipe deliveries might have suffered
2070 PH/06 A failure to get a lock for a hints database would erroneously always say
2071 "Failed to get write lock", even when it was really a read lock.
2073 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2074 of 0600. This has been changed to use the value of the lockfile_mode
2075 option (which defaults to 0600).
2077 PH/08 Applied small patch from the Sieve maintainer.
2079 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2080 folder from quota calculations, a direct delivery into this folder messed
2081 up the contents of the maildirsize file. This was because the regex was
2082 used only to exclude .Trash (or whatever) when the size of the mailbox
2083 was calculated. There was no check that a delivery was happening into an
2084 excluded directory. This bug has been fixed by ignoring all quota
2085 processing for deliveries into excluded directories.
2087 PH/10 Added the maildirfolder_create_regex option to appendfile.
2093 PH/01 The code for finding all the local interface addresses on a FreeBSD
2094 system running IPv6 was broken. This may well have applied to all BSD
2095 systems, as well as to others that have similar system calls. The broken
2096 code found IPv4 interfaces correctly, but gave incorrect values for the
2097 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2098 that it would not match correctly against @[] and not recognize the IPv6
2101 PH/02 The ipliteral router was not recognizing addresses of the form user@
2102 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2104 PH/03 Added disable_ipv6.
2106 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2107 decoded form, because it is most often used to construct To: headers
2108 lines in autoreplies, and the decoded form may well be syntactically
2109 invalid. However, $reply_address has leading white space removed, and all
2110 newlines turned into spaces so that the autoreply transport does not
2113 PH/05 If group was specified without a user on a router, and no group or user
2114 was specified on a transport, the group from the router was ignored.
2116 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2117 for visible compile-time settings that can be used to change these
2118 numbers, for those that want even more. Backwards compatibility with old
2119 spool files has been maintained. However, going back to a previous Exim
2120 release will lost any variables that are in spool files.
2122 PH/07 Two small changes when running in the test harness: increase delay when
2123 passing a TCP/IP connection to a new process, in case the original
2124 process has to generate a bounce, and remove special handling of
2125 127.0.0.2 (sic), which is no longer necessary.
2127 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2128 be the same on different OS.
2130 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2133 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2134 whether --show-vars was specified or not
2136 JJ/02 exipick: Added support for new ACL variable spool format introduced
2139 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2140 syntactically invalid From: or Reply-to: line, and a filter used this to
2141 generate an autoreply, and therefore failed to obtain an address for the
2142 autoreply, Exim could try to deliver to a non-existent relative file
2143 name, causing unrelated and misleading errors. What now happens is that
2144 it logs this as a hard delivery error, but does not attempt to create a
2147 PH/11 The exinext utility has a -C option for testing purposes, but although
2148 the given file was scanned by exinext itself; it wasn't being passed on
2149 when Exim was called.
2151 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2152 an end-of-file indication when reading a command response.
2154 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2155 compiled. In many other places in Exim, IPv6 addresses are always
2156 recognized, so I have changed this. It also means that IPv4 domain
2157 literals of the form [IPV4:n.n.n.n] are now always recognized.
2159 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2160 used if the router is not running as root, for example, when verifying at
2161 ACL time, or when using -bh. The debugging output from this situation was
2162 non-existent - all you got was a failure to exec. I have made two
2165 (a) Failures to set uid/gid, the current directory, or a process leader
2166 in a subprocess such as that created by queryprogram now generate
2167 suitable debugging ouput when -d is set.
2169 (b) The queryprogram router detects when it is not running as root,
2170 outputs suitable debugging information if -d is set, and then runs
2171 the subprocess without attempting to change uid/gid.
2173 PH/15 Minor change to Makefile for building test_host (undocumented testing
2176 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2177 additional section of a DNS packet that returns MX or SRV records.
2178 Instead, it always explicitly searches for A/AAAA records. This avoids
2179 major problems that occur when a DNS server includes only records of one
2180 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2181 fixed another bug: if SRV records were looked up and the corresponding
2182 address records were *not* found in the additional section, the port
2183 values from the SRV records were lost.
2185 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2186 using the correct key (the original address) when searching the retry
2187 rules in order to find which one to use for generating the retry hint.
2189 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2190 adding the default one. Similarly, if it contains a Reply-To: header, the
2191 errors_reply_to option, if set, is not used.
2193 PH/19 When calculating a retry time, Exim used to measure the "time since
2194 failure" by looking at the "first failed" field in the retry record. Now
2195 it does not use this if it is later than than the arrival time of the
2196 message. Instead it uses the arrival time. This makes for better
2197 behaviour in cases where some deliveries succeed, thus re-setting the
2198 "first failed" field. An example is a quota failure for a huge message
2199 when small messages continue to be delivered. Without this change, the
2200 "time since failure" will always be short, possible causing more frequent
2201 delivery attempts for the huge message than are intended.
2202 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2204 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2205 $1, $2, $3) because the numerical variables can be reset during some
2206 expansion items (e.g. "match"), thereby losing the authentication data.
2208 PH/21 Make -bV show the size of off_t variables so that the test suite can
2209 decide whether to run tests for quotas > 2G.
2211 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2212 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2213 filecount value is greater than 2G or if a quota value is greater than 2G
2214 on a system where the size of off_t is not greater than 4, a panic error
2217 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2218 never match. The debug and -bh output now contains an explicit error
2219 message indicating a malformed IPv4 address or mask.
2221 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2222 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2223 PH/23 above applies.
2225 PH/25 Do not write to syslog when running in the test harness. The only
2226 occasion when this arises is a failure to open the main or panic logs
2227 (for which there is an explicit test).
2229 PH/26 Added the /no_tell option to "control=freeze".
2231 PH/27 If a host name lookup failed very early in a connection, for example, if
2232 the IP address matched host_lookup and the reverse lookup yielded a name
2233 that did not have a forward lookup, an error message of the form "no IP
2234 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2235 could be logged. Now it outputs the IP address instead of "NULL".
2237 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2238 allows the sender and the authenticated sender to be set when
2239 submitting a message from within Exim. Since child_open_exim() is
2240 documented for local_scan(), the new function should be too.
2242 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2243 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2244 results in an empty string is now treated as unset.
2246 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2248 PH/31 Added sender_verify_fail logging option.
2250 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2251 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2252 tidied the source and removed it altogether.
2254 PH/33 When a queue run was abandoned because the load average was too high, a
2255 log line was always written; now it is written only if the queue_run log
2256 selector is set. In addition, the log line for abandonment now contains
2257 information about the queue run such as the pid. This is always present
2258 in "start" and "stop" lines but was omitted from the "abandon" line.
2260 PH/34 Omit spaces between a header name and the colon in the error message that
2261 is given when verify = headers_syntax fails (if there are lots of them,
2262 the message gets confusing).
2264 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2265 names, as there are now some PTR records that contain slashes. This check
2266 is only to protect against broken name servers that fall over on strange
2267 characters, so the fact that it applies to all lookups doesn't matter.
2269 PH/36 Now that the new test suite is complete, we can remove some of the
2270 special code in Exim that was needed for the old test suite. For example,
2271 sorting DNS records because real resolvers return them in an arbitrary
2272 order. The new test suite's fake resolver always returns records in the
2275 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2276 bounces) except when queue_only is set, to avoid logging races between
2277 the different processes.
2279 PH/38 Panic-die if .include specifies a non-absolute path.
2281 PH/39 A tweak to the "H" retry rule from its user.
2283 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2284 a label. They prevented compilation on older perls.
2286 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2287 a warning to be raised on newish perls.
2289 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2290 on queue. Changes to match documented behaviour of showing count of
2291 messages matching specified criteria.
2293 PH/40 Changed the default ident timeout from 30s to 5s.
2295 PH/41 Added support for the use of login_cap features, on those BSD systems
2296 that have them, for controlling the resources used by pipe deliveries.
2298 PH/42 The content-scanning code uses fopen() to create files in which to put
2299 message data. Previously it was not paying any attention to the mode of
2300 the files. Exim runs with umask(0) because the rest of the code creates
2301 files with open(), and sets the required mode explicitly. Thus, these
2302 files were ending up world-writeable. This was not a big issue, because,
2303 being within the spool directory, they were not world-accessible. I have
2304 created a function called modefopen, which takes an additional mode
2305 argument. It sets umask(777), creates the file, chmods it to the required
2306 mode, then resets the umask. All the relevant calls to fopen() in the
2307 content scanning code have been changed to use this function.
2309 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2310 to 24 hours. This avoids potential overflow problems when processing G
2311 and H retry rules. I suspect nobody ever tinkers with this value.
2313 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2315 PH/45 When the plaintext authenticator is running as a client, the server's
2316 challenges are checked to ensure they are valid base64 strings. By
2317 default, the authentication attempt is cancelled if an invalid string is
2318 received. Setting client_ignore_invalid_base64 true ignores these errors.
2319 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2320 they are received. Thus, the responses can be made to depend on the
2321 challenges. If an invalid string is ignored, an empty string is placed in
2324 PH/46 Messages that are created by the autoreply transport now contains a
2325 References: header, in accordance with RFCs 2822 and 3834.
2327 PH/47 Added authenticated_sender_force to the smtp transport.
2329 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2331 PH/49 Installed latest patch from the Sieve maintainer.
2333 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2334 was also set, the appendfile transport was unnecessarily scanning a
2335 directory of message files (e.g. for maildir delivery) to find the count
2336 of files (along with the size), even though it did not need this
2337 information. It now does the scan only if it needs to find either the
2338 size of the count of files.
2340 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2342 PH/52 Two bugs concerned with error handling when the smtp transport is
2345 (i) Exim was not creating retry information for temporary errors given
2346 for individual recipients after the DATA command when the smtp transport
2347 was used in LMTP mode. This meant that they could be retried too
2348 frequently, and not timed out correctly.
2350 (ii) Exim was setting the flag that allows error details to be returned
2351 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2352 recipients that were returned after the DATA command.
2354 PH/53 This is related to PH/52, but is more general: for any failing address,
2355 when detailed error information was permitted to be returned to the
2356 sender, but the error was temporary, then after the final timeout, only
2357 "retry timeout exceeded" was returned. Now it returns the full error as
2358 well as "retry timeout exceeded".
2360 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2361 do this, and (what is worse) MTAs that accept it.
2363 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2364 will now be deprecated.
2366 PH/56 New os.c-cygwin from the Cygwin maintainer.
2368 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2369 formats (previously only available in exim formats via -bpr, -bpru,
2370 and -bpra. Now also available in native and exiqgrep formats)
2372 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2373 with very large, slow to parse queues
2375 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2377 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2379 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2380 responses to authentication challenges, though it was showing the
2381 challenges; (ii) I've removed the CR characters from the debug output for
2384 PH/58 Allow for the insertion of a newline as well as a space when a string
2385 is turned into more than one encoded-word during RFC 2047 encoding. The
2386 Sieve code now uses this.
2388 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2389 data_4xx, lost_connection, tls_required.
2391 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2392 message was being sent as an SMTP response.
2394 PH/61 Add -l and -k options to exicyclog.
2396 PH/62 When verifying, if an address was redirected to one new address, so that
2397 verification continued, and the new address failed or deferred after
2398 having set something in $address_data, the value of $address_data was not
2399 passed back to the ACL. This was different to the case when no
2400 redirection occurred. The value is now passed back in both cases.
2402 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2403 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2404 use login_cap.h, so on its own it isn't the distinguishing feature. The
2405 new name refers directly to the setclassresources() function.
2407 PH/65 Added configuration files for NetBSD3.
2409 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2411 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2412 is preferred over IPv4.
2414 PH/68 The bounce_return_message and bounce_return_body options were not being
2415 honoured for bounces generated during the reception of non-SMTP messages.
2416 In particular, this applied to messages rejected by the ACL. This bug has
2417 been fixed. However, if bounce_return_message is true and bounce_return_
2418 body is false, the headers that are returned for a non-SMTP message
2419 include only those that have been read before the error was detected.
2420 (In the case of an ACL rejection, they have all been read.)
2422 PH/69 The HTML version of the specification is now built in a directory called
2423 spec_html instead of spec.html, because the latter looks like a path with
2424 a MIME-type, and this confuses some software.
2426 PH/70 Catch two compiler warnings in sieve.c.
2428 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2429 function verify_get_ident() calls ip_connect() to connect a socket, but
2430 if the "connect()" function timed out, ip_connect() used to close the
2431 socket. However, verify_get_ident() also closes the socket later, and in
2432 between Exim writes to the log, which may get opened at this point. When
2433 the socket was closed in ip_connect(), the log could get the same file
2434 descriptor number as the socket. This naturally causes chaos. The fix is
2435 not to close the socket in ip_connect(); the socket should be closed by
2436 the function that creates it. There was only one place in the code where
2437 this was missing, in the iplookup router, which I don't think anybody now
2438 uses, but I've fixed it anyway.
2440 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2441 well as to direct DNS lookups. Otherwise the handling of names in host
2442 lists is inconsistent and therefore confusing.
2448 PH/01 Two changes to the default runtime configuration:
2450 (1) Move the checks for relay_from_hosts and authenticated clients from
2451 after to before the (commented out) DNS black list checks.
2453 (2) Add control=submission to the relay_from_hosts and authenticated
2454 clients checks, on the grounds that messages accepted by these
2455 statements are most likely to be submissions.
2457 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2459 (1) Generate an error if the third argument for the ${prvs expansion is
2462 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2465 (3) Reset the variables that are obtained from the first argument of
2466 ${prvscheck and used in the second argument before leaving the code,
2467 because their memory is reclaimed, so using them afterwards may do
2470 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2471 one (it's much easier than Tom thought :-).
2473 (5) Because of (4), we can now allow for the use of $prvscheck_result
2474 inside the third argument.
2476 PH/03 For some reason, the default setting of PATH when running a command from
2477 a pipe transport was just "/usr/bin". I have changed it to
2480 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2481 anything to be listed in the output from -bV.
2483 PH/05 When a filter generated an autoreply, the entire To: header line was
2484 quoted in the delivery log line, like this:
2486 => >A.N.Other <ano@some.domain> <original@ddress> ...
2488 This has been changed so that it extracts the operative address. There
2489 may be more than one such address. If so, they are comma-separated, like
2492 => >ano@some.domain,ona@other.domain <original@ddress> ...
2494 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2495 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2496 not being looked up in the rDNS to get a host name, Exim was showing the
2497 IP address twice in Received: lines, even though the IP addresses were
2498 identical. For example:
2500 Received: from [1.2.3.4] (helo=[1.2.3.4])
2502 However, if the real host name was known, it was omitting the HELO data
2503 if it matched the actual IP address. This has been tidied up so that it
2504 doesn't show the same IP address twice.
2506 PH/07 When both +timestamp and +memory debugging was on, the value given by
2507 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2508 called by the debug printing, thereby overwriting the timestamp buffer.
2509 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2511 PH/08 When the original message was included in an autoreply transport, it
2512 always said "this is a copy of the message, including all the headers",
2513 even if body_only or headers_only was set. It now gives an appropriate
2516 PH/09 Applied a patch from the Sieve maintainer which:
2518 o fixes some comments
2519 o adds the (disabled) notify extension core
2520 o adds some debug output for the result of if/elsif tests
2521 o points to the current vacation draft in the documentation
2522 and documents the missing references header update
2526 o fixes a bug in processing the envelope test (when testing
2527 multiple envelope elements, the last element determinted the
2530 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2531 Electronic Mail") by including:
2533 Auto-submitted: auto-generated
2535 in the messages that it generates (bounce messages and others, such as
2536 warnings). In the case of bounce messages for non-SMTP mesages, there was
2537 also a typo: it was using "Auto_submitted" (underscore instead of
2538 hyphen). Since every message generated by Exim is necessarily in response
2539 to another message, thes have all been changed to:
2541 Auto-Submitted: auto-replied
2543 in accordance with these statements in the RFC:
2545 The auto-replied keyword:
2547 - SHOULD be used on messages sent in direct response to another
2548 message by an automatic process,
2550 - MUST NOT be used on manually-generated messages,
2552 - MAY be used on Delivery Status Notifications (DSNs) and Message
2553 Disposition Notifications (MDNs),
2555 - MUST NOT be used on messages generated by automatic or periodic
2556 processes, except for messages which are automatic responses to
2559 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2560 to the default Received: header definition.
2562 PH/12 Added log selector acl_warn_skipped (default on).
2564 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2565 variables because (a) they are in the wrong storage pool and (b) even if
2566 they were copied, it wouldn't work properly because of the caching.
2568 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2569 checking when decoding. Apparently there are clients that generate
2570 overlong encoded strings. Why am I not surprised?
2572 PH/15 If the first argument of "${if match_address" was not empty, but did not
2573 contain an "@" character, Exim crashed. Now it writes a panic log message
2574 and treats the condition as false.
2576 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2578 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2579 extension "envelope-auth". The code is finished and in agreement with
2580 other implementations, but there is no documentation so far and in fact,
2581 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2582 not changing the active code.
2584 Print executed "if" and "elsif" statements when debugging is used. This
2585 helps a great deal to understand what a filter does.
2587 Document more things not specified clearly in RFC3028. I had all this
2588 sorted out, when out of a sudden new issues came to my mind. Oops."
2590 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2593 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2594 canonical form (as documented). However, after a host name lookup from
2595 the IP address, check_host() was doing a simple string comparison with
2596 addresses acquired from the DNS when checking that the found name did
2597 have the original IP as one of its addresses. Since any found IPv6
2598 addresses are likely to be in abbreviated form, the comparison could
2599 fail. Luckily, there already exists a function for doing the comparison
2600 by converting both addresses to binary, so now that is used instead of
2601 the text comparison.
2603 PH/20 There was another similar case to PH/19, when a complete host name was
2604 given in a host list; looking up its IP address could give an abbreviated
2605 form, whereas the current host's name might or might not be abbreviated.
2606 The same fix has been applied.
2612 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2613 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2616 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2617 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2619 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2621 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2622 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2623 Also, for "Auto-Submitted", treat anything other than "no" as
2624 non-personal, in accordance with RFC 3834. (Previously it treated
2625 anything starting "auto-" as non-personal.)
2627 TF/01 The control=submission/name=... option had a problem with syntax
2628 errors if the name included a slash character. The /name= option
2629 now slurps the rest of the string, so it can include any characters
2630 but it must come last in the list of options (after /sender_retain
2633 PH/05 Some modifications to the interface to the fake nameserver for the new
2641 TK/01 Added the "success_on_redirect" address verification option. See
2642 NewStuff for rationale and an example.
2644 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2646 PH/02 Patch to exigrep to allow it to work on syslog lines.
2648 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2649 fread() to skip over the body file's header line, because in Cygwin the
2650 header line is locked and is inaccessible.
2652 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2653 co-exist for some time) to make it clear that it is the Exim ID that is
2654 referenced, not the Message-ID: header line.
2656 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2657 string_format() function, because snprintf() does not exist on all
2660 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2661 ${stat: expansion item.
2663 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2664 protocol synchronization error", to keep the pedants happy.
2666 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2667 well as for IRIX systems, when gcc is being used. See the host.c source
2670 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2672 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2675 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2676 transport and to the smtp transport in LMTP mode.
2678 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2680 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2681 run a filter in a subprocess. This could lead to confusion in subsequent
2682 lookups in the parent process. There should also be a search_tidyup() at
2683 the end of the subprocess.
2685 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2686 only if the host matched helo_try_verify_hosts, which caused the
2687 verification to occur when the EHLO/HELO command was issued. The ACL just
2688 tested the remembered result. Now, if a previous verification attempt has
2689 not happened, "verify = helo" does it there and then.
2691 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2693 TK/03 Fix log output including CR from clamd.
2695 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2696 exist provoked a memory error which could cause a segfault.
2698 PH/15 Installed PCRE 6.2
2700 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2702 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2703 of the problem. Specifically, suggested +O2 rather than +O1 for the
2706 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2708 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2709 record, the port was not correctly logged when the outgoing_port log
2710 selector was set (it logged the transort's default port).
2712 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2713 fallback_hosts, and "hosts" in the smtp transport.
2715 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2716 host errors such as "Connection refused".
2718 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2719 authentication with radiusclient 0.4.9:
2721 - Error returned from rc_read_config was caught wrongly
2722 - Username/password not passed on to radius server due to wrong length.
2724 The presumption is that some radiusclient API changes for 4.51/PH/17
2725 were not taken care of correctly. The code is still untested by me (my
2726 Linux distribution still has 0.3.2 of radiusclient), but it was
2727 contributed by a Radius user.
2729 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2730 expanding the "port" option of the smtp transport.
2732 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2733 while reading a MIME header. Thanks to Tom Hughes for a patch.
2735 PH/24 Include config.h inside local_scan.h so that configuration settings are
2738 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2739 that in an ACL for RCPT (for example), you can examine exactly what was
2742 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2743 commands, but it was not correctly comparing the address with the actual
2744 client host address. Thus, it would show the EHLO address in Received:
2745 header lines when this was not necessary.
2747 PH/27 Added the % operator to ${eval:}.
2749 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2750 it should be ignoring failures (because with -C, for example, it has lost
2751 privilege). It wasn't ignoring creation failures other than "already
2754 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2757 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2758 because an input line was too long, either on its own, or by virtue of
2759 too many continuations, the temporary file was not being removed, and the
2760 return code was incorrect.
2762 PH/31 Missing "BOOL" in function definition in filtertest.c.
2764 PH/32 Applied Sieve patches from the maintainer.
2766 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2768 PH/33 Added "verify = not_blind".
2770 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2771 Local/Makefile (with some defaults set). These are used in built scripts
2772 such as exicyclog, but they have never been used in the exim_install
2773 script (though there are many overriding facilities there). I have
2774 arranged that the exim_install script now takes note of these two
2777 PH/35 Installed configuration files for Dragonfly.
2779 PH/36 When a locally submitted message by a trusted user did not contain a
2780 From: header, and the sender address was obtained from -f or from an SMTP
2781 MAIL command, and the trusted user did not use -F to supply a sender
2782 name, $originator_name was incorrectly used when constructing a From:
2783 header. Furthermore, $originator_name was used for submission mode
2784 messages from external hosts without From: headers in a similar way,
2785 which is clearly wrong.
2787 PH/37 Added control=suppress_local_fixups.
2789 PH/38 When log_selector = +received_sender was set, and the addition of the
2790 sender made the log line's construction buffer exactly full, or one byte
2791 less than full, an overflow happened when the terminating "\n" was
2794 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2795 when the result of a list match is failure because a DNS lookup failed.
2797 PH/40 RM_COMMAND is now used in the building process.
2799 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2800 the "build-* directories that it finds.
2802 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2803 address in a domain literal was a prefix of an interface address.
2805 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2806 when verifying a sender address, unless rewrite_headers is false.
2808 PH/44 Wrote a long comment about why errors_to addresses are verified as
2809 recipients, not senders.
2811 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2812 the ratelimit ACL was added.
2814 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2816 PH/46 Added extra information about PostgreSQL errors to the error string.
2818 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2819 avoiding the need to install special zones in a real server. This is
2820 backwards compatible; if it can't find the fake resolver, it drops back.
2821 Thus, both old and new test suites can be run.
2823 TF/02 Added util/ratelimit.pl
2825 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2826 clock is set back in time.
2828 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2829 Candler <B.Candler@pobox.com>.
2831 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2832 OK for addresses that are the result of redirections.
2834 PH/48 A number of further additions for the benefit of the new test suite,
2835 including a fake gethostbyname() that interfaces to the fake DNS resolver
2838 TF/06 The fix for widen_domains has also been applied to qualify_single and
2839 search_parents which are the other dnslookup options that can cause
2842 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2845 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2847 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2848 service" if it encountered an MX record with an empty target hostname.
2849 The message is now "an MX or SRV record indicated no SMTP service".
2851 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2852 if the DNS of the sending site is misconfigured. This is quite a
2853 common situation. This change restores the behaviour of treating a
2854 helo verification defer as a failure.
2856 PH/51 If self=fail was set on a router, the bounce message did not include the
2857 actual error message.
2863 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2865 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2866 command itself ended in error, the underlying message about the transport
2867 filter timeout was being overwritten with the pipe command error. Now the
2868 underlying error message should be appended to the second error message.
2870 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2872 PH/02 Reduce the amount of output that "make" produces by default. Full output
2873 can still be requested.
2875 PH/03 The warning log line about a condition test deferring for a "warn" verb
2876 was being output only once per connection, rather than after each
2877 occurrence (because it was using the same function as for successful
2878 "warn" verbs). This seems wrong, so I have changed it.
2880 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2881 it should not have, which might have caused a crash in the right
2882 circumstances, but probably never did.
2884 PH/04 Installed a modified version of Tony Finch's patch to make submission
2885 mode fix the return path as well as the Sender: header line, and to
2886 add a /name= option so that you can make the user's friendly name appear
2889 TF/03 Added the control = fakedefer ACL modifier.
2891 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2892 Mark Lowes for thorough testing.
2894 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2896 TK/03 Merged latest SRS patch from Miles Wilton.
2898 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2899 with the definition in sysexits.h (which is #included earlier).
2900 Fortunately, Exim does not actually use EX_OK. The code used to try to
2901 preserve the sysexits.h value, by assumimg that macro definitions were
2902 scanned for macro replacements. I have been disabused of this notion,
2903 so now the code just undefines EX_OK before #including unistd.h.
2905 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2906 in the smtp transport. When a block could not be written in a single
2907 write() function, the timeout was being re-applied to each part-write.
2908 This seems wrong - if the receiver was accepting one byte at a time it
2909 would take for ever. The timeout is now adjusted when this happens. It
2910 doesn't have to be particularly precise.
2912 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2913 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2915 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2916 without a domain if the check to <postmaster@domain> fails.
2918 SC/01 Eximstats: added -xls and the ability to specify output files
2919 (patch written by Frank Heydlauf).
2921 SC/02 Eximstats: use FileHandles for outputing results.
2923 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2925 SC/04 Eximstats: fixed display of large numbers with -nvr option
2927 SC/05 Eximstats: fixed merging of reports with empty tables.
2929 SC/06 Eximstats: added the -include_original_destination flag
2931 SC/07 Eximstats: removed tabs and trailing whitespace.
2933 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2935 TK/06 MBOX spool code: Add real "From " MBOX separator line
2936 so the .eml file is really in mbox format (even though
2937 most programs do not really care). Patch from Alex Miller.
2939 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2940 The latter is generated from $received_to and is only set if the
2941 message has one envelope recipient. SA can use these headers,
2942 obviously out-of-the-box. Patch from Alex Miller.
2944 PH/08 The ${def test on a variable was returning false if the variable's
2945 value was "0", contrary to what the specification has always said!
2946 The result should be true unless the variable is empty.
2948 PH/09 The syntax error of a character other than { following "${if
2949 def:variable_name" (after optional whitespace) was not being diagnosed.
2950 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2951 accidental colon was present, for example, could give incorrect results.
2953 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2954 result is used (not including appendfile, where other changes are about
2957 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2958 This involved changing a lot of size variables from int to off_t. It
2959 should work with maildirs and everything.
2961 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2962 spamd dying while we are connected to it.
2964 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2967 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2968 function for BDB 4.3.
2970 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2972 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2973 the sources for PCRE have been re-arranged and more files are now
2976 PH/15 The code I had for printing potentially long long variables in PH/11
2977 above was not the best (it lost precision). The length of off_t variables
2978 is now inspected at build time, and an appropriate printing format (%ld
2979 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2980 to be "long long int" or "long int". This is needed for the internal
2981 formatting function string_vformat().
2983 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2984 the configuration file to be ":syslog", then the script "guesses" where
2985 the logs files are, rather than using the compiled in default. In our
2986 case the guess is not the same as the compiled default, so the script
2987 suddenly stopped working when I started to use syslog. The patch checks
2988 to see if log_file_path is "". If so, it attempts to read it from exim
2989 with no configuration file to get the compiled in version, before it
2990 falls back to the previous guessing code."
2992 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2993 implementing BATV in an Exim configuration. See NewStuff for the gory
2996 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2997 Makefile that are specific to HP-UX.
2999 PH/18 If the "use_postmaster" option was set for a recipient callout together
3000 with the "random" option, the postmaster address was used as the MAIL
3001 FROM address for the random test, but not for the subsequent recipient
3002 test. It is now used for both.
3004 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3005 patch removes a few documentation additions to RFC 3028, because the
3006 latest draft now contains them. It adds the new en;ascii-case comparator
3007 and a new error check for 8bit text in MIME parts. Comparator and
3008 require names are now matched exactly. I enabled the subaddress
3009 extension, but it is not well tested yet (read: it works for me)."
3011 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3012 rework some of the code of TK/09 above to avoid the hardwired use of
3013 "%lld" and "long long". Replaced the call to snprintf() with a call to
3016 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3017 records point to non-existent hosts", "retry timeout exceeded", and
3018 "retry time not reached for any host after a long failure period".
3020 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3021 experimental DomainKeys support:
3023 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3024 (2) On an error such as an illegally used "control", the wrong name for
3025 the control was given.
3027 These problems did NOT occur unless DomainKeys support was compiled.
3029 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3031 PH/24 Added ${if match_ip condition.
3033 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3034 databases so that it will be absolutely obvious if a crash occurs in the
3035 DB library. This is a regular occurrence (often caused by mis-matched
3038 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3039 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3040 picked up on a user's system that detects such things. There doesn't seem
3041 to be a gcc warning option for this - only an attribute that has to be
3042 put on the function's prototype. It seems that in Fedora Core 4 they have
3043 set this on a number of new functions. No doubt there will be more in due
3046 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3047 specify a transport. However, if an address that was verified by such a
3048 router was the subject of a callout, Exim crashed because it tried to
3049 read the rcpt_include_affixes from the non-existent transport. Now it
3050 just assumes that the setting of that option is false. This bug was
3051 introduced by 4.51/PH/31.
3053 PH/28 Changed -d+all to exclude +memory, because that information is very
3054 rarely of interest, but it makes the output a lot bigger. People tend to
3055 do -d+all out of habit.
3057 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3058 code in os-type was giving problems when libc.so lives in lib64, like on
3061 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3062 aren't the modern standard, and it seems that some systems' include files
3063 don't always have them. Exim was already checking for some of the newer
3064 ones like T_AAAA, and defining it itself. I've added checks for all the
3065 record types that Exim uses.
3067 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3068 not automatically generating a new one, as it is supposed to. This
3069 prevented TLS from working. If the file did exist, but contained invalid
3070 data, a new version was generated, as expected. It was only the case of a
3071 non-existent file that was broken.
3073 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3074 with a change in libdomainkeys > 0.64.
3076 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3077 from DNS. If the selector record carries the flag, it now has
3078 precedence over the domain-wide flag.
3080 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3082 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3083 the use of a port name that isn't defined in /etc/services) occurred, the
3084 message was deferred as in a normal delivery, and thus remained on the
3085 spool, instead of being failed because of the mua_wrapper setting. This
3086 is now fixed, and I tidied up some of the mua_wrapper messages at the
3089 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3090 the messages in a hash of arrays rather than using individual hashes.
3091 This is a bit cleaner and results in dramatic memory savings, albeit
3092 at a slight CPU cost.
3094 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3095 as requested by Marc Sherman.
3097 SC/10 Eximstats: added histograms for user specified patterns as requested
3100 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3102 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3103 fopen() in the content-scanning modules that did not already have it.
3109 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3110 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3112 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3114 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3116 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3117 to test Sieve filters that use "vacation".
3119 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3120 that changes the way the GnuTLS parameters are stored in the cache file.
3121 The new format can be generated externally. For backward compatibility,
3122 if the data in the cache doesn't make sense, Exim assumes it has read an
3123 old-format file, and it generates new data and writes a new file. This
3124 means that you can't go back to an older release without removing the
3127 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3128 work if there are any delivery delays because "one_time" forces the
3129 parent to be marked "delivered", so its unseen clone is never tried
3130 again. For this reason, Exim now forbids the simultaneous setting of
3133 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3134 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3135 entry if you want to know the details. The fix, however, neglected to
3136 consider the case when local delivery batching is involved. The test for
3137 "previously delivered" was not happening when checking to see if an
3138 address could be batched with a previous (undelivered) one; under
3139 certain circumstances this could lead to multiple deliveries to the same
3142 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3143 in its include files, and this causes problems building Exim.
3145 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3146 header_syntax) but Exim was just ignoring anything given after a slash.
3147 In particular, this caused confusion with an attempt to use "verify =
3148 reverse_host_lookup/defer_ok". An error is now given when options are
3149 supplied for verify items that do not have them. (Maybe reverse_host_
3150 lookup should have a defer_ok option, but that's a different point.)
3152 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3153 defined by RFC 821) to 2048, because there were problems with some AUTH
3154 commands, and RFC 1869 says the size should be increased for extended
3155 SMTP commands that take arguments.
3157 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3160 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3161 "unknown" error; now it says that the functionality isn't in the binary.
3163 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3164 an address' error message when a string expansion fails (syntax or
3165 whatever). Otherwise the password may appear in the log. Following change
3166 PH/42 below, there is no longer a chance of it appearing in a bounce
3169 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3171 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3172 of its addresses was ever tried. (Bugzilla bug #2.)
3174 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3175 the result incorrectly in the debug output. (It correctly added a newline
3176 to what was transported.)
3178 TF/01 Added $received_time.
3180 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3181 commented out examples of how to interface to a virus scanner and to
3182 SpamAssassin. Also added commented examples of av_scanner and
3183 spamd_address settings.
3185 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3186 and controls are allowed in which ACLs. There were a couple of minor
3187 errors. Some of the entries in the conditions table (which is a table of
3188 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3189 negation of where the condition IS allowed.
3191 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3193 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3194 header file does not have a version number, so I've had to invent a new
3195 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3196 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3197 radiusclient), but it was contributed by a Radius user.
3199 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3200 files or directories, for OpenSSL.
3202 PH/19 When an Exim process that is running as root has to create an Exim log
3203 file, it does so in a subprocess that runs as exim:exim so as to get the
3204 ownership right at creation (otherwise, other Exim processes might see
3205 the file with the wrong ownership). There was no test for failure of this
3206 fork() call, which would lead to the process getting stuck as it waited
3207 for a non-existent subprocess. Forks do occasionally fail when resources
3208 run out. I reviewed all the other calls to fork(); they all seem to check
3211 PH/20 When checking for unexpected SMTP input at connect time (before writing
3212 the banner), Exim was not dealing correctly with a non-positive return
3213 from the read() function. If the client had disconnected by this time,
3214 the result was a log entry for a synchronization error with an empty
3215 string after "input=" when read() returned zero. If read() returned -1
3216 (an event I could not check), uninitialized data bytes were printed.
3217 There were reports of junk text (parts of files, etc) appearing after
3220 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3222 PH/22 Added support for macro redefinition, and (re)definition in between
3223 driver and ACL definitions.
3225 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3226 forgetting to use the resulting value; it was using the unexpanded value.
3228 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3229 hadn't been configured. The fix is from Juergen Kreileder, who
3230 understands it better than I do:
3232 "Here's what I see happening with three configured cyrus_sasl
3233 authenticators configured (plain, login, cram-md5):
3235 On startup auth_cyrus_sasl_init() gets called for each of these.
3236 This means three calls to sasl_listmech() without a specified mech_list.
3237 => SASL tests which mechs of all available mechs actually work
3238 => three warnings about OTP not working
3239 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3241 With the patch, sasl_listmech() also gets called three times. But now
3242 SASL's mech_list option is set to the server_mech specified in the the
3243 authenticator. Or in other words, the answer from sasl_listmech()
3244 gets limited to just the mech you're testing for (which is different
3246 => the return list contains just 'plain' or 'login', 'cram-md5' or
3247 nothing depending on the value of ob->server_mech.
3249 I've just tested the patch: Authentication still works fine,
3250 unavailable mechs specified in the exim configuration are still
3251 caught, and the auth.log warnings about OTP are gone."
3253 PH/25 When debugging is enabled, the contents of the command line are added
3254 to the debugging output, even when log_selector=+arguments is not
3257 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3258 answer is "GNU", and only if the return is "GNU/something" is the answer
3261 PH/27 $acl_verify_message is now set immediately after the failure of a
3262 verification in an ACL, and so is available in subsequent modifiers. In
3263 particular, the message can be preserved by coding like this:
3265 warn !verify = sender
3266 set acl_m0 = $acl_verify_message
3268 Previously, $acl_verify_message was set only while expanding "message"
3269 and "log_message" when a very denied access.
3271 PH/28 Modified OS/os.c-Linux with
3273 -#ifndef OS_LOAD_AVERAGE
3274 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3276 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3277 nomenclature these days.)
3279 PH/29 Installed patch from the Sieve maintainer that adds the options
3280 sieve_useraddress and sieve_subaddress to the redirect router.
3282 PH/30 In these circumstances:
3283 . Two addresses routed to the same list of hosts;
3284 . First host does not offer TLS;
3285 . First host accepts first address;
3286 . First host gives temporary error to second address;
3287 . Second host offers TLS and a TLS session is established;
3288 . Second host accepts second address.
3289 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3290 and peerdn, if requested) that were in fact used only for the second
3293 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3294 attention to any local part prefix or suffix that was matched by the
3295 router that accepted the address. It now behaves in the same way as it
3296 does for delivery: the affixes are removed from the local part unless
3297 rcpt_include_affixes is set on the transport.
3299 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3300 timeout during the DATA phase of an incoming message.
3302 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3303 applied a patch sanctioned by the Sieve maintainer.
3305 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3306 the uid or gid is negative. A case of a negative gid caused this to be
3307 noticed. The fix allows for either to be negative.
3309 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3310 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3313 PH/36 The change PH/12 above was broken. Fixed it.
3315 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3316 the grounds that routing the same address twice would always produce the
3317 same answer. This might have been true once, but it is certainly no
3318 longer true now. Routing a child address may depend on the previous
3319 routing that produced that child. Some complicated redirection strategies
3320 went wrong when messages had multiple recipients, and made Exim's
3321 behaviour dependent on the order in which the addresses were given.
3323 I have moved the duplicate checking until after the routing is complete.
3324 Exim scans the addresses that are assigned to local and remote
3325 transports, and removes any duplicates. This means that more work will be
3326 done, as duplicates will always all be routed, but duplicates are
3327 presumably rare, so I don't expect this is of any significance.
3329 For deliveries to pipes, files, and autoreplies, the duplicate checking
3330 still happens during the routing process, since they are not going to be
3333 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3334 It corrects a timeout issue with spamd. This is Ian's comment: "The
3335 background is that sometimes spamd either never reads data from a
3336 connection it has accepted, or it never writes response data. The exiscan
3337 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3338 blindly assumes that writes won't block so it may never time out."
3340 PH/39 Allow G after quota size as well as K and M.
3342 PH/40 The value set for $authenticated_id in an authenticator may not contain
3343 binary zeroes or newlines because the value is written to log lines and
3344 to spool files. There was no check on this. Now the value is run through
3345 the string_printing() function so that such characters are converted to
3346 printable escape sequences.
3348 PH/41 $message_linecount is a new variable that contains the total number of
3349 lines in the message. Compare $body_linecount, which is the count for the
3352 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3353 bounce and delay warning messages, except in certain special cases, which
3356 (a) An SMTP error message from a remote host;
3357 (b) A message specified in a :fail: redirection;
3358 (c) A message specified in a "fail" command in a system filter;
3359 (d) A message specified in a FAIL return from the queryprogram router;
3360 (e) A message specified by the cannot_route_message router option.
3362 In these cases only, Exim does include the error details in bounce and
3363 warning messages. There are also a few cases where bland messages such
3364 as "unrouteable address" or "local delivery error" are given.
3366 PH/43 $value is now also set for the "else" part of a ${run expansion.
3368 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3369 being worked on, but at least Exim now implements the latest version to
3372 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3373 process to complete was treated as a delivery failure, a timeout while
3374 writing the message to the pipe was logged, but erroneously treated as a
3375 successful delivery. Such timeouts include transport filter timeouts. For
3376 consistency with the overall process timeout, these timeouts are now
3377 treated as errors, giving rise to delivery failures by default. However,
3378 there is now a new Boolean option for the pipe transport called
3379 timeout_defer, which, if set TRUE, converts the failures into defers for
3380 both kinds of timeout. A transport filter timeout is now identified in
3383 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3384 systems where "make" and "gmake" are different, calling "gmake" at top
3385 level broke things. I've arranged for the value of $(MAKE) to be passed
3386 from the Makefile to this script so that it can call the same version of
3390 A note about Exim versions 4.44 and 4.50
3391 ----------------------------------------
3393 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3394 changes of various kinds. As a consequence, a big documentation update was
3395 needed. This delayed the release for rather longer than seemed good, especially
3396 in the light of a couple of (minor) security issues. Therefore, the changes
3397 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3398 release. So 4.44 and 4.50 are in effect two different branches that both start
3401 I have left the 4.50 change log unchanged; it contains all the changes since
3402 4.43. The change log for 4.44 is below; many of its items are identical to
3403 those for 4.50. This seems to be the most sensible way to preserve the
3404 historical information.
3410 1. Minor wording change to the doc/README.SIEVE file.
3412 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3413 computation of the current number of files was incorrect.
3415 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3416 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3419 4. Give more explanation in the error message when the command for a transport
3420 filter fails to execute.
3422 5. There are several places where Exim runs a non-Exim command in a
3423 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3424 was being done only for the command run by the queryprogram router. It is
3425 now done for all such subprocesses. The other cases are: ${run, transport
3426 filters, and the commands run by the lmtp and pipe transports.
3428 6. Added CONFIGURE_GROUP build-time option.
3430 7. Some older OS have a limit of 256 on the maximum number of file
3431 descriptors. Exim was using setrlimit() to set 1000 as a large value
3432 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3433 systems. I've change it so that if it can't get 1000, it tries for 256.
3435 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3436 was an oversight, and furthermore, ever since the addition of extra
3437 controls (e.g. 4.43/32), the checks on when to allow different forms of
3438 "control" were broken. There should now be diagnostics for all cases when a
3439 control that does not make sense is encountered.
3441 9. Added the /retain_sender option to "control=submission".
3443 10. $recipients is now available in the predata ACL (oversight).
3445 11. Tidy the search cache before the fork to do a delivery from a message
3446 received from the command line. Otherwise the child will trigger a lookup
3447 failure and thereby defer the delivery if it tries to use (for example) a
3448 cached ldap connection that the parent has called unbind on.
3450 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3451 of $address_data from the recipient verification was clobbered by the
3452 sender verification.
3454 13. The value of address_data from a sender verification is now available in
3455 $sender_address_data in subsequent conditions in the ACL statement.
3457 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3459 15. Added a new option "connect=<time>" to callout options, to set a different
3462 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3463 was its contents. (It was OK if the option was not defined at all.)
3465 17. A "Completed" log line is now written for messages that are removed from
3466 the spool by the -Mrm option.
3468 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3469 information about exactly what failed.
3471 19. Added -dd to debug only the daemon process.
3473 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3474 handles timeouts, both on the server side and network timeouts. Renamed the
3475 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3477 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3478 It is now set to "smtps".
3480 22. $host_address is now set to the target address during the checking of
3481 ignore_target_hosts.
3483 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3484 being passed; this would have caused $sender_host_name to have been used if
3485 matching the list had actually called for a host name (not very likely,
3486 since this list is usually IP addresses). A host name is now passed as
3489 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3490 code that specifies a non-restarting handler (typically sigaction() in
3491 modern systems) in an attempt to fix a rare and obscure crash bug.
3493 25. Narrowed the window for a race in the daemon that could cause it to ignore
3494 SIGCHLD signals. This is not a major problem, because they are used only to
3495 wake it up if nothing else does.
3497 26. A malformed maildirsize file could cause Exim to calculate negative values
3498 for the mailbox size or file count. Odd effects could occur as a result.
3499 The maildirsize information is now recalculated if the size or filecount
3502 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3503 support for a long time. Removed HAVE_SYS_VFS_H.
3505 28. Installed the latest version of exipick from John Jetmore.
3507 29. In an address list, if the pattern was not a regular expression, an empty
3508 subject address (from a bounce message) matched only if the pattern was an
3509 empty string. Non-empty patterns were not even tested. This was the wrong
3510 because it is perfectly reasonable to use an empty address as part of a
3511 database query. An empty address is now tested by patterns that are
3512 lookups. However, all the other forms of pattern expect the subject to
3513 contain a local part and a domain, and therefore, for them, an empty
3514 address still always fails if the pattern is not itself empty.
3516 30. Exim went into a mad DNS loop when attempting to do a callout where the
3517 host was specified on an smtp transport, and looking it up yielded more
3518 than one IP address.
3520 31. Re-factored the code for checking spool and log partition space into a
3521 function that finds that data and another that does the check. The former
3522 is then used to implement four new variables: $spool_space, $log_space,
3523 $spool_inodes, and $log_inodes.
3525 32. The RFC2047 encoding function was originally intended for short strings
3526 such as real names; it was not keeping to the 75-character limit for
3527 encoded words that the RFC imposes. It now respects the limit, and
3528 generates multiple encoded words if necessary. To be on the safe side, I
3529 have increased the buffer size for the ${rfc2047: expansion operator from
3532 33. It is now permitted to omit both strings after an "if" condition; if the
3533 condition is true, the result is "true". As before, when the second string
3534 is omitted, a false condition yields an empty string. This makes it less
3535 cumbersome to write custom ACL and router conditions.
3537 34. Failure to deliver a bounce message always caused it to be frozen, even if
3538 there was an errors_to setting on the router. The errors_to setting is now
3541 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3542 canonical form (fully expanded) before being placed in
3543 $sender_host_address.
3545 36. The table in the code that translates DNS record types into text (T_A to
3546 "A" for instance) was missing entries for NS and CNAME. It is just possible
3547 that this could have caused confusion if both these types were looked up
3548 for the same domain, because the text type is used as part of Exim's
3549 per-process caching. But the chance of anyone hitting this buglet seems
3552 37. The dnsdb lookup has been extended in a number of ways.
3554 (1) There is a new type, "zns", which walks up the domain tree until it
3555 finds some nameserver records. It should be used with care.
3557 (2) There is a new type, "mxh", which is like "mx" except that it returns
3558 just the host names, not the priorities.
3560 (3) It is now possible to give a list of domains (or IP addresses) to be
3561 looked up. The behaviour when one of the lookups defers can be
3562 controlled by a keyword.
3564 (4) It is now possible to specify the separator character for use when
3565 multiple records are returned.
3567 38. The dnslists ACL condition has been extended: it is now possible to supply
3568 a list of IP addresses and/or domains to be looked up in a particular DNS
3571 39. Added log_selector=+queue_time_overall.
3573 40. When running the queue in the test harness, wait just a tad after forking a
3574 delivery process, to get repeatability of debugging output.
3576 41. Include certificate and key file names in error message when GnuTLS fails
3577 to set them up, because the GnuTLS error message doesn't include the name
3578 of the failing file when there is a problem reading it.
3580 42. Allow both -bf and -bF in the same test run.
3582 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3584 44. The "Exiscan patch" is now merged into the mainline Exim source.
3586 45. Sometimes the final signoff response after QUIT could fail to get
3587 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3588 < 0 before doing a fflush(). This bug looks as though it goes back to the
3589 introduction of TLS in release 3.20, but "sometimes" must have been rare
3590 because the tests only now provoked it.
3592 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3593 (this can affect the format of dates).
3595 47. exim_tidydb, when checking for the continued existence of a message for
3596 which it has found a message-specific retry record, was not finding
3597 messages that were in split spool directories. Consequently, it was
3598 deleting retry records that should have stayed in existence.
3600 48. Steve fixed some bugs in eximstats.
3602 49. The SPA authentication driver was not abandoning authentication and moving
3603 on to the next authenticator when an expansion was forced to fail,
3604 contradicting the general specification for all authenticators. Instead it
3605 was generating a temporary error. It now behaves as specified.
3607 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3608 (the order specifies the preference for clients). The order is now AES256,
3609 AES128, 3DES, ARCFOUR128.
3611 51. Small patch to Sieve code - explicitly set From: when generating an
3614 52. Exim crashed if a remote delivery caused a very long error message to be
3615 recorded - for instance if somebody sent an entire SpamAssassin report back
3616 as a large number of 550 error lines. This bug was coincidentally fixed by
3617 increasing the size of one of Exim's internal buffers (big_buffer) that
3618 happened as part of the Exiscan merge. However, to be on the safe side, I
3619 have made the code more robust (and fixed the comments that describe what
3622 53. Now that there can be additional text after "Completed" in log lines (if
3623 the queue_time_overall log selector is set), a one-byte patch to exigrep
3624 was needed to allow it to recognize "Completed" as not the last thing in
3627 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3628 patch that reportedly fixes this has been added. I am not expert enough to
3629 create a test for it. This is what the patch creator wrote:
3631 "I found a little strange behaviour of ldap code when working with
3632 Windows 2003 AD Domain, where users was placed in more than one
3633 Organization Units. When I tried to give exim partial DN, the exit code
3634 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3635 But simultaneously result of request was absolutely normal ldap result,
3636 so I produce this patch..."
3638 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3639 REFERENCE, so I have modified the code to exclude the patch when that macro
3642 55. Some experimental protocols are using DNS PTR records for new purposes. The
3643 keys for these records are domain names, not reversed IP addresses. The
3644 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3645 leaves it alone. Component reversal etc. now happens only for IP addresses.
3648 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3650 57. Double the size of the debug message buffer (to 2048) so that more of very
3651 long debug lines gets shown.
3653 58. The exicyclog utility now does better if the number of log files to keep
3654 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3656 59. Two changes related to the smtp_active_hostname option:
3658 (1) $smtp_active_hostname is now available as a variable.
3659 (2) The default for smtp_banner uses $smtp_active_hostname instead
3660 of $primary_hostname.
3662 60. The host_aton() function is supposed to be passed a string that is known
3663 to be a valid IP address. However, in the case of IPv6 addresses, it was
3664 not checking this. This is a hostage to fortune. Exim now panics and dies
3665 if the condition is not met. A case was found where this could be provoked
3666 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3667 components; fortuitously, this particular loophole had already been fixed
3668 by change 4.50/55 above.
3670 If there are any other similar loopholes, the new check in host_aton()
3671 itself should stop them being exploited. The report I received stated that
3672 data on the command line could provoke the exploit when Exim was running as
3673 exim, but did not say which command line option was involved. All I could
3674 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3675 running as the user.
3678 61. There was a buffer overflow vulnerability in the SPA authentication code
3679 (which came originally from the Samba project). I have added a test to the
3680 spa_base64_to_bits() function which I hope fixes it.
3683 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3684 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3686 63. The daemon start-up calls getloadavg() while still root for those OS that
3687 need the first call to be done as root, but it missed one case: when
3688 deliver_queue_load_max is set with deliver_drop_privilege. This is
3689 necessary for the benefit of the queue runner, because there is no re-exec
3690 when deliver_drop_privilege is set.
3692 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3693 This has been fixed.
3695 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3696 was in use, was not putting the data itself into the right store pool;
3697 consequently, it could be overwritten for a subsequent message in the same
3698 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3701 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3703 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3704 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3705 the function were treating the return as a boolean value, which happened to
3706 work because 0=false and not-0=true, but is not correct code.
3708 68. The host_aton() function was not handling scoped IPv6 addresses (those
3709 with, for example, "%eth0" on the end) correctly.
3711 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3712 negated items (that is, ~something) in unsigned ints. Some compilers
3713 apparently mutter when there is no cast.
3715 70. If an address verification called from an ACL failed, and did not produce a
3716 user-specific message (i.e. there was only a "system" message), nothing was
3717 put in $acl_verify_message. In this situation, it now puts the system
3720 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3721 session; change 4.31/43 added the unwanted input to the log line - except
3722 that it did not do this in the start of session case. It now does.
3724 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3725 This is wrong and can cause the other end to generate a synchronization
3726 error if it is another Exim or anything else that does the synchronization
3727 check. A QUIT command is no longer sent after a timeout.
3729 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3730 during host lookups.
3732 74. The defer_ok option of callout verification was not working if it was used
3733 when verifying addresses in header lines, that is, for this case:
3735 verify = header_sender/callout=defer_ok
3737 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3738 those file descriptors could be used for SMTP connections. If anything
3739 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3740 could be sent to the SMTP client, causing chaos. The daemon now opens
3741 stdin, stdout, and stderr to /dev/null when it puts itself into the
3744 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3745 log by default. The user can override this with suitable Perl magic.
3747 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3748 the log message when discard triggers, was not working for the DATA ACL or
3749 for the non-SMTP ACL.
3751 78. Error message wording change in sieve.c.
3753 79. If smtp_accept_max_per_host was set, the number of connections could be
3754 restricted to fewer than expected, because the daemon was trying to set up
3755 a new connection before checking whether the processes handling previous
3756 connections had finished. The check for completed processes is now done
3757 earlier. On busy systems, this bug wouldn't be noticed because something
3758 else would have woken the daemon, and it would have reaped the completed
3761 80. If a message was submitted locally by a user whose login name contained one
3762 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3763 It caused a spool format error. I have fixed the spool reading code. A
3764 related problem was that the "from" clause in the Received: line became
3765 illegal because of the space(s). It is now covered by ${quote_local_part.
3767 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3770 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3772 83. Patch from Sieve maintainer to fix unterminated string problem in
3773 "vacation" handling.
3775 84. Some minor changes to the Linux configuration files to help with other
3776 OS variants using glibc.
3778 85. One more patch for Sieve to update vacation handling to latest spec.
3781 ----------------------------------------------------
3782 See the note above about the 4.44 and 4.50 releases.
3783 ----------------------------------------------------
3789 1. Change 4.43/35 introduced a bug that caused file counts to be
3790 incorrectly computed when quota_filecount was set in an appendfile
3793 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3794 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3797 3. Give more explanation in the error message when the command for a transport
3798 filter fails to execute.
3800 4. There are several places where Exim runs a non-Exim command in a
3801 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3802 was being done only for the command run by the queryprogram router. It is
3803 now done for all such subprocesses. The other cases are: ${run, transport
3804 filters, and the commands run by the lmtp and pipe transports.
3806 5. Some older OS have a limit of 256 on the maximum number of file
3807 descriptors. Exim was using setrlimit() to set 1000 as a large value
3808 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3809 systems. I've change it so that if it can't get 1000, it tries for 256.
3811 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3812 was an oversight, and furthermore, ever since the addition of extra
3813 controls (e.g. 4.43/32), the checks on when to allow different forms of
3814 "control" were broken. There should now be diagnostics for all cases when a
3815 control that does not make sense is encountered.
3817 7. $recipients is now available in the predata ACL (oversight).
3819 8. Tidy the search cache before the fork to do a delivery from a message
3820 received from the command line. Otherwise the child will trigger a lookup
3821 failure and thereby defer the delivery if it tries to use (for example) a
3822 cached ldap connection that the parent has called unbind on.
3824 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3825 of $address_data from the recipient verification was clobbered by the
3826 sender verification.
3828 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3829 was its contents. (It was OK if the option was not defined at all.)
3831 11. A "Completed" log line is now written for messages that are removed from
3832 the spool by the -Mrm option.
3834 12. $host_address is now set to the target address during the checking of
3835 ignore_target_hosts.
3837 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3838 being passed; this would have caused $sender_host_name to have been used if
3839 matching the list had actually called for a host name (not very likely,
3840 since this list is usually IP addresses). A host name is now passed as
3843 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3844 code that specifies a non-restarting handler (typically sigaction() in
3845 modern systems) in an attempt to fix a rare and obscure crash bug.
3847 15. Narrowed the window for a race in the daemon that could cause it to ignore
3848 SIGCHLD signals. This is not a major problem, because they are used only to
3849 wake it up if nothing else does.
3851 16. A malformed maildirsize file could cause Exim to calculate negative values
3852 for the mailbox size or file count. Odd effects could occur as a result.
3853 The maildirsize information is now recalculated if the size or filecount
3856 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3857 support for a long time. Removed HAVE_SYS_VFS_H.
3859 18. Updated exipick to current release from John Jetmore.
3861 19. Allow an empty sender to be matched against a lookup in an address list.
3862 Previously the only cases considered were a regular expression, or an
3865 20. Exim went into a mad DNS lookup loop when doing a callout where the
3866 host was specified on the transport, if the DNS lookup yielded more than
3869 21. The RFC2047 encoding function was originally intended for short strings
3870 such as real names; it was not keeping to the 75-character limit for
3871 encoded words that the RFC imposes. It now respects the limit, and
3872 generates multiple encoded words if necessary. To be on the safe side, I
3873 have increased the buffer size for the ${rfc2047: expansion operator from
3876 22. Failure to deliver a bounce message always caused it to be frozen, even if
3877 there was an errors_to setting on the router. The errors_to setting is now
3880 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3881 canonical form (fully expanded) before being placed in
3882 $sender_host_address.
3884 24. Updated eximstats to version 1.33
3886 25. Include certificate and key file names in error message when GnuTLS fails
3887 to set them up, because the GnuTLS error message doesn't include the name
3888 of the failing file when there is a problem reading it.
3890 26. Expand error message when OpenSSL has problems setting up cert/key files.
3893 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3894 (this can affect the format of dates).
3896 28. exim_tidydb, when checking for the continued existence of a message for
3897 which it has found a message-specific retry record, was not finding
3898 messages that were in split spool directories. Consequently, it was
3899 deleting retry records that should have stayed in existence.
3901 29. eximstats updated to version 1.35
3902 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3903 1.35 - bugfix such that pie charts by volume are generated correctly
3905 30. The SPA authentication driver was not abandoning authentication and moving
3906 on to the next authenticator when an expansion was forced to fail,
3907 contradicting the general specification for all authenticators. Instead it
3908 was generating a temporary error. It now behaves as specified.
3910 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3911 (the order specifies the preference for clients). The order is now AES256,
3912 AES128, 3DES, ARCFOUR128.
3914 31. Small patch to Sieve code - explicitly set From: when generating an
3917 32. Exim crashed if a remote delivery caused a very long error message to be
3918 recorded - for instance if somebody sent an entire SpamAssassin report back
3919 as a large number of 550 error lines. This bug was coincidentally fixed by
3920 increasing the size of one of Exim's internal buffers (big_buffer) that
3921 happened as part of the Exiscan merge. However, to be on the safe side, I
3922 have made the code more robust (and fixed the comments that describe what
3925 33. Some experimental protocols are using DNS PTR records for new purposes. The
3926 keys for these records are domain names, not reversed IP addresses. The
3927 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3928 leaves it alone. Component reversal etc. now happens only for IP addresses.
3931 34. The host_aton() function is supposed to be passed a string that is known
3932 to be a valid IP address. However, in the case of IPv6 addresses, it was
3933 not checking this. This is a hostage to fortune. Exim now panics and dies
3934 if the condition is not met. A case was found where this could be provoked
3935 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3936 components; fortuitously, this particular loophole had already been fixed
3937 by change 4.50/55 or 4.44/33 above.
3939 If there are any other similar loopholes, the new check in host_aton()
3940 itself should stop them being exploited. The report I received stated that
3941 data on the command line could provoke the exploit when Exim was running as
3942 exim, but did not say which command line option was involved. All I could
3943 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3944 running as the user.
3947 35. There was a buffer overflow vulnerability in the SPA authentication code
3948 (which came originally from the Samba project). I have added a test to the
3949 spa_base64_to_bits() function which I hope fixes it.
3952 36. The daemon start-up calls getloadavg() while still root for those OS that
3953 need the first call to be done as root, but it missed one case: when
3954 deliver_queue_load_max is set with deliver_drop_privilege. This is
3955 necessary for the benefit of the queue runner, because there is no re-exec
3956 when deliver_drop_privilege is set.
3958 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3959 was in use, was not putting the data itself into the right store pool;
3960 consequently, it could be overwritten for a subsequent message in the same
3961 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3964 38. Sometimes the final signoff response after QUIT could fail to get
3965 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3966 < 0 before doing a fflush(). This bug looks as though it goes back to the
3967 introduction of TLS in release 3.20, but "sometimes" must have been rare
3968 because the tests only now provoked it.
3974 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3975 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3976 Now it returns an integer. A number of calls were still expecting a T/F
3977 return. Fortuitously, in all cases, the tests worked in OK situations,
3978 which is the norm. However, things would have gone wrong on any write
3979 failures on the smtp file descriptor. This function is used when sending
3980 messages over SMTP and also when doing verify callouts.
3982 2. When Exim is called to do synchronous delivery of a locally submitted
3983 message (the -odf or -odi options), it no longer closes stderr before doing
3986 3. Implemented the mua_wrapper option.
3988 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3990 5. Implemented the functions header_remove(), header_testname(),
3991 header_add_at_position(), and receive_remove_recipient(), and exported them
3994 6. If an ACL "warn" statement specified the addition of headers, Exim already
3995 inserted X-ACL-Warn: at the start if there was no header name. However, it
3996 was not making this test for the second and subsequent header lines if
3997 there were newlines in the string. This meant that an invalid header could
3998 be inserted if Exim was badly configured.
4000 7. Allow an ACL "warn" statement to add header lines at the start or after all
4001 the Received: headers, as well as at the end.
4003 8. Added the rcpt_4xx retry error code.
4005 9. Added postmaster_mailfrom=xxx to callout verification option.
4007 10. Added mailfrom=xxxx to the callout verification option, for verify=
4010 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4011 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4013 12. Inserted some casts to stop certain compilers warning when using pointer
4014 differences as field lengths or precisions in printf-type calls (mostly
4015 affecting debugging statements).
4017 13. Added optional readline() support for -be (dynamically loaded).
4019 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4020 same clock tick as a message's arrival, so that its received time was the
4021 same as the "first fail" time on the retry record, and that message
4022 remained on the queue past the ultimate address timeout, every queue runner
4023 would try a delivery (because it was past the ultimate address timeout) but
4024 after another failure, the ultimate address timeout, which should have then
4025 bounced the address, did not kick in. This was a "< instead of <=" error;
4026 in most cases the first failure would have been in the next clock tick
4027 after the received time, and all would be well.
4029 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4030 being recognized when the domain list was tested by the match_domain
4031 condition in an expansion string.
4033 16. Added the ${str2b64: operator.
4035 17. Exim was always calling setrlimit() to set a large limit for the number of
4036 processes, without checking whether the existing limit was already
4037 adequate. (It did check for the limit on file descriptors.) Furthermore,
4038 errors from getrlimit() and setrlimit() were being ignored. Now they are
4039 logged to the main and panic logs, but Exim does carry on, to try to do its
4040 job under whatever limits there are.
4042 18. Imported PCRE 5.0.
4044 19. Trivial typo in log message " temporarily refused connection" (the leading
4047 20. If the log selector return_path_on_delivery was set and an address was
4048 redirected to /dev/null, the delivery process crashed because it assumed
4049 that a return path would always be set for a "successful" delivery. In this
4050 case, the whole delivery is bypassed as an optimization, and therefore no
4053 21. Internal re-arrangement: the function for sending a challenge and reading
4054 a response while authentication was assuming a zero-terminated challenge
4055 string. It's now changed to take a pointer and a length, to allow for
4056 binary data in such strings.
4058 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4060 23. Exim was not respecting finduser_retries when seeking the login of the
4061 uid under which it was called; it was always trying 10 times. (The default
4062 setting of finduser_retries is zero.) Also, it was sleeping after the final
4063 failure, which is pointless.
4065 24. Implemented tls_on_connect_ports.
4067 25. Implemented acl_smtp_predata.
4069 26. If the domain in control=submission is set empty, Exim assumes that the
4070 authenticated id is a complete email address when it generates From: or
4071 Sender: header lines.
4073 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4074 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4075 chown and chgrp in /bin and hostname in /usr/bin.
4077 28. Exim was keeping the "process log" file open after each use, just as it
4078 does for the main log. This opens the possibility of it remaining open for
4079 long periods when the USR1 signal hits a daemon. Occasional processlog
4080 errors were reported, that could have been caused by this. Anyway, it seems
4081 much more sensible not to leave this file open at all, so that is what now
4084 29. The long-running daemon process does not normally write to the log once it
4085 has entered its main loop, and it closes the log before doing so. This is
4086 so that log files can straightforwardly be renamed and moved. However,
4087 there are a couple of unusual error situations where the daemon does write
4088 log entries, and I had neglected to close the log afterwards.
4090 30. The text of an SMTP error response that was received during a remote
4091 delivery was being truncated at 512 bytes. This is too short for some of
4092 the long messages that one sometimes sees. I've increased the limit to
4095 31. It is now possible to make retry rules that apply only when a message has a
4096 specific sender, in particular, an empty sender.
4098 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4099 it possible to be selective about when SMTP synchronization is enforced.
4101 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4103 32. Implemented hosts_connection_nolog.
4105 33. Added an ACL for QUIT.
4107 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4110 35. Added mailbox_size and mailbox_filecount to appendfile.
4112 36. Added control = no_multiline_responses to ACLs.
4114 37. There was a bug in the logic of the code that waits for the clock to tick
4115 in the case where the clock went backwards by a substantial amount such
4116 that the microsecond fraction of "now" was more than the microsecond
4117 fraction of "then" (but the whole seconds number was less).
4119 38. Added support for the libradius Radius client library this is found on
4120 FreeBSD (previously only the radiusclient library was supported).
4126 1. When certain lookups returned multiple values in the form name=value, the
4127 quoting of the values was not always being done properly. Specifically:
4128 (a) If the value started with a double quote, but contained no whitespace,
4130 (b) If the value contained whitespace other than a space character (i.e.
4131 tabs or newlines or carriage returns) it was not quoted.
4132 This fix has been applied to the mysql and pgsql lookups by writing a
4133 separate quoting function and calling it from the lookup code. The fix
4134 should probably also be applied to nisplus, ibase and oracle lookups, but
4135 since I cannot test any of those, I have not disturbed their existing code.
4137 2. A hit in the callout cache for a specific address caused a log line with no
4138 reason for rejecting RCPT. Now it says "Previous (cached) callout
4139 verification failure".
4141 3. There was an off-by-one bug in the queryprogram router. An over-long
4142 return line was truncated at 256 instead of 255 characters, thereby
4143 overflowing its buffer with the terminating zero. As well as fixing this, I
4144 have increased the buffer size to 1024 (and made a note to document this).
4146 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4147 when Exim is waiting for an SMTP response from a remote server, Exim
4148 restarts its select() call on the socket, thereby resetting its timeout.
4149 This is not a problem when such interrupts are rare. Somebody set up a cron
4150 job to run exiwhat every 2 minutes, which is less than the normal select()
4151 timeout (5 or 10 minutes). This meant that the select() timeout never
4152 kicked in because it was always reset. I have fixed this by comparing the
4153 time when an interrupt arrives with the time at the start of the first call
4154 to select(). If more time than the timeout has elapsed, the interrupt is
4155 treated as a timeout.
4157 5. Some internal re-factoring in preparation for the addition of Sieve
4158 extensions (by MH). In particular, the "personal" test is moved to a
4159 separate function, and given an option for scanning Cc: and Bcc: (which is
4160 not set for Exim filters).
4162 6. When Exim created an email address using the login of the caller as the
4163 local part (e.g. when creating a From: or Sender: header line), it was not
4164 quoting the local part when it contained special characters such as @.
4166 7. Installed new OpenBSD configuration files.
4168 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4169 try to make them clearer.
4171 9. Callout options, other than the timeout value, were being ignored when
4172 verifying sender addresses in header lines. For example, when using
4174 verify = header_sender/callout=no_cache
4176 the cache was (incorrectly) being used.
4178 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4179 only the Cygwin environment.
4181 11. When return_path_on_delivery was set as a log selector, if different remote
4182 addresses in the same message used different return paths and parallel
4183 remote delivery occurred, the wrong values would sometimes be logged.
4184 (Whenever a remote delivery process finished, the return path value from
4185 the most recently started remote delivery process was logged.)
4187 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4188 lines when AUTH and/or TLS are in use. This is the "received protocol"
4189 field. Exim used to use "asmtp" for authenticated SMTP, without any
4190 indication (in the protocol name) for TLS use. Now it follows the RFC and
4191 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4192 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4193 names appear in log lines as well as in Received: header lines.
4195 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4196 extensions, and comparison tests, and to fix some bugs.
4198 14. Changes to the "personal" filter test:
4200 (1) The test was buggy in that it was just doing the equivalent of
4201 "contains" tests on header lines. For example, if a user's address was
4202 anne@some.where, the "personal" test would incorrectly be true for
4204 To: susanne@some.where
4206 This test is now done by extracting each address from the header in turn,
4207 and checking the entire address. Other tests that are part of "personal"
4208 are now done using regular expressions (for example, to check local parts
4209 of addresses in From: header lines).
4211 (2) The list of non-personal local parts in From: addresses has been
4212 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4213 taken from the Sieve specification recommendations.
4215 (3) If the message contains any header line starting with "List-" it is
4216 treated as non-personal.
4218 (4) The test for "circular" in the Subject: header line has been removed
4219 because it now seems ill-conceived.
4221 15. Minor typos in src/EDITME comments corrected.
4223 16. Installed latest exipick from John Jetmore.
4225 17. If headers_add on a router specified a text string that was too long for
4226 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4227 of string_sprintf() is now avoided.
4229 18. $message_body_size was not set (it was always zero) when running the DATA
4230 ACL and the local_scan() function.
4232 19. For the "mail" command in an Exim filter, no default was being set for
4233 the once_repeat time, causing a random time value to be used if "once" was
4234 specified. (If the value happened to be <= 0, no repeat happened.) The
4235 default is now 0s, meaning "never repeat". The "vacation" command was OK
4236 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4237 (I found it when inspecting the code).
4239 20. There is now an overall timeout for performing a callout verification. It
4240 defaults to 4 times the callout timeout, which applies to individual SMTP
4241 commands during the callout. The overall timeout applies when there is more
4242 than one host that can be tried. The timeout is checked before trying the
4243 next host. This prevents very long delays if there are a large number of
4244 hosts and all are timing out (e.g. when the network connections are timing
4245 out). The value of the overall timeout can be changed by specifying an
4246 additional sub-option for "callout", called "maxwait". For example:
4248 verify = sender/callout=5s,maxwait=20s
4250 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4251 to the end before writing, but this should make it even safer).
4253 22. Exim was forgetting that it had advertised PIPELINING for the second and
4254 subsequent messages on an SMTP connection. It was also not resetting its
4255 memory on STARTTLS and an internal HELO.
4257 23. When Exim logs an SMTP synchronization error within a session, it now
4258 records whether PIPELINING has been advertised or not.
4260 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4261 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4262 rather than long int.
4264 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4266 26. Added the never_mail option to autoreply.
4272 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4273 crash if the getsockname() call failed; this can happen if a connection is
4274 closed very soon after it is established. The problem was simply in the
4275 order in which certain operations were done, causing Exim to try to write
4276 to the SMTP stream before it had set up the file descriptor. The bug has
4277 been fixed by making things happen in the correct order.
4283 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4284 before the connection was closed, thus losing the rejection response.
4286 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4287 some early Solaris releases, but causes trouble in current releases where
4288 socklen_t is defined.
4290 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4293 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4296 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4297 that releasing the top of it at the end releases what was used for sub-
4298 expansions (unless the block got too big). However, discard this block if
4299 the first thing is a variable or header, so that we can use its block when
4300 it is dynamic (useful for very large $message_headers, for example).
4302 6. Lookups now cache *every* query, not just the most recent. A new, separate
4303 store pool is used for this. It can be recovered when all lookup caches are
4304 flushed. Lookups now release memory at the end of their result strings.
4305 This has involved some general refactoring of the lookup sources.
4307 7. Some code has been added to the store_xxx() functions to reduce the amount
4308 of flapping under certain conditions.
4310 8. log_incoming_interface used to affect only the <= reception log lines. Now
4311 it causes the local interface and port to be added to several more SMTP log
4312 lines, for example "SMTP connection from", and rejection lines.
4314 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4316 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4318 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4319 could be overwritten at the end of the current message (or the start of a
4320 new message if it was set in a HELO ACL). The value is now preserved for
4321 the duration of the SMTP connection.
4323 12. If a transport had a headers_rewrite setting, and a matching header line
4324 contained an unqualified address, that address was qualified, even if it
4325 did not match any rewriting rules. The underlying bug was that the values
4326 of the flags that permit the existence of unqualified sender and recipient
4327 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4328 non-local messages, and by -bnq for local messages) were not being
4329 preserved with the message after it was received.
4331 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4332 "next input=" as part of the text comprising the host identity instead of
4333 the correct text. The code was using the same buffer for two different
4334 strings. However, depending on which order the printing function evaluated
4335 its arguments, the bug did not always show up. Under Linux, for example, my
4336 test suite worked just fine.
4338 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4339 doesn't work with some older versions of Perl. It has been changed to "my",
4340 which in any case is probably the better facility to use.
4342 15. A really picky compiler found some instances of statements for creating
4343 error messages that either had too many or two few arguments for the format
4346 16. The size of the buffer for calls to the DNS resolver has been increased
4347 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4348 for addresses that have a lot of PTR records. This alleviates a problem; it
4349 does not fully solve it.
4351 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4352 buffer now truncates the list and logs the incident, which is the same
4353 action as happens when Exim is looking up a host name and its aliases.
4354 Previously in this situation something unpredictable would happen;
4355 sometimes it was "internal error: store_reset failed".
4357 18. If a server dropped the connection unexpectedly when an Exim client was
4358 using GnuTLS and trying to read a response, the client delivery process
4359 crashed while trying to generate an error log message.
4361 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4362 string, for example:
4364 warn message = H1: something\nH2: something
4366 the text was added as a single header line from Exim's point of view
4367 though it ended up OK in the delivered message. However, searching for the
4368 second and subsequent header lines using $h_h2: did not work. This has been
4369 fixed. Similarly, if a system filter added multiple headers in this way,
4370 the routers could not see them.
4372 20. Expanded the error message when iplsearch is called with an invalid key to
4373 suggest using net-iplsearch in a host list.
4375 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4376 ACLs are no longer actually imposed (and a message to that effect is
4379 22. If a "gecos" field in a passwd entry contained escaped characters, in
4380 particular, if it contained a \" sequence, Exim got it wrong when building
4381 a From: or a Sender: header from that name. A second bug also caused
4382 incorrect handling when an unquoted " was present following a character
4383 that needed quoting.
4385 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4386 was not being matched caselessly.
4388 24. Arranged for all hyphens in the exim.8 source to be escaped with
4391 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4392 properly. Recipient callout cache records were still being keyed to include
4393 the sender, even when use_sender was set false. This led to far more
4394 callouts that were necessary. The sender is no longer included in the key
4395 when use_sender is false.
4397 26. Added "control = submission" modifier to ACLs.
4399 27. Added the ${base62d: operator to decode base 62 numbers.
4401 28. dnsdb lookups can now access SRV records.
4403 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4404 the configuration file.
4406 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4407 (-v) mode. This makes the output for a verbose queue run more intelligible.
4409 31. Added a use_postmaster feature to recipient callouts.
4411 32. Added the $body_zerocount variable, containing the number of binary zero
4412 bytes in the message body.
4414 33. The time of last modification of the "new" subdirectory is now used as the
4415 "mailbox time last read" when there is a quota error for a maildir
4418 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4420 35. Added +ignore_unknown as a special item in host lists.
4422 36. Code for decoding IPv6 addresses in host lists is now included, even if
4423 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4424 address was recognized as an IP address, but was then not correctly decoded
4425 into binary, causing unexpected and incorrect effects when compared with
4432 1. Very minor rewording of debugging text in manualroute to say "list of
4433 hosts" instead of "hostlist".
4435 2. If verify=header_syntax was set, and a header line with an unqualified
4436 address (no domain) and a large number of spaces between the end of the
4437 name and the colon was received, the reception process suffered a buffer
4438 overflow, and (when I tested it) crashed. This was caused by some obsolete
4439 code that should have been removed. The fix is to remove it!
4441 3. When running in the test harness, delay a bit after writing a bounce
4442 message to get a bit more predictability in the log output.
4444 4. Added a call to search_tidyup() just before forking a reception process. In
4445 theory, someone could use a lookup in the expansion of smtp_accept_max_
4446 per_host which, without the tidyup, could leave open a database connection.
4448 5. Added the variables $recipient_data and $sender_data which get set from a
4449 lookup success in an ACL "recipients" or "senders" condition, or a router
4450 "senders" option, similar to $domain_data and $local_part_data.
4452 6. Moved the writing of debug_print from before to after the "senders" test
4455 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4456 problems for message scanning, either using a data ACL, or using
4457 local_scan() because the Received: header was not generated till after they
4458 were called (in order to set the time as the time of reception completion).
4459 I have revised the way this works. The header is now generated after the
4460 body is received, but before the ACL or local_scan() are called. After they
4461 are run, the timestamp in the header is updated.
4467 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4468 before starting a queue runner without re-exec. This happened only when
4469 deliver_drop_privilege was set or when the Exim user was set to root. The
4470 effect of the bug was that timeouts during subsequent deliveries caused
4471 crashes instead of being properly handled. The handler is now left at its
4472 default (and expected) setting.
4474 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4475 message, again when deliver_drop_privilege is set or Exim is run as root.
4476 The bug described in (1) was not present in this case, but the tidying up
4477 of the other signals was missing. I have made the two cases consistent.
4479 3. The ignore_target_hosts setting on a manualroute router was being ignored
4480 for hosts that were looked up using the /MX notation.
4482 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4485 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4486 operated on the sender address. After changing the $sender_address to <>
4487 for the sender address verify, Exim was re-instated it as the original
4488 (before rewriting) address, but remembering that it had rewritten it, so it
4489 wasn't rewriting it again. This bug also had the effect of breaking the
4490 sender address verification caching when the sender address was rewritten.
4492 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4493 This has been changed so that if the ip literal address matches
4494 ignore_target_hosts, the router declines.
4496 7. Added expansion conditions match_domain, match_address, and match_local_
4497 part (NOT match_host).
4499 8. The placeholder for the Received: header didn't have a length field set.
4501 9. Added code to Exim itself and to exim_lock to test for a specific race
4502 condition that could lead to file corruption when using MBX delivery. The
4503 issue is with the lockfile that is created in /tmp. If this file is removed
4504 after a process has opened it but before that process has acquired a lock,
4505 there is the potential for a second process to recreate the file and also
4506 acquire a lock. This could lead to two Exim processes writing to the file
4507 at the same time. The added code performs the same test as UW imapd; it
4508 checks after acquiring the lock that its file descriptor still refers to
4509 the same named file.
4511 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4512 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4515 11. Added the smtp_active_hostname option. If used, this will typically be made
4516 to depend on the incoming interface address. Because $interface_address is
4517 not set up until the daemon has forked a reception process, error responses
4518 that can happen earlier (such as "too many connections") no longer contain
4521 12. If an expansion in a condition on a "warn" statement fails because a lookup
4522 defers, the "warn" statement is abandoned, and the next ACL statement is
4523 processed. Previously this caused the whole ACL to be aborted.
4525 13. Added the iplsearch lookup type.
4527 14. Added ident_timeout as a log selector.
4529 15. Added tls_certificate_verified as a log selector.
4531 16. Added a global option tls_require_ciphers (compare the smtp transport
4532 option of the same name). This controls incoming TLS connections.
4534 17. I finally figured out how to make tls_require_ciphers do a similar thing
4535 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4536 before starting the TLS session.
4538 18. Tabs are now shown as \t in -bP output.
4540 19. If the log selector return_path_on_delivery was set, Exim crashed when
4541 bouncing a message because it had too many Received: header lines.
4543 20. If two routers both had headers_remove settings, and the first one included
4544 a superfluous trailing colon, the final name in the first list and the
4545 first name in the second list were incorrectly joined into one item (with a
4546 colon in the middle).
4552 1. Added -C and -D options to the exinext utility, mainly to make it easier
4553 to include in the automated testing, but these could be helpful when
4554 multiple configurations are in use.
4556 2. The exinext utility was not formatting the output nicely when there was
4557 an alternate port involved in the retry record key, nor when there was a
4558 message id as well (for retries that were specific to a specific message
4559 and a specific host). It was also confused by IPv6 addresses, because of
4560 the additional colons they contain. I have fixed the IPv4 problem, and
4561 patched it up to do a reasonable job for IPv6.
4563 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4564 delivery, the log line now contains "pipelined" if PIPELINING was used.
4566 4. An SMTP transport process used to panic and die if the bind() call to set
4567 an explicit outgoing interface failed. This has been changed; it is now
4568 treated in the same way as a connect() failure.
4570 5. A reference to $sender_host_name in the part of a conditional expansion
4571 that was being skipped was still causing a DNS lookup. This no longer
4574 6. The def: expansion condition was not recognizing references to header lines
4575 that used bh_ and bheader_.
4577 7. Added the _cache feature to named lists.
4579 8. The code for checking quota_filecount in the appendfile transport was
4580 allowing one more file than it should have been.
4582 9. For compatibility with Sendmail, the command line option
4590 and sets the incoming protocol and host name (for trusted callers). The
4591 host name and its colon can be omitted when only the protocol is to be set.
4592 Note the Exim already has two private options, -pd and -ps, that refer to
4593 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4594 "s", but I don't think that's a major issue.
4596 10. A number of refactoring changes to the code, none of which should affect
4599 (a) The number of logging options was getting close to filling up the
4600 32-bit word that was used as a bit map. I have split them into two classes:
4601 those that are passed in the argument to log_write(), and those that are
4602 only ever tested independently outside of that function. These are now in
4603 separate 32-bit words, so there is plenty of room for expansion again.
4604 There is no change in the user interface or the logging behaviour.
4606 (b) When building, for example, log lines, the code previously used a
4607 macro that called string_cat() twice, in order to add two strings. This is
4608 not really sufficiently general. Furthermore, there was one instance where
4609 it was actually wrong because one of the argument was used twice, and in
4610 one call a function was used. (As it happened, calling the function twice
4611 did not affect the overall behaviour.) The macro has been replaced by a
4612 function that can join an arbitrary number of extra strings onto a growing
4615 (c) The code for expansion conditions now uses a table and a binary chop
4616 instead of a serial search (which was left over from when there were very
4617 few conditions). Also, it now recognizes conditions like "pam" even when
4618 the relevant support is not compiled in: a suitably worded error message is
4619 given if an attempt is made to use such a condition.
4621 11. Added ${time_interval:xxxxx}.
4623 12. A bug was causing one of the ddress fields not to be passed back correctly
4624 from remote delivery subprocesses. The field in question was not being
4625 subsequently used, so this caused to problems in practice.
4627 13. Added new log selectors queue_time and deliver_time.
4629 14. Might have fixed a bug in maildirsizefile handling that threw up
4630 "unexpected character" debug warnings, and recalculated the data
4631 unnecessarily. In any case, I expanded the warning message to give more
4634 15. Added the message "Restricted characters in address" to the statements in
4635 the default ACL that block characters like @ and % in local parts.
4637 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4638 Three changes have been made:
4640 (a) There was a serious bug; a negative response to MAIL caused the whole
4641 recipient domain to be cached as invalid, thereby blocking all messages
4642 to all local parts at the same domain, from all senders. This bug has
4643 been fixed. The domain is no longer cached after a negative response to
4644 MAIL if the sender used is not empty.
4646 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4649 (c) A new callout option, "use_sender" has been added for people who want
4650 the modified behaviour.
4656 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4659 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4660 indeed breaks things for older releases.
4662 3. Added additional logging to the case where there is a problem reading data
4663 from a filter that is running in a subprocess using a pipe, in order to
4664 try to track down a specific problem.
4666 4. Testing facility fudge: when running in the test harness and attempting
4667 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4668 getting "No route to host". Convert this to a timeout.
4670 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4673 6. Some OS don't have socklen_t but use size_t instead. This affects the
4674 fifth argument of getsockopt() amongst other things. This is now
4675 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4676 can be set for individual OS. I have set it for SunOS5, OSF1, and
4677 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4678 some earlier ones do not.
4680 7. Change 4.30/15 was not doing the test caselessly.
4682 8. The standard form for an IPv6 address literal was being rejected by address
4683 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4684 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4685 this, as well as the form without the "IPv6" on the front (but only when
4686 address literals are enabled, of course).
4688 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4690 10. Exim crashed if a message with an empty sender address specified by -f
4691 encountered a router with an errors_to setting. This could be provoked only
4692 by a command such as
4696 where an empty string was supplied; "<>" did not hit this bug.
4698 11. Installed PCRE release 4.5.
4700 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4701 remained set. It is now erased.
4703 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4704 times from message ids (which are base 36 rather than the normal 62).
4706 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4707 were being counted as actual protocol errors, and logged if the log
4708 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4709 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4710 and DATA following a set of rejected RCPTs do not count as protocol errors.
4711 In other words, Exim assumes they were pipelined, though this may not
4712 actually be the case. Of course, in all cases the client gets an
4713 appropriate error code.
4715 15. If a lookup fails in an ACL condition, a message about the failure may
4716 be available; it is used if testing the ACL cannot continue, because most
4717 such messages specify what the cause of the deferral is. However, some
4718 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4719 that caused an old message to be retained and used if a later statement
4720 caused a defer, replacing the real cause of the deferral.
4722 16. If an IP address had so many PTR records that the DNS lookup buffer
4723 was not large enough to hold them, Exim could crash while trying to process
4724 the truncated data. It now detects and logs this case.
4726 17. Further to 4.21/58, another change has been made: if (and only if) the
4727 first line of a message (the first header line) ends with CRLF, a bare LF
4728 in a subsequent header line has a space inserted after it, so as not to
4729 terminate the header.
4731 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4732 unnecessarily, used atoi() instead of strtol(), and didn't check the
4733 termination when getting file sizes from file names by regex.
4735 19. Completely re-implemented the support for maildirsize files, in the light
4736 of a number of problems with the previous contributed implementation
4737 (4.30/29). In particular:
4739 . If the quota is zero, the maildirsize file is maintained, but no quota is
4742 . If the maildir directory does not exist, it is created before any attempt
4743 to write a maildirsize file.
4745 . The quota value in the file is just a cache; if the quota is changed in
4746 the transport, the new value overrides.
4748 . A regular expression is available for excluding directories from the
4751 20. The autoreply transport checks the characters in options that define the
4752 message's headers; it allows continued headers, but it was checking with
4753 isspace() after an embedded newline instead of explicitly looking for a
4756 21. If all the "regular" hosts to which an address was routed had passed their
4757 expiry times, and had not reached their retry times, the address was
4758 bounced, even if fallback hosts were defined. Now Exim should go on to try
4761 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4762 equivalent code in the SMTP transport. Some hosts send humungous responses
4763 to HELO/EHLO, more than 1024 it seems.
4765 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4766 gives compiler warnings in some environments. I've now done it "properly",
4769 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4770 (because of problems with the built-in one) was declared to return uschar *
4771 instead of char *, causing compiler failure.
4773 25. Fixed a file descriptor leak when processing alias/forward files.
4775 26. Fixed a minor format string issue in dbfn.c.
4777 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4779 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4780 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4783 29. When a sender address is verified, it is cached, to save repeating the test
4784 when there is more than one recipient in a message. However, when the
4785 verification involves a callout, it is possible for different callout
4786 options to be set for different recipients. It is too complicated to keep
4787 track of this in the cache, so now Exim always runs a verification when a
4788 callout is required, relying on the callout cache for the optimization.
4789 The overhead is duplication of the address routing, but this should not be
4792 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4793 to be verified with callout=postmaster, and the main callout worked but the
4794 postmaster check failed, the verification correctly failed. However, if a
4795 subsequent RCPT command asked for sender verification *without* the
4796 postmaster check, incorrect caching caused this verification also to fail,
4799 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4800 it was not caching the DNS options (qualify_single, search_parents) that
4801 were used when the lookup failed. A subsequent lookup with different
4802 options therefore always gave the same answer, though there were cases
4803 where it should not have. (Example: a "domains = !$mx_any" option on a
4804 dnslookup router: the "domains" option is always processed without any
4805 widening, but the router might have qualify_single set.) Now Exim uses the
4806 cached value only when the same options are set.
4808 32. Added John Jetmore's "exipick" utility to the distribution.
4810 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4811 than a timeout (e.g. a certificate is required, and is not provided), an
4812 Exim server now closes the connection immediately. Previously it waited for
4813 the client to close - but if the client is SSL, it seems that they each
4814 wait for each other, leading to a delay before one of them times out.
4816 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4817 maintained 0.8.x compatibility because I don't think many are using it, and
4818 it is clearly obsolete.
4820 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4823 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4824 client certificate was expired. A simple patch fixes this, though I don't
4825 understand the full logic of why the verify callback is called multiple
4828 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4829 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4830 which causes problems with some clients (such as the Certicom SSL Plus
4831 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4832 disables the coutermeasure allowing Eudora to connect."
4834 38. Exim was not checking that a write() to a log file succeeded. This could
4835 lead to Bad Things if a log got too big, in particular if it hit a file
4836 size limit. Exim now panics and dies if it cannot write to a log file, just
4837 as it does if it cannot open a log file.
4839 39. Modified OS/Makefile-Linux so that it now contains
4841 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4843 The two -D definitions ensure that Exim is compiled with large file
4844 support, which makes it possible to handle log files that are bigger than
4847 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4848 instance) a domain was checked against a named list that involved a lookup,
4849 causing $domain_data to be set, then another domain was checked against the
4850 same list, then the first domain was re-checked, the value of $domain_data
4851 after the final check could be wrong. In particular, if the second check
4852 failed, it could be set empty. This bug probably also applied to
4855 41. The strip_trailing_dot option was not being applied to the address given
4856 with the -f command-line option.
4858 42. The code for reading a message's header from the spool was incrementing
4859 $received_count, but never initializing it. This meant that the value was
4860 incorrect (doubled) while delivering a message in the same process in which
4861 it was received. In the most common configuration of Exim, this never
4862 happens - a fresh exec is done - but it can happen when
4863 deliver_drop_privilege is set.
4865 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4866 it now includes up to 150 characters of the unexpected data in the log
4869 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4870 and building data strings. The size of both of these buffers was 10 000
4871 bytes - far larger than anybody would *ever* want, thought I. Needless to
4872 say, somebody hit the limit. I have increased the maximum line length to
4873 20 000 and the maximum data length of concatenated lines to 100 000. I have
4874 also fixed two bugs, because there was no checking on these buffers. Tsk,
4875 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4876 buffer is too small.
4878 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4879 lsearch lookups. Now it does.
4881 46. When parsing a route_list item in a manualroute router, a fixed-length
4882 buffer was used for the list of hosts. I made this 1024 bytes long,
4883 thinking that nobody would ever have a list of hosts that long. Wrong.
4884 Somebody had a whole pile of complicated expansion conditions, and the
4885 string was silently truncated, leading to an expansion error. It turns out
4886 that it is easier to change to an unlimited length (owing to other changes
4887 that have happened since this code was originally written) than to build
4888 structure for giving a limitation error. The length of the item that
4889 expands into the list of hosts is now unlimited.
4891 47. The lsearch lookup could not handle data where the length of text line was
4892 more than 4095 characters. Such lines were truncated, leading to shortened
4893 data being returned. It should now handle lines of any length.
4895 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4896 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4899 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4900 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4901 no longer happens in comment lines. A list of replacements is now placed
4902 at the head of all of the source files, except those whose only change is
4903 to replace PERL_COMMAND in the very first #! line.
4905 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4906 messages on the queue, with a bottom-up merge sort, using code contributed
4907 by Michael Haardt. This should make operations like -bp somewhat faster on
4908 large queues. It won't affect queue runners, except when queue_run_in_order
4911 51. Installed eximstats 1.31 in the distribution.
4913 52. Added support for SRV lookups to the dnslookup router.
4915 53. If an ACL referred to $message_body or $message_body_end, the value was not
4916 reset for any messages that followed in the same SMTP session.
4918 54. The store-handling optimization for building very long strings was not
4919 differentiating between the different store pools. I don't think this
4920 actually made any difference in practice, but I've tidied it.
4922 55. While running the routers to verify a sender address, $sender_address
4923 was still set to the sender address. This is wrong, because when routing to
4924 send a bounce to the sender, it would be empty. Therefore, I have changed
4925 it so that, while verifying a sender address, $sender_address is set to <>.
4926 (There is no change to what happens when verifying a recipient address.)
4928 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4929 target A or AAAA records (if not already returned) without resetting the
4930 qualify_single or search_parents options of the DNS resolver. These are
4931 inappropriate in this case because the targets of MX and SRV records must
4932 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4933 target that, when qualified, matched something in the local domain. These
4934 two options are now turned off when doing these lookups.
4936 57. It seems that at least some releases of Reiserfs (which does not have the
4937 concept of a fixed number of inodes) returns zero and not -1 for the
4938 number of available inodes. This interacted badly with check_spool_inodes,
4939 which assumed that -1 was the "no such thing" setting. What I have done is
4940 to check that the total number of inodes is greater than zero before doing
4941 the test of how many are available.
4943 58. When a "warn" ACL statement has a log_message modifier, the message is
4944 remembered, and not repeated. This is to avoid a lot of repetition when a
4945 message has many recipients that cause the same warning to be written.
4946 Howewer, Exim was preserving the list of already written lines for an
4947 entire SMTP session, which doesn't seem right. The memory is now reset if a
4948 new message is started.
4950 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4951 debugging output unless log_rewrite was also set.
4953 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4954 of (int)(handle) when we know that handle contains (void *)(-1).
4956 61. The Exim daemon panic-logs an error return when it closes the incoming
4957 connection. However "connection reset by peer" seems to be common, and
4958 isn't really an error worthy of noting specially, so that particular error
4961 62. When Exim is trying to find all the local interfaces, it used to panic and
4962 die if the ioctl to get the interface flags failed. However, it seems that
4963 on at least one OS (Solaris 9) it is possible to have an interface that is
4964 included in the list of interfaces, but for which you get a failure error
4965 for this call. This happens when the interface is not "plumbed" into a
4966 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4967 failure of the "get flags" call assumes that the interface is down.
4969 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4970 makes life easier for people who are doing arithmetic on fields extracted
4971 from dates, where you often get leading zeros that should not be
4972 interpreted as octal.
4974 64. Added qualify_domain to the redirect router, to override the global
4977 65. If a pathologically long header line contained very many addresses (the
4978 report of this problem mentioned 10 000) and each of them was rewritten,
4979 Exim could use up a very large amount of memory. (It kept on making new
4980 copies of the header line as it rewrote, and never released the old ones.)
4981 At the expense of a bit more processing, the header rewriting function has
4982 been changed so that it no longer eats memory in this way.
4984 66. The generation of the Received: header has been moved from the time that a
4985 message starts to be received, to the time that it finishes. The timestamp
4986 in the Received: header should now be very close to that of the <= log
4987 line. There are two side-effects of this change:
4989 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4990 logged header lines no longer include the local Received: line, because
4991 it has not yet been created. The same applies to a copy of the message
4992 that is returned to a non-SMTP sender when a message is rejected.
4994 (b) When a filter file is tested using -bf, no additional Received: header
4995 is added to the test message. After some thought, I decided that this
4998 This change does not affect the value of $received_for. It is still set
4999 after address rewriting, but before local_scan() is called.
5001 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5003 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5004 gave an unhelpful panic error message, and a defer error. I have managed to
5005 change this behaviour so that it now rejects any supplied certificate,
5006 which seems right, as the list of acceptable certificates is empty.
5008 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5009 gave an unhelpful defer error. I have not managed to make this reject any
5010 supplied certificates, but the error message it gives is "no certificate
5011 supplied", which is not helpful.
5013 70. exigrep's output now also includes lines that are not associated with any
5014 message, but which match the given pattern. Implemented by a patch from
5015 Martin Sluka, which also tidied up the Perl a bit.
5017 71. Recipient callout verification, like sender verification, was using <> in
5018 the MAIL FROM command. This isn't really the right thing, since the actual
5019 sender may affect whether the remote host accepts the recipient or not. I
5020 have changed it to use the actual sender in the callout; this means that
5021 the cache record is now keyed on a recipient/sender pair, not just the
5022 recipient address. There doesn't seem to be a real danger of callout loops,
5023 since a callout by the remote host to check the sender would use <>.
5024 [SEE ABOVE: changed after hitting problems.]
5026 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5027 temporary errors. However, in the case of such a code being given after
5028 the end of a data transmission (i.e. after ".") Exim was failing to write
5029 a retry record for the message. (Yes, there was some broken host that was
5030 actually sending 8xx at this point.)
5032 73. An unknown lookup type in a host list could cause Exim to panic-die when
5033 the list was checked. (An example that provoked this was putting <; in the
5034 middle of a list instead of at the start.) If this happened during a DATA
5035 ACL check, a -D file could be left lying around. This kind of configuration
5036 error no longer causes Exim to die; instead it causes a defer errror. The
5037 incident is still logged to the main and panic logs.
5039 74. Buglet left over from Exim 3 conversion. The message "too many messages
5040 in one connection" was written to the rejectlog but not the mainlog, except
5041 when address rewriting (yes!) was being logged.
5043 75. Added write_rejectlog option.
5045 76. When a system filter was run not as root (that is, when system_filter_user
5046 was set), the values of the $n variables were not being returned to the
5047 main process; thus, they were not subsequently available in the $sn
5050 77. Added +return_path_on_delivery log selector.
5052 78. A connection timeout was being treated differently from recipients deferred
5053 when testing hosts_max_try with a message that was older than the host's
5054 retry timeout. (The host should not be counted, thus allowing all hosts to
5055 be tried at least once before bouncing.) This may have been the cause of an
5056 occasionally reported bug whereby a message would remain on the queue
5057 longer than the retry timeout, but would be bounced if a delivery was
5058 forced. I say "may" because I never totally pinned down the problem;
5059 setting up timeout/retry tests is difficult. See also the next item.
5061 79. The ultimate address timeout was not being applied to errors that involved
5062 a combination of host plus message (for example, a timeout on a MAIL
5063 command). When an address resolved to a number of possible hosts, and they
5064 were not all tried for each delivery (e.g. because of hosts_max_try), a
5065 message could remain on the queue longer than the retry timeout.
5067 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5070 81. Fixed an obscure SMTP outgoing bug which required at least the following
5071 conditions: (a) there was another message waiting for the same server;
5072 (b) the server returned 5xx to all RCPT commands in the first message so
5073 that the message was not completed; (c) the server dropped the connection
5074 or gave a negative response to the RSET that Exim sends to abort the
5075 transaction. The observed case was a dropped connection after DATA that had
5076 been sent in pipelining mode. That is, the server had advertised PIPELINING
5077 but was not implementing it correctly. The effect of the bug was incorrect
5078 behaviour, such as trying another host, and this could lead to a crash.
5084 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5085 and daemon.c were passed as pointers to ints; they should have been
5086 pointers to socklen_t variables (which are typically unsigned ints).
5088 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5091 3. Fixed a really odd bug that affected only the testing scheme; patching a
5092 certain fixed string in the binary changed the value of another string that
5093 happened to be identical to the end of the original first string.
5095 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5096 name", it returns that address as the IP address. On some operating
5097 systems (e.g. Solaris), it also passes back the IP address string as the
5098 "host name". However, on others (e.g. Linux), it passes back an empty
5099 string. Exim wasn't checking for this, and was changing the host name to an
5100 empty string, assuming it had been canonicized.
5102 5. Although rare, it is permitted to have more than one PTR record for a given
5103 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5104 all the names associated with an address, because they do in Solaris.
5105 However, it seems that they do not in Linux for data that comes from the
5106 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5107 I found this out when I moved to a new Linux workstation and tried to run
5108 the Exim test suite.
5110 To get round this problem I have changed the code so that it now does its
5111 own call to the DNS to look up PTR records when searching for a host name.
5112 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5113 addresses that are only in /etc/hosts are still found.
5115 This behaviour is, however, controlled by an option called host_lookup_
5116 order, which defaults to "bydns:byaddr". If people want to use the other
5117 order, or indeed, just use one or the other means of lookup, they can
5118 specify it in this variable.
5120 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5121 some operating systems, this comes back from gethostbyaddr() as an empty
5122 string, and this is what Exim used to test for. However, it seems that in
5123 other systems, "." is yielded. Exim now tests for this case too.
5125 7. The values of check_spool_space and check_log_space are now held internally
5126 as a number of kilobytes instead of an absolute number of bytes. If a
5127 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5128 kilobyte. This means that much larger values can be stored.
5130 8. Exim monitor: an attempt to get the action menu when not actually pointing
5131 at a message produces an empty menu entitled "No message selected". This
5132 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5133 no entries in it ("Shell widget menu has zero width and/or height"). So I
5134 have added a single, blank menu entry in this case.
5136 9. Added ${quote_local_part.
5138 10. MIME decoding is now applied to the contents of Subject: header lines when
5141 11. Now that a reference to $sender_host_address automatically causes a reverse
5142 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5143 host lookup before query-style lookups in lists that might use this
5144 variable. This has therefore been abolished, and the "net-" prefix is no
5145 longer necessary for query-style lookups.
5147 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5148 have been a typo for LFLAGS, so it has been changed.
5150 13. The install script calls Exim with "-C /dev/null" in order to find the
5151 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5152 to be output. Howeve, since Exim outputs its version number before the
5153 error, it didn't break the script. It just looked ugly. I fixed this by
5154 always allowing "-C /dev/null" if the caller is root.
5156 14. Ignore overlarge ACL variable number when reading spool file - insurance
5157 against a later release with more variables having written the file.
5159 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5160 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5161 this, as well as the form without the "IPv6" on the front.
5163 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5164 OS/Makefile-Darwin file.
5166 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5167 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5169 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5170 variable. However, some OS use ints for inodes. I've added cast to long int
5171 to get rid of the compiler warning.
5173 19. I had forgotten to lock out "/../" in configuration file names when
5174 ALT_CONFIG_PREFIX was set.
5176 20. Routers used for verification do not need to specify transports. However,
5177 if such a router generated a host list, and callout was configured, Exim
5178 crashed, because it could not find a port number from the (non-existent)
5179 transport. It now assumes port 25 in this circumstance.
5181 21. Added the -t option to exigrep.
5183 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5184 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5185 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5186 them is set, LOOKUP_LSEARCH is forced.
5188 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5190 24. Added sender and host information to the "rejected by local_scan()" log
5191 line; previously there was no indication of these.
5193 25. Added .include_if_exists.
5195 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5196 Linux. It turns out that not all file systems support this. Apparently some
5197 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5198 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5199 not supported on the file descriptor, is now ignored when Exim is trying to
5200 sync a directory. This applies only to Linux.
5202 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5204 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5205 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5208 29. Added support for maildirsize files from supplied patch (modified a bit).
5210 30. The use of :fail: followed by an empty string could lead Exim to respond to
5211 sender verification failures with (e.g.):
5213 550 Verification failed for <xxx>
5214 550 Sender verify failed
5216 where the first response line was missing the '-' that indicates it is not
5217 the final line of the response.
5219 31. The loop for finding the name of the user that called Exim had a hardwired
5220 limit of 10; it now uses the value of finduser_retries, which is used for
5221 all other user lookups.
5223 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5226 33. Exim was neglecting to zero errno before one call of strtol() when
5227 expanding a string and expecting an integer value. On some systems this
5228 resulted in spurious "integer overflow" errors. Also, it was casting the
5229 result into an int without checking.
5231 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5232 did not work. The code looks as if it has *never* worked, though it appears
5233 to have been documented since at least releast 1.62. I have made it work.
5235 35. The "timeout_DNS" error in retry rules, also documented since at least
5236 1.62, also never worked. As it isn't clear exactly what this means, and
5237 clearly it isn't a major issue, I have abolished the feature by treating it
5238 as "timeout", and writing a warning to the main and panic logs.
5240 36. The display of retry rules for -brt wasn't always showing the error code
5243 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5244 timeout_connect_A, timeout_connect_MX.
5246 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5247 to the empty sender.
5249 39. The daemon was not analysing the content of -oX till after it had closed
5250 stderr and disconnected from the controlling terminal. This meant that any
5251 syntax errors were only noted on the panic log, and the return code from
5252 the command was 0. By re-arranging the code a little, I've made the
5253 decoding happen first, so such errors now appear on stderr, and the return
5254 code is 1. However, the actual setting up of the sockets still happens in
5255 the disconnected process, so errors there are still only recorded on the
5258 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5259 connections (as happens on some IP stacks) was logged at start up time as
5260 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5261 it from "IPv6 and IPv4", which means that two separate sockets are being
5264 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5265 says whether AF_INET or AF_INET6 was passed as an argument.
5267 42. Exiwhat output was messed up when time zones were included in log
5270 43. Exiwhat now gives more information about the daemon's listening ports,
5271 and whether -tls-on-connect was used.
5273 44. The "port" option of the smtp transport is now expanded.
5275 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5276 silently ignored. Now an error message is written to the main and panic
5279 46. There's a new ACL modifier called "logwrite" which writes to a log file
5280 as soon as it is encountered.
5282 47. Added $local_user_uid and $local_user_gid at routing time.
5284 48. Exim crashed when trying to verify a sender address that was being
5287 49. Exim was recognizing only a space character after ".include". It now also
5288 recognizes a tab character.
5290 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5291 extracting the relevant information from the specification. The man page no
5292 longer contains scrambled data for the -d option, and I've added a section
5293 at the front about calling Exim under different names.
5295 51. Added "extra_headers" argument to the "mail" command in filter files.
5297 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5300 53. Installed eximstats 1.29.
5302 54. Added transport_filter_timeout as a generic transport option.
5304 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5305 Cc: header lines. This was required by RFC 822, but it not required by RFC
5308 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5309 incoming messages that did not have them. Now it does so only if the
5310 message originates locally, that is, if there is no associated remote host
5311 address. When Resent- header lines are present, this applies to the Resent-
5312 lines rather than the non-Resent- lines.
5314 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5315 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5317 58. Messages for configuration errors now include the name of the main
5318 configuration file - useful now that there may be more than one file in a
5319 list (.included file names were always shown).
5321 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5322 for those rare installations that do not start the daemon as root or run it
5323 setuid root. I've cut out the call to initgroups() if the daemon is not
5326 60. The Exim user and group can now be bound into the binary as text strings
5327 that are looked up at the start of Exim's processing.
5329 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5331 62. Added $mailstore_basename variable.
5333 63. Installed patch to sieve.c from Michael Haardt.
5335 64. When Exim failed to open the panic log after failing to open the main log,
5336 the original message it was trying to log was written to stderr and debug
5337 output, but if they were not available (the usual case in production), it
5338 was lost. Now it is written to syslog before the two lines that record the
5339 failures to open the logs.
5341 65. Users' Exim filters run in subprocesses under the user's uid. It is
5342 possible for a "deliver" command or an alias in a "personal" command to
5343 provoke an address rewrite. If logging of address rewriting is configured,
5344 this fails because the process is not running as root or exim. There may be
5345 a better way of dealing with this, but for the moment (because 4.30 needs
5346 to be released), I have disabled address rewrite logging when running a
5347 filter in a non-root, non-exim process.
5353 1. The buildconfig auxiliary program wasn't quoting the value set for
5354 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5355 not defined. This bug was masked in 4.22 by the effect that was fixed in
5358 2. Some messages that were rejected after a message id was allocated were
5359 shown as "incomplete" by exigrep. It no longer does this for messages that
5360 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5362 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5363 have allow_domain_literals set, the ID did not get logged in the <= line.
5364 Domain literals are now always recognized in Message-ID: header lines.
5366 4. The first argument for a ${extract expansion item is the key name or field
5367 number. Leading and trailing spaces in this item were not being ignored,
5368 causing some misleading effects.
5370 5. When deliver_drop_privilege was set, single queue runner processes started
5371 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5372 same command in the process it spins off) were not dropping privilege.
5374 6. When the daemon running as "exim" started a queue runner, it always
5375 re-executed Exim in the spun-off process. This is a waste of effort when
5376 deliver_drop_privilege is set. The new process now just calls the
5377 queue-runner function directly.
5383 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5386 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5387 ignored. Though the use of -oP was forcing the writing of a pid file, it
5388 was always written to the default place.
5390 3. If the message "no IP address found for host xxxx" is generated during
5391 incoming verification, it is now followed by identification of the incoming
5392 connection (so you can more easily find what provoked it).
5394 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5396 5. Added some features to "harden" Exim a bit more against certain attacks:
5398 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5399 be put in Local/Makefile. This is like the never_users runtime option,
5400 but it cannot be overridden. The default setting is "root".
5402 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5403 prefix string with which any file named in a -C command line option
5406 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5407 is retained for -C and -D only if the caller of Exim is root. Without
5408 it, the exim user may also use -C and -D and retain privilege.
5410 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5411 command line option is disabled.
5413 6. Macro names set by the -D option must start with an upper case letter, just
5414 like macro names defined in the configuration file.
5416 7. Added "dereference=" facility to LDAP.
5418 8. Two instances of the typo "uknown" in the source files are fixed.
5420 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5421 the Configure-Makefile script screwed up while processing it.
5423 10. Incorporated PCRE 4.4.
5425 11. The SMTP synchronization check was not operating right at the start of an
5426 SMTP session. For example, it could not catch a HELO sent before the client
5427 waited for the greeting. There is now a check for outstanding input at the
5428 point when the greeting is written. Because of the duplex, asynchronous
5429 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5430 way, but not yet received, when the check is performed.
5432 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5433 on TCP/IP sockets, because this apparently causes some broken clients to
5436 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5437 unchanged) from the Cygwin maintainer.
5439 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5440 was supported instead of testing for mbx. Effectively a typo.
5442 15. The spa authenticator server code was not checking that the input it
5443 received was valid base64.
5445 16. The debug output line for the "set" modifier in ACLs was not showing the
5446 name of the variable that was being set.
5448 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5450 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5451 lookup on its own. Something else was needed to trigger the lookup. For
5452 example, a match in host_lookup or the need for a host name in a host list.
5453 Now, if $sender_host_name is referenced and the host name has not yet been
5454 looked up, a lookup is performed. If the lookup fails, the variable remains
5455 empty, and $host_lookup_failed is set to "1".
5457 19. Added "eqi" as a case-independent comparison operator.
5459 20. The saslauthd authentication condition could segfault if neither service
5460 nor realm was specified.
5462 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5463 error message that was logged was misleading, and incoming SMTP
5464 connections were dropped. The message is now more accurate, and temporary
5465 errors are given to SMTP connections.
5467 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5468 (except RSET) with a 421 error, until QUIT is received. However, it was
5469 failing to send a response to QUIT.
5471 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5472 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5473 ACL is now run after the helo_try_verify_hosts code.
5475 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5476 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5477 case-independent, so other case variants are also recognized). Apparently
5478 some systems use these upper case variants.
5480 25. If more than two messages were waiting for the same host, and a transport
5481 filter was specified for the transport, Exim sent two messages over the
5482 same TCP/IP connection, and then failed with "socket operation on non-
5483 socket" when it tried to send the third.
5485 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5487 27. The extern definition of crypt16() in expand.c was not being excluded when
5488 the OS had its own crypt16() function.
5490 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5491 as a preferred synonym for return_size_limit, both as an option and as an
5494 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5496 30. Changed the default configuration ACL to relax the local part checking rule
5497 for addresses that are not in any local domains. For these addresses,
5498 slashes and pipe symbols are allowed within local parts, but the sequence
5499 /../ is explicitly forbidden.
5501 31. SPA server authentication was not clearing the challenge buffer before
5504 32. log_message in a "warn" ACL statement was writing to the reject log as
5505 well as to the main log, which contradicts the documentation and doesn't
5506 seem right (because no rejection is happening). So I have stopped it.
5508 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5509 However, I am unable to do any testing of this.
5511 34. Fixed an infelicity in the appendfile transport. When checking directories
5512 for a mailbox, to see if any needed to be created, it was accidentally
5513 using path names with one or more superfluous leading slashes; tracing
5514 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5516 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5517 contents are added to the log line that is written for every discarded
5518 recipient. (Previously a log_message setting was ignored.)
5520 36. The ${quote: operator now quotes the string if it is empty.
5522 37. The install script runs exim in order to find its version number. If for
5523 some reason other than non-existence or emptiness, which it checks, it
5524 could not run './exim', it was installing it with an empty version number,
5525 i.e. as "exim-". This error state is now caught, and the installation is
5528 38. An argument was missing from the function that creates an error message
5529 when Exim fails to connect to the socket for saslauthd authentication.
5530 This could cause Exim to crash, or give a corrupted message.
5532 39. Added isip, isip4, and isip6 to ${if conditions.
5534 40. The ACL variables $acl_xx are now saved with the message, and can be
5535 accessed later in routers, transports, and filters.
5537 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5538 strings in the file are not string-expanded.
5540 42. If a MAIL command specified a SIZE value that was too large to fit into an
5541 int variable, the check against message_size_limit failed. Such values are
5542 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5543 day this will have to be increased, but I don't think I want to be around
5544 when emails are that large.
5551 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5552 iconv() is not standard in FreeBSD.
5554 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5555 IPv6 enabled. The observed symptom was a segmentation fault on return from
5556 the function os_common_find_running_interfaces() in src/os.c.
5558 3. In the check_special_case() function in daemon.c I had used "errno" as an
5559 argument name, which causes warnings on some systems. This was basically a
5560 typo, since it was named "eno" in the comments!
5562 4. The code that waits for the clock to tick (at a resolution of some fraction
5563 of a second) so as to ensure message-id uniqueness was always waiting for
5564 at least one whole tick, when it could have waited for less. [This is
5565 almost certainly not relevant at current processor speeds, where it is
5566 unlikely to ever wait at all. But we try to future-proof.]
5568 5. The function that sleeps for a time interval that includes fractions of a
5569 second contained a race. It did not block SIGALRM between setting the
5570 timer, and suspending (a couple of lines later). If the interval was short
5571 and the sigsuspend() was delayed until after it had expired, the suspension
5572 never ended. On busy systems this could lead to processes getting stuck for
5575 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5576 process, before it forks any delivery processes. The open lookup caching
5577 mechanism meant that the open file or database connection was passed into
5578 the delivery process. The problem was that delivery processes always tidy
5579 up cached lookup data. This could cause a problem for the next delivery
5580 process started by the queue runner, because the external queue runner
5581 process does not know about the closure. So the next delivery process
5582 still has data in the lookup cache. In the case of a file lookup, there was
5583 no problem because closing a file descriptor in a subprocess doesn't affect
5584 the parent. However, if the lookup was caching a connection to a database,
5585 the connection was closed, and the second delivery process was likely to
5586 see errors such as "PGSQL: query failed: server closed the connection
5587 unexpectedly". The problem has been fixed by closing all cached lookups
5588 in a queue runner before running a delivery process.
5590 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5591 seem to have the "const" qualifier which it has on other OS. I've
5594 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5595 *of the same type* (client or server) with the same public name that an
5596 error should be diagnosed.
5598 9. When Exim looked up a host name for an IP address, but failed to find the
5599 original IP address when looking up the host name (a safety check), it
5600 output the message "<ip address> does not match any IP for NULL", which was
5601 confusing, to say the least. The bug was that the host name should have
5602 appeared instead of "NULL".
5604 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5605 user that is built into the binary, and the -C or -D options is used, root
5606 privilege is dropped before the configuration file is read. In addition,
5607 logging is switched to stderr instead of the normal log files. If the
5608 configuration then re-defines the Exim user, the unprivileged environment
5609 is probably not what is expected, so Exim logs a panic warning message (but
5612 However, if deliver_drop_privilege is set, the unprivileged state may well
5613 be exactly what is intended, so the warning has been cut out in that case,
5614 and Exim is allowed to try to write to its normal log files.
5620 1. smtp_return_error_details was not giving details for temporary sender
5621 or receiver verification errors.
5623 2. Diagnose a configuration error if two authenticators have the same public
5626 3. Exim used not to create the message log file for a message until the first
5627 delivery attempt. This could be confusing when incoming messages were held
5628 for policy or load reasons. The message log file is now created at the time
5629 the message is received, and an initial "Received" line is written to it.
5631 4. The automatically generated man page for command line options had a minor
5632 bug that caused no ill effects; however, a more serious problem was that
5633 the procedure for building the man page automatically didn't always
5634 operate. Consequently, release 4.20 contains an out-of-date version. This
5635 shouldn't happen again.
5637 5. When building Exim with embedded Perl support, the script that builds the
5638 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5639 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5641 6. The freeze_tell option was not being used for messages that were frozen on
5642 arrival, either by an ACL or by local_scan().
5644 7. Added the smtp_incomplete_transaction log selector.
5646 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5647 was accepting AUTH without a new EHLO.
5649 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5650 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5653 10. Logging of TCP/IP connections (when configured) now happens in the main
5654 daemon process instead of the child process, so that the TCP/IP connection
5655 count is more accurate (but it can never be perfect).
5657 11. The use of "drop" in a nested ACL was not being handled correctly in the
5658 outer ACL. Now, if condition failure induced by the nested "drop" causes
5659 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5660 or "require"), the connection is dropped.
5662 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5663 that yield "discard" can now be used with an "accept" or a "discard" verb,
5664 but an error is generated for any others (because I can't see a useful way
5665 to define what should happen).
5667 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5668 are now processed in the same way as lines in the Exim configuration file.
5669 In particular, continuation lines are supported.
5671 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5673 15. Added -ti meaning -t -i.
5675 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5676 domains, and warn by logging if others are found.
5678 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5679 returned from GIFCONF when Exim is trying to find the list of interfaces on
5680 a host. The code in os.c has been modified to copy each ifreq to an aligned
5681 structure in all cases.
5683 Also, in some cases, the returned ifreq's were being copied to a 'struct
5684 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5685 means the last couple of bytes of an IPv6 address could be chopped if the
5686 ifreq contained only a normal sockaddr (14 bytes storage).
5688 18. Named domain lists were not supported in the hosts_treat_as_local option.
5689 An entry such as +xxxx was not recognized, and was treated as a literal
5692 19. Ensure that header lines added by a DATA ACL are included in the reject log
5693 if the ACL subsequently rejects the message.
5695 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5696 MD5 (which is deprecated).
5698 21. When testing a filter file using -bf, Exim was writing a message when it
5699 took the sender from a "From " line in the message, but it was not doing so
5700 when it took $return_path from a Return-Path: header line. It now does.
5702 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5703 with a valid header line field name (a series of printing characters
5704 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5706 23. Changed "disc" in the source to "disk" to conform to the documentation and
5707 the book and for uniformity.
5709 24. Ignore Sendmail's -Ooption=value command line item.
5711 25. When execve() failed while trying to run a command in a pipe transport,
5712 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5713 could be confused with a return value of 69 from the command itself. This
5714 has been changed to 127, the value the shell returns if it is asked to run
5715 a non-existent command. The wording for the related log line suggests a
5716 non-existent command as the problem.
5718 26. If received_header_text expands to an empty string, do not add a Received:
5719 header line to the message. (Well, it adds a token one on the spool, but
5720 marks it "old" so that it doesn't get used or transmitted.)
5722 27. Installed eximstats 1.28 (addition of -nt option).
5724 28. There was no check for failure on the call to getsockname() in the daemon
5725 code. This can fail if there is a shortage of resources on the system, with
5726 ENOMEM, for example. A temporary error is now given on failure.
5728 29. Contrary to the C standard, it seems that in some environments, the
5729 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5730 program. Exim now does this explicitly; it affects the formatting of
5731 timestamps using strftime().
5733 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5734 complaints. I've now initialized all the variables, to avoid this.
5736 32. Header lines added by a system filter were not being "seen" during
5737 transport-time rewrites.
5739 33. The info_callback() function passed to OpenSSL is set up with type void
5740 (*)(SSL *, int, int), as described somewhere. However, when calling the
5741 function (actually a macro) that sets it up, the type void(*)() is
5742 expected. I've put in a cast to prevent warnings from picky compilers.
5744 34. If a DNS black list lookup found a CNAME record, but there were no A
5745 records associated with the domain it pointed at, Exim crashed.
5747 35. If a DNS black list lookup returned more than one A record, Exim ignored
5748 all but the first. It now scans all returned addresses if a particular IP
5749 value is being sought. In this situation, the contents of the
5750 $dnslist_value variable are a list of all the addresses, separated by a
5753 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5754 to accept a host name and all its aliases if the forward lookup for any of
5755 them yielded the IP address of the incoming connection. Now it accepts only
5756 those names whose forward lookup yields the correct IP address. Any other
5757 names are discarded. This closes a loophole whereby a rogue DNS
5758 administrator could create reverse DNS records to break through a
5759 wildcarded host restriction in an ACL.
5761 37. If a user filter or a system filter that ran in a subprocess used any of
5762 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5763 the wrong values were passed to the pipe command ($thisaddress had the
5764 value of $0, $0 had the value of $1, etc). This bug was introduced by
5765 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5767 38. Improved the line breaking for long SMTP error messages from ACLs.
5768 Previously, if there was no break point between 40 and 75 characters, Exim
5769 left the rest of the message alone. Two changes have been made: (a) I've
5770 reduced the minimum length to 35 characters; (b) if it can't find a break
5771 point between 35 and 75 characters, it looks ahead and uses the first one
5772 that it finds. This may give the occasional overlong line, but at least the
5773 remaining text gets split now.
5775 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5776 file descriptors might be low, and that setting 1000 would always raise it.
5777 It turns out that in some environments, the limit is already over 1000 and
5778 that lowering it causes trouble. So now Exim takes care not to decrease it.
5780 40. When delivering a message, the value of $return_path is set to $sender_
5781 address at the start of routing (routers may change the value). By an
5782 oversight, this default was not being set up when an address was tested by
5783 -bt or -bv, which affected the outcome if any router or filter referred to
5786 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5787 log only when "message" or "log_message" are set. However, if one of the
5788 conditions was an address verification, or a call to a nested ACL, the
5789 messages generated by the underlying test were being passed through. This
5790 no longer happens. The underlying message is available in $acl_verify_
5791 message for both "message" and "log_message" expansions, so it can be
5792 passed through if needed.
5794 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5795 new expansion $bh_ to give the encoded byte string without charset
5796 translation. Translation happens only if iconv() is available; HAVE_ICONV
5797 indicates this at build time. HEADERS_CHARSET gives the charset to
5798 translate to; headers_charset can change it in the configuration, and
5799 "headers charset" can change it in an individual filter file.
5801 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5802 that creates RFC 2047 encoded "words" labels them as that charset instead
5803 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5804 expansion operator; (ii) when Exim creates a From: line for a local
5805 message; (iii) when a header line is rewritten to include a "phrase" part.
5807 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5808 buggy, causing it to skip the first lines of messages whose message ID
5809 ended in 'D'. This would not have bitten before Exim release 4.14, because
5810 message IDs were unlikely to end in 'D' before then. The effect was to have
5811 incorrect size information for certain domains.
5813 45. #include "config.h" was missing at the start of the crypt16.c module. This
5814 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5817 46. If there was a timeout during a "random" callout check, Exim treated it as
5818 a failure of the random address, and carried on sending RSET and the real
5819 address. If the delay was just some slowness somewhere, the response to the
5820 original RCPT would be taken as a response to RSET and so on, causing
5821 mayhem of various kinds.
5823 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5824 when I implemented it. It didn't allow for the fact that some option values
5825 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5826 the right test for positive values.
5828 48. Domain names in DNS records are case-independent. Exim always looks them up
5829 in lower case. Some resolvers return domain names in exactly the case they
5830 appear in the zone file, that is, they may contain uppercase letters. Not
5831 all resolvers do this - some return always lower case. Exim was treating a
5832 change of case by a resolver as a change of domain, similar to a widening
5833 of a domain abbreviation. This triggered its re-routing code and so it was
5834 trying to route what was effectively the same domain again. This normally
5835 caused routing to fail (because the router wouldn't handle the domain
5836 twice). Now Exim checks for this case specially, and just changes the
5837 casing of the domain that it ultimately uses when it transmits the message
5840 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5843 50. If a filter generated a file delivery with a non-absolute name (possible if
5844 no home directory exists for the router), the forbid_file option was not
5847 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5848 the existing equality matching.
5850 52. Exim was using ints instead of ino_t variables in some places where it was
5851 dealing with inode numbers.
5853 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5854 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5855 called TMPDIR, and if it finds it is different, it changes its value.
5857 54. The smtp_printf() function is now made available to local_scan() so
5858 additional output lines can be written before returning. There is also an
5859 smtp_fflush() function to enable the detection of a dropped connection.
5860 The variables smtp_input and smtp_batched_input are exported to
5863 55. Changed the default runtime configuration: the message "Unknown user"
5864 has been removed from the ACL, and instead placed on the localuser router,
5865 using the cannot_route_message feature. This means that any verification
5866 failures that generate their own messages won't get overridden. Similarly,
5867 the "Unrouteable address" message that was in the ACL for unverifiable
5868 relay addresses has also been removed.
5870 56. Added hosts_avoid_esmtp to the smtp transport.
5872 57. The exicyclog script was not checking for the esoteric option
5873 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5874 will work only if exicyclog is run under the appropriate euid.
5876 58. Following a discussion on the list, the rules by which Exim recognises line
5877 endings on incoming messages have been changed. The -dropcr and drop_cr
5878 options are now no-ops, retained only for backwards compatibility. The
5879 following line terminators are recognized: LF CRLF CR. However, special
5880 processing applies to CR:
5882 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5883 nor a local message in the state where . is a terminator.
5885 (ii) If a bare CR is encountered in a header line, an extra space is added
5886 after the line terminator so as not to end the header. The reasoning
5887 behind this is that bare CRs in header lines are most likely either
5888 to be mistakes, or people trying to play silly games.
5890 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5891 was being incorrectly given as 18 bytes larger than it should have been.
5892 This is a VOB (very old bug).
5894 60. This may never have affected anything current, but just in case it has:
5895 When the local host is found other than at the start of a list of hosts,
5896 the local host, those with the same MX, and any that follow, are discarded.
5897 When the list in question was part of a longer list of hosts, the following
5898 hosts (not currently being processed) were also being discarded. This no
5899 longer happens. I'm not sure if this situation could ever has previously
5902 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5905 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5906 header. This is something that is recommended in a new Internet Draft, and
5907 is something that is documented as being done by Sendmail. There are two
5908 possible values. For messages generated by the autoreply transport, Exim
5911 Auto-Submitted: auto-replied
5913 whereas for all other generated messages (e.g. bounces) it adds
5915 Auto-Submitted: auto-generated
5917 63. The "personal" condition in filters now includes a test for the
5918 Auto-Submitted: header. If it contains the string "auto-" the message it
5919 not considered personal.
5921 64. Added rcpt_include_affixes as a generic transport option.
5923 65. Added queue_only_override (default true).
5925 66. Added the syslog_duplication option.
5927 67. If what should have been the first header line of a message consisted of
5928 a space followed by a colon, Exim was mis-interpreting it as a header line.
5929 It isn't of course - it is syntactically invalid and should therefore be
5930 treated as the start of the message body. The misbehaviour could have
5931 caused a number of strange effects, including loss of data in subsequent
5932 header lines, and spool format errors.
5934 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5935 client host had authenticated. This control can now be exercised by an ACL
5936 for more flexibility.
5938 69. By default, callouts do not happen when testing with -bh. There is now a
5939 variant, -bhc, which does actually run the callout code, including
5940 consulting and updating the callout cache.
5942 70. Added support for saslauthd authentication, courtesy of Alexander
5945 71. If statvfs() failed on the spool or log directories while checking their
5946 size for availability, Exim confusingly gave the error "space shortage".
5947 Furthermore, in debugging mode it crashed with a floating point exception.
5948 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5949 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5950 serious problem, Exim now writes to the main and panic logs when this
5951 happens, with details of the failure. It then refuses to accept the
5952 incoming message, giving the message "spool directory problem" or "log
5953 directory problem" with a 421 code for SMTP messages.
5955 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5956 0, 1, and 2 exist, because some OS complain for execs without them (see
5957 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5958 descriptors. However, the code omitted to check that the open succeeded,
5959 causing mysterious errors if for some reason the permissions on /dev/null
5960 got screwed. Now Exim writes a message to the main and panic logs, and
5961 bombs out if it can't open /dev/null.
5963 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5964 interact so that it is all more flexible. It is supposed to remain
5965 backwards compatible. Also added extra_local_interfaces.
5967 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5968 to bomb out with an assertion failure - to the client this appears as a
5969 connection drop. This problem occurs in the part of the code that was taken
5970 from the Samba project. Fortunately, the assertion is in a very simple
5971 function, so I have fixed this by reproducing the function inline in the
5972 one place where it is called, and arranging for authentication to fail
5973 instead of killing the process with assert().
5975 75. The SPA client code was not working when the server requested OEM rather
5976 than Unicode encoding.
5978 76. Added code to make require_files with a specific uid setting more usable in
5979 the case where statting the file as root fails - usually a non-root-mounted
5980 NFS file system. When this happens and the failure is EACCES, Exim now
5981 forks a subprocess and does the per-uid checking as the relevant uid.
5983 77. Added process_log_path.
5985 78. If log_file_path was not explicitly set, a setting of check_log_space or
5986 check_log_inodes was ignored.
5988 79. If a space check for the spool or log partitions fails, the incident is now
5989 logged. Of course, in the latter case the data may get lost...
5991 80. Added the %p formatting code to string_format() so that it can be used to
5992 print addresses in debug_print(). Adjusted all the address printing in the
5993 debugging in store.c to use %p rather than %d.
5995 81. There was a concern that a line of code in smtp_in.c could overflow a
5996 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5997 initially expressed, the concern was not well-founded, because trailing
5998 spaces are removed early. However, if the trailing spaces were followed by
5999 a NULL, they did not get removed, so the overflow was possible. Two fixes
6002 (a) I re-wrote the offending code in a cleaner fashion.
6003 (b) If an incoming SMTP command contains a NULL character, it is rejected
6006 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6007 runs initgroups(), so that if the Exim user is in any additional groups,
6008 they will be used during message reception.
6014 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob