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 and the smtp transport (applying to the forward
91 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
92 of ldap servers used for a specific lookup. Patch provided by Heiko
95 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 New variable $lookup_dnssec_authenticated for observability.
98 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
99 Patch submitted by Lars Timman.
101 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
103 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
104 Requires trusted mode and valid format message id, aborts otherwise.
105 Patch contributed by Heiko Schlichting.
107 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
108 certextract with support for various fields. Bug 1358.
114 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
116 PP/02 Make -n do something, by making it not do something.
117 When combined with -bP, the name of an option is not output.
119 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
122 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
123 $sender_host_name and config options to manage this, and basic check
126 PP/05 DSCP support for outbound connections and control modifier for inbound.
128 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
129 (Only plugin which currently uses this is kerberos4, which nobody should
130 be using, but we should make it available and other future plugins might
131 conceivably use it, even though it would break NAT; stuff *should* be
132 using channel bindings instead).
134 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
135 name; added for Sendmail compatibility; requires admin caller.
136 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
137 just ignore it); requires trusted caller.
138 Also parse but ignore: -Ac -Am -X<logfile>
141 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
143 TL/02 Add +smtp_confirmation as a default logging option.
145 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
146 Patch by Magnus Holmgren from 2007-02-20.
148 TL/04 Bugzilla 1281 - Spec typo.
149 Bugzilla 1283 - Spec typo.
150 Bugzilla 1290 - Spec grammar fixes.
152 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
154 TL/06 Add Experimental DMARC support using libopendmarc libraries.
156 TL/07 Fix an out of order global option causing a segfault. Reported to dev
157 mailing list by by Dmitry Isaikin.
159 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
161 JH/02 Support "G" suffix to numbers in ${if comparisons.
163 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
165 NM/01 Bugzilla 1197 - Spec typo
166 Bugzilla 1196 - Spec examples corrections
168 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
170 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
171 gnutls_enable_pkcs11, but renamed to more accurately indicate its
174 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
175 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
177 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
178 "acl {{name}{arg}...}", and optional args on acl condition
181 JH/05 Permit multiple router/transport headers_add/remove lines.
183 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
185 JH/07 Avoid using a waiting database for a single-message-only transport.
186 Performance patch from Paul Fisher. Bugzilla 1262.
188 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
191 JH/09 Add $headers_added variable, with content from use of ACL modifier
192 add_header (but not yet added to the message). Bugzilla 199.
194 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
195 Pulled from Bugzilla 817 by Wolfgang Breyha.
197 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
199 (nb: this is the same fix as in Exim 4.80.1)
201 JH/11 Add A= logging on delivery lines, and a client_set_id option on
204 JH/12 Add optional authenticated_sender logging to A= and a log_selector
207 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
209 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
210 advertise SMTP AUTH mechanism to us, instead of a generic
211 protocol violation error. Also, make Exim more robust to bad
212 data from the Dovecot auth socket.
214 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
216 When a queue runner is handling a message, Exim first routes the
217 recipient addresses, during which it prunes them based on the retry
218 hints database. After that it attempts to deliver the message to
219 any remaining recipients. It then updates the hints database using
222 So if a recipient address works intermittently, it can get repeatedly
223 deferred at routing time. The retry hints record remains fresh so the
224 address never reaches the final cutoff time.
226 This is a fairly common occurrence when a user is bumping up against
227 their storage quota. Exim had some logic in its local delivery code
228 to deal with this. However it did not apply to per-recipient defers
229 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
231 This change adds a proper retry rule check during routing so that the
232 final cutoff time is checked against the message's age. We only do
233 this check if there is an address retry record and there is not a
234 domain retry record; this implies that previous attempts to handle
235 the address had the retry_use_local_parts option turned on. We use
236 this as an approximation for the destination being like a local
237 delivery, as in LMTP.
239 I suspect this new check makes the old local delivery cutoff check
240 redundant, but I have not verified this so I left the code in place.
242 TF/02 Correct gecos expansion when From: is a prefix of the username.
244 Test 0254 submits a message to Exim with the header
248 When I ran the test suite under the user fanf2, Exim expanded
249 the header to contain my full name, whereas it should have added
250 a Resent-Sender: header. It erroneously treats any prefix of the
251 username as equal to the username.
253 This change corrects that bug.
255 GF/01 DCC debug and logging tidyup
256 Error conditions log to paniclog rather than rejectlog.
257 Debug lines prefixed by "DCC: " to remove any ambiguity.
259 TF/03 Avoid unnecessary rebuilds of lookup-related code.
261 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
262 Bug spotted by Jeremy Harris; was flawed since initial commit.
263 Would have resulted in OCSP responses post-SNI triggering an Exim
264 NULL dereference and crash.
266 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
268 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
269 Bug detection, analysis and fix by Samuel Thibault.
270 Bugzilla 1331, Debian bug #698092.
272 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
274 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
275 Server implementation by Todd Lyons, client by JH.
276 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
277 config variable "prdr_enable" controls whether the server
278 advertises the facility. If the client requests PRDR a new
279 acl_data_smtp_prdr ACL is called once for each recipient, after
280 the body content is received and before the acl_smtp_data ACL.
281 The client is controlled by bolth of: a hosts_try_prdr option
282 on the smtp transport, and the server advertisement.
283 Default client logging of deliveries and rejections involving
284 PRDR are flagged with the string "PRDR".
286 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
287 fclose(). Diagnosis by Todd Lyons.
289 PP/17 Update configure.default to handle IPv6 localhost better.
290 Patch by Alain Williams (plus minor tweaks).
293 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
294 This is now consistent with GnuTLS, and is now documented: the
295 previous undocumented portable approach to treating the option as
296 unset was to force an expansion failure. That still works, and
297 an empty string is now equivalent.
299 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
300 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
301 not performing validation itself.
303 PP/20 Added force_command boolean option to pipe transport.
304 Patch from Nick Koston, of cPanel Inc.
306 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
309 TF/04 Added udpsend ACL modifer and hexquote expansion operator
311 PP/21 Fix eximon continuous updating with timestamped log-files.
312 Broken in a format-string cleanup in 4.80, missed when I repaired the
313 other false fix of the same issue.
314 Report and fix from Heiko Schlichting.
317 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
318 Report from Prashanth Katuri.
320 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
321 It's SecureTransport, so affects any MacOS clients which use the
322 system-integrated TLS libraries, including email clients.
324 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
325 using a MIME ACL for non-SMTP local injection.
326 Report and assistance in diagnosis by Warren Baker.
328 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
330 JH/16 Fix comparisons for 64b. Bugzilla 1385.
332 TL/09 Add expansion variable $authenticated_fail_id to keep track of
333 last id that failed so it may be referenced in subsequent ACL's.
335 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
338 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
339 ldap library initialization, allowing self-signed CA's to be
340 used. Also properly sets require_cert option later in code by
341 using NULL (global ldap config) instead of ldap handle (per
342 session). Bug diagnosis and testing by alxgomz.
344 TL/12 Enhanced documentation in the ratelimit.pl script provided in
345 the src/util/ subdirectory.
347 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
348 renamed to Transport Post Delivery Action by Jeremy Harris, as
351 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
352 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
353 redis_servers = needs to be configured which will be used by the redis
354 lookup. Patch from Warren Baker, of The Packet Hub.
356 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
358 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
359 hostname or reverse DNS when processing a host list. Used suggestions
360 from multiple comments on this bug.
362 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
364 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
365 Missed a few lines, added it to make the runtest require no keyboard
368 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
369 contains upper case chars. Make router use caseful_local_part.
371 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
372 support when GnuTLS has been built with p11-kit.
378 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
380 This, or similar/improved, will also be change PP/11 of 4.82.
386 PP/01 Handle short writes when writing local log-files.
387 In practice, only affects FreeBSD (8 onwards).
388 Bugzilla 1053, with thanks to Dmitry Isaikin.
390 NM/01 Bugzilla 949 - Documentation tweak
392 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
395 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
397 PP/02 Implemented gsasl authenticator.
399 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
401 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
402 `pkg-config foo` for cflags/libs.
404 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
405 with rest of GSASL and with heimdal_gssapi.
407 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
408 `pkg-config foo` for cflags/libs for the TLS implementation.
410 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
411 properties get this fed in as external SSF. A number of robustness
412 and debugging improvements to the cyrus_sasl authenticator.
414 PP/08 cyrus_sasl server now expands the server_realm option.
416 PP/09 Bugzilla 1214 - Log authentication information in reject log.
417 Patch by Jeremy Harris.
419 PP/10 Added dbmjz lookup type.
421 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
423 PP/12 MAIL args handles TAB as well as SP, for better interop with
424 non-compliant senders.
425 Analysis and variant patch by Todd Lyons.
427 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
428 Bug report from Lars Müller <lars@samba.org> (via SUSE),
429 Patch from Dirk Mueller <dmueller@suse.com>
431 PP/13 tls_peerdn now print-escaped for spool files.
432 Observed some $tls_peerdn in wild which contained \n, which resulted
433 in spool file corruption.
435 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
436 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
437 or write after TLS renegotiation, which otherwise led to messages
440 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
441 as a tracking header (ie: a signed header comes before the signature).
442 Patch from Wolfgang Breyha.
444 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
445 comma-sep list; embedded commas doubled.
447 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
449 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
451 Report and patch from Dmitry Banschikov.
453 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
454 Removed SSL_clear() after SSL_new() which led to protocol negotiation
455 failures. We appear to now support TLS1.1+ with Exim.
457 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
458 lets Exim select keys and certificates based upon TLS SNI from client.
459 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
460 before an outbound SMTP session. New log_selector, +tls_sni.
462 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
463 NULL dereference. Report and patch from Alun Jones.
465 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
466 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
467 Not seeing resolver debug output on NetBSD, but suspect this is a
468 resolver implementation change.
470 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
471 Left warnings. Added "eximon gdb" invocation mode.
473 PP/21 Defaulting "accept_8bitmime" to true, not false.
475 PP/22 Added -bw for inetd wait mode support.
477 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
478 locate the relevant includes and libraries. Made this the default.
480 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
481 Bugzilla 1246, report and most of solution from Tomasz Kusy.
483 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
484 This may cause build issues on older platforms.
486 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
487 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
488 gnutls_require_mac & gnutls_require_protocols (no longer supported).
489 Added SNI support via GnuTLS too.
490 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
492 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
494 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
495 Applied second patch from Janne, implementing suggestion to default
496 multiple-strings-in-record handling to match SPF spec.
498 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
500 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
501 read-only, out of scope).
502 Patch from Wolfgang Breyha, report from Stuart Northfield.
504 PP/29 Fix three issues highlighted by clang analyser static analysis.
505 Only crash-plausible issue would require the Cambridge-specific
506 iplookup router and a misconfiguration.
507 Report from Marcin Mirosław.
509 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
511 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
512 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
513 As part of this, removing so much warning spew let me fix some minor
514 real issues in debug logging.
516 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
517 assignment on my part. Fixed.
519 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
520 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
521 Janne Snabb (who went above and beyond: thank you).
523 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
524 string otherwise requires a connection and a bunch more work and it's
525 relatively easy to get wrong. Should also expose TLS library linkage
528 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
529 64-bit ${eval} (JH/03).
531 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
532 GNU libc to support some of the 64-bit stuff, should not lead to
533 conflicts. Defined before os.h is pulled in, so if a given platform
534 needs to override this, it can.
536 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
537 protection layer was required, which is not implemented.
538 Bugzilla 1254, patch from Wolfgang Breyha.
540 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
541 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
542 tls_dhparam take prime identifiers. Also unbreak combination of
543 OpenSSL+DH_params+TLSSNI.
545 PP/39 Disable SSLv2 by default in OpenSSL support.
551 PP/01 Solaris build fix for Oracle's LDAP libraries.
552 Bugzilla 1109, patch from Stephen Usher.
554 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
556 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
559 TF/02 Fix a couple more cases where we did not log the error message
560 when unlink() failed. See also change 4.74-TF/03.
562 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
563 lock up or crash if it happened to be inside a call to libc when it
564 got a SIGUSR1 from exiwhat.
566 The SIGUSR1 handler appends the current process status to the process
567 log which is later printed by exiwhat. It used to use the general
568 purpose logging code to do this, but several functions it calls are
569 not safe for signals.
571 The new output code in the SIGUSR1 handler is specific to the process
572 log, and simple enough that it's easy to inspect for signal safety.
573 Removing some special cases also simplifies the general logging code.
574 Removing the spurious timestamps from the process log simplifies
577 TF/04 Improved ratelimit ACL condition.
579 The /noupdate option has been deprecated in favour of /readonly which
580 has clearer semantics. The /leaky, /strict, and /readonly update modes
581 are mutually exclusive. The update mode is no longer included in the
582 database key; it just determines when the database is updated. (This
583 means that when you upgrde Exim will forget old rate measurements.)
585 Exim now checks that the per_* options are used with an update mode that
586 makes sense for the current ACL. For example, when Exim is processing a
587 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
588 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
589 must specify per_mail/readonly. If you omit the update mode it defaults to
590 /leaky where that makes sense (as before) or /readonly where required.
592 The /noupdate option is now undocumented but still supported for
593 backwards compatibility. It is equivalent to /readonly except that in
594 ACLs where /readonly is required you may specify /leaky/noupdate or
595 /strict/noupdate which are treated the same as /readonly.
597 A useful new feature is the /count= option. This is a generalization
598 of the per_byte option, so that you can measure the throughput of other
599 aggregate values. For example, the per_byte option is now equivalent
600 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
602 The per_rcpt option has been generalized using the /count= mechanism
603 (though it's more complicated than the per_byte equivalence). When it is
604 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
605 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
606 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
607 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
608 Note that using per_rcpt with a non-readonly update mode in more than
609 one ACL will cause the recipients to be double-counted. (The per_mail
610 and per_byte options don't have this problem.)
612 The handling of very low rates has changed slightly. If the computed rate
613 is less than the event's count (usually one) then this event is the first
614 after a long gap. In this case the rate is set to the same as this event's
615 count, so that the first message of a spam run is counted properly.
617 The major new feature is a mechanism for counting the rate of unique
618 events. The new per_addr option counts the number of different
619 recipients that someone has sent messages to in the last time period. It
620 behaves like per_rcpt if all the recipient addresses are different, but
621 duplicate recipient addresses do not increase the measured rate. Like
622 the /count= option this is a general mechanism, so the per_addr option
623 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
624 example, measure the rate that a client uses different sender addresses
625 with the options per_mail/unique=$sender_address. There are further
626 details in the main documentation.
628 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
630 TF/06 Removed a few PCRE remnants.
632 TF/07 Automatically extract Exim's version number from tags in the git
633 repository when doing development or release builds.
635 PP/02 Raise smtp_cmd_buffer_size to 16kB.
636 Bugzilla 879. Patch from Paul Fisher.
638 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
639 Heavily based on revision 40f9a89a from Simon Arlott's tree.
642 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
644 PP/05 Variable $av_failed, true if the AV scanner deferred.
645 Bugzilla 1078. Patch from John Horne.
647 PP/06 Stop make process more reliably on build failure.
648 Bugzilla 1087. Patch from Heiko Schlittermann.
650 PP/07 Make maildir_use_size_file an _expandable_ boolean.
651 Bugzilla 1089. Patch from Heiko Schlittermann.
653 PP/08 Handle ${run} returning more data than OS pipe buffer size.
654 Bugzilla 1131. Patch from Holger Weiß.
656 PP/09 Handle IPv6 addresses with SPF.
657 Bugzilla 860. Patch from Wolfgang Breyha.
659 PP/10 GnuTLS: support TLS 1.2 & 1.1.
661 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
664 PP/11 match_* no longer expand right-hand-side by default.
665 New compile-time build option, EXPAND_LISTMATCH_RHS.
666 New expansion conditions, "inlist", "inlisti".
668 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
670 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
672 PP/14 fix log_write() format string regression from TF/03.
673 Bugzilla 1152. Patch from Dmitry Isaikin.
679 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
681 PP/02 Harmonised TLS library version reporting; only show if debugging.
682 Layout now matches that introduced for other libraries in 4.74 PP/03.
684 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
686 PP/04 New "dns_use_edns0" global option.
688 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
691 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
692 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
694 TK/01 Updated PolarSSL code to 0.14.2.
695 Bugzilla 1097. Patch from Andreas Metzler.
697 PP/07 Catch divide-by-zero in ${eval:...}.
700 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
703 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
704 format-string attack -- SECURITY: remote arbitrary code execution.
706 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
707 time unintentionally subject to list matching rules, letting the header
708 cause arbitrary Exim lookups (of items which can occur in lists, *not*
709 arbitrary string expansion). This allowed for information disclosure.
711 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
712 INT_MIN/-1 -- value coerced to INT_MAX.
718 NM/01 Workround for PCRE version dependency in version reporting
721 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
722 This fixes portability to compilers other than gcc, notably
723 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
725 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
726 makefiles for portability to HP-UX and POSIX correctness.
728 PP/01 Permit LOOKUP_foo enabling on the make command-line.
729 Also via indented variable definition in the Makefile.
730 (Debugging by Oliver Heesakkers).
732 PP/02 Restore caching of spamd results with expanded spamd_address.
733 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
735 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
736 Improves build reliability. Fix from: Frank Elsner
738 NM/02 Fix wide character breakage in the rfc2047 coding
739 Fixes bug 1064. Patch from Andrey N. Oktyabrski
741 NM/03 Allow underscore in dnslist lookups
742 Fixes bug 1026. Patch from Graeme Fowler
744 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
745 Code patches from Adam Ciarcinski of NetBSD.
747 NM/04 Fixed exiqgrep to cope with mailq missing size issue
750 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
751 is logged, to avoid truncation. Patch from John Horne.
753 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
754 Patch from Jakob Hirsch.
756 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
757 SQL string expansion failure details.
758 Patch from Andrey Oktyabrski.
760 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
761 Patch from Simon Arlott.
763 PP/09 New lookups functionality failed to compile on old gcc which rejects
764 extern declarations in function scope.
765 Patch from Oliver Fleischmann
767 PP/10 Use sig_atomic_t for flags set from signal handlers.
768 Check getgroups() return and improve debugging.
769 Fixed developed for diagnosis in bug 927 (which turned out to be
772 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
773 Patch from Mark Zealey.
775 PP/12 Bugzilla 1056: Improved spamd server selection.
776 Patch from Mark Zealey.
778 PP/13 Bugzilla 1086: Deal with maildir quota file races.
779 Based on patch from Heiko Schlittermann.
781 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
782 Patch from Uwe Doering, sign-off by Michael Haardt.
784 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
785 variable declaration deep within a block. Bug and patch from
788 PP/15 lookups-Makefile IRIX compatibilty coercion.
790 PP/16 Make DISABLE_DKIM build knob functional.
792 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
793 Patch by Simon Arlott
795 TF/03 Fix valgrind.h portability to C89 compilers that do not support
796 variable argument macros. Our copy now differs from upstream.
802 TF/01 Failure to get a lock on a hints database can have serious
803 consequences so log it to the panic log.
805 TF/02 Log LMTP confirmation messages in the same way as SMTP,
806 controlled using the smtp_confirmation log selector.
808 TF/03 Include the error message when we fail to unlink a spool file.
810 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
811 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
812 for maintaining out-of-tree patches for some time.
814 PP/01 Bugzilla 139: Documentation and portability issues.
815 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
816 Handle per-OS dynamic-module compilation flags.
818 PP/02 Let /dev/null have normal permissions.
819 The 4.73 fixes were a little too stringent and complained about the
820 permissions on /dev/null. Exempt it from some checks.
821 Reported by Andreas M. Kirchwitz.
823 PP/03 Report version information for many libraries, including
824 Exim version information for dynamically loaded libraries. Created
825 version.h, now support a version extension string for distributors
826 who patch heavily. Dynamic module ABI change.
828 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
829 privilege escalation vulnerability whereby the Exim run-time user
830 can cause root to append content of the attacker's choosing to
833 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
836 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
837 If dropping privileges for untrusted macros, we disabled normal logging
838 on the basis that it would fail; for the Exim run-time user, this is not
839 the case, and it resulted in successful deliveries going unlogged.
840 Fixed. Reported by Andreas Metzler.
846 PP/01 Date: & Message-Id: revert to normally being appended to a message,
847 only prepend for the Resent-* case. Fixes regression introduced in
848 Exim 4.70 by NM/22 for Bugzilla 607.
850 PP/02 Include check_rfc2047_length in configure.default because we're seeing
851 increasing numbers of administrators be bitten by this.
853 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
855 PP/03 Bugzilla 994: added openssl_options main configuration option.
857 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
859 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
861 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
863 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
864 without a peer certificate, leading to a segfault because of an
865 assumption that peers always have certificates. Be a little more
866 paranoid. Problem reported by Martin Tscholak.
868 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
869 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
870 NB: ClamAV planning to remove STREAM in "middle of 2010".
871 CL also introduces -bmalware, various -d+acl logging additions and
872 more caution in buffer sizes.
874 PP/09 Implemented reverse_ip expansion operator.
876 PP/10 Bugzilla 937: provide a "debug" ACL control.
878 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
880 PP/12 Bugzilla 973: Implement --version.
882 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
884 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
886 PP/15 Bugzilla 816: support multiple condition rules on Routers.
888 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
889 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
890 ignore trailing whitespace.
892 JJ/02 prevent non-panic DKIM error from being sent to paniclog
894 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
897 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
898 Notification from Dr Andrew Aitchison.
900 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
901 ExtendedDetectionInfo response format.
902 Notification from John Horne.
904 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
907 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
908 XSL and documented dependency on system catalogs, with examples of how
911 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
914 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
915 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
916 configuration file which is writeable by the Exim user or group.
918 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
919 of configuration files to cover files specified with the -C option if
920 they are going to be used with root privileges, not just the default
923 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
924 option (effectively making it always true).
926 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
927 files to be used while preserving root privileges.
929 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
930 that rogue child processes cannot use them.
932 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
933 run-time user, instead of root.
935 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
936 Exim run-time user without dropping privileges.
938 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
939 result string, instead of calling string_vformat() twice with the same
942 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
943 for other users. Others should always drop root privileges if they use
944 -C on the command line, even for a whitelisted configure file.
946 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
948 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
954 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
955 $data_path, and $header_path variables; fixed documentation bugs and
958 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
959 exipick to access non-standard spools, including the "frozen" queue
962 NM/01 Bugzilla 965: Support mysql stored procedures.
963 Patch from Alain Williams
965 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
967 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
968 Patch from Andreas Metzler
970 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
971 Patch from Kirill Miazine
973 NM/05 Bugzilla 671: Added umask to procmail example.
975 JJ/03 installed exipick 20100323.0, fixing doc bug
977 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
978 directory. Notification and patch from Dan Rosenberg.
980 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
982 TK/02 Improve log output when DKIM signing operation fails.
984 MH/01 Treat the transport option dkim_domain as a colon separated
985 list, not as a single string, and sign the message with each element,
986 omitting multiple occurences of the same signer.
988 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
989 Bugzilla 985, 986. Patch by Simon Arlott
991 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
992 Patch by Simon Arlott
994 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
995 MBX locking. Notification from Dan Rosenberg.
1001 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1003 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1005 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1007 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1009 NM/04 Bugzilla 915: Flag broken perl installation during build.
1015 TK/01 Added patch by Johannes Berg that expands the main option
1016 "spamd_address" if it starts with a dollar sign.
1018 TK/02 Write list of recipients to X-Envelope-Sender header when building
1019 the mbox-format spool file for content scanning (suggested by Jakob
1022 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1023 (http://www.dcc-servers.net/) support via dccifd. Activated by
1024 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1026 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1027 by Mark Daniel Reidel <mr@df.eu>.
1029 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1030 When building exim an external PCRE library is now needed -
1031 PCRE is a system library on the majority of modern systems.
1032 See entry on PCRE_LIBS in EDITME file.
1034 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1035 conversation. Added nologin parameter to request.
1036 Patch contributed by Kirill Miazine.
1038 TF/01 Do not log submission mode rewrites if they do not change the address.
1040 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1042 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1043 log files in place. Contributed by Roberto Lima.
1045 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1047 TF/03 Bugzilla 615: When checking the local_parts router precondition
1048 after a local_part_suffix or local_part_prefix option, Exim now
1049 does not use the address's named list lookup cache, since this
1050 contains cached lookups for the whole local part.
1052 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1053 Robert Millan. Documentation is in experimental-spec.txt.
1055 TF/04 Bugzilla 668: Fix parallel build (make -j).
1057 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1059 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1060 Patch provided by Jan Srzednicki.
1062 TF/05 Leading white space used to be stripped from $spam_report which
1063 wrecked the formatting. Now it is preserved.
1065 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1066 that they are available at delivery time.
1068 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1070 TF/08 TLS error reporting now respects the incoming_interface and
1071 incoming_port log selectors.
1073 TF/09 Produce a more useful error message if an SMTP transport's hosts
1074 setting expands to an empty string.
1076 NM/06 Bugzilla 744: EXPN did not work under TLS.
1077 Patch provided by Phil Pennock.
1079 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1080 Patch provided by Richard Godbee.
1082 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1083 acl_smtp_notquit, added index entry.
1085 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1086 Patch provided by Eugene Bujak.
1088 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1089 accept(). Patch provided by Maxim Dounin.
1091 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1092 Patch provided by Phil Pennock.
1094 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1096 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1097 Patch provided by Brad "anomie" Jorsch.
1099 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1100 Patch provided by Dean Brooks.
1102 TK/05 Add native DKIM support (does not depend on external libraries).
1104 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1105 Patch provided by Graeme Fowler.
1107 NM/16 Bugzilla 851: Documentation example syntax fix.
1109 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1111 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1114 NM/19 Bugzilla 745: TLS version reporting.
1115 Patch provided by Phil Pennock.
1117 NM/20 Bugzilla 167: bool: condition support.
1118 Patch provided by Phil Pennock.
1120 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1121 clients. Patch provided by Phil Pennock.
1123 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1124 Patch provided by Brad "anomie" Jorsch.
1126 NM/23 Bugzilla 687: Fix misparses in eximstats.
1127 Patch provided by Heiko Schlittermann.
1129 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1130 Patch provided by Heiko Schlittermann.
1132 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1133 plus update to original patch.
1135 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1137 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1138 Patch provided by David Brownlee.
1140 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1142 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1144 NM/30 Bugzilla 888: TLS documentation bugfixes.
1146 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1148 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1149 Unlike the original bugzilla I have changed all shell scripts in src tree.
1151 NM/33 Bugzilla 898: Transport filter timeout fix.
1152 Patch by Todd Rinaldo.
1154 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1155 Patch by Serge Demonchaux.
1157 NM/35 Bugzilla 39: Base64 decode bug fixes.
1158 Patch by Jakob Hirsch.
1160 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1162 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1164 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1166 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1172 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1173 ALT-N's libdkim that I have put here:
1174 http://duncanthrax.net/exim-experimental/
1176 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1177 were called 'true' and it seems that C99 defines that as a reserved
1178 keyword to be used with 'bool' variable types. That means you could
1179 not include C99-style headers which use bools without triggering
1180 build errors in sieve.c.
1182 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1183 as mailq or other aliases. Changed the --help handling significantly
1184 to do whats expected. exim_usage() emits usage/help information.
1186 SC/01 Added the -bylocaldomain option to eximstats.
1188 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1190 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1192 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1198 PH/01 Another patch from the Sieve maintainer.
1200 PH/02 When an IPv6 address is converted to a string for single-key lookup
1201 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1202 dots are used instead of colons so that keys in lsearch files need not
1203 contain colons. This was done some time before quoting was made available
1204 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1205 (notated using the quote facility) so as to distinguish them from IPv4
1206 keys. This meant that lookups for IP addresses in host lists did not work
1207 for iplsearch lookups.
1209 This has been fixed by arranging for IPv6 addresses to be expressed with
1210 colons if the lookup type is iplsearch. This is not incompatible, because
1211 previously such lookups could never work.
1213 The situation is now rather anomolous, since one *can* have colons in
1214 ordinary lsearch keys. However, making the change in all cases is
1215 incompatible and would probably break a number of configurations.
1217 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1220 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1221 conversion specification without a maximum field width, thereby enabling
1222 a rogue spamd server to cause a buffer overflow. While nobody in their
1223 right mind would setup Exim to query an untrusted spamd server, an
1224 attacker that gains access to a server running spamd could potentially
1225 exploit this vulnerability to run arbitrary code as the Exim user.
1227 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1228 $primary_hostname instead of what libspf2 thinks the hosts name is.
1230 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1231 a directory entry by the name of the lookup key. Previously, if a
1232 symlink pointed to a non-existing file or a file in a directory that
1233 Exim lacked permissions to read, a lookup for a key matching that
1234 symlink would fail. Now it is enough that a matching directory entry
1235 exists, symlink or not. (Bugzilla 503.)
1237 PH/03 The body_linecount and body_zerocount variables are now exported in the
1240 PH/04 Added the $dnslist_matched variable.
1242 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1243 This means they are set thereafter only if the connection becomes
1246 PH/06 Added the client_condition to authenticators so that some can be skipped
1247 by clients under certain conditions.
1249 PH/07 The error message for a badly-placed control=no_multiline_responses left
1250 "_responses" off the end of the name.
1252 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1254 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1255 (without spaces) instead of just copying the configuration text.
1257 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1259 PH/11 Added $max_received_linelength.
1261 PH/12 Added +ignore_defer and +include_defer to host lists.
1263 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1264 way in which PCRE > 7.0 is built.
1266 PH/14 Implemented queue_only_load_latch.
1268 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1269 MAIL command. The effect was to mangle the value on 64-bit systems.
1271 PH/16 Another patch from the Sieve maintainer.
1273 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1275 PH/18 If a system quota error occurred while trying to create the file for
1276 a maildir delivery, the message "Mailbox is full" was not appended to the
1277 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1278 only to a quota excession during the actual writing of the file.
1280 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1281 characters?) which causes problems in log lines. The DN values are now
1282 passed through string_printing() before being added to log lines.
1284 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1285 and InterBase are left for another time.)
1287 PH/21 Added message_body_newlines option.
1289 PH/22 Guard against possible overflow in moan_check_errorcopy().
1291 PH/23 POSIX allows open() to be a macro; guard against that.
1293 PH/24 If the recipient of an error message contained an @ in the local part
1294 (suitably quoted, of course), incorrect values were put in $domain and
1295 $local_part during the evaluation of errors_copy.
1301 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1302 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1305 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1306 issue a MAIL command.
1308 PH/02 In an ACL statement such as
1310 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1312 if a client was not listed at all, or was listed with a value other than
1313 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1314 the condition was not true (as it should be), so access was not denied.
1315 The bug was that the ! inversion was incorrectly passed on to the second
1316 item. This has been fixed.
1318 PH/03 Added additional dnslists conditions == and =& which are different from
1319 = and & when the dns lookup returns more than one IP address.
1321 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1322 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1324 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1325 FSYNC, which compiles an option called disable_fsync that allows for
1326 bypassing fsync(). The documentation is heavily laced with warnings.
1328 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1330 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1331 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1332 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1333 including adding "make clean"; (3) Added -fPIC when compiling the test
1334 dynamically loaded module, to get rid of a warning.
1336 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1337 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1338 = 0s. The bug is otherwise harmless.
1340 PH/07 There was a bug in the dovecot authenticator such that the value of
1341 $auth1 could be overwritten, and so not correctly preserved, after a
1342 successful authentication. This usually meant that the value preserved by
1343 the server_setid option was incorrect.
1345 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1347 PH/09 Installed PCRE release 7.0.
1349 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1350 run for batched SMTP input. It is now run at the start of every message
1351 in the batch. While fixing this I discovered that the process information
1352 (output by running exiwhat) was not always getting set for -bs and -bS
1353 input. This is fixed, and it now also says "batched" for BSMTP.
1355 PH/11 Added control=no_pipelining.
1357 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1358 patch, slightly modified), and move the expansion of helo_data till after
1359 the connection is made in the smtp transport (so it can use these
1362 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1364 PH/14 Added log_selector = +pid.
1366 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1368 PH/16 Add ${if forany and ${if forall.
1370 PH/17 Added dsn_from option to vary the From: line in DSNs.
1372 PH/18 Flush SMTP output before performing a callout, unless control =
1373 no_callout_flush is set.
1375 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1376 was true (the default) a successful delivery failed to delete the retry
1377 item, thus causing premature timeout of the address. The bug is now
1380 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1382 PH/21 Long custom messages for fakedefer and fakereject are now split up
1383 into multiline reponses in the same way that messages for "deny" and
1384 other ACL rejections are.
1386 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1387 with slight modification.
1389 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1390 draft, changing the syntax and factoring some duplicate code".
1392 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1393 for deliveries of the second and subsequent messages over the same SMTP
1396 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1397 ${reduce, with only minor "tidies".
1399 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1401 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1402 expansion side effects.
1404 PH/27 When a message times out after an over-quota error from an Exim-imposed
1405 quota, the bounce message says "mailbox is full". This message was not
1406 being given when it was a system quota that was exceeded. It now should
1409 MH/03 Made $recipients available in local_scan(). local_scan() already has
1410 better access to the recipient list through recipients_list[], but
1411 $recipients can be useful in postmaster-provided expansion strings.
1413 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1414 in the case of a MAIL command with additional options following the
1415 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1416 were accidentally chopped off.
1418 PH/29 SMTP synchronization checks are implemented when a command is read -
1419 there is a check that no more input is waiting when there shouldn't be
1420 any. However, for some commands, a delay in an ACL can mean that it is
1421 some time before the response is written. In this time, more input might
1422 arrive, invalidly. So now there are extra checks after an ACL has run for
1423 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1424 pipelining has not been advertised.
1426 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1428 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1429 returned in the SMTP response when smtp_return_error_details was false.
1430 This has been fixed.
1432 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1433 instead of the C I/O that was originally supplied, because problems were
1434 reported on Solaris.
1436 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1437 Exim which did not show up earlier: it was assuming that a call to
1438 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1439 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1440 was a macro that became an assignment, so it seemed to work. This has
1441 changed to a proper function call with a void return, hence the compile
1442 error. Exim's code has been fixed.
1444 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1447 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1449 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1451 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1452 because it turns out to be too big or there is a timeout) there is no
1453 "Completed" line in the log. When some messages of this type were
1454 selected by exigrep, they were listed as "not completed". Others were
1455 picked up by some special patterns. I have improved the selection
1456 criteria to be more general.
1458 PH/38 The host_find_failed option in the manualroute router can now be set
1459 to "ignore", to completely ignore a host whose IP address cannot be
1460 found. If all hosts are ignored, the behaviour is controlled by the new
1461 host_all_ignored option.
1463 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1464 homing or because of multiple MX records with /mx) generated more than
1465 one IP address, and the following item turned out to be the local host,
1466 all the secondary addresses of the first item were incorrectly removed
1467 from the list, along with the local host and any following hosts (which
1468 is what is supposed to happen).
1470 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1471 whoever called Exim into the -H file. In the case of the daemon it was
1472 behaving confusingly. When first started, it used values for whoever
1473 started the daemon, but after a SIGHUP it used the Exim user (because it
1474 calls itself on a restart). I have changed the code so that it now always
1477 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1478 message are rejected with the same error (e.g. no authentication or bad
1479 sender address), and a DATA command is nevertheless sent (as can happen
1480 with PIPELINING or a stupid MUA), the error message that was given to the
1481 RCPT commands is included in the rejection of the DATA command. This is
1482 intended to be helpful for MUAs that show only the final error to their
1485 PH/42 Another patch from the Sieve maintainer.
1487 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1488 Eximstats - Fixed some broken HTML links and added missing column headers
1490 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1491 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1493 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1495 PH/43 Yet another patch from the Sieve maintainer.
1497 PH/44 I found a way to check for a TCP/IP connection going away before sending
1498 the response to the final '.' that terminates a message, but only in the
1499 case where the client has not sent further data following the '.'
1500 (unfortunately, this is allowed). However, in many cases there won't be
1501 any further data because there won't be any more messages to send. A call
1502 to select() can be used: if it shows that the input is "ready", there is
1503 either input waiting, or the socket has been closed. An attempt to read
1504 the next input character can distinguish the two cases. Previously, Exim
1505 would have sent an OK response which the client would never have see.
1506 This could lead to message repetition. This fix should cure that, at
1507 least in a lot of common cases.
1509 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1510 advertised in response to EHLO.
1516 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1517 fixed by 4.65/MH/01 (is this a record?) are fixed:
1519 (i) An empty string was always treated as zero by the numeric comparison
1520 operators. This behaviour has been restored.
1522 (ii) It is documented that the numeric comparison operators always treat
1523 their arguments as decimal numbers. This was broken in that numbers
1524 starting with 0 were being interpreted as octal.
1526 While fixing these problems I realized that there was another issue that
1527 hadn't been noticed. Values of message_size_limit (both the global option
1528 and the transport option) were treated as octal if they started with 0.
1529 The documentation was vague. These values are now always treated as
1530 decimal, and I will make that clear in the documentation.
1536 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1537 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1540 MH/01 Don't check that the operands of numeric comparison operators are
1541 integers when their expansion is in "skipping" mode (fixes bug
1542 introduced by 4.64-PH/07).
1544 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1545 child addresses, Exim now panics and dies. Previously, because the count
1546 is held in a short int, deliveries were likely to be lost. As such a
1547 large number of recipients for a single message is ridiculous
1548 (performance will be very, very poor), I have chosen to impose a limit
1549 rather than extend the field.
1555 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1556 leftover -K file (the existence of which was triggered by #402).
1557 While we were at it, introduced process PID as part of the -K
1558 filename. This should rule out race conditions when creating
1561 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1562 processing considerably. Previous code took too long for large mails,
1563 triggering a timeout which in turn triggers #401.
1565 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1566 in the DK code in transports.c. sendfile() is not really portable,
1567 hence the _LINUX specificness.
1569 TF/01 In the add_headers option to the mail command in an Exim filter,
1570 there was a bug that Exim would claim a syntax error in any
1571 header after the first one which had an odd number of characters
1574 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1575 callout verification, Exim cached a "reject" for the entire domain. This
1576 is correct for most verifications, but it is not correct for a recipient
1577 verification with use_sender or use_postmaster set, because in that case
1578 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1579 case of MAIL FROM:<> rejection from other early rejections (e.g.
1580 rejection of HELO). When verifying a recipient using a non-null MAIL
1581 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1582 Whatever the result of the callout, the value of the domain cache is
1583 left unchanged (for any other kind of callout, getting as far as trying
1584 RCPT means that the domain itself is ok).
1586 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1589 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1590 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1591 session, the daemon ignores ECONNECT errors and logs others; it now
1592 ignores EPIPE as well.
1594 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1595 (quoted-printable decoding).
1597 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1598 later the small subsequent patch to fix an introduced bug.
1600 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1602 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1604 PH/08 An error is now given if message_size_limit is specified negative.
1606 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1607 to be given (somewhat) arbitrary names.
1609 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1612 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1613 miscellaneous code fixes
1615 PH/10 Added the log_reject_target ACL modifier to specify where to log
1618 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1619 hostname. This is wrong, because it relates to the incoming message (and
1620 probably the interface on which it is arriving) and not to the outgoing
1621 callout (which could be using a different interface). This has been
1622 changed to use the value of the helo_data option from the smtp transport
1623 instead - this is what is used when a message is actually being sent. If
1624 there is no remote transport (possible with a router that sets up host
1625 addresses), $smtp_active_hostname is used.
1627 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1628 tweaks were necessary in order to get it to work (see also 21 below):
1629 (a) The code assumed that strncpy() returns a negative number on buffer
1630 overflow, which isn't the case. Replaced with Exim's string_format()
1632 (b) There were several signed/unsigned issues. I just did the minimum
1633 hacking in of casts. There is scope for a larger refactoring.
1634 (c) The code used strcasecmp() which is not a standard C function.
1635 Replaced with Exim's strcmpic() function.
1636 (d) The code set only $1; it now sets $auth1 as well.
1637 (e) A simple test gave the error "authentication client didn't specify
1638 service in request". It would seem that Dovecot has changed its
1639 interface. Fortunately there's a specification; I followed it and
1640 changed what the client sends and it appears to be working now.
1642 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1645 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1646 address A is aliased to B and C, where B exists and C does not. Without
1647 -v the output is "A verified" because verification stops after a
1648 successful redirection if more than one address is generated. However,
1649 with -v the child addresses are also verified. Exim was outputting "A
1650 failed to verify" and then showing the successful verification for C,
1651 with its parentage. It now outputs "B failed to verify", showing B's
1652 parentage before showing the successful verification of C.
1654 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1655 look up a TXT record in a specific list after matching in a combined
1658 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1659 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1660 they consult the DNS. I had assumed they would set it the way they
1661 wanted; and indeed my experiments on Linux seem to show that in some
1662 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1663 To be on the safe side, however, I have now made the interface to
1664 host_find_byname() similar to host_find_bydns(), with an argument
1665 containing the DNS resolver options. The host_find_byname() function now
1666 sets these options at its start, just as host_find_bydns() does. The smtp
1667 transport options dns_qualify_single and dns_search_parents are passed to
1668 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1669 of host_find_byname() use the default settings of RES_DEFNAMES
1670 (qualify_single) but not RES_DNSRCH (search_parents).
1672 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1673 spool_read_header() do less string testing, by means of a preliminary
1674 switch on the second character of optional "-foo" lines. (This is
1675 overdue, caused by the large number of possibilities that now exist.
1676 Originally there were few.) While I was there, I also converted the
1677 str(n)cmp tests so they don't re-test the leading "-" and the first
1678 character, in the hope this might squeeze out yet more improvement.
1680 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1681 flag allowing group syntax was set by the header_syntax check but not
1682 turned off, possible causing trouble later; (2) The flag was not being
1683 set at all for the header_verify test, causing "group"-style headers to
1684 be rejected. I have now set it in this case, and also caused header_
1685 verify to ignore an empty address taken from a group. While doing this, I
1686 came across some other cases where the code for allowing group syntax
1687 while scanning a header line wasn't quite right (mostly, not resetting
1688 the flag correctly in the right place). These bugs could have caused
1689 trouble for malformed header lines. I hope it is now all correct.
1691 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1692 with the "reply" argument non-NULL. The code, however (which originally
1693 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1694 but it didn't always do it. This confused somebody who was copying the
1695 code for some other use. I have removed all the tests.
1697 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1698 feature that was used to support insecure browsers during the U.S. crypto
1699 embargo. It requires special client support, and Exim is probably the
1700 only MTA that supported it -- and would never use it because real RSA is
1701 always available. This code has been removed, because it had the bad
1702 effect of slowing Exim down by computing (never used) parameters for the
1703 RSA_EXPORT functionality.
1705 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1706 authenticator to fail if there's a tab character in the incoming data
1707 (there should never be unless someone is messing about, as it's supposed
1708 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1709 if the connection is using TLS or if the remote IP is the same as the
1710 local IP, and the "valid-client-cert option" if a client certificate has
1713 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1714 authenticators. This can be used for authorization after authentication
1715 succeeds. (In the case of plaintext, it servers for both authentication
1718 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1719 if any retry times were supplied.
1721 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1722 connection, where there is no client IP address to check. In this
1723 situation, the verify now always succeeds.
1725 PH/25 Applied John Jetmore's -Mset patch.
1727 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1729 PH/27 In a string expansion for a processed (not raw) header when multiple
1730 headers of the same name were present, leading whitespace was being
1731 removed from all of them, but trailing whitespace was being removed only
1732 from the last one. Now trailing whitespace is removed from each header
1733 before concatenation. Completely empty headers in a concatenation (as
1734 before) are ignored.
1736 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1737 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1739 PH/29 [Removed. This was a change that I later backed out, and forgot to
1740 correct the ChangeLog entry (that I had efficiently created) before
1741 committing the later change.]
1743 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1744 address errors (4xx response to RCPT) over the same connection as other
1745 messages routed to the same hosts. Such deliveries are always "forced",
1746 so retry times are not inspected. This resulted in far too many retries
1747 for the affected addresses. The effect occurred only when there were more
1748 hosts than the hosts_max_try setting in the smtp transport when it had
1749 the 4xx errors. Those hosts that it had tried were not added to the list
1750 of hosts for which the message was waiting, so if all were tried, there
1751 was no problem. Two fixes have been applied:
1753 (i) If there are any address or message errors in an SMTP delivery, none
1754 of the hosts (tried or untried) are now added to the list of hosts
1755 for which the message is waiting, so the message should not be a
1756 candidate for sending over the same connection that was used for a
1757 successful delivery of some other message. This seems entirely
1758 reasonable: after all the message is NOT "waiting for some host".
1759 This is so "obvious" that I'm not sure why it wasn't done
1760 previously. Hope I haven't missed anything, but it can't do any
1761 harm, as the worst effect is to miss an optimization.
1763 (ii) If, despite (i), such a delivery is accidentally attempted, the
1764 routing retry time is respected, so at least it doesn't keep
1765 hammering the server.
1767 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1768 in ${readsocket because some servers need this prod.
1770 PH/32 Added some extra debug output when updating a wait-xxx database.
1772 PH/33 The hint "could be header name not terminated by colon", which has been
1773 given for certain expansion errors for a long time, was not being given
1774 for the ${if def:h_colon_omitted{... case.
1776 PH/34 The spec says: "With one important exception, whenever a domain list is
1777 being scanned, $domain contains the subject domain." There was at least
1778 one case where this was not true.
1780 PH/35 The error "getsockname() failed: connection reset by peer" was being
1781 written to the panic log as well as the main log, but it isn't really
1782 panic-worthy as it just means the connection died rather early on. I have
1783 removed the panic log writing for the ECONNRESET error when getsockname()
1786 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1787 runs only) independently of the message's sender address. This meant
1788 that, if the 4xx error was in fact related to the sender, a different
1789 message to the same recipient with a different sender could confuse
1790 things. In particualar, this can happen when sending to a greylisting
1791 server, but other circumstances could also provoke similar problems.
1792 I have changed the default so that the retry time for these errors is now
1793 based a combination of the sender and recipient addresses. This change
1794 can be overridden by setting address_retry_include_sender=false in the
1797 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1798 remote server are returned as part of bounce messages. This was not
1799 happening for LMTP over a pipe (the lmtp transport), but now it is the
1800 same for both kinds of LMTP.
1802 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1803 in header lines that were in fact CNAMEs. This is no longer the case.
1805 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1806 and queue runs started by the daemon processed all messages. This has
1807 been fixed so that -R and -S can now usefully be given with -q<time>.
1809 PH/40 Import PCRE release 6.7 (fixes some bugs).
1811 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1813 PH/42 Give an error if -q is specified more than once.
1815 PH/43 Renamed the variables $interface_address and $interface_port as
1816 $received_ip_address and $received_port, to make it clear that these
1817 values apply to message reception, and not to the outgoing interface when
1818 a message is delivered. (The old names remain recognized, of course.)
1820 PH/44 There was no timeout on the connect() call when using a Unix domain
1821 socket in the ${readsocket expansion. There now is.
1823 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1824 be meaningful with "accept".
1826 SC/01 Eximstats V1.43
1827 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1829 SC/02 Eximstats V1.44
1830 Use a glob alias rather than an array ref in the generated
1831 parser. This improves both readability and performance.
1833 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1834 Collect SpamAssassin and rejection statistics.
1835 Don't display local sender or destination tables unless
1836 there is data to show.
1837 Added average volumes into the top table text output.
1839 SC/04 Eximstats V1.46
1840 Collect data on the number of addresses (recipients)
1841 as well as the number of messages.
1843 SC/05 Eximstats V1.47
1844 Added 'Message too big' to the list of mail rejection
1845 reasons (thanks to Marco Gaiarin).
1847 SC/06 Eximstats V1.48
1848 Mainlog lines which have GMT offsets and are too short to
1849 have a flag are now skipped.
1851 SC/07 Eximstats V1.49 (Alain Williams)
1852 Added the -emptyok flag.
1854 SC/08 Eximstats V1.50
1855 Fixes for obtaining the IP address from reject messages.
1857 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1858 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1859 whitesspace changes from 4.64-PH/27
1861 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1864 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1865 are found, allow negative numbers in numeric criteria)
1867 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1869 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1870 to match changes made in 4.64-PH/43
1872 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1874 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1875 "domain" is actually an IP address, and force a failure. This locks out
1876 those revolvers/nameservers that support "A-for-A" lookups, in
1877 contravention of the specifications.
1879 PH/48 When a host name was looked up from an IP address, and the subsequent
1880 forward lookup of the name timed out, the host name was left in
1881 $sender_host_name, contrary to the specification.
1883 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1884 restricted to single-key lookups, Exim was not diagnosing an error if
1885 * or *@ was used with a query-style lookup.
1887 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1889 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1890 long ago, but noone interested enough thought of it. Let's just say that
1891 the "1.1" means that there are some new functions that weren't there at
1892 some point in the past.
1894 PH/51 Error processing for expansion failure of helo_data from an smtp
1895 transport during callout processing was broken.
1897 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1898 tested/used via the -bh/-bhc/-bs options.
1900 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1901 bug, fixed in subsequent PCRE releases).
1903 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1904 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1906 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1912 SC/01 Use a glob alias rather than an array ref in eximstats generated
1913 parser. This improves both readability and performance.
1915 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1916 Don't display local sender or destination tables in eximstats unless
1917 there is data to show.
1918 Added average volumes into the eximstats top table text output.
1920 SC/03 Collect data on the number of addresses (recipients) as well
1921 as the number of messages in eximstats.
1923 TF/01 Correct an error in the documentation for the redirect router. Exim
1924 does (usually) call initgroups() when daemonizing.
1926 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1927 with consistent privilege compared to when running as a daemon.
1929 TF/03 Note in the spec that $authenticated_id is not set for local
1930 submissions from trusted users.
1932 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1933 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1935 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1936 by adding some example configuration directives to the default
1937 configuration file. A little bit of work is required to uncomment the
1938 directives and define how usernames and passwords are checked, but
1939 there is now a framework to start from.
1941 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1942 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1943 without this. I don't know how relevant this is to other LDAP libraries.
1945 PH/02 Add the verb name to the "unknown ACL verb" error.
1947 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1949 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1951 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1952 directory not expanded when it should be if an expanded home directory
1953 was set for the address (which is overridden by the transport).
1955 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1958 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1959 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1960 because it is too late at that time, and has no effect.
1962 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1963 security issue with \' (bugzilla #107). I could not use the
1964 PQescapeStringConn() function, because it needs a PGconn value as one of
1967 PH/08 When testing addresses using -bt, indicate those final addresses that
1968 are duplicates that would not cause an additional delivery. At least one
1969 person was confused, thinking that -bt output corresponded to deliveries.
1970 (Suppressing duplicates isn't a good idea as you lose the information
1971 about possibly different redirections that led to the duplicates.)
1973 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1974 systems where poll() doesn't work, in particular OS X.
1976 PH/10 Added more information to debugging output for retry time not reached.
1978 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1979 operations in malware.c.
1981 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1984 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1985 syslog_duplication set false, log lines that would normally be written
1986 both the the main log and to the reject log were not written to syslog at
1989 PH/14 In the default configuration, change the use of "message" in ACL warn
1990 statements to "add_header".
1992 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1993 not followed by a command (e.g. "seen endif").
1995 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1996 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1999 PH/17 Added extra conditions to the default value of delay_warning_condition
2003 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2004 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2005 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2008 The Auto-Submitted: and various List- headers are standardised, whereas I
2009 don't think Precedence: ever was.
2011 PH/18 Refactored debugging code in route_finduser() to show more information,
2012 in particular, the error code if getpwnam() issues one.
2014 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2015 This is apparently needed in addition to the PH/07 change above to avoid
2016 any possible encoding problems.
2018 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2019 but not after initializing Perl.
2021 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2022 output them only if debugging. By default they are written stderr,
2023 apparently, which is not desirable.
2025 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2028 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2031 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2033 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2034 authenticated or an ident call has been made. Suppress the default
2035 values for $authenticated_id and $authenticated_sender (but permit -oMai
2036 and -oMas) when testing with -bh.
2038 PH/24 Re-jigged the order of the tests in the default configuration so that the
2039 tests for valid domains and recipients precede the DNS black list and CSA
2040 tests, on the grounds that those ones are more expensive.
2042 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2043 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2044 This bug exists in every version of Exim that I still have, right back to
2047 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2048 However, an attempt to turn on case-sensitivity in a regex key by
2049 including (?-i) didn't work because the subject string was already
2050 lowercased, and the effects were non-intuitive. It turns out that a
2051 one-line patch can be used to allow (?-i) to work as expected.
2057 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2058 other effects) broke the use of negated acl sub-conditions.
2060 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2063 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2064 "Deny" causes Exim to reject the incoming connection with a 554 error.
2065 Unfortunately, if there is a major crisis, such as a disk failure,
2066 tcp-wrappers gives "deny", whereas what one would like would be some
2067 kind of temporary error. A kludge has been added to help with this.
2068 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2069 554 error is used if errno is still zero or contains ENOENT (which occurs
2070 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2073 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2075 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2076 errors. Otherwise a message that provokes a temporary error (when other
2077 messages do not) can cause a whole host to time out.
2079 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2080 addresses were routed directly to files or pipes from a redirect router.
2081 File deliveries just didn't batch; pipe deliveries might have suffered
2084 PH/06 A failure to get a lock for a hints database would erroneously always say
2085 "Failed to get write lock", even when it was really a read lock.
2087 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2088 of 0600. This has been changed to use the value of the lockfile_mode
2089 option (which defaults to 0600).
2091 PH/08 Applied small patch from the Sieve maintainer.
2093 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2094 folder from quota calculations, a direct delivery into this folder messed
2095 up the contents of the maildirsize file. This was because the regex was
2096 used only to exclude .Trash (or whatever) when the size of the mailbox
2097 was calculated. There was no check that a delivery was happening into an
2098 excluded directory. This bug has been fixed by ignoring all quota
2099 processing for deliveries into excluded directories.
2101 PH/10 Added the maildirfolder_create_regex option to appendfile.
2107 PH/01 The code for finding all the local interface addresses on a FreeBSD
2108 system running IPv6 was broken. This may well have applied to all BSD
2109 systems, as well as to others that have similar system calls. The broken
2110 code found IPv4 interfaces correctly, but gave incorrect values for the
2111 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2112 that it would not match correctly against @[] and not recognize the IPv6
2115 PH/02 The ipliteral router was not recognizing addresses of the form user@
2116 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2118 PH/03 Added disable_ipv6.
2120 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2121 decoded form, because it is most often used to construct To: headers
2122 lines in autoreplies, and the decoded form may well be syntactically
2123 invalid. However, $reply_address has leading white space removed, and all
2124 newlines turned into spaces so that the autoreply transport does not
2127 PH/05 If group was specified without a user on a router, and no group or user
2128 was specified on a transport, the group from the router was ignored.
2130 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2131 for visible compile-time settings that can be used to change these
2132 numbers, for those that want even more. Backwards compatibility with old
2133 spool files has been maintained. However, going back to a previous Exim
2134 release will lost any variables that are in spool files.
2136 PH/07 Two small changes when running in the test harness: increase delay when
2137 passing a TCP/IP connection to a new process, in case the original
2138 process has to generate a bounce, and remove special handling of
2139 127.0.0.2 (sic), which is no longer necessary.
2141 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2142 be the same on different OS.
2144 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2147 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2148 whether --show-vars was specified or not
2150 JJ/02 exipick: Added support for new ACL variable spool format introduced
2153 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2154 syntactically invalid From: or Reply-to: line, and a filter used this to
2155 generate an autoreply, and therefore failed to obtain an address for the
2156 autoreply, Exim could try to deliver to a non-existent relative file
2157 name, causing unrelated and misleading errors. What now happens is that
2158 it logs this as a hard delivery error, but does not attempt to create a
2161 PH/11 The exinext utility has a -C option for testing purposes, but although
2162 the given file was scanned by exinext itself; it wasn't being passed on
2163 when Exim was called.
2165 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2166 an end-of-file indication when reading a command response.
2168 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2169 compiled. In many other places in Exim, IPv6 addresses are always
2170 recognized, so I have changed this. It also means that IPv4 domain
2171 literals of the form [IPV4:n.n.n.n] are now always recognized.
2173 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2174 used if the router is not running as root, for example, when verifying at
2175 ACL time, or when using -bh. The debugging output from this situation was
2176 non-existent - all you got was a failure to exec. I have made two
2179 (a) Failures to set uid/gid, the current directory, or a process leader
2180 in a subprocess such as that created by queryprogram now generate
2181 suitable debugging ouput when -d is set.
2183 (b) The queryprogram router detects when it is not running as root,
2184 outputs suitable debugging information if -d is set, and then runs
2185 the subprocess without attempting to change uid/gid.
2187 PH/15 Minor change to Makefile for building test_host (undocumented testing
2190 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2191 additional section of a DNS packet that returns MX or SRV records.
2192 Instead, it always explicitly searches for A/AAAA records. This avoids
2193 major problems that occur when a DNS server includes only records of one
2194 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2195 fixed another bug: if SRV records were looked up and the corresponding
2196 address records were *not* found in the additional section, the port
2197 values from the SRV records were lost.
2199 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2200 using the correct key (the original address) when searching the retry
2201 rules in order to find which one to use for generating the retry hint.
2203 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2204 adding the default one. Similarly, if it contains a Reply-To: header, the
2205 errors_reply_to option, if set, is not used.
2207 PH/19 When calculating a retry time, Exim used to measure the "time since
2208 failure" by looking at the "first failed" field in the retry record. Now
2209 it does not use this if it is later than than the arrival time of the
2210 message. Instead it uses the arrival time. This makes for better
2211 behaviour in cases where some deliveries succeed, thus re-setting the
2212 "first failed" field. An example is a quota failure for a huge message
2213 when small messages continue to be delivered. Without this change, the
2214 "time since failure" will always be short, possible causing more frequent
2215 delivery attempts for the huge message than are intended.
2216 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2218 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2219 $1, $2, $3) because the numerical variables can be reset during some
2220 expansion items (e.g. "match"), thereby losing the authentication data.
2222 PH/21 Make -bV show the size of off_t variables so that the test suite can
2223 decide whether to run tests for quotas > 2G.
2225 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2226 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2227 filecount value is greater than 2G or if a quota value is greater than 2G
2228 on a system where the size of off_t is not greater than 4, a panic error
2231 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2232 never match. The debug and -bh output now contains an explicit error
2233 message indicating a malformed IPv4 address or mask.
2235 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2236 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2237 PH/23 above applies.
2239 PH/25 Do not write to syslog when running in the test harness. The only
2240 occasion when this arises is a failure to open the main or panic logs
2241 (for which there is an explicit test).
2243 PH/26 Added the /no_tell option to "control=freeze".
2245 PH/27 If a host name lookup failed very early in a connection, for example, if
2246 the IP address matched host_lookup and the reverse lookup yielded a name
2247 that did not have a forward lookup, an error message of the form "no IP
2248 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2249 could be logged. Now it outputs the IP address instead of "NULL".
2251 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2252 allows the sender and the authenticated sender to be set when
2253 submitting a message from within Exim. Since child_open_exim() is
2254 documented for local_scan(), the new function should be too.
2256 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2257 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2258 results in an empty string is now treated as unset.
2260 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2262 PH/31 Added sender_verify_fail logging option.
2264 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2265 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2266 tidied the source and removed it altogether.
2268 PH/33 When a queue run was abandoned because the load average was too high, a
2269 log line was always written; now it is written only if the queue_run log
2270 selector is set. In addition, the log line for abandonment now contains
2271 information about the queue run such as the pid. This is always present
2272 in "start" and "stop" lines but was omitted from the "abandon" line.
2274 PH/34 Omit spaces between a header name and the colon in the error message that
2275 is given when verify = headers_syntax fails (if there are lots of them,
2276 the message gets confusing).
2278 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2279 names, as there are now some PTR records that contain slashes. This check
2280 is only to protect against broken name servers that fall over on strange
2281 characters, so the fact that it applies to all lookups doesn't matter.
2283 PH/36 Now that the new test suite is complete, we can remove some of the
2284 special code in Exim that was needed for the old test suite. For example,
2285 sorting DNS records because real resolvers return them in an arbitrary
2286 order. The new test suite's fake resolver always returns records in the
2289 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2290 bounces) except when queue_only is set, to avoid logging races between
2291 the different processes.
2293 PH/38 Panic-die if .include specifies a non-absolute path.
2295 PH/39 A tweak to the "H" retry rule from its user.
2297 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2298 a label. They prevented compilation on older perls.
2300 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2301 a warning to be raised on newish perls.
2303 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2304 on queue. Changes to match documented behaviour of showing count of
2305 messages matching specified criteria.
2307 PH/40 Changed the default ident timeout from 30s to 5s.
2309 PH/41 Added support for the use of login_cap features, on those BSD systems
2310 that have them, for controlling the resources used by pipe deliveries.
2312 PH/42 The content-scanning code uses fopen() to create files in which to put
2313 message data. Previously it was not paying any attention to the mode of
2314 the files. Exim runs with umask(0) because the rest of the code creates
2315 files with open(), and sets the required mode explicitly. Thus, these
2316 files were ending up world-writeable. This was not a big issue, because,
2317 being within the spool directory, they were not world-accessible. I have
2318 created a function called modefopen, which takes an additional mode
2319 argument. It sets umask(777), creates the file, chmods it to the required
2320 mode, then resets the umask. All the relevant calls to fopen() in the
2321 content scanning code have been changed to use this function.
2323 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2324 to 24 hours. This avoids potential overflow problems when processing G
2325 and H retry rules. I suspect nobody ever tinkers with this value.
2327 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2329 PH/45 When the plaintext authenticator is running as a client, the server's
2330 challenges are checked to ensure they are valid base64 strings. By
2331 default, the authentication attempt is cancelled if an invalid string is
2332 received. Setting client_ignore_invalid_base64 true ignores these errors.
2333 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2334 they are received. Thus, the responses can be made to depend on the
2335 challenges. If an invalid string is ignored, an empty string is placed in
2338 PH/46 Messages that are created by the autoreply transport now contains a
2339 References: header, in accordance with RFCs 2822 and 3834.
2341 PH/47 Added authenticated_sender_force to the smtp transport.
2343 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2345 PH/49 Installed latest patch from the Sieve maintainer.
2347 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2348 was also set, the appendfile transport was unnecessarily scanning a
2349 directory of message files (e.g. for maildir delivery) to find the count
2350 of files (along with the size), even though it did not need this
2351 information. It now does the scan only if it needs to find either the
2352 size of the count of files.
2354 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2356 PH/52 Two bugs concerned with error handling when the smtp transport is
2359 (i) Exim was not creating retry information for temporary errors given
2360 for individual recipients after the DATA command when the smtp transport
2361 was used in LMTP mode. This meant that they could be retried too
2362 frequently, and not timed out correctly.
2364 (ii) Exim was setting the flag that allows error details to be returned
2365 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2366 recipients that were returned after the DATA command.
2368 PH/53 This is related to PH/52, but is more general: for any failing address,
2369 when detailed error information was permitted to be returned to the
2370 sender, but the error was temporary, then after the final timeout, only
2371 "retry timeout exceeded" was returned. Now it returns the full error as
2372 well as "retry timeout exceeded".
2374 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2375 do this, and (what is worse) MTAs that accept it.
2377 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2378 will now be deprecated.
2380 PH/56 New os.c-cygwin from the Cygwin maintainer.
2382 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2383 formats (previously only available in exim formats via -bpr, -bpru,
2384 and -bpra. Now also available in native and exiqgrep formats)
2386 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2387 with very large, slow to parse queues
2389 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2391 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2393 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2394 responses to authentication challenges, though it was showing the
2395 challenges; (ii) I've removed the CR characters from the debug output for
2398 PH/58 Allow for the insertion of a newline as well as a space when a string
2399 is turned into more than one encoded-word during RFC 2047 encoding. The
2400 Sieve code now uses this.
2402 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2403 data_4xx, lost_connection, tls_required.
2405 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2406 message was being sent as an SMTP response.
2408 PH/61 Add -l and -k options to exicyclog.
2410 PH/62 When verifying, if an address was redirected to one new address, so that
2411 verification continued, and the new address failed or deferred after
2412 having set something in $address_data, the value of $address_data was not
2413 passed back to the ACL. This was different to the case when no
2414 redirection occurred. The value is now passed back in both cases.
2416 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2417 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2418 use login_cap.h, so on its own it isn't the distinguishing feature. The
2419 new name refers directly to the setclassresources() function.
2421 PH/65 Added configuration files for NetBSD3.
2423 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2425 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2426 is preferred over IPv4.
2428 PH/68 The bounce_return_message and bounce_return_body options were not being
2429 honoured for bounces generated during the reception of non-SMTP messages.
2430 In particular, this applied to messages rejected by the ACL. This bug has
2431 been fixed. However, if bounce_return_message is true and bounce_return_
2432 body is false, the headers that are returned for a non-SMTP message
2433 include only those that have been read before the error was detected.
2434 (In the case of an ACL rejection, they have all been read.)
2436 PH/69 The HTML version of the specification is now built in a directory called
2437 spec_html instead of spec.html, because the latter looks like a path with
2438 a MIME-type, and this confuses some software.
2440 PH/70 Catch two compiler warnings in sieve.c.
2442 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2443 function verify_get_ident() calls ip_connect() to connect a socket, but
2444 if the "connect()" function timed out, ip_connect() used to close the
2445 socket. However, verify_get_ident() also closes the socket later, and in
2446 between Exim writes to the log, which may get opened at this point. When
2447 the socket was closed in ip_connect(), the log could get the same file
2448 descriptor number as the socket. This naturally causes chaos. The fix is
2449 not to close the socket in ip_connect(); the socket should be closed by
2450 the function that creates it. There was only one place in the code where
2451 this was missing, in the iplookup router, which I don't think anybody now
2452 uses, but I've fixed it anyway.
2454 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2455 well as to direct DNS lookups. Otherwise the handling of names in host
2456 lists is inconsistent and therefore confusing.
2462 PH/01 Two changes to the default runtime configuration:
2464 (1) Move the checks for relay_from_hosts and authenticated clients from
2465 after to before the (commented out) DNS black list checks.
2467 (2) Add control=submission to the relay_from_hosts and authenticated
2468 clients checks, on the grounds that messages accepted by these
2469 statements are most likely to be submissions.
2471 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2473 (1) Generate an error if the third argument for the ${prvs expansion is
2476 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2479 (3) Reset the variables that are obtained from the first argument of
2480 ${prvscheck and used in the second argument before leaving the code,
2481 because their memory is reclaimed, so using them afterwards may do
2484 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2485 one (it's much easier than Tom thought :-).
2487 (5) Because of (4), we can now allow for the use of $prvscheck_result
2488 inside the third argument.
2490 PH/03 For some reason, the default setting of PATH when running a command from
2491 a pipe transport was just "/usr/bin". I have changed it to
2494 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2495 anything to be listed in the output from -bV.
2497 PH/05 When a filter generated an autoreply, the entire To: header line was
2498 quoted in the delivery log line, like this:
2500 => >A.N.Other <ano@some.domain> <original@ddress> ...
2502 This has been changed so that it extracts the operative address. There
2503 may be more than one such address. If so, they are comma-separated, like
2506 => >ano@some.domain,ona@other.domain <original@ddress> ...
2508 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2509 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2510 not being looked up in the rDNS to get a host name, Exim was showing the
2511 IP address twice in Received: lines, even though the IP addresses were
2512 identical. For example:
2514 Received: from [1.2.3.4] (helo=[1.2.3.4])
2516 However, if the real host name was known, it was omitting the HELO data
2517 if it matched the actual IP address. This has been tidied up so that it
2518 doesn't show the same IP address twice.
2520 PH/07 When both +timestamp and +memory debugging was on, the value given by
2521 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2522 called by the debug printing, thereby overwriting the timestamp buffer.
2523 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2525 PH/08 When the original message was included in an autoreply transport, it
2526 always said "this is a copy of the message, including all the headers",
2527 even if body_only or headers_only was set. It now gives an appropriate
2530 PH/09 Applied a patch from the Sieve maintainer which:
2532 o fixes some comments
2533 o adds the (disabled) notify extension core
2534 o adds some debug output for the result of if/elsif tests
2535 o points to the current vacation draft in the documentation
2536 and documents the missing references header update
2540 o fixes a bug in processing the envelope test (when testing
2541 multiple envelope elements, the last element determinted the
2544 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2545 Electronic Mail") by including:
2547 Auto-submitted: auto-generated
2549 in the messages that it generates (bounce messages and others, such as
2550 warnings). In the case of bounce messages for non-SMTP mesages, there was
2551 also a typo: it was using "Auto_submitted" (underscore instead of
2552 hyphen). Since every message generated by Exim is necessarily in response
2553 to another message, thes have all been changed to:
2555 Auto-Submitted: auto-replied
2557 in accordance with these statements in the RFC:
2559 The auto-replied keyword:
2561 - SHOULD be used on messages sent in direct response to another
2562 message by an automatic process,
2564 - MUST NOT be used on manually-generated messages,
2566 - MAY be used on Delivery Status Notifications (DSNs) and Message
2567 Disposition Notifications (MDNs),
2569 - MUST NOT be used on messages generated by automatic or periodic
2570 processes, except for messages which are automatic responses to
2573 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2574 to the default Received: header definition.
2576 PH/12 Added log selector acl_warn_skipped (default on).
2578 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2579 variables because (a) they are in the wrong storage pool and (b) even if
2580 they were copied, it wouldn't work properly because of the caching.
2582 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2583 checking when decoding. Apparently there are clients that generate
2584 overlong encoded strings. Why am I not surprised?
2586 PH/15 If the first argument of "${if match_address" was not empty, but did not
2587 contain an "@" character, Exim crashed. Now it writes a panic log message
2588 and treats the condition as false.
2590 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2592 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2593 extension "envelope-auth". The code is finished and in agreement with
2594 other implementations, but there is no documentation so far and in fact,
2595 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2596 not changing the active code.
2598 Print executed "if" and "elsif" statements when debugging is used. This
2599 helps a great deal to understand what a filter does.
2601 Document more things not specified clearly in RFC3028. I had all this
2602 sorted out, when out of a sudden new issues came to my mind. Oops."
2604 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2607 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2608 canonical form (as documented). However, after a host name lookup from
2609 the IP address, check_host() was doing a simple string comparison with
2610 addresses acquired from the DNS when checking that the found name did
2611 have the original IP as one of its addresses. Since any found IPv6
2612 addresses are likely to be in abbreviated form, the comparison could
2613 fail. Luckily, there already exists a function for doing the comparison
2614 by converting both addresses to binary, so now that is used instead of
2615 the text comparison.
2617 PH/20 There was another similar case to PH/19, when a complete host name was
2618 given in a host list; looking up its IP address could give an abbreviated
2619 form, whereas the current host's name might or might not be abbreviated.
2620 The same fix has been applied.
2626 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2627 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2630 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2631 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2633 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2635 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2636 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2637 Also, for "Auto-Submitted", treat anything other than "no" as
2638 non-personal, in accordance with RFC 3834. (Previously it treated
2639 anything starting "auto-" as non-personal.)
2641 TF/01 The control=submission/name=... option had a problem with syntax
2642 errors if the name included a slash character. The /name= option
2643 now slurps the rest of the string, so it can include any characters
2644 but it must come last in the list of options (after /sender_retain
2647 PH/05 Some modifications to the interface to the fake nameserver for the new
2655 TK/01 Added the "success_on_redirect" address verification option. See
2656 NewStuff for rationale and an example.
2658 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2660 PH/02 Patch to exigrep to allow it to work on syslog lines.
2662 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2663 fread() to skip over the body file's header line, because in Cygwin the
2664 header line is locked and is inaccessible.
2666 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2667 co-exist for some time) to make it clear that it is the Exim ID that is
2668 referenced, not the Message-ID: header line.
2670 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2671 string_format() function, because snprintf() does not exist on all
2674 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2675 ${stat: expansion item.
2677 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2678 protocol synchronization error", to keep the pedants happy.
2680 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2681 well as for IRIX systems, when gcc is being used. See the host.c source
2684 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2686 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2689 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2690 transport and to the smtp transport in LMTP mode.
2692 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2694 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2695 run a filter in a subprocess. This could lead to confusion in subsequent
2696 lookups in the parent process. There should also be a search_tidyup() at
2697 the end of the subprocess.
2699 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2700 only if the host matched helo_try_verify_hosts, which caused the
2701 verification to occur when the EHLO/HELO command was issued. The ACL just
2702 tested the remembered result. Now, if a previous verification attempt has
2703 not happened, "verify = helo" does it there and then.
2705 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2707 TK/03 Fix log output including CR from clamd.
2709 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2710 exist provoked a memory error which could cause a segfault.
2712 PH/15 Installed PCRE 6.2
2714 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2716 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2717 of the problem. Specifically, suggested +O2 rather than +O1 for the
2720 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2722 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2723 record, the port was not correctly logged when the outgoing_port log
2724 selector was set (it logged the transort's default port).
2726 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2727 fallback_hosts, and "hosts" in the smtp transport.
2729 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2730 host errors such as "Connection refused".
2732 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2733 authentication with radiusclient 0.4.9:
2735 - Error returned from rc_read_config was caught wrongly
2736 - Username/password not passed on to radius server due to wrong length.
2738 The presumption is that some radiusclient API changes for 4.51/PH/17
2739 were not taken care of correctly. The code is still untested by me (my
2740 Linux distribution still has 0.3.2 of radiusclient), but it was
2741 contributed by a Radius user.
2743 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2744 expanding the "port" option of the smtp transport.
2746 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2747 while reading a MIME header. Thanks to Tom Hughes for a patch.
2749 PH/24 Include config.h inside local_scan.h so that configuration settings are
2752 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2753 that in an ACL for RCPT (for example), you can examine exactly what was
2756 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2757 commands, but it was not correctly comparing the address with the actual
2758 client host address. Thus, it would show the EHLO address in Received:
2759 header lines when this was not necessary.
2761 PH/27 Added the % operator to ${eval:}.
2763 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2764 it should be ignoring failures (because with -C, for example, it has lost
2765 privilege). It wasn't ignoring creation failures other than "already
2768 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2771 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2772 because an input line was too long, either on its own, or by virtue of
2773 too many continuations, the temporary file was not being removed, and the
2774 return code was incorrect.
2776 PH/31 Missing "BOOL" in function definition in filtertest.c.
2778 PH/32 Applied Sieve patches from the maintainer.
2780 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2782 PH/33 Added "verify = not_blind".
2784 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2785 Local/Makefile (with some defaults set). These are used in built scripts
2786 such as exicyclog, but they have never been used in the exim_install
2787 script (though there are many overriding facilities there). I have
2788 arranged that the exim_install script now takes note of these two
2791 PH/35 Installed configuration files for Dragonfly.
2793 PH/36 When a locally submitted message by a trusted user did not contain a
2794 From: header, and the sender address was obtained from -f or from an SMTP
2795 MAIL command, and the trusted user did not use -F to supply a sender
2796 name, $originator_name was incorrectly used when constructing a From:
2797 header. Furthermore, $originator_name was used for submission mode
2798 messages from external hosts without From: headers in a similar way,
2799 which is clearly wrong.
2801 PH/37 Added control=suppress_local_fixups.
2803 PH/38 When log_selector = +received_sender was set, and the addition of the
2804 sender made the log line's construction buffer exactly full, or one byte
2805 less than full, an overflow happened when the terminating "\n" was
2808 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2809 when the result of a list match is failure because a DNS lookup failed.
2811 PH/40 RM_COMMAND is now used in the building process.
2813 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2814 the "build-* directories that it finds.
2816 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2817 address in a domain literal was a prefix of an interface address.
2819 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2820 when verifying a sender address, unless rewrite_headers is false.
2822 PH/44 Wrote a long comment about why errors_to addresses are verified as
2823 recipients, not senders.
2825 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2826 the ratelimit ACL was added.
2828 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2830 PH/46 Added extra information about PostgreSQL errors to the error string.
2832 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2833 avoiding the need to install special zones in a real server. This is
2834 backwards compatible; if it can't find the fake resolver, it drops back.
2835 Thus, both old and new test suites can be run.
2837 TF/02 Added util/ratelimit.pl
2839 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2840 clock is set back in time.
2842 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2843 Candler <B.Candler@pobox.com>.
2845 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2846 OK for addresses that are the result of redirections.
2848 PH/48 A number of further additions for the benefit of the new test suite,
2849 including a fake gethostbyname() that interfaces to the fake DNS resolver
2852 TF/06 The fix for widen_domains has also been applied to qualify_single and
2853 search_parents which are the other dnslookup options that can cause
2856 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2859 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2861 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2862 service" if it encountered an MX record with an empty target hostname.
2863 The message is now "an MX or SRV record indicated no SMTP service".
2865 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2866 if the DNS of the sending site is misconfigured. This is quite a
2867 common situation. This change restores the behaviour of treating a
2868 helo verification defer as a failure.
2870 PH/51 If self=fail was set on a router, the bounce message did not include the
2871 actual error message.
2877 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2879 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2880 command itself ended in error, the underlying message about the transport
2881 filter timeout was being overwritten with the pipe command error. Now the
2882 underlying error message should be appended to the second error message.
2884 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2886 PH/02 Reduce the amount of output that "make" produces by default. Full output
2887 can still be requested.
2889 PH/03 The warning log line about a condition test deferring for a "warn" verb
2890 was being output only once per connection, rather than after each
2891 occurrence (because it was using the same function as for successful
2892 "warn" verbs). This seems wrong, so I have changed it.
2894 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2895 it should not have, which might have caused a crash in the right
2896 circumstances, but probably never did.
2898 PH/04 Installed a modified version of Tony Finch's patch to make submission
2899 mode fix the return path as well as the Sender: header line, and to
2900 add a /name= option so that you can make the user's friendly name appear
2903 TF/03 Added the control = fakedefer ACL modifier.
2905 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2906 Mark Lowes for thorough testing.
2908 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2910 TK/03 Merged latest SRS patch from Miles Wilton.
2912 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2913 with the definition in sysexits.h (which is #included earlier).
2914 Fortunately, Exim does not actually use EX_OK. The code used to try to
2915 preserve the sysexits.h value, by assumimg that macro definitions were
2916 scanned for macro replacements. I have been disabused of this notion,
2917 so now the code just undefines EX_OK before #including unistd.h.
2919 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2920 in the smtp transport. When a block could not be written in a single
2921 write() function, the timeout was being re-applied to each part-write.
2922 This seems wrong - if the receiver was accepting one byte at a time it
2923 would take for ever. The timeout is now adjusted when this happens. It
2924 doesn't have to be particularly precise.
2926 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2927 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2929 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2930 without a domain if the check to <postmaster@domain> fails.
2932 SC/01 Eximstats: added -xls and the ability to specify output files
2933 (patch written by Frank Heydlauf).
2935 SC/02 Eximstats: use FileHandles for outputing results.
2937 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2939 SC/04 Eximstats: fixed display of large numbers with -nvr option
2941 SC/05 Eximstats: fixed merging of reports with empty tables.
2943 SC/06 Eximstats: added the -include_original_destination flag
2945 SC/07 Eximstats: removed tabs and trailing whitespace.
2947 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2949 TK/06 MBOX spool code: Add real "From " MBOX separator line
2950 so the .eml file is really in mbox format (even though
2951 most programs do not really care). Patch from Alex Miller.
2953 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2954 The latter is generated from $received_to and is only set if the
2955 message has one envelope recipient. SA can use these headers,
2956 obviously out-of-the-box. Patch from Alex Miller.
2958 PH/08 The ${def test on a variable was returning false if the variable's
2959 value was "0", contrary to what the specification has always said!
2960 The result should be true unless the variable is empty.
2962 PH/09 The syntax error of a character other than { following "${if
2963 def:variable_name" (after optional whitespace) was not being diagnosed.
2964 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2965 accidental colon was present, for example, could give incorrect results.
2967 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2968 result is used (not including appendfile, where other changes are about
2971 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2972 This involved changing a lot of size variables from int to off_t. It
2973 should work with maildirs and everything.
2975 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2976 spamd dying while we are connected to it.
2978 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2981 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2982 function for BDB 4.3.
2984 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2986 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2987 the sources for PCRE have been re-arranged and more files are now
2990 PH/15 The code I had for printing potentially long long variables in PH/11
2991 above was not the best (it lost precision). The length of off_t variables
2992 is now inspected at build time, and an appropriate printing format (%ld
2993 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2994 to be "long long int" or "long int". This is needed for the internal
2995 formatting function string_vformat().
2997 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2998 the configuration file to be ":syslog", then the script "guesses" where
2999 the logs files are, rather than using the compiled in default. In our
3000 case the guess is not the same as the compiled default, so the script
3001 suddenly stopped working when I started to use syslog. The patch checks
3002 to see if log_file_path is "". If so, it attempts to read it from exim
3003 with no configuration file to get the compiled in version, before it
3004 falls back to the previous guessing code."
3006 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3007 implementing BATV in an Exim configuration. See NewStuff for the gory
3010 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3011 Makefile that are specific to HP-UX.
3013 PH/18 If the "use_postmaster" option was set for a recipient callout together
3014 with the "random" option, the postmaster address was used as the MAIL
3015 FROM address for the random test, but not for the subsequent recipient
3016 test. It is now used for both.
3018 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3019 patch removes a few documentation additions to RFC 3028, because the
3020 latest draft now contains them. It adds the new en;ascii-case comparator
3021 and a new error check for 8bit text in MIME parts. Comparator and
3022 require names are now matched exactly. I enabled the subaddress
3023 extension, but it is not well tested yet (read: it works for me)."
3025 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3026 rework some of the code of TK/09 above to avoid the hardwired use of
3027 "%lld" and "long long". Replaced the call to snprintf() with a call to
3030 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3031 records point to non-existent hosts", "retry timeout exceeded", and
3032 "retry time not reached for any host after a long failure period".
3034 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3035 experimental DomainKeys support:
3037 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3038 (2) On an error such as an illegally used "control", the wrong name for
3039 the control was given.
3041 These problems did NOT occur unless DomainKeys support was compiled.
3043 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3045 PH/24 Added ${if match_ip condition.
3047 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3048 databases so that it will be absolutely obvious if a crash occurs in the
3049 DB library. This is a regular occurrence (often caused by mis-matched
3052 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3053 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3054 picked up on a user's system that detects such things. There doesn't seem
3055 to be a gcc warning option for this - only an attribute that has to be
3056 put on the function's prototype. It seems that in Fedora Core 4 they have
3057 set this on a number of new functions. No doubt there will be more in due
3060 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3061 specify a transport. However, if an address that was verified by such a
3062 router was the subject of a callout, Exim crashed because it tried to
3063 read the rcpt_include_affixes from the non-existent transport. Now it
3064 just assumes that the setting of that option is false. This bug was
3065 introduced by 4.51/PH/31.
3067 PH/28 Changed -d+all to exclude +memory, because that information is very
3068 rarely of interest, but it makes the output a lot bigger. People tend to
3069 do -d+all out of habit.
3071 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3072 code in os-type was giving problems when libc.so lives in lib64, like on
3075 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3076 aren't the modern standard, and it seems that some systems' include files
3077 don't always have them. Exim was already checking for some of the newer
3078 ones like T_AAAA, and defining it itself. I've added checks for all the
3079 record types that Exim uses.
3081 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3082 not automatically generating a new one, as it is supposed to. This
3083 prevented TLS from working. If the file did exist, but contained invalid
3084 data, a new version was generated, as expected. It was only the case of a
3085 non-existent file that was broken.
3087 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3088 with a change in libdomainkeys > 0.64.
3090 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3091 from DNS. If the selector record carries the flag, it now has
3092 precedence over the domain-wide flag.
3094 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3096 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3097 the use of a port name that isn't defined in /etc/services) occurred, the
3098 message was deferred as in a normal delivery, and thus remained on the
3099 spool, instead of being failed because of the mua_wrapper setting. This
3100 is now fixed, and I tidied up some of the mua_wrapper messages at the
3103 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3104 the messages in a hash of arrays rather than using individual hashes.
3105 This is a bit cleaner and results in dramatic memory savings, albeit
3106 at a slight CPU cost.
3108 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3109 as requested by Marc Sherman.
3111 SC/10 Eximstats: added histograms for user specified patterns as requested
3114 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3116 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3117 fopen() in the content-scanning modules that did not already have it.
3123 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3124 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3126 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3128 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3130 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3131 to test Sieve filters that use "vacation".
3133 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3134 that changes the way the GnuTLS parameters are stored in the cache file.
3135 The new format can be generated externally. For backward compatibility,
3136 if the data in the cache doesn't make sense, Exim assumes it has read an
3137 old-format file, and it generates new data and writes a new file. This
3138 means that you can't go back to an older release without removing the
3141 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3142 work if there are any delivery delays because "one_time" forces the
3143 parent to be marked "delivered", so its unseen clone is never tried
3144 again. For this reason, Exim now forbids the simultaneous setting of
3147 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3148 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3149 entry if you want to know the details. The fix, however, neglected to
3150 consider the case when local delivery batching is involved. The test for
3151 "previously delivered" was not happening when checking to see if an
3152 address could be batched with a previous (undelivered) one; under
3153 certain circumstances this could lead to multiple deliveries to the same
3156 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3157 in its include files, and this causes problems building Exim.
3159 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3160 header_syntax) but Exim was just ignoring anything given after a slash.
3161 In particular, this caused confusion with an attempt to use "verify =
3162 reverse_host_lookup/defer_ok". An error is now given when options are
3163 supplied for verify items that do not have them. (Maybe reverse_host_
3164 lookup should have a defer_ok option, but that's a different point.)
3166 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3167 defined by RFC 821) to 2048, because there were problems with some AUTH
3168 commands, and RFC 1869 says the size should be increased for extended
3169 SMTP commands that take arguments.
3171 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3174 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3175 "unknown" error; now it says that the functionality isn't in the binary.
3177 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3178 an address' error message when a string expansion fails (syntax or
3179 whatever). Otherwise the password may appear in the log. Following change
3180 PH/42 below, there is no longer a chance of it appearing in a bounce
3183 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3185 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3186 of its addresses was ever tried. (Bugzilla bug #2.)
3188 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3189 the result incorrectly in the debug output. (It correctly added a newline
3190 to what was transported.)
3192 TF/01 Added $received_time.
3194 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3195 commented out examples of how to interface to a virus scanner and to
3196 SpamAssassin. Also added commented examples of av_scanner and
3197 spamd_address settings.
3199 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3200 and controls are allowed in which ACLs. There were a couple of minor
3201 errors. Some of the entries in the conditions table (which is a table of
3202 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3203 negation of where the condition IS allowed.
3205 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3207 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3208 header file does not have a version number, so I've had to invent a new
3209 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3210 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3211 radiusclient), but it was contributed by a Radius user.
3213 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3214 files or directories, for OpenSSL.
3216 PH/19 When an Exim process that is running as root has to create an Exim log
3217 file, it does so in a subprocess that runs as exim:exim so as to get the
3218 ownership right at creation (otherwise, other Exim processes might see
3219 the file with the wrong ownership). There was no test for failure of this
3220 fork() call, which would lead to the process getting stuck as it waited
3221 for a non-existent subprocess. Forks do occasionally fail when resources
3222 run out. I reviewed all the other calls to fork(); they all seem to check
3225 PH/20 When checking for unexpected SMTP input at connect time (before writing
3226 the banner), Exim was not dealing correctly with a non-positive return
3227 from the read() function. If the client had disconnected by this time,
3228 the result was a log entry for a synchronization error with an empty
3229 string after "input=" when read() returned zero. If read() returned -1
3230 (an event I could not check), uninitialized data bytes were printed.
3231 There were reports of junk text (parts of files, etc) appearing after
3234 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3236 PH/22 Added support for macro redefinition, and (re)definition in between
3237 driver and ACL definitions.
3239 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3240 forgetting to use the resulting value; it was using the unexpanded value.
3242 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3243 hadn't been configured. The fix is from Juergen Kreileder, who
3244 understands it better than I do:
3246 "Here's what I see happening with three configured cyrus_sasl
3247 authenticators configured (plain, login, cram-md5):
3249 On startup auth_cyrus_sasl_init() gets called for each of these.
3250 This means three calls to sasl_listmech() without a specified mech_list.
3251 => SASL tests which mechs of all available mechs actually work
3252 => three warnings about OTP not working
3253 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3255 With the patch, sasl_listmech() also gets called three times. But now
3256 SASL's mech_list option is set to the server_mech specified in the the
3257 authenticator. Or in other words, the answer from sasl_listmech()
3258 gets limited to just the mech you're testing for (which is different
3260 => the return list contains just 'plain' or 'login', 'cram-md5' or
3261 nothing depending on the value of ob->server_mech.
3263 I've just tested the patch: Authentication still works fine,
3264 unavailable mechs specified in the exim configuration are still
3265 caught, and the auth.log warnings about OTP are gone."
3267 PH/25 When debugging is enabled, the contents of the command line are added
3268 to the debugging output, even when log_selector=+arguments is not
3271 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3272 answer is "GNU", and only if the return is "GNU/something" is the answer
3275 PH/27 $acl_verify_message is now set immediately after the failure of a
3276 verification in an ACL, and so is available in subsequent modifiers. In
3277 particular, the message can be preserved by coding like this:
3279 warn !verify = sender
3280 set acl_m0 = $acl_verify_message
3282 Previously, $acl_verify_message was set only while expanding "message"
3283 and "log_message" when a very denied access.
3285 PH/28 Modified OS/os.c-Linux with
3287 -#ifndef OS_LOAD_AVERAGE
3288 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3290 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3291 nomenclature these days.)
3293 PH/29 Installed patch from the Sieve maintainer that adds the options
3294 sieve_useraddress and sieve_subaddress to the redirect router.
3296 PH/30 In these circumstances:
3297 . Two addresses routed to the same list of hosts;
3298 . First host does not offer TLS;
3299 . First host accepts first address;
3300 . First host gives temporary error to second address;
3301 . Second host offers TLS and a TLS session is established;
3302 . Second host accepts second address.
3303 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3304 and peerdn, if requested) that were in fact used only for the second
3307 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3308 attention to any local part prefix or suffix that was matched by the
3309 router that accepted the address. It now behaves in the same way as it
3310 does for delivery: the affixes are removed from the local part unless
3311 rcpt_include_affixes is set on the transport.
3313 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3314 timeout during the DATA phase of an incoming message.
3316 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3317 applied a patch sanctioned by the Sieve maintainer.
3319 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3320 the uid or gid is negative. A case of a negative gid caused this to be
3321 noticed. The fix allows for either to be negative.
3323 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3324 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3327 PH/36 The change PH/12 above was broken. Fixed it.
3329 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3330 the grounds that routing the same address twice would always produce the
3331 same answer. This might have been true once, but it is certainly no
3332 longer true now. Routing a child address may depend on the previous
3333 routing that produced that child. Some complicated redirection strategies
3334 went wrong when messages had multiple recipients, and made Exim's
3335 behaviour dependent on the order in which the addresses were given.
3337 I have moved the duplicate checking until after the routing is complete.
3338 Exim scans the addresses that are assigned to local and remote
3339 transports, and removes any duplicates. This means that more work will be
3340 done, as duplicates will always all be routed, but duplicates are
3341 presumably rare, so I don't expect this is of any significance.
3343 For deliveries to pipes, files, and autoreplies, the duplicate checking
3344 still happens during the routing process, since they are not going to be
3347 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3348 It corrects a timeout issue with spamd. This is Ian's comment: "The
3349 background is that sometimes spamd either never reads data from a
3350 connection it has accepted, or it never writes response data. The exiscan
3351 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3352 blindly assumes that writes won't block so it may never time out."
3354 PH/39 Allow G after quota size as well as K and M.
3356 PH/40 The value set for $authenticated_id in an authenticator may not contain
3357 binary zeroes or newlines because the value is written to log lines and
3358 to spool files. There was no check on this. Now the value is run through
3359 the string_printing() function so that such characters are converted to
3360 printable escape sequences.
3362 PH/41 $message_linecount is a new variable that contains the total number of
3363 lines in the message. Compare $body_linecount, which is the count for the
3366 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3367 bounce and delay warning messages, except in certain special cases, which
3370 (a) An SMTP error message from a remote host;
3371 (b) A message specified in a :fail: redirection;
3372 (c) A message specified in a "fail" command in a system filter;
3373 (d) A message specified in a FAIL return from the queryprogram router;
3374 (e) A message specified by the cannot_route_message router option.
3376 In these cases only, Exim does include the error details in bounce and
3377 warning messages. There are also a few cases where bland messages such
3378 as "unrouteable address" or "local delivery error" are given.
3380 PH/43 $value is now also set for the "else" part of a ${run expansion.
3382 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3383 being worked on, but at least Exim now implements the latest version to
3386 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3387 process to complete was treated as a delivery failure, a timeout while
3388 writing the message to the pipe was logged, but erroneously treated as a
3389 successful delivery. Such timeouts include transport filter timeouts. For
3390 consistency with the overall process timeout, these timeouts are now
3391 treated as errors, giving rise to delivery failures by default. However,
3392 there is now a new Boolean option for the pipe transport called
3393 timeout_defer, which, if set TRUE, converts the failures into defers for
3394 both kinds of timeout. A transport filter timeout is now identified in
3397 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3398 systems where "make" and "gmake" are different, calling "gmake" at top
3399 level broke things. I've arranged for the value of $(MAKE) to be passed
3400 from the Makefile to this script so that it can call the same version of
3404 A note about Exim versions 4.44 and 4.50
3405 ----------------------------------------
3407 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3408 changes of various kinds. As a consequence, a big documentation update was
3409 needed. This delayed the release for rather longer than seemed good, especially
3410 in the light of a couple of (minor) security issues. Therefore, the changes
3411 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3412 release. So 4.44 and 4.50 are in effect two different branches that both start
3415 I have left the 4.50 change log unchanged; it contains all the changes since
3416 4.43. The change log for 4.44 is below; many of its items are identical to
3417 those for 4.50. This seems to be the most sensible way to preserve the
3418 historical information.
3424 1. Minor wording change to the doc/README.SIEVE file.
3426 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3427 computation of the current number of files was incorrect.
3429 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3430 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3433 4. Give more explanation in the error message when the command for a transport
3434 filter fails to execute.
3436 5. There are several places where Exim runs a non-Exim command in a
3437 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3438 was being done only for the command run by the queryprogram router. It is
3439 now done for all such subprocesses. The other cases are: ${run, transport
3440 filters, and the commands run by the lmtp and pipe transports.
3442 6. Added CONFIGURE_GROUP build-time option.
3444 7. Some older OS have a limit of 256 on the maximum number of file
3445 descriptors. Exim was using setrlimit() to set 1000 as a large value
3446 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3447 systems. I've change it so that if it can't get 1000, it tries for 256.
3449 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3450 was an oversight, and furthermore, ever since the addition of extra
3451 controls (e.g. 4.43/32), the checks on when to allow different forms of
3452 "control" were broken. There should now be diagnostics for all cases when a
3453 control that does not make sense is encountered.
3455 9. Added the /retain_sender option to "control=submission".
3457 10. $recipients is now available in the predata ACL (oversight).
3459 11. Tidy the search cache before the fork to do a delivery from a message
3460 received from the command line. Otherwise the child will trigger a lookup
3461 failure and thereby defer the delivery if it tries to use (for example) a
3462 cached ldap connection that the parent has called unbind on.
3464 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3465 of $address_data from the recipient verification was clobbered by the
3466 sender verification.
3468 13. The value of address_data from a sender verification is now available in
3469 $sender_address_data in subsequent conditions in the ACL statement.
3471 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3473 15. Added a new option "connect=<time>" to callout options, to set a different
3476 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3477 was its contents. (It was OK if the option was not defined at all.)
3479 17. A "Completed" log line is now written for messages that are removed from
3480 the spool by the -Mrm option.
3482 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3483 information about exactly what failed.
3485 19. Added -dd to debug only the daemon process.
3487 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3488 handles timeouts, both on the server side and network timeouts. Renamed the
3489 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3491 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3492 It is now set to "smtps".
3494 22. $host_address is now set to the target address during the checking of
3495 ignore_target_hosts.
3497 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3498 being passed; this would have caused $sender_host_name to have been used if
3499 matching the list had actually called for a host name (not very likely,
3500 since this list is usually IP addresses). A host name is now passed as
3503 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3504 code that specifies a non-restarting handler (typically sigaction() in
3505 modern systems) in an attempt to fix a rare and obscure crash bug.
3507 25. Narrowed the window for a race in the daemon that could cause it to ignore
3508 SIGCHLD signals. This is not a major problem, because they are used only to
3509 wake it up if nothing else does.
3511 26. A malformed maildirsize file could cause Exim to calculate negative values
3512 for the mailbox size or file count. Odd effects could occur as a result.
3513 The maildirsize information is now recalculated if the size or filecount
3516 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3517 support for a long time. Removed HAVE_SYS_VFS_H.
3519 28. Installed the latest version of exipick from John Jetmore.
3521 29. In an address list, if the pattern was not a regular expression, an empty
3522 subject address (from a bounce message) matched only if the pattern was an
3523 empty string. Non-empty patterns were not even tested. This was the wrong
3524 because it is perfectly reasonable to use an empty address as part of a
3525 database query. An empty address is now tested by patterns that are
3526 lookups. However, all the other forms of pattern expect the subject to
3527 contain a local part and a domain, and therefore, for them, an empty
3528 address still always fails if the pattern is not itself empty.
3530 30. Exim went into a mad DNS loop when attempting to do a callout where the
3531 host was specified on an smtp transport, and looking it up yielded more
3532 than one IP address.
3534 31. Re-factored the code for checking spool and log partition space into a
3535 function that finds that data and another that does the check. The former
3536 is then used to implement four new variables: $spool_space, $log_space,
3537 $spool_inodes, and $log_inodes.
3539 32. The RFC2047 encoding function was originally intended for short strings
3540 such as real names; it was not keeping to the 75-character limit for
3541 encoded words that the RFC imposes. It now respects the limit, and
3542 generates multiple encoded words if necessary. To be on the safe side, I
3543 have increased the buffer size for the ${rfc2047: expansion operator from
3546 33. It is now permitted to omit both strings after an "if" condition; if the
3547 condition is true, the result is "true". As before, when the second string
3548 is omitted, a false condition yields an empty string. This makes it less
3549 cumbersome to write custom ACL and router conditions.
3551 34. Failure to deliver a bounce message always caused it to be frozen, even if
3552 there was an errors_to setting on the router. The errors_to setting is now
3555 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3556 canonical form (fully expanded) before being placed in
3557 $sender_host_address.
3559 36. The table in the code that translates DNS record types into text (T_A to
3560 "A" for instance) was missing entries for NS and CNAME. It is just possible
3561 that this could have caused confusion if both these types were looked up
3562 for the same domain, because the text type is used as part of Exim's
3563 per-process caching. But the chance of anyone hitting this buglet seems
3566 37. The dnsdb lookup has been extended in a number of ways.
3568 (1) There is a new type, "zns", which walks up the domain tree until it
3569 finds some nameserver records. It should be used with care.
3571 (2) There is a new type, "mxh", which is like "mx" except that it returns
3572 just the host names, not the priorities.
3574 (3) It is now possible to give a list of domains (or IP addresses) to be
3575 looked up. The behaviour when one of the lookups defers can be
3576 controlled by a keyword.
3578 (4) It is now possible to specify the separator character for use when
3579 multiple records are returned.
3581 38. The dnslists ACL condition has been extended: it is now possible to supply
3582 a list of IP addresses and/or domains to be looked up in a particular DNS
3585 39. Added log_selector=+queue_time_overall.
3587 40. When running the queue in the test harness, wait just a tad after forking a
3588 delivery process, to get repeatability of debugging output.
3590 41. Include certificate and key file names in error message when GnuTLS fails
3591 to set them up, because the GnuTLS error message doesn't include the name
3592 of the failing file when there is a problem reading it.
3594 42. Allow both -bf and -bF in the same test run.
3596 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3598 44. The "Exiscan patch" is now merged into the mainline Exim source.
3600 45. Sometimes the final signoff response after QUIT could fail to get
3601 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3602 < 0 before doing a fflush(). This bug looks as though it goes back to the
3603 introduction of TLS in release 3.20, but "sometimes" must have been rare
3604 because the tests only now provoked it.
3606 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3607 (this can affect the format of dates).
3609 47. exim_tidydb, when checking for the continued existence of a message for
3610 which it has found a message-specific retry record, was not finding
3611 messages that were in split spool directories. Consequently, it was
3612 deleting retry records that should have stayed in existence.
3614 48. Steve fixed some bugs in eximstats.
3616 49. The SPA authentication driver was not abandoning authentication and moving
3617 on to the next authenticator when an expansion was forced to fail,
3618 contradicting the general specification for all authenticators. Instead it
3619 was generating a temporary error. It now behaves as specified.
3621 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3622 (the order specifies the preference for clients). The order is now AES256,
3623 AES128, 3DES, ARCFOUR128.
3625 51. Small patch to Sieve code - explicitly set From: when generating an
3628 52. Exim crashed if a remote delivery caused a very long error message to be
3629 recorded - for instance if somebody sent an entire SpamAssassin report back
3630 as a large number of 550 error lines. This bug was coincidentally fixed by
3631 increasing the size of one of Exim's internal buffers (big_buffer) that
3632 happened as part of the Exiscan merge. However, to be on the safe side, I
3633 have made the code more robust (and fixed the comments that describe what
3636 53. Now that there can be additional text after "Completed" in log lines (if
3637 the queue_time_overall log selector is set), a one-byte patch to exigrep
3638 was needed to allow it to recognize "Completed" as not the last thing in
3641 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3642 patch that reportedly fixes this has been added. I am not expert enough to
3643 create a test for it. This is what the patch creator wrote:
3645 "I found a little strange behaviour of ldap code when working with
3646 Windows 2003 AD Domain, where users was placed in more than one
3647 Organization Units. When I tried to give exim partial DN, the exit code
3648 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3649 But simultaneously result of request was absolutely normal ldap result,
3650 so I produce this patch..."
3652 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3653 REFERENCE, so I have modified the code to exclude the patch when that macro
3656 55. Some experimental protocols are using DNS PTR records for new purposes. The
3657 keys for these records are domain names, not reversed IP addresses. The
3658 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3659 leaves it alone. Component reversal etc. now happens only for IP addresses.
3662 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3664 57. Double the size of the debug message buffer (to 2048) so that more of very
3665 long debug lines gets shown.
3667 58. The exicyclog utility now does better if the number of log files to keep
3668 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3670 59. Two changes related to the smtp_active_hostname option:
3672 (1) $smtp_active_hostname is now available as a variable.
3673 (2) The default for smtp_banner uses $smtp_active_hostname instead
3674 of $primary_hostname.
3676 60. The host_aton() function is supposed to be passed a string that is known
3677 to be a valid IP address. However, in the case of IPv6 addresses, it was
3678 not checking this. This is a hostage to fortune. Exim now panics and dies
3679 if the condition is not met. A case was found where this could be provoked
3680 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3681 components; fortuitously, this particular loophole had already been fixed
3682 by change 4.50/55 above.
3684 If there are any other similar loopholes, the new check in host_aton()
3685 itself should stop them being exploited. The report I received stated that
3686 data on the command line could provoke the exploit when Exim was running as
3687 exim, but did not say which command line option was involved. All I could
3688 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3689 running as the user.
3692 61. There was a buffer overflow vulnerability in the SPA authentication code
3693 (which came originally from the Samba project). I have added a test to the
3694 spa_base64_to_bits() function which I hope fixes it.
3697 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3698 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3700 63. The daemon start-up calls getloadavg() while still root for those OS that
3701 need the first call to be done as root, but it missed one case: when
3702 deliver_queue_load_max is set with deliver_drop_privilege. This is
3703 necessary for the benefit of the queue runner, because there is no re-exec
3704 when deliver_drop_privilege is set.
3706 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3707 This has been fixed.
3709 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3710 was in use, was not putting the data itself into the right store pool;
3711 consequently, it could be overwritten for a subsequent message in the same
3712 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3715 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3717 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3718 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3719 the function were treating the return as a boolean value, which happened to
3720 work because 0=false and not-0=true, but is not correct code.
3722 68. The host_aton() function was not handling scoped IPv6 addresses (those
3723 with, for example, "%eth0" on the end) correctly.
3725 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3726 negated items (that is, ~something) in unsigned ints. Some compilers
3727 apparently mutter when there is no cast.
3729 70. If an address verification called from an ACL failed, and did not produce a
3730 user-specific message (i.e. there was only a "system" message), nothing was
3731 put in $acl_verify_message. In this situation, it now puts the system
3734 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3735 session; change 4.31/43 added the unwanted input to the log line - except
3736 that it did not do this in the start of session case. It now does.
3738 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3739 This is wrong and can cause the other end to generate a synchronization
3740 error if it is another Exim or anything else that does the synchronization
3741 check. A QUIT command is no longer sent after a timeout.
3743 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3744 during host lookups.
3746 74. The defer_ok option of callout verification was not working if it was used
3747 when verifying addresses in header lines, that is, for this case:
3749 verify = header_sender/callout=defer_ok
3751 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3752 those file descriptors could be used for SMTP connections. If anything
3753 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3754 could be sent to the SMTP client, causing chaos. The daemon now opens
3755 stdin, stdout, and stderr to /dev/null when it puts itself into the
3758 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3759 log by default. The user can override this with suitable Perl magic.
3761 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3762 the log message when discard triggers, was not working for the DATA ACL or
3763 for the non-SMTP ACL.
3765 78. Error message wording change in sieve.c.
3767 79. If smtp_accept_max_per_host was set, the number of connections could be
3768 restricted to fewer than expected, because the daemon was trying to set up
3769 a new connection before checking whether the processes handling previous
3770 connections had finished. The check for completed processes is now done
3771 earlier. On busy systems, this bug wouldn't be noticed because something
3772 else would have woken the daemon, and it would have reaped the completed
3775 80. If a message was submitted locally by a user whose login name contained one
3776 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3777 It caused a spool format error. I have fixed the spool reading code. A
3778 related problem was that the "from" clause in the Received: line became
3779 illegal because of the space(s). It is now covered by ${quote_local_part.
3781 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3784 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3786 83. Patch from Sieve maintainer to fix unterminated string problem in
3787 "vacation" handling.
3789 84. Some minor changes to the Linux configuration files to help with other
3790 OS variants using glibc.
3792 85. One more patch for Sieve to update vacation handling to latest spec.
3795 ----------------------------------------------------
3796 See the note above about the 4.44 and 4.50 releases.
3797 ----------------------------------------------------
3803 1. Change 4.43/35 introduced a bug that caused file counts to be
3804 incorrectly computed when quota_filecount was set in an appendfile
3807 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3808 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3811 3. Give more explanation in the error message when the command for a transport
3812 filter fails to execute.
3814 4. There are several places where Exim runs a non-Exim command in a
3815 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3816 was being done only for the command run by the queryprogram router. It is
3817 now done for all such subprocesses. The other cases are: ${run, transport
3818 filters, and the commands run by the lmtp and pipe transports.
3820 5. Some older OS have a limit of 256 on the maximum number of file
3821 descriptors. Exim was using setrlimit() to set 1000 as a large value
3822 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3823 systems. I've change it so that if it can't get 1000, it tries for 256.
3825 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3826 was an oversight, and furthermore, ever since the addition of extra
3827 controls (e.g. 4.43/32), the checks on when to allow different forms of
3828 "control" were broken. There should now be diagnostics for all cases when a
3829 control that does not make sense is encountered.
3831 7. $recipients is now available in the predata ACL (oversight).
3833 8. Tidy the search cache before the fork to do a delivery from a message
3834 received from the command line. Otherwise the child will trigger a lookup
3835 failure and thereby defer the delivery if it tries to use (for example) a
3836 cached ldap connection that the parent has called unbind on.
3838 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3839 of $address_data from the recipient verification was clobbered by the
3840 sender verification.
3842 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3843 was its contents. (It was OK if the option was not defined at all.)
3845 11. A "Completed" log line is now written for messages that are removed from
3846 the spool by the -Mrm option.
3848 12. $host_address is now set to the target address during the checking of
3849 ignore_target_hosts.
3851 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3852 being passed; this would have caused $sender_host_name to have been used if
3853 matching the list had actually called for a host name (not very likely,
3854 since this list is usually IP addresses). A host name is now passed as
3857 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3858 code that specifies a non-restarting handler (typically sigaction() in
3859 modern systems) in an attempt to fix a rare and obscure crash bug.
3861 15. Narrowed the window for a race in the daemon that could cause it to ignore
3862 SIGCHLD signals. This is not a major problem, because they are used only to
3863 wake it up if nothing else does.
3865 16. A malformed maildirsize file could cause Exim to calculate negative values
3866 for the mailbox size or file count. Odd effects could occur as a result.
3867 The maildirsize information is now recalculated if the size or filecount
3870 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3871 support for a long time. Removed HAVE_SYS_VFS_H.
3873 18. Updated exipick to current release from John Jetmore.
3875 19. Allow an empty sender to be matched against a lookup in an address list.
3876 Previously the only cases considered were a regular expression, or an
3879 20. Exim went into a mad DNS lookup loop when doing a callout where the
3880 host was specified on the transport, if the DNS lookup yielded more than
3883 21. The RFC2047 encoding function was originally intended for short strings
3884 such as real names; it was not keeping to the 75-character limit for
3885 encoded words that the RFC imposes. It now respects the limit, and
3886 generates multiple encoded words if necessary. To be on the safe side, I
3887 have increased the buffer size for the ${rfc2047: expansion operator from
3890 22. Failure to deliver a bounce message always caused it to be frozen, even if
3891 there was an errors_to setting on the router. The errors_to setting is now
3894 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3895 canonical form (fully expanded) before being placed in
3896 $sender_host_address.
3898 24. Updated eximstats to version 1.33
3900 25. Include certificate and key file names in error message when GnuTLS fails
3901 to set them up, because the GnuTLS error message doesn't include the name
3902 of the failing file when there is a problem reading it.
3904 26. Expand error message when OpenSSL has problems setting up cert/key files.
3907 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3908 (this can affect the format of dates).
3910 28. exim_tidydb, when checking for the continued existence of a message for
3911 which it has found a message-specific retry record, was not finding
3912 messages that were in split spool directories. Consequently, it was
3913 deleting retry records that should have stayed in existence.
3915 29. eximstats updated to version 1.35
3916 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3917 1.35 - bugfix such that pie charts by volume are generated correctly
3919 30. The SPA authentication driver was not abandoning authentication and moving
3920 on to the next authenticator when an expansion was forced to fail,
3921 contradicting the general specification for all authenticators. Instead it
3922 was generating a temporary error. It now behaves as specified.
3924 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3925 (the order specifies the preference for clients). The order is now AES256,
3926 AES128, 3DES, ARCFOUR128.
3928 31. Small patch to Sieve code - explicitly set From: when generating an
3931 32. Exim crashed if a remote delivery caused a very long error message to be
3932 recorded - for instance if somebody sent an entire SpamAssassin report back
3933 as a large number of 550 error lines. This bug was coincidentally fixed by
3934 increasing the size of one of Exim's internal buffers (big_buffer) that
3935 happened as part of the Exiscan merge. However, to be on the safe side, I
3936 have made the code more robust (and fixed the comments that describe what
3939 33. Some experimental protocols are using DNS PTR records for new purposes. The
3940 keys for these records are domain names, not reversed IP addresses. The
3941 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3942 leaves it alone. Component reversal etc. now happens only for IP addresses.
3945 34. The host_aton() function is supposed to be passed a string that is known
3946 to be a valid IP address. However, in the case of IPv6 addresses, it was
3947 not checking this. This is a hostage to fortune. Exim now panics and dies
3948 if the condition is not met. A case was found where this could be provoked
3949 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3950 components; fortuitously, this particular loophole had already been fixed
3951 by change 4.50/55 or 4.44/33 above.
3953 If there are any other similar loopholes, the new check in host_aton()
3954 itself should stop them being exploited. The report I received stated that
3955 data on the command line could provoke the exploit when Exim was running as
3956 exim, but did not say which command line option was involved. All I could
3957 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3958 running as the user.
3961 35. There was a buffer overflow vulnerability in the SPA authentication code
3962 (which came originally from the Samba project). I have added a test to the
3963 spa_base64_to_bits() function which I hope fixes it.
3966 36. The daemon start-up calls getloadavg() while still root for those OS that
3967 need the first call to be done as root, but it missed one case: when
3968 deliver_queue_load_max is set with deliver_drop_privilege. This is
3969 necessary for the benefit of the queue runner, because there is no re-exec
3970 when deliver_drop_privilege is set.
3972 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3973 was in use, was not putting the data itself into the right store pool;
3974 consequently, it could be overwritten for a subsequent message in the same
3975 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3978 38. Sometimes the final signoff response after QUIT could fail to get
3979 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3980 < 0 before doing a fflush(). This bug looks as though it goes back to the
3981 introduction of TLS in release 3.20, but "sometimes" must have been rare
3982 because the tests only now provoked it.
3988 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3989 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3990 Now it returns an integer. A number of calls were still expecting a T/F
3991 return. Fortuitously, in all cases, the tests worked in OK situations,
3992 which is the norm. However, things would have gone wrong on any write
3993 failures on the smtp file descriptor. This function is used when sending
3994 messages over SMTP and also when doing verify callouts.
3996 2. When Exim is called to do synchronous delivery of a locally submitted
3997 message (the -odf or -odi options), it no longer closes stderr before doing
4000 3. Implemented the mua_wrapper option.
4002 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4004 5. Implemented the functions header_remove(), header_testname(),
4005 header_add_at_position(), and receive_remove_recipient(), and exported them
4008 6. If an ACL "warn" statement specified the addition of headers, Exim already
4009 inserted X-ACL-Warn: at the start if there was no header name. However, it
4010 was not making this test for the second and subsequent header lines if
4011 there were newlines in the string. This meant that an invalid header could
4012 be inserted if Exim was badly configured.
4014 7. Allow an ACL "warn" statement to add header lines at the start or after all
4015 the Received: headers, as well as at the end.
4017 8. Added the rcpt_4xx retry error code.
4019 9. Added postmaster_mailfrom=xxx to callout verification option.
4021 10. Added mailfrom=xxxx to the callout verification option, for verify=
4024 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4025 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4027 12. Inserted some casts to stop certain compilers warning when using pointer
4028 differences as field lengths or precisions in printf-type calls (mostly
4029 affecting debugging statements).
4031 13. Added optional readline() support for -be (dynamically loaded).
4033 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4034 same clock tick as a message's arrival, so that its received time was the
4035 same as the "first fail" time on the retry record, and that message
4036 remained on the queue past the ultimate address timeout, every queue runner
4037 would try a delivery (because it was past the ultimate address timeout) but
4038 after another failure, the ultimate address timeout, which should have then
4039 bounced the address, did not kick in. This was a "< instead of <=" error;
4040 in most cases the first failure would have been in the next clock tick
4041 after the received time, and all would be well.
4043 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4044 being recognized when the domain list was tested by the match_domain
4045 condition in an expansion string.
4047 16. Added the ${str2b64: operator.
4049 17. Exim was always calling setrlimit() to set a large limit for the number of
4050 processes, without checking whether the existing limit was already
4051 adequate. (It did check for the limit on file descriptors.) Furthermore,
4052 errors from getrlimit() and setrlimit() were being ignored. Now they are
4053 logged to the main and panic logs, but Exim does carry on, to try to do its
4054 job under whatever limits there are.
4056 18. Imported PCRE 5.0.
4058 19. Trivial typo in log message " temporarily refused connection" (the leading
4061 20. If the log selector return_path_on_delivery was set and an address was
4062 redirected to /dev/null, the delivery process crashed because it assumed
4063 that a return path would always be set for a "successful" delivery. In this
4064 case, the whole delivery is bypassed as an optimization, and therefore no
4067 21. Internal re-arrangement: the function for sending a challenge and reading
4068 a response while authentication was assuming a zero-terminated challenge
4069 string. It's now changed to take a pointer and a length, to allow for
4070 binary data in such strings.
4072 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4074 23. Exim was not respecting finduser_retries when seeking the login of the
4075 uid under which it was called; it was always trying 10 times. (The default
4076 setting of finduser_retries is zero.) Also, it was sleeping after the final
4077 failure, which is pointless.
4079 24. Implemented tls_on_connect_ports.
4081 25. Implemented acl_smtp_predata.
4083 26. If the domain in control=submission is set empty, Exim assumes that the
4084 authenticated id is a complete email address when it generates From: or
4085 Sender: header lines.
4087 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4088 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4089 chown and chgrp in /bin and hostname in /usr/bin.
4091 28. Exim was keeping the "process log" file open after each use, just as it
4092 does for the main log. This opens the possibility of it remaining open for
4093 long periods when the USR1 signal hits a daemon. Occasional processlog
4094 errors were reported, that could have been caused by this. Anyway, it seems
4095 much more sensible not to leave this file open at all, so that is what now
4098 29. The long-running daemon process does not normally write to the log once it
4099 has entered its main loop, and it closes the log before doing so. This is
4100 so that log files can straightforwardly be renamed and moved. However,
4101 there are a couple of unusual error situations where the daemon does write
4102 log entries, and I had neglected to close the log afterwards.
4104 30. The text of an SMTP error response that was received during a remote
4105 delivery was being truncated at 512 bytes. This is too short for some of
4106 the long messages that one sometimes sees. I've increased the limit to
4109 31. It is now possible to make retry rules that apply only when a message has a
4110 specific sender, in particular, an empty sender.
4112 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4113 it possible to be selective about when SMTP synchronization is enforced.
4115 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4117 32. Implemented hosts_connection_nolog.
4119 33. Added an ACL for QUIT.
4121 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4124 35. Added mailbox_size and mailbox_filecount to appendfile.
4126 36. Added control = no_multiline_responses to ACLs.
4128 37. There was a bug in the logic of the code that waits for the clock to tick
4129 in the case where the clock went backwards by a substantial amount such
4130 that the microsecond fraction of "now" was more than the microsecond
4131 fraction of "then" (but the whole seconds number was less).
4133 38. Added support for the libradius Radius client library this is found on
4134 FreeBSD (previously only the radiusclient library was supported).
4140 1. When certain lookups returned multiple values in the form name=value, the
4141 quoting of the values was not always being done properly. Specifically:
4142 (a) If the value started with a double quote, but contained no whitespace,
4144 (b) If the value contained whitespace other than a space character (i.e.
4145 tabs or newlines or carriage returns) it was not quoted.
4146 This fix has been applied to the mysql and pgsql lookups by writing a
4147 separate quoting function and calling it from the lookup code. The fix
4148 should probably also be applied to nisplus, ibase and oracle lookups, but
4149 since I cannot test any of those, I have not disturbed their existing code.
4151 2. A hit in the callout cache for a specific address caused a log line with no
4152 reason for rejecting RCPT. Now it says "Previous (cached) callout
4153 verification failure".
4155 3. There was an off-by-one bug in the queryprogram router. An over-long
4156 return line was truncated at 256 instead of 255 characters, thereby
4157 overflowing its buffer with the terminating zero. As well as fixing this, I
4158 have increased the buffer size to 1024 (and made a note to document this).
4160 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4161 when Exim is waiting for an SMTP response from a remote server, Exim
4162 restarts its select() call on the socket, thereby resetting its timeout.
4163 This is not a problem when such interrupts are rare. Somebody set up a cron
4164 job to run exiwhat every 2 minutes, which is less than the normal select()
4165 timeout (5 or 10 minutes). This meant that the select() timeout never
4166 kicked in because it was always reset. I have fixed this by comparing the
4167 time when an interrupt arrives with the time at the start of the first call
4168 to select(). If more time than the timeout has elapsed, the interrupt is
4169 treated as a timeout.
4171 5. Some internal re-factoring in preparation for the addition of Sieve
4172 extensions (by MH). In particular, the "personal" test is moved to a
4173 separate function, and given an option for scanning Cc: and Bcc: (which is
4174 not set for Exim filters).
4176 6. When Exim created an email address using the login of the caller as the
4177 local part (e.g. when creating a From: or Sender: header line), it was not
4178 quoting the local part when it contained special characters such as @.
4180 7. Installed new OpenBSD configuration files.
4182 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4183 try to make them clearer.
4185 9. Callout options, other than the timeout value, were being ignored when
4186 verifying sender addresses in header lines. For example, when using
4188 verify = header_sender/callout=no_cache
4190 the cache was (incorrectly) being used.
4192 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4193 only the Cygwin environment.
4195 11. When return_path_on_delivery was set as a log selector, if different remote
4196 addresses in the same message used different return paths and parallel
4197 remote delivery occurred, the wrong values would sometimes be logged.
4198 (Whenever a remote delivery process finished, the return path value from
4199 the most recently started remote delivery process was logged.)
4201 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4202 lines when AUTH and/or TLS are in use. This is the "received protocol"
4203 field. Exim used to use "asmtp" for authenticated SMTP, without any
4204 indication (in the protocol name) for TLS use. Now it follows the RFC and
4205 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4206 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4207 names appear in log lines as well as in Received: header lines.
4209 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4210 extensions, and comparison tests, and to fix some bugs.
4212 14. Changes to the "personal" filter test:
4214 (1) The test was buggy in that it was just doing the equivalent of
4215 "contains" tests on header lines. For example, if a user's address was
4216 anne@some.where, the "personal" test would incorrectly be true for
4218 To: susanne@some.where
4220 This test is now done by extracting each address from the header in turn,
4221 and checking the entire address. Other tests that are part of "personal"
4222 are now done using regular expressions (for example, to check local parts
4223 of addresses in From: header lines).
4225 (2) The list of non-personal local parts in From: addresses has been
4226 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4227 taken from the Sieve specification recommendations.
4229 (3) If the message contains any header line starting with "List-" it is
4230 treated as non-personal.
4232 (4) The test for "circular" in the Subject: header line has been removed
4233 because it now seems ill-conceived.
4235 15. Minor typos in src/EDITME comments corrected.
4237 16. Installed latest exipick from John Jetmore.
4239 17. If headers_add on a router specified a text string that was too long for
4240 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4241 of string_sprintf() is now avoided.
4243 18. $message_body_size was not set (it was always zero) when running the DATA
4244 ACL and the local_scan() function.
4246 19. For the "mail" command in an Exim filter, no default was being set for
4247 the once_repeat time, causing a random time value to be used if "once" was
4248 specified. (If the value happened to be <= 0, no repeat happened.) The
4249 default is now 0s, meaning "never repeat". The "vacation" command was OK
4250 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4251 (I found it when inspecting the code).
4253 20. There is now an overall timeout for performing a callout verification. It
4254 defaults to 4 times the callout timeout, which applies to individual SMTP
4255 commands during the callout. The overall timeout applies when there is more
4256 than one host that can be tried. The timeout is checked before trying the
4257 next host. This prevents very long delays if there are a large number of
4258 hosts and all are timing out (e.g. when the network connections are timing
4259 out). The value of the overall timeout can be changed by specifying an
4260 additional sub-option for "callout", called "maxwait". For example:
4262 verify = sender/callout=5s,maxwait=20s
4264 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4265 to the end before writing, but this should make it even safer).
4267 22. Exim was forgetting that it had advertised PIPELINING for the second and
4268 subsequent messages on an SMTP connection. It was also not resetting its
4269 memory on STARTTLS and an internal HELO.
4271 23. When Exim logs an SMTP synchronization error within a session, it now
4272 records whether PIPELINING has been advertised or not.
4274 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4275 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4276 rather than long int.
4278 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4280 26. Added the never_mail option to autoreply.
4286 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4287 crash if the getsockname() call failed; this can happen if a connection is
4288 closed very soon after it is established. The problem was simply in the
4289 order in which certain operations were done, causing Exim to try to write
4290 to the SMTP stream before it had set up the file descriptor. The bug has
4291 been fixed by making things happen in the correct order.
4297 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4298 before the connection was closed, thus losing the rejection response.
4300 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4301 some early Solaris releases, but causes trouble in current releases where
4302 socklen_t is defined.
4304 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4307 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4310 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4311 that releasing the top of it at the end releases what was used for sub-
4312 expansions (unless the block got too big). However, discard this block if
4313 the first thing is a variable or header, so that we can use its block when
4314 it is dynamic (useful for very large $message_headers, for example).
4316 6. Lookups now cache *every* query, not just the most recent. A new, separate
4317 store pool is used for this. It can be recovered when all lookup caches are
4318 flushed. Lookups now release memory at the end of their result strings.
4319 This has involved some general refactoring of the lookup sources.
4321 7. Some code has been added to the store_xxx() functions to reduce the amount
4322 of flapping under certain conditions.
4324 8. log_incoming_interface used to affect only the <= reception log lines. Now
4325 it causes the local interface and port to be added to several more SMTP log
4326 lines, for example "SMTP connection from", and rejection lines.
4328 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4330 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4332 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4333 could be overwritten at the end of the current message (or the start of a
4334 new message if it was set in a HELO ACL). The value is now preserved for
4335 the duration of the SMTP connection.
4337 12. If a transport had a headers_rewrite setting, and a matching header line
4338 contained an unqualified address, that address was qualified, even if it
4339 did not match any rewriting rules. The underlying bug was that the values
4340 of the flags that permit the existence of unqualified sender and recipient
4341 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4342 non-local messages, and by -bnq for local messages) were not being
4343 preserved with the message after it was received.
4345 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4346 "next input=" as part of the text comprising the host identity instead of
4347 the correct text. The code was using the same buffer for two different
4348 strings. However, depending on which order the printing function evaluated
4349 its arguments, the bug did not always show up. Under Linux, for example, my
4350 test suite worked just fine.
4352 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4353 doesn't work with some older versions of Perl. It has been changed to "my",
4354 which in any case is probably the better facility to use.
4356 15. A really picky compiler found some instances of statements for creating
4357 error messages that either had too many or two few arguments for the format
4360 16. The size of the buffer for calls to the DNS resolver has been increased
4361 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4362 for addresses that have a lot of PTR records. This alleviates a problem; it
4363 does not fully solve it.
4365 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4366 buffer now truncates the list and logs the incident, which is the same
4367 action as happens when Exim is looking up a host name and its aliases.
4368 Previously in this situation something unpredictable would happen;
4369 sometimes it was "internal error: store_reset failed".
4371 18. If a server dropped the connection unexpectedly when an Exim client was
4372 using GnuTLS and trying to read a response, the client delivery process
4373 crashed while trying to generate an error log message.
4375 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4376 string, for example:
4378 warn message = H1: something\nH2: something
4380 the text was added as a single header line from Exim's point of view
4381 though it ended up OK in the delivered message. However, searching for the
4382 second and subsequent header lines using $h_h2: did not work. This has been
4383 fixed. Similarly, if a system filter added multiple headers in this way,
4384 the routers could not see them.
4386 20. Expanded the error message when iplsearch is called with an invalid key to
4387 suggest using net-iplsearch in a host list.
4389 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4390 ACLs are no longer actually imposed (and a message to that effect is
4393 22. If a "gecos" field in a passwd entry contained escaped characters, in
4394 particular, if it contained a \" sequence, Exim got it wrong when building
4395 a From: or a Sender: header from that name. A second bug also caused
4396 incorrect handling when an unquoted " was present following a character
4397 that needed quoting.
4399 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4400 was not being matched caselessly.
4402 24. Arranged for all hyphens in the exim.8 source to be escaped with
4405 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4406 properly. Recipient callout cache records were still being keyed to include
4407 the sender, even when use_sender was set false. This led to far more
4408 callouts that were necessary. The sender is no longer included in the key
4409 when use_sender is false.
4411 26. Added "control = submission" modifier to ACLs.
4413 27. Added the ${base62d: operator to decode base 62 numbers.
4415 28. dnsdb lookups can now access SRV records.
4417 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4418 the configuration file.
4420 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4421 (-v) mode. This makes the output for a verbose queue run more intelligible.
4423 31. Added a use_postmaster feature to recipient callouts.
4425 32. Added the $body_zerocount variable, containing the number of binary zero
4426 bytes in the message body.
4428 33. The time of last modification of the "new" subdirectory is now used as the
4429 "mailbox time last read" when there is a quota error for a maildir
4432 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4434 35. Added +ignore_unknown as a special item in host lists.
4436 36. Code for decoding IPv6 addresses in host lists is now included, even if
4437 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4438 address was recognized as an IP address, but was then not correctly decoded
4439 into binary, causing unexpected and incorrect effects when compared with
4446 1. Very minor rewording of debugging text in manualroute to say "list of
4447 hosts" instead of "hostlist".
4449 2. If verify=header_syntax was set, and a header line with an unqualified
4450 address (no domain) and a large number of spaces between the end of the
4451 name and the colon was received, the reception process suffered a buffer
4452 overflow, and (when I tested it) crashed. This was caused by some obsolete
4453 code that should have been removed. The fix is to remove it!
4455 3. When running in the test harness, delay a bit after writing a bounce
4456 message to get a bit more predictability in the log output.
4458 4. Added a call to search_tidyup() just before forking a reception process. In
4459 theory, someone could use a lookup in the expansion of smtp_accept_max_
4460 per_host which, without the tidyup, could leave open a database connection.
4462 5. Added the variables $recipient_data and $sender_data which get set from a
4463 lookup success in an ACL "recipients" or "senders" condition, or a router
4464 "senders" option, similar to $domain_data and $local_part_data.
4466 6. Moved the writing of debug_print from before to after the "senders" test
4469 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4470 problems for message scanning, either using a data ACL, or using
4471 local_scan() because the Received: header was not generated till after they
4472 were called (in order to set the time as the time of reception completion).
4473 I have revised the way this works. The header is now generated after the
4474 body is received, but before the ACL or local_scan() are called. After they
4475 are run, the timestamp in the header is updated.
4481 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4482 before starting a queue runner without re-exec. This happened only when
4483 deliver_drop_privilege was set or when the Exim user was set to root. The
4484 effect of the bug was that timeouts during subsequent deliveries caused
4485 crashes instead of being properly handled. The handler is now left at its
4486 default (and expected) setting.
4488 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4489 message, again when deliver_drop_privilege is set or Exim is run as root.
4490 The bug described in (1) was not present in this case, but the tidying up
4491 of the other signals was missing. I have made the two cases consistent.
4493 3. The ignore_target_hosts setting on a manualroute router was being ignored
4494 for hosts that were looked up using the /MX notation.
4496 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4499 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4500 operated on the sender address. After changing the $sender_address to <>
4501 for the sender address verify, Exim was re-instated it as the original
4502 (before rewriting) address, but remembering that it had rewritten it, so it
4503 wasn't rewriting it again. This bug also had the effect of breaking the
4504 sender address verification caching when the sender address was rewritten.
4506 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4507 This has been changed so that if the ip literal address matches
4508 ignore_target_hosts, the router declines.
4510 7. Added expansion conditions match_domain, match_address, and match_local_
4511 part (NOT match_host).
4513 8. The placeholder for the Received: header didn't have a length field set.
4515 9. Added code to Exim itself and to exim_lock to test for a specific race
4516 condition that could lead to file corruption when using MBX delivery. The
4517 issue is with the lockfile that is created in /tmp. If this file is removed
4518 after a process has opened it but before that process has acquired a lock,
4519 there is the potential for a second process to recreate the file and also
4520 acquire a lock. This could lead to two Exim processes writing to the file
4521 at the same time. The added code performs the same test as UW imapd; it
4522 checks after acquiring the lock that its file descriptor still refers to
4523 the same named file.
4525 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4526 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4529 11. Added the smtp_active_hostname option. If used, this will typically be made
4530 to depend on the incoming interface address. Because $interface_address is
4531 not set up until the daemon has forked a reception process, error responses
4532 that can happen earlier (such as "too many connections") no longer contain
4535 12. If an expansion in a condition on a "warn" statement fails because a lookup
4536 defers, the "warn" statement is abandoned, and the next ACL statement is
4537 processed. Previously this caused the whole ACL to be aborted.
4539 13. Added the iplsearch lookup type.
4541 14. Added ident_timeout as a log selector.
4543 15. Added tls_certificate_verified as a log selector.
4545 16. Added a global option tls_require_ciphers (compare the smtp transport
4546 option of the same name). This controls incoming TLS connections.
4548 17. I finally figured out how to make tls_require_ciphers do a similar thing
4549 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4550 before starting the TLS session.
4552 18. Tabs are now shown as \t in -bP output.
4554 19. If the log selector return_path_on_delivery was set, Exim crashed when
4555 bouncing a message because it had too many Received: header lines.
4557 20. If two routers both had headers_remove settings, and the first one included
4558 a superfluous trailing colon, the final name in the first list and the
4559 first name in the second list were incorrectly joined into one item (with a
4560 colon in the middle).
4566 1. Added -C and -D options to the exinext utility, mainly to make it easier
4567 to include in the automated testing, but these could be helpful when
4568 multiple configurations are in use.
4570 2. The exinext utility was not formatting the output nicely when there was
4571 an alternate port involved in the retry record key, nor when there was a
4572 message id as well (for retries that were specific to a specific message
4573 and a specific host). It was also confused by IPv6 addresses, because of
4574 the additional colons they contain. I have fixed the IPv4 problem, and
4575 patched it up to do a reasonable job for IPv6.
4577 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4578 delivery, the log line now contains "pipelined" if PIPELINING was used.
4580 4. An SMTP transport process used to panic and die if the bind() call to set
4581 an explicit outgoing interface failed. This has been changed; it is now
4582 treated in the same way as a connect() failure.
4584 5. A reference to $sender_host_name in the part of a conditional expansion
4585 that was being skipped was still causing a DNS lookup. This no longer
4588 6. The def: expansion condition was not recognizing references to header lines
4589 that used bh_ and bheader_.
4591 7. Added the _cache feature to named lists.
4593 8. The code for checking quota_filecount in the appendfile transport was
4594 allowing one more file than it should have been.
4596 9. For compatibility with Sendmail, the command line option
4604 and sets the incoming protocol and host name (for trusted callers). The
4605 host name and its colon can be omitted when only the protocol is to be set.
4606 Note the Exim already has two private options, -pd and -ps, that refer to
4607 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4608 "s", but I don't think that's a major issue.
4610 10. A number of refactoring changes to the code, none of which should affect
4613 (a) The number of logging options was getting close to filling up the
4614 32-bit word that was used as a bit map. I have split them into two classes:
4615 those that are passed in the argument to log_write(), and those that are
4616 only ever tested independently outside of that function. These are now in
4617 separate 32-bit words, so there is plenty of room for expansion again.
4618 There is no change in the user interface or the logging behaviour.
4620 (b) When building, for example, log lines, the code previously used a
4621 macro that called string_cat() twice, in order to add two strings. This is
4622 not really sufficiently general. Furthermore, there was one instance where
4623 it was actually wrong because one of the argument was used twice, and in
4624 one call a function was used. (As it happened, calling the function twice
4625 did not affect the overall behaviour.) The macro has been replaced by a
4626 function that can join an arbitrary number of extra strings onto a growing
4629 (c) The code for expansion conditions now uses a table and a binary chop
4630 instead of a serial search (which was left over from when there were very
4631 few conditions). Also, it now recognizes conditions like "pam" even when
4632 the relevant support is not compiled in: a suitably worded error message is
4633 given if an attempt is made to use such a condition.
4635 11. Added ${time_interval:xxxxx}.
4637 12. A bug was causing one of the ddress fields not to be passed back correctly
4638 from remote delivery subprocesses. The field in question was not being
4639 subsequently used, so this caused to problems in practice.
4641 13. Added new log selectors queue_time and deliver_time.
4643 14. Might have fixed a bug in maildirsizefile handling that threw up
4644 "unexpected character" debug warnings, and recalculated the data
4645 unnecessarily. In any case, I expanded the warning message to give more
4648 15. Added the message "Restricted characters in address" to the statements in
4649 the default ACL that block characters like @ and % in local parts.
4651 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4652 Three changes have been made:
4654 (a) There was a serious bug; a negative response to MAIL caused the whole
4655 recipient domain to be cached as invalid, thereby blocking all messages
4656 to all local parts at the same domain, from all senders. This bug has
4657 been fixed. The domain is no longer cached after a negative response to
4658 MAIL if the sender used is not empty.
4660 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4663 (c) A new callout option, "use_sender" has been added for people who want
4664 the modified behaviour.
4670 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4673 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4674 indeed breaks things for older releases.
4676 3. Added additional logging to the case where there is a problem reading data
4677 from a filter that is running in a subprocess using a pipe, in order to
4678 try to track down a specific problem.
4680 4. Testing facility fudge: when running in the test harness and attempting
4681 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4682 getting "No route to host". Convert this to a timeout.
4684 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4687 6. Some OS don't have socklen_t but use size_t instead. This affects the
4688 fifth argument of getsockopt() amongst other things. This is now
4689 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4690 can be set for individual OS. I have set it for SunOS5, OSF1, and
4691 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4692 some earlier ones do not.
4694 7. Change 4.30/15 was not doing the test caselessly.
4696 8. The standard form for an IPv6 address literal was being rejected by address
4697 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4698 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4699 this, as well as the form without the "IPv6" on the front (but only when
4700 address literals are enabled, of course).
4702 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4704 10. Exim crashed if a message with an empty sender address specified by -f
4705 encountered a router with an errors_to setting. This could be provoked only
4706 by a command such as
4710 where an empty string was supplied; "<>" did not hit this bug.
4712 11. Installed PCRE release 4.5.
4714 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4715 remained set. It is now erased.
4717 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4718 times from message ids (which are base 36 rather than the normal 62).
4720 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4721 were being counted as actual protocol errors, and logged if the log
4722 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4723 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4724 and DATA following a set of rejected RCPTs do not count as protocol errors.
4725 In other words, Exim assumes they were pipelined, though this may not
4726 actually be the case. Of course, in all cases the client gets an
4727 appropriate error code.
4729 15. If a lookup fails in an ACL condition, a message about the failure may
4730 be available; it is used if testing the ACL cannot continue, because most
4731 such messages specify what the cause of the deferral is. However, some
4732 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4733 that caused an old message to be retained and used if a later statement
4734 caused a defer, replacing the real cause of the deferral.
4736 16. If an IP address had so many PTR records that the DNS lookup buffer
4737 was not large enough to hold them, Exim could crash while trying to process
4738 the truncated data. It now detects and logs this case.
4740 17. Further to 4.21/58, another change has been made: if (and only if) the
4741 first line of a message (the first header line) ends with CRLF, a bare LF
4742 in a subsequent header line has a space inserted after it, so as not to
4743 terminate the header.
4745 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4746 unnecessarily, used atoi() instead of strtol(), and didn't check the
4747 termination when getting file sizes from file names by regex.
4749 19. Completely re-implemented the support for maildirsize files, in the light
4750 of a number of problems with the previous contributed implementation
4751 (4.30/29). In particular:
4753 . If the quota is zero, the maildirsize file is maintained, but no quota is
4756 . If the maildir directory does not exist, it is created before any attempt
4757 to write a maildirsize file.
4759 . The quota value in the file is just a cache; if the quota is changed in
4760 the transport, the new value overrides.
4762 . A regular expression is available for excluding directories from the
4765 20. The autoreply transport checks the characters in options that define the
4766 message's headers; it allows continued headers, but it was checking with
4767 isspace() after an embedded newline instead of explicitly looking for a
4770 21. If all the "regular" hosts to which an address was routed had passed their
4771 expiry times, and had not reached their retry times, the address was
4772 bounced, even if fallback hosts were defined. Now Exim should go on to try
4775 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4776 equivalent code in the SMTP transport. Some hosts send humungous responses
4777 to HELO/EHLO, more than 1024 it seems.
4779 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4780 gives compiler warnings in some environments. I've now done it "properly",
4783 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4784 (because of problems with the built-in one) was declared to return uschar *
4785 instead of char *, causing compiler failure.
4787 25. Fixed a file descriptor leak when processing alias/forward files.
4789 26. Fixed a minor format string issue in dbfn.c.
4791 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4793 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4794 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4797 29. When a sender address is verified, it is cached, to save repeating the test
4798 when there is more than one recipient in a message. However, when the
4799 verification involves a callout, it is possible for different callout
4800 options to be set for different recipients. It is too complicated to keep
4801 track of this in the cache, so now Exim always runs a verification when a
4802 callout is required, relying on the callout cache for the optimization.
4803 The overhead is duplication of the address routing, but this should not be
4806 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4807 to be verified with callout=postmaster, and the main callout worked but the
4808 postmaster check failed, the verification correctly failed. However, if a
4809 subsequent RCPT command asked for sender verification *without* the
4810 postmaster check, incorrect caching caused this verification also to fail,
4813 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4814 it was not caching the DNS options (qualify_single, search_parents) that
4815 were used when the lookup failed. A subsequent lookup with different
4816 options therefore always gave the same answer, though there were cases
4817 where it should not have. (Example: a "domains = !$mx_any" option on a
4818 dnslookup router: the "domains" option is always processed without any
4819 widening, but the router might have qualify_single set.) Now Exim uses the
4820 cached value only when the same options are set.
4822 32. Added John Jetmore's "exipick" utility to the distribution.
4824 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4825 than a timeout (e.g. a certificate is required, and is not provided), an
4826 Exim server now closes the connection immediately. Previously it waited for
4827 the client to close - but if the client is SSL, it seems that they each
4828 wait for each other, leading to a delay before one of them times out.
4830 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4831 maintained 0.8.x compatibility because I don't think many are using it, and
4832 it is clearly obsolete.
4834 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4837 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4838 client certificate was expired. A simple patch fixes this, though I don't
4839 understand the full logic of why the verify callback is called multiple
4842 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4843 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4844 which causes problems with some clients (such as the Certicom SSL Plus
4845 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4846 disables the coutermeasure allowing Eudora to connect."
4848 38. Exim was not checking that a write() to a log file succeeded. This could
4849 lead to Bad Things if a log got too big, in particular if it hit a file
4850 size limit. Exim now panics and dies if it cannot write to a log file, just
4851 as it does if it cannot open a log file.
4853 39. Modified OS/Makefile-Linux so that it now contains
4855 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4857 The two -D definitions ensure that Exim is compiled with large file
4858 support, which makes it possible to handle log files that are bigger than
4861 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4862 instance) a domain was checked against a named list that involved a lookup,
4863 causing $domain_data to be set, then another domain was checked against the
4864 same list, then the first domain was re-checked, the value of $domain_data
4865 after the final check could be wrong. In particular, if the second check
4866 failed, it could be set empty. This bug probably also applied to
4869 41. The strip_trailing_dot option was not being applied to the address given
4870 with the -f command-line option.
4872 42. The code for reading a message's header from the spool was incrementing
4873 $received_count, but never initializing it. This meant that the value was
4874 incorrect (doubled) while delivering a message in the same process in which
4875 it was received. In the most common configuration of Exim, this never
4876 happens - a fresh exec is done - but it can happen when
4877 deliver_drop_privilege is set.
4879 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4880 it now includes up to 150 characters of the unexpected data in the log
4883 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4884 and building data strings. The size of both of these buffers was 10 000
4885 bytes - far larger than anybody would *ever* want, thought I. Needless to
4886 say, somebody hit the limit. I have increased the maximum line length to
4887 20 000 and the maximum data length of concatenated lines to 100 000. I have
4888 also fixed two bugs, because there was no checking on these buffers. Tsk,
4889 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4890 buffer is too small.
4892 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4893 lsearch lookups. Now it does.
4895 46. When parsing a route_list item in a manualroute router, a fixed-length
4896 buffer was used for the list of hosts. I made this 1024 bytes long,
4897 thinking that nobody would ever have a list of hosts that long. Wrong.
4898 Somebody had a whole pile of complicated expansion conditions, and the
4899 string was silently truncated, leading to an expansion error. It turns out
4900 that it is easier to change to an unlimited length (owing to other changes
4901 that have happened since this code was originally written) than to build
4902 structure for giving a limitation error. The length of the item that
4903 expands into the list of hosts is now unlimited.
4905 47. The lsearch lookup could not handle data where the length of text line was
4906 more than 4095 characters. Such lines were truncated, leading to shortened
4907 data being returned. It should now handle lines of any length.
4909 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4910 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4913 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4914 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4915 no longer happens in comment lines. A list of replacements is now placed
4916 at the head of all of the source files, except those whose only change is
4917 to replace PERL_COMMAND in the very first #! line.
4919 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4920 messages on the queue, with a bottom-up merge sort, using code contributed
4921 by Michael Haardt. This should make operations like -bp somewhat faster on
4922 large queues. It won't affect queue runners, except when queue_run_in_order
4925 51. Installed eximstats 1.31 in the distribution.
4927 52. Added support for SRV lookups to the dnslookup router.
4929 53. If an ACL referred to $message_body or $message_body_end, the value was not
4930 reset for any messages that followed in the same SMTP session.
4932 54. The store-handling optimization for building very long strings was not
4933 differentiating between the different store pools. I don't think this
4934 actually made any difference in practice, but I've tidied it.
4936 55. While running the routers to verify a sender address, $sender_address
4937 was still set to the sender address. This is wrong, because when routing to
4938 send a bounce to the sender, it would be empty. Therefore, I have changed
4939 it so that, while verifying a sender address, $sender_address is set to <>.
4940 (There is no change to what happens when verifying a recipient address.)
4942 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4943 target A or AAAA records (if not already returned) without resetting the
4944 qualify_single or search_parents options of the DNS resolver. These are
4945 inappropriate in this case because the targets of MX and SRV records must
4946 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4947 target that, when qualified, matched something in the local domain. These
4948 two options are now turned off when doing these lookups.
4950 57. It seems that at least some releases of Reiserfs (which does not have the
4951 concept of a fixed number of inodes) returns zero and not -1 for the
4952 number of available inodes. This interacted badly with check_spool_inodes,
4953 which assumed that -1 was the "no such thing" setting. What I have done is
4954 to check that the total number of inodes is greater than zero before doing
4955 the test of how many are available.
4957 58. When a "warn" ACL statement has a log_message modifier, the message is
4958 remembered, and not repeated. This is to avoid a lot of repetition when a
4959 message has many recipients that cause the same warning to be written.
4960 Howewer, Exim was preserving the list of already written lines for an
4961 entire SMTP session, which doesn't seem right. The memory is now reset if a
4962 new message is started.
4964 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4965 debugging output unless log_rewrite was also set.
4967 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4968 of (int)(handle) when we know that handle contains (void *)(-1).
4970 61. The Exim daemon panic-logs an error return when it closes the incoming
4971 connection. However "connection reset by peer" seems to be common, and
4972 isn't really an error worthy of noting specially, so that particular error
4975 62. When Exim is trying to find all the local interfaces, it used to panic and
4976 die if the ioctl to get the interface flags failed. However, it seems that
4977 on at least one OS (Solaris 9) it is possible to have an interface that is
4978 included in the list of interfaces, but for which you get a failure error
4979 for this call. This happens when the interface is not "plumbed" into a
4980 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4981 failure of the "get flags" call assumes that the interface is down.
4983 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4984 makes life easier for people who are doing arithmetic on fields extracted
4985 from dates, where you often get leading zeros that should not be
4986 interpreted as octal.
4988 64. Added qualify_domain to the redirect router, to override the global
4991 65. If a pathologically long header line contained very many addresses (the
4992 report of this problem mentioned 10 000) and each of them was rewritten,
4993 Exim could use up a very large amount of memory. (It kept on making new
4994 copies of the header line as it rewrote, and never released the old ones.)
4995 At the expense of a bit more processing, the header rewriting function has
4996 been changed so that it no longer eats memory in this way.
4998 66. The generation of the Received: header has been moved from the time that a
4999 message starts to be received, to the time that it finishes. The timestamp
5000 in the Received: header should now be very close to that of the <= log
5001 line. There are two side-effects of this change:
5003 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5004 logged header lines no longer include the local Received: line, because
5005 it has not yet been created. The same applies to a copy of the message
5006 that is returned to a non-SMTP sender when a message is rejected.
5008 (b) When a filter file is tested using -bf, no additional Received: header
5009 is added to the test message. After some thought, I decided that this
5012 This change does not affect the value of $received_for. It is still set
5013 after address rewriting, but before local_scan() is called.
5015 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5017 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5018 gave an unhelpful panic error message, and a defer error. I have managed to
5019 change this behaviour so that it now rejects any supplied certificate,
5020 which seems right, as the list of acceptable certificates is empty.
5022 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5023 gave an unhelpful defer error. I have not managed to make this reject any
5024 supplied certificates, but the error message it gives is "no certificate
5025 supplied", which is not helpful.
5027 70. exigrep's output now also includes lines that are not associated with any
5028 message, but which match the given pattern. Implemented by a patch from
5029 Martin Sluka, which also tidied up the Perl a bit.
5031 71. Recipient callout verification, like sender verification, was using <> in
5032 the MAIL FROM command. This isn't really the right thing, since the actual
5033 sender may affect whether the remote host accepts the recipient or not. I
5034 have changed it to use the actual sender in the callout; this means that
5035 the cache record is now keyed on a recipient/sender pair, not just the
5036 recipient address. There doesn't seem to be a real danger of callout loops,
5037 since a callout by the remote host to check the sender would use <>.
5038 [SEE ABOVE: changed after hitting problems.]
5040 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5041 temporary errors. However, in the case of such a code being given after
5042 the end of a data transmission (i.e. after ".") Exim was failing to write
5043 a retry record for the message. (Yes, there was some broken host that was
5044 actually sending 8xx at this point.)
5046 73. An unknown lookup type in a host list could cause Exim to panic-die when
5047 the list was checked. (An example that provoked this was putting <; in the
5048 middle of a list instead of at the start.) If this happened during a DATA
5049 ACL check, a -D file could be left lying around. This kind of configuration
5050 error no longer causes Exim to die; instead it causes a defer errror. The
5051 incident is still logged to the main and panic logs.
5053 74. Buglet left over from Exim 3 conversion. The message "too many messages
5054 in one connection" was written to the rejectlog but not the mainlog, except
5055 when address rewriting (yes!) was being logged.
5057 75. Added write_rejectlog option.
5059 76. When a system filter was run not as root (that is, when system_filter_user
5060 was set), the values of the $n variables were not being returned to the
5061 main process; thus, they were not subsequently available in the $sn
5064 77. Added +return_path_on_delivery log selector.
5066 78. A connection timeout was being treated differently from recipients deferred
5067 when testing hosts_max_try with a message that was older than the host's
5068 retry timeout. (The host should not be counted, thus allowing all hosts to
5069 be tried at least once before bouncing.) This may have been the cause of an
5070 occasionally reported bug whereby a message would remain on the queue
5071 longer than the retry timeout, but would be bounced if a delivery was
5072 forced. I say "may" because I never totally pinned down the problem;
5073 setting up timeout/retry tests is difficult. See also the next item.
5075 79. The ultimate address timeout was not being applied to errors that involved
5076 a combination of host plus message (for example, a timeout on a MAIL
5077 command). When an address resolved to a number of possible hosts, and they
5078 were not all tried for each delivery (e.g. because of hosts_max_try), a
5079 message could remain on the queue longer than the retry timeout.
5081 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5084 81. Fixed an obscure SMTP outgoing bug which required at least the following
5085 conditions: (a) there was another message waiting for the same server;
5086 (b) the server returned 5xx to all RCPT commands in the first message so
5087 that the message was not completed; (c) the server dropped the connection
5088 or gave a negative response to the RSET that Exim sends to abort the
5089 transaction. The observed case was a dropped connection after DATA that had
5090 been sent in pipelining mode. That is, the server had advertised PIPELINING
5091 but was not implementing it correctly. The effect of the bug was incorrect
5092 behaviour, such as trying another host, and this could lead to a crash.
5098 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5099 and daemon.c were passed as pointers to ints; they should have been
5100 pointers to socklen_t variables (which are typically unsigned ints).
5102 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5105 3. Fixed a really odd bug that affected only the testing scheme; patching a
5106 certain fixed string in the binary changed the value of another string that
5107 happened to be identical to the end of the original first string.
5109 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5110 name", it returns that address as the IP address. On some operating
5111 systems (e.g. Solaris), it also passes back the IP address string as the
5112 "host name". However, on others (e.g. Linux), it passes back an empty
5113 string. Exim wasn't checking for this, and was changing the host name to an
5114 empty string, assuming it had been canonicized.
5116 5. Although rare, it is permitted to have more than one PTR record for a given
5117 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5118 all the names associated with an address, because they do in Solaris.
5119 However, it seems that they do not in Linux for data that comes from the
5120 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5121 I found this out when I moved to a new Linux workstation and tried to run
5122 the Exim test suite.
5124 To get round this problem I have changed the code so that it now does its
5125 own call to the DNS to look up PTR records when searching for a host name.
5126 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5127 addresses that are only in /etc/hosts are still found.
5129 This behaviour is, however, controlled by an option called host_lookup_
5130 order, which defaults to "bydns:byaddr". If people want to use the other
5131 order, or indeed, just use one or the other means of lookup, they can
5132 specify it in this variable.
5134 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5135 some operating systems, this comes back from gethostbyaddr() as an empty
5136 string, and this is what Exim used to test for. However, it seems that in
5137 other systems, "." is yielded. Exim now tests for this case too.
5139 7. The values of check_spool_space and check_log_space are now held internally
5140 as a number of kilobytes instead of an absolute number of bytes. If a
5141 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5142 kilobyte. This means that much larger values can be stored.
5144 8. Exim monitor: an attempt to get the action menu when not actually pointing
5145 at a message produces an empty menu entitled "No message selected". This
5146 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5147 no entries in it ("Shell widget menu has zero width and/or height"). So I
5148 have added a single, blank menu entry in this case.
5150 9. Added ${quote_local_part.
5152 10. MIME decoding is now applied to the contents of Subject: header lines when
5155 11. Now that a reference to $sender_host_address automatically causes a reverse
5156 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5157 host lookup before query-style lookups in lists that might use this
5158 variable. This has therefore been abolished, and the "net-" prefix is no
5159 longer necessary for query-style lookups.
5161 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5162 have been a typo for LFLAGS, so it has been changed.
5164 13. The install script calls Exim with "-C /dev/null" in order to find the
5165 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5166 to be output. Howeve, since Exim outputs its version number before the
5167 error, it didn't break the script. It just looked ugly. I fixed this by
5168 always allowing "-C /dev/null" if the caller is root.
5170 14. Ignore overlarge ACL variable number when reading spool file - insurance
5171 against a later release with more variables having written the file.
5173 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5174 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5175 this, as well as the form without the "IPv6" on the front.
5177 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5178 OS/Makefile-Darwin file.
5180 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5181 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5183 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5184 variable. However, some OS use ints for inodes. I've added cast to long int
5185 to get rid of the compiler warning.
5187 19. I had forgotten to lock out "/../" in configuration file names when
5188 ALT_CONFIG_PREFIX was set.
5190 20. Routers used for verification do not need to specify transports. However,
5191 if such a router generated a host list, and callout was configured, Exim
5192 crashed, because it could not find a port number from the (non-existent)
5193 transport. It now assumes port 25 in this circumstance.
5195 21. Added the -t option to exigrep.
5197 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5198 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5199 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5200 them is set, LOOKUP_LSEARCH is forced.
5202 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5204 24. Added sender and host information to the "rejected by local_scan()" log
5205 line; previously there was no indication of these.
5207 25. Added .include_if_exists.
5209 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5210 Linux. It turns out that not all file systems support this. Apparently some
5211 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5212 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5213 not supported on the file descriptor, is now ignored when Exim is trying to
5214 sync a directory. This applies only to Linux.
5216 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5218 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5219 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5222 29. Added support for maildirsize files from supplied patch (modified a bit).
5224 30. The use of :fail: followed by an empty string could lead Exim to respond to
5225 sender verification failures with (e.g.):
5227 550 Verification failed for <xxx>
5228 550 Sender verify failed
5230 where the first response line was missing the '-' that indicates it is not
5231 the final line of the response.
5233 31. The loop for finding the name of the user that called Exim had a hardwired
5234 limit of 10; it now uses the value of finduser_retries, which is used for
5235 all other user lookups.
5237 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5240 33. Exim was neglecting to zero errno before one call of strtol() when
5241 expanding a string and expecting an integer value. On some systems this
5242 resulted in spurious "integer overflow" errors. Also, it was casting the
5243 result into an int without checking.
5245 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5246 did not work. The code looks as if it has *never* worked, though it appears
5247 to have been documented since at least releast 1.62. I have made it work.
5249 35. The "timeout_DNS" error in retry rules, also documented since at least
5250 1.62, also never worked. As it isn't clear exactly what this means, and
5251 clearly it isn't a major issue, I have abolished the feature by treating it
5252 as "timeout", and writing a warning to the main and panic logs.
5254 36. The display of retry rules for -brt wasn't always showing the error code
5257 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5258 timeout_connect_A, timeout_connect_MX.
5260 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5261 to the empty sender.
5263 39. The daemon was not analysing the content of -oX till after it had closed
5264 stderr and disconnected from the controlling terminal. This meant that any
5265 syntax errors were only noted on the panic log, and the return code from
5266 the command was 0. By re-arranging the code a little, I've made the
5267 decoding happen first, so such errors now appear on stderr, and the return
5268 code is 1. However, the actual setting up of the sockets still happens in
5269 the disconnected process, so errors there are still only recorded on the
5272 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5273 connections (as happens on some IP stacks) was logged at start up time as
5274 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5275 it from "IPv6 and IPv4", which means that two separate sockets are being
5278 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5279 says whether AF_INET or AF_INET6 was passed as an argument.
5281 42. Exiwhat output was messed up when time zones were included in log
5284 43. Exiwhat now gives more information about the daemon's listening ports,
5285 and whether -tls-on-connect was used.
5287 44. The "port" option of the smtp transport is now expanded.
5289 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5290 silently ignored. Now an error message is written to the main and panic
5293 46. There's a new ACL modifier called "logwrite" which writes to a log file
5294 as soon as it is encountered.
5296 47. Added $local_user_uid and $local_user_gid at routing time.
5298 48. Exim crashed when trying to verify a sender address that was being
5301 49. Exim was recognizing only a space character after ".include". It now also
5302 recognizes a tab character.
5304 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5305 extracting the relevant information from the specification. The man page no
5306 longer contains scrambled data for the -d option, and I've added a section
5307 at the front about calling Exim under different names.
5309 51. Added "extra_headers" argument to the "mail" command in filter files.
5311 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5314 53. Installed eximstats 1.29.
5316 54. Added transport_filter_timeout as a generic transport option.
5318 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5319 Cc: header lines. This was required by RFC 822, but it not required by RFC
5322 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5323 incoming messages that did not have them. Now it does so only if the
5324 message originates locally, that is, if there is no associated remote host
5325 address. When Resent- header lines are present, this applies to the Resent-
5326 lines rather than the non-Resent- lines.
5328 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5329 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5331 58. Messages for configuration errors now include the name of the main
5332 configuration file - useful now that there may be more than one file in a
5333 list (.included file names were always shown).
5335 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5336 for those rare installations that do not start the daemon as root or run it
5337 setuid root. I've cut out the call to initgroups() if the daemon is not
5340 60. The Exim user and group can now be bound into the binary as text strings
5341 that are looked up at the start of Exim's processing.
5343 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5345 62. Added $mailstore_basename variable.
5347 63. Installed patch to sieve.c from Michael Haardt.
5349 64. When Exim failed to open the panic log after failing to open the main log,
5350 the original message it was trying to log was written to stderr and debug
5351 output, but if they were not available (the usual case in production), it
5352 was lost. Now it is written to syslog before the two lines that record the
5353 failures to open the logs.
5355 65. Users' Exim filters run in subprocesses under the user's uid. It is
5356 possible for a "deliver" command or an alias in a "personal" command to
5357 provoke an address rewrite. If logging of address rewriting is configured,
5358 this fails because the process is not running as root or exim. There may be
5359 a better way of dealing with this, but for the moment (because 4.30 needs
5360 to be released), I have disabled address rewrite logging when running a
5361 filter in a non-root, non-exim process.
5367 1. The buildconfig auxiliary program wasn't quoting the value set for
5368 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5369 not defined. This bug was masked in 4.22 by the effect that was fixed in
5372 2. Some messages that were rejected after a message id was allocated were
5373 shown as "incomplete" by exigrep. It no longer does this for messages that
5374 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5376 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5377 have allow_domain_literals set, the ID did not get logged in the <= line.
5378 Domain literals are now always recognized in Message-ID: header lines.
5380 4. The first argument for a ${extract expansion item is the key name or field
5381 number. Leading and trailing spaces in this item were not being ignored,
5382 causing some misleading effects.
5384 5. When deliver_drop_privilege was set, single queue runner processes started
5385 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5386 same command in the process it spins off) were not dropping privilege.
5388 6. When the daemon running as "exim" started a queue runner, it always
5389 re-executed Exim in the spun-off process. This is a waste of effort when
5390 deliver_drop_privilege is set. The new process now just calls the
5391 queue-runner function directly.
5397 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5400 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5401 ignored. Though the use of -oP was forcing the writing of a pid file, it
5402 was always written to the default place.
5404 3. If the message "no IP address found for host xxxx" is generated during
5405 incoming verification, it is now followed by identification of the incoming
5406 connection (so you can more easily find what provoked it).
5408 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5410 5. Added some features to "harden" Exim a bit more against certain attacks:
5412 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5413 be put in Local/Makefile. This is like the never_users runtime option,
5414 but it cannot be overridden. The default setting is "root".
5416 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5417 prefix string with which any file named in a -C command line option
5420 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5421 is retained for -C and -D only if the caller of Exim is root. Without
5422 it, the exim user may also use -C and -D and retain privilege.
5424 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5425 command line option is disabled.
5427 6. Macro names set by the -D option must start with an upper case letter, just
5428 like macro names defined in the configuration file.
5430 7. Added "dereference=" facility to LDAP.
5432 8. Two instances of the typo "uknown" in the source files are fixed.
5434 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5435 the Configure-Makefile script screwed up while processing it.
5437 10. Incorporated PCRE 4.4.
5439 11. The SMTP synchronization check was not operating right at the start of an
5440 SMTP session. For example, it could not catch a HELO sent before the client
5441 waited for the greeting. There is now a check for outstanding input at the
5442 point when the greeting is written. Because of the duplex, asynchronous
5443 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5444 way, but not yet received, when the check is performed.
5446 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5447 on TCP/IP sockets, because this apparently causes some broken clients to
5450 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5451 unchanged) from the Cygwin maintainer.
5453 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5454 was supported instead of testing for mbx. Effectively a typo.
5456 15. The spa authenticator server code was not checking that the input it
5457 received was valid base64.
5459 16. The debug output line for the "set" modifier in ACLs was not showing the
5460 name of the variable that was being set.
5462 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5464 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5465 lookup on its own. Something else was needed to trigger the lookup. For
5466 example, a match in host_lookup or the need for a host name in a host list.
5467 Now, if $sender_host_name is referenced and the host name has not yet been
5468 looked up, a lookup is performed. If the lookup fails, the variable remains
5469 empty, and $host_lookup_failed is set to "1".
5471 19. Added "eqi" as a case-independent comparison operator.
5473 20. The saslauthd authentication condition could segfault if neither service
5474 nor realm was specified.
5476 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5477 error message that was logged was misleading, and incoming SMTP
5478 connections were dropped. The message is now more accurate, and temporary
5479 errors are given to SMTP connections.
5481 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5482 (except RSET) with a 421 error, until QUIT is received. However, it was
5483 failing to send a response to QUIT.
5485 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5486 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5487 ACL is now run after the helo_try_verify_hosts code.
5489 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5490 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5491 case-independent, so other case variants are also recognized). Apparently
5492 some systems use these upper case variants.
5494 25. If more than two messages were waiting for the same host, and a transport
5495 filter was specified for the transport, Exim sent two messages over the
5496 same TCP/IP connection, and then failed with "socket operation on non-
5497 socket" when it tried to send the third.
5499 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5501 27. The extern definition of crypt16() in expand.c was not being excluded when
5502 the OS had its own crypt16() function.
5504 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5505 as a preferred synonym for return_size_limit, both as an option and as an
5508 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5510 30. Changed the default configuration ACL to relax the local part checking rule
5511 for addresses that are not in any local domains. For these addresses,
5512 slashes and pipe symbols are allowed within local parts, but the sequence
5513 /../ is explicitly forbidden.
5515 31. SPA server authentication was not clearing the challenge buffer before
5518 32. log_message in a "warn" ACL statement was writing to the reject log as
5519 well as to the main log, which contradicts the documentation and doesn't
5520 seem right (because no rejection is happening). So I have stopped it.
5522 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5523 However, I am unable to do any testing of this.
5525 34. Fixed an infelicity in the appendfile transport. When checking directories
5526 for a mailbox, to see if any needed to be created, it was accidentally
5527 using path names with one or more superfluous leading slashes; tracing
5528 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5530 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5531 contents are added to the log line that is written for every discarded
5532 recipient. (Previously a log_message setting was ignored.)
5534 36. The ${quote: operator now quotes the string if it is empty.
5536 37. The install script runs exim in order to find its version number. If for
5537 some reason other than non-existence or emptiness, which it checks, it
5538 could not run './exim', it was installing it with an empty version number,
5539 i.e. as "exim-". This error state is now caught, and the installation is
5542 38. An argument was missing from the function that creates an error message
5543 when Exim fails to connect to the socket for saslauthd authentication.
5544 This could cause Exim to crash, or give a corrupted message.
5546 39. Added isip, isip4, and isip6 to ${if conditions.
5548 40. The ACL variables $acl_xx are now saved with the message, and can be
5549 accessed later in routers, transports, and filters.
5551 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5552 strings in the file are not string-expanded.
5554 42. If a MAIL command specified a SIZE value that was too large to fit into an
5555 int variable, the check against message_size_limit failed. Such values are
5556 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5557 day this will have to be increased, but I don't think I want to be around
5558 when emails are that large.
5565 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5566 iconv() is not standard in FreeBSD.
5568 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5569 IPv6 enabled. The observed symptom was a segmentation fault on return from
5570 the function os_common_find_running_interfaces() in src/os.c.
5572 3. In the check_special_case() function in daemon.c I had used "errno" as an
5573 argument name, which causes warnings on some systems. This was basically a
5574 typo, since it was named "eno" in the comments!
5576 4. The code that waits for the clock to tick (at a resolution of some fraction
5577 of a second) so as to ensure message-id uniqueness was always waiting for
5578 at least one whole tick, when it could have waited for less. [This is
5579 almost certainly not relevant at current processor speeds, where it is
5580 unlikely to ever wait at all. But we try to future-proof.]
5582 5. The function that sleeps for a time interval that includes fractions of a
5583 second contained a race. It did not block SIGALRM between setting the
5584 timer, and suspending (a couple of lines later). If the interval was short
5585 and the sigsuspend() was delayed until after it had expired, the suspension
5586 never ended. On busy systems this could lead to processes getting stuck for
5589 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5590 process, before it forks any delivery processes. The open lookup caching
5591 mechanism meant that the open file or database connection was passed into
5592 the delivery process. The problem was that delivery processes always tidy
5593 up cached lookup data. This could cause a problem for the next delivery
5594 process started by the queue runner, because the external queue runner
5595 process does not know about the closure. So the next delivery process
5596 still has data in the lookup cache. In the case of a file lookup, there was
5597 no problem because closing a file descriptor in a subprocess doesn't affect
5598 the parent. However, if the lookup was caching a connection to a database,
5599 the connection was closed, and the second delivery process was likely to
5600 see errors such as "PGSQL: query failed: server closed the connection
5601 unexpectedly". The problem has been fixed by closing all cached lookups
5602 in a queue runner before running a delivery process.
5604 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5605 seem to have the "const" qualifier which it has on other OS. I've
5608 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5609 *of the same type* (client or server) with the same public name that an
5610 error should be diagnosed.
5612 9. When Exim looked up a host name for an IP address, but failed to find the
5613 original IP address when looking up the host name (a safety check), it
5614 output the message "<ip address> does not match any IP for NULL", which was
5615 confusing, to say the least. The bug was that the host name should have
5616 appeared instead of "NULL".
5618 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5619 user that is built into the binary, and the -C or -D options is used, root
5620 privilege is dropped before the configuration file is read. In addition,
5621 logging is switched to stderr instead of the normal log files. If the
5622 configuration then re-defines the Exim user, the unprivileged environment
5623 is probably not what is expected, so Exim logs a panic warning message (but
5626 However, if deliver_drop_privilege is set, the unprivileged state may well
5627 be exactly what is intended, so the warning has been cut out in that case,
5628 and Exim is allowed to try to write to its normal log files.
5634 1. smtp_return_error_details was not giving details for temporary sender
5635 or receiver verification errors.
5637 2. Diagnose a configuration error if two authenticators have the same public
5640 3. Exim used not to create the message log file for a message until the first
5641 delivery attempt. This could be confusing when incoming messages were held
5642 for policy or load reasons. The message log file is now created at the time
5643 the message is received, and an initial "Received" line is written to it.
5645 4. The automatically generated man page for command line options had a minor
5646 bug that caused no ill effects; however, a more serious problem was that
5647 the procedure for building the man page automatically didn't always
5648 operate. Consequently, release 4.20 contains an out-of-date version. This
5649 shouldn't happen again.
5651 5. When building Exim with embedded Perl support, the script that builds the
5652 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5653 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5655 6. The freeze_tell option was not being used for messages that were frozen on
5656 arrival, either by an ACL or by local_scan().
5658 7. Added the smtp_incomplete_transaction log selector.
5660 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5661 was accepting AUTH without a new EHLO.
5663 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5664 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5667 10. Logging of TCP/IP connections (when configured) now happens in the main
5668 daemon process instead of the child process, so that the TCP/IP connection
5669 count is more accurate (but it can never be perfect).
5671 11. The use of "drop" in a nested ACL was not being handled correctly in the
5672 outer ACL. Now, if condition failure induced by the nested "drop" causes
5673 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5674 or "require"), the connection is dropped.
5676 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5677 that yield "discard" can now be used with an "accept" or a "discard" verb,
5678 but an error is generated for any others (because I can't see a useful way
5679 to define what should happen).
5681 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5682 are now processed in the same way as lines in the Exim configuration file.
5683 In particular, continuation lines are supported.
5685 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5687 15. Added -ti meaning -t -i.
5689 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5690 domains, and warn by logging if others are found.
5692 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5693 returned from GIFCONF when Exim is trying to find the list of interfaces on
5694 a host. The code in os.c has been modified to copy each ifreq to an aligned
5695 structure in all cases.
5697 Also, in some cases, the returned ifreq's were being copied to a 'struct
5698 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5699 means the last couple of bytes of an IPv6 address could be chopped if the
5700 ifreq contained only a normal sockaddr (14 bytes storage).
5702 18. Named domain lists were not supported in the hosts_treat_as_local option.
5703 An entry such as +xxxx was not recognized, and was treated as a literal
5706 19. Ensure that header lines added by a DATA ACL are included in the reject log
5707 if the ACL subsequently rejects the message.
5709 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5710 MD5 (which is deprecated).
5712 21. When testing a filter file using -bf, Exim was writing a message when it
5713 took the sender from a "From " line in the message, but it was not doing so
5714 when it took $return_path from a Return-Path: header line. It now does.
5716 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5717 with a valid header line field name (a series of printing characters
5718 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5720 23. Changed "disc" in the source to "disk" to conform to the documentation and
5721 the book and for uniformity.
5723 24. Ignore Sendmail's -Ooption=value command line item.
5725 25. When execve() failed while trying to run a command in a pipe transport,
5726 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5727 could be confused with a return value of 69 from the command itself. This
5728 has been changed to 127, the value the shell returns if it is asked to run
5729 a non-existent command. The wording for the related log line suggests a
5730 non-existent command as the problem.
5732 26. If received_header_text expands to an empty string, do not add a Received:
5733 header line to the message. (Well, it adds a token one on the spool, but
5734 marks it "old" so that it doesn't get used or transmitted.)
5736 27. Installed eximstats 1.28 (addition of -nt option).
5738 28. There was no check for failure on the call to getsockname() in the daemon
5739 code. This can fail if there is a shortage of resources on the system, with
5740 ENOMEM, for example. A temporary error is now given on failure.
5742 29. Contrary to the C standard, it seems that in some environments, the
5743 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5744 program. Exim now does this explicitly; it affects the formatting of
5745 timestamps using strftime().
5747 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5748 complaints. I've now initialized all the variables, to avoid this.
5750 32. Header lines added by a system filter were not being "seen" during
5751 transport-time rewrites.
5753 33. The info_callback() function passed to OpenSSL is set up with type void
5754 (*)(SSL *, int, int), as described somewhere. However, when calling the
5755 function (actually a macro) that sets it up, the type void(*)() is
5756 expected. I've put in a cast to prevent warnings from picky compilers.
5758 34. If a DNS black list lookup found a CNAME record, but there were no A
5759 records associated with the domain it pointed at, Exim crashed.
5761 35. If a DNS black list lookup returned more than one A record, Exim ignored
5762 all but the first. It now scans all returned addresses if a particular IP
5763 value is being sought. In this situation, the contents of the
5764 $dnslist_value variable are a list of all the addresses, separated by a
5767 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5768 to accept a host name and all its aliases if the forward lookup for any of
5769 them yielded the IP address of the incoming connection. Now it accepts only
5770 those names whose forward lookup yields the correct IP address. Any other
5771 names are discarded. This closes a loophole whereby a rogue DNS
5772 administrator could create reverse DNS records to break through a
5773 wildcarded host restriction in an ACL.
5775 37. If a user filter or a system filter that ran in a subprocess used any of
5776 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5777 the wrong values were passed to the pipe command ($thisaddress had the
5778 value of $0, $0 had the value of $1, etc). This bug was introduced by
5779 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5781 38. Improved the line breaking for long SMTP error messages from ACLs.
5782 Previously, if there was no break point between 40 and 75 characters, Exim
5783 left the rest of the message alone. Two changes have been made: (a) I've
5784 reduced the minimum length to 35 characters; (b) if it can't find a break
5785 point between 35 and 75 characters, it looks ahead and uses the first one
5786 that it finds. This may give the occasional overlong line, but at least the
5787 remaining text gets split now.
5789 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5790 file descriptors might be low, and that setting 1000 would always raise it.
5791 It turns out that in some environments, the limit is already over 1000 and
5792 that lowering it causes trouble. So now Exim takes care not to decrease it.
5794 40. When delivering a message, the value of $return_path is set to $sender_
5795 address at the start of routing (routers may change the value). By an
5796 oversight, this default was not being set up when an address was tested by
5797 -bt or -bv, which affected the outcome if any router or filter referred to
5800 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5801 log only when "message" or "log_message" are set. However, if one of the
5802 conditions was an address verification, or a call to a nested ACL, the
5803 messages generated by the underlying test were being passed through. This
5804 no longer happens. The underlying message is available in $acl_verify_
5805 message for both "message" and "log_message" expansions, so it can be
5806 passed through if needed.
5808 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5809 new expansion $bh_ to give the encoded byte string without charset
5810 translation. Translation happens only if iconv() is available; HAVE_ICONV
5811 indicates this at build time. HEADERS_CHARSET gives the charset to
5812 translate to; headers_charset can change it in the configuration, and
5813 "headers charset" can change it in an individual filter file.
5815 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5816 that creates RFC 2047 encoded "words" labels them as that charset instead
5817 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5818 expansion operator; (ii) when Exim creates a From: line for a local
5819 message; (iii) when a header line is rewritten to include a "phrase" part.
5821 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5822 buggy, causing it to skip the first lines of messages whose message ID
5823 ended in 'D'. This would not have bitten before Exim release 4.14, because
5824 message IDs were unlikely to end in 'D' before then. The effect was to have
5825 incorrect size information for certain domains.
5827 45. #include "config.h" was missing at the start of the crypt16.c module. This
5828 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5831 46. If there was a timeout during a "random" callout check, Exim treated it as
5832 a failure of the random address, and carried on sending RSET and the real
5833 address. If the delay was just some slowness somewhere, the response to the
5834 original RCPT would be taken as a response to RSET and so on, causing
5835 mayhem of various kinds.
5837 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5838 when I implemented it. It didn't allow for the fact that some option values
5839 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5840 the right test for positive values.
5842 48. Domain names in DNS records are case-independent. Exim always looks them up
5843 in lower case. Some resolvers return domain names in exactly the case they
5844 appear in the zone file, that is, they may contain uppercase letters. Not
5845 all resolvers do this - some return always lower case. Exim was treating a
5846 change of case by a resolver as a change of domain, similar to a widening
5847 of a domain abbreviation. This triggered its re-routing code and so it was
5848 trying to route what was effectively the same domain again. This normally
5849 caused routing to fail (because the router wouldn't handle the domain
5850 twice). Now Exim checks for this case specially, and just changes the
5851 casing of the domain that it ultimately uses when it transmits the message
5854 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5857 50. If a filter generated a file delivery with a non-absolute name (possible if
5858 no home directory exists for the router), the forbid_file option was not
5861 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5862 the existing equality matching.
5864 52. Exim was using ints instead of ino_t variables in some places where it was
5865 dealing with inode numbers.
5867 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5868 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5869 called TMPDIR, and if it finds it is different, it changes its value.
5871 54. The smtp_printf() function is now made available to local_scan() so
5872 additional output lines can be written before returning. There is also an
5873 smtp_fflush() function to enable the detection of a dropped connection.
5874 The variables smtp_input and smtp_batched_input are exported to
5877 55. Changed the default runtime configuration: the message "Unknown user"
5878 has been removed from the ACL, and instead placed on the localuser router,
5879 using the cannot_route_message feature. This means that any verification
5880 failures that generate their own messages won't get overridden. Similarly,
5881 the "Unrouteable address" message that was in the ACL for unverifiable
5882 relay addresses has also been removed.
5884 56. Added hosts_avoid_esmtp to the smtp transport.
5886 57. The exicyclog script was not checking for the esoteric option
5887 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5888 will work only if exicyclog is run under the appropriate euid.
5890 58. Following a discussion on the list, the rules by which Exim recognises line
5891 endings on incoming messages have been changed. The -dropcr and drop_cr
5892 options are now no-ops, retained only for backwards compatibility. The
5893 following line terminators are recognized: LF CRLF CR. However, special
5894 processing applies to CR:
5896 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5897 nor a local message in the state where . is a terminator.
5899 (ii) If a bare CR is encountered in a header line, an extra space is added
5900 after the line terminator so as not to end the header. The reasoning
5901 behind this is that bare CRs in header lines are most likely either
5902 to be mistakes, or people trying to play silly games.
5904 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5905 was being incorrectly given as 18 bytes larger than it should have been.
5906 This is a VOB (very old bug).
5908 60. This may never have affected anything current, but just in case it has:
5909 When the local host is found other than at the start of a list of hosts,
5910 the local host, those with the same MX, and any that follow, are discarded.
5911 When the list in question was part of a longer list of hosts, the following
5912 hosts (not currently being processed) were also being discarded. This no
5913 longer happens. I'm not sure if this situation could ever has previously
5916 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5919 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5920 header. This is something that is recommended in a new Internet Draft, and
5921 is something that is documented as being done by Sendmail. There are two
5922 possible values. For messages generated by the autoreply transport, Exim
5925 Auto-Submitted: auto-replied
5927 whereas for all other generated messages (e.g. bounces) it adds
5929 Auto-Submitted: auto-generated
5931 63. The "personal" condition in filters now includes a test for the
5932 Auto-Submitted: header. If it contains the string "auto-" the message it
5933 not considered personal.
5935 64. Added rcpt_include_affixes as a generic transport option.
5937 65. Added queue_only_override (default true).
5939 66. Added the syslog_duplication option.
5941 67. If what should have been the first header line of a message consisted of
5942 a space followed by a colon, Exim was mis-interpreting it as a header line.
5943 It isn't of course - it is syntactically invalid and should therefore be
5944 treated as the start of the message body. The misbehaviour could have
5945 caused a number of strange effects, including loss of data in subsequent
5946 header lines, and spool format errors.
5948 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5949 client host had authenticated. This control can now be exercised by an ACL
5950 for more flexibility.
5952 69. By default, callouts do not happen when testing with -bh. There is now a
5953 variant, -bhc, which does actually run the callout code, including
5954 consulting and updating the callout cache.
5956 70. Added support for saslauthd authentication, courtesy of Alexander
5959 71. If statvfs() failed on the spool or log directories while checking their
5960 size for availability, Exim confusingly gave the error "space shortage".
5961 Furthermore, in debugging mode it crashed with a floating point exception.
5962 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5963 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5964 serious problem, Exim now writes to the main and panic logs when this
5965 happens, with details of the failure. It then refuses to accept the
5966 incoming message, giving the message "spool directory problem" or "log
5967 directory problem" with a 421 code for SMTP messages.
5969 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5970 0, 1, and 2 exist, because some OS complain for execs without them (see
5971 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5972 descriptors. However, the code omitted to check that the open succeeded,
5973 causing mysterious errors if for some reason the permissions on /dev/null
5974 got screwed. Now Exim writes a message to the main and panic logs, and
5975 bombs out if it can't open /dev/null.
5977 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5978 interact so that it is all more flexible. It is supposed to remain
5979 backwards compatible. Also added extra_local_interfaces.
5981 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5982 to bomb out with an assertion failure - to the client this appears as a
5983 connection drop. This problem occurs in the part of the code that was taken
5984 from the Samba project. Fortunately, the assertion is in a very simple
5985 function, so I have fixed this by reproducing the function inline in the
5986 one place where it is called, and arranging for authentication to fail
5987 instead of killing the process with assert().
5989 75. The SPA client code was not working when the server requested OEM rather
5990 than Unicode encoding.
5992 76. Added code to make require_files with a specific uid setting more usable in
5993 the case where statting the file as root fails - usually a non-root-mounted
5994 NFS file system. When this happens and the failure is EACCES, Exim now
5995 forks a subprocess and does the per-uid checking as the relevant uid.
5997 77. Added process_log_path.
5999 78. If log_file_path was not explicitly set, a setting of check_log_space or
6000 check_log_inodes was ignored.
6002 79. If a space check for the spool or log partitions fails, the incident is now
6003 logged. Of course, in the latter case the data may get lost...
6005 80. Added the %p formatting code to string_format() so that it can be used to
6006 print addresses in debug_print(). Adjusted all the address printing in the
6007 debugging in store.c to use %p rather than %d.
6009 81. There was a concern that a line of code in smtp_in.c could overflow a
6010 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6011 initially expressed, the concern was not well-founded, because trailing
6012 spaces are removed early. However, if the trailing spaces were followed by
6013 a NULL, they did not get removed, so the overflow was possible. Two fixes
6016 (a) I re-wrote the offending code in a cleaner fashion.
6017 (b) If an incoming SMTP command contains a NULL character, it is rejected
6020 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6021 runs initgroups(), so that if the Exim user is in any additional groups,
6022 they will be used during message reception.
6028 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob