1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router (applying to the forward lookup).
90 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
91 of ldap servers used for a specific lookup. Patch provided by Heiko
94 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
95 New variable $lookup_dnssec_authenticated for observability.
97 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
98 Patch submitted by Lars Timman.
100 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
106 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
108 PP/02 Make -n do something, by making it not do something.
109 When combined with -bP, the name of an option is not output.
111 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
114 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
115 $sender_host_name and config options to manage this, and basic check
118 PP/05 DSCP support for outbound connections and control modifier for inbound.
120 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
121 (Only plugin which currently uses this is kerberos4, which nobody should
122 be using, but we should make it available and other future plugins might
123 conceivably use it, even though it would break NAT; stuff *should* be
124 using channel bindings instead).
126 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
127 name; added for Sendmail compatibility; requires admin caller.
128 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
129 just ignore it); requires trusted caller.
130 Also parse but ignore: -Ac -Am -X<logfile>
133 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
135 TL/02 Add +smtp_confirmation as a default logging option.
137 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
138 Patch by Magnus Holmgren from 2007-02-20.
140 TL/04 Bugzilla 1281 - Spec typo.
141 Bugzilla 1283 - Spec typo.
142 Bugzilla 1290 - Spec grammar fixes.
144 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
146 TL/06 Add Experimental DMARC support using libopendmarc libraries.
148 TL/07 Fix an out of order global option causing a segfault. Reported to dev
149 mailing list by by Dmitry Isaikin.
151 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
153 JH/02 Support "G" suffix to numbers in ${if comparisons.
155 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
157 NM/01 Bugzilla 1197 - Spec typo
158 Bugzilla 1196 - Spec examples corrections
160 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
162 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
163 gnutls_enable_pkcs11, but renamed to more accurately indicate its
166 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
167 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
169 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
170 "acl {{name}{arg}...}", and optional args on acl condition
173 JH/05 Permit multiple router/transport headers_add/remove lines.
175 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
177 JH/07 Avoid using a waiting database for a single-message-only transport.
178 Performance patch from Paul Fisher. Bugzilla 1262.
180 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
183 JH/09 Add $headers_added variable, with content from use of ACL modifier
184 add_header (but not yet added to the message). Bugzilla 199.
186 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
187 Pulled from Bugzilla 817 by Wolfgang Breyha.
189 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
191 (nb: this is the same fix as in Exim 4.80.1)
193 JH/11 Add A= logging on delivery lines, and a client_set_id option on
196 JH/12 Add optional authenticated_sender logging to A= and a log_selector
199 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
201 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
202 advertise SMTP AUTH mechanism to us, instead of a generic
203 protocol violation error. Also, make Exim more robust to bad
204 data from the Dovecot auth socket.
206 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
208 When a queue runner is handling a message, Exim first routes the
209 recipient addresses, during which it prunes them based on the retry
210 hints database. After that it attempts to deliver the message to
211 any remaining recipients. It then updates the hints database using
214 So if a recipient address works intermittently, it can get repeatedly
215 deferred at routing time. The retry hints record remains fresh so the
216 address never reaches the final cutoff time.
218 This is a fairly common occurrence when a user is bumping up against
219 their storage quota. Exim had some logic in its local delivery code
220 to deal with this. However it did not apply to per-recipient defers
221 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
223 This change adds a proper retry rule check during routing so that the
224 final cutoff time is checked against the message's age. We only do
225 this check if there is an address retry record and there is not a
226 domain retry record; this implies that previous attempts to handle
227 the address had the retry_use_local_parts option turned on. We use
228 this as an approximation for the destination being like a local
229 delivery, as in LMTP.
231 I suspect this new check makes the old local delivery cutoff check
232 redundant, but I have not verified this so I left the code in place.
234 TF/02 Correct gecos expansion when From: is a prefix of the username.
236 Test 0254 submits a message to Exim with the header
240 When I ran the test suite under the user fanf2, Exim expanded
241 the header to contain my full name, whereas it should have added
242 a Resent-Sender: header. It erroneously treats any prefix of the
243 username as equal to the username.
245 This change corrects that bug.
247 GF/01 DCC debug and logging tidyup
248 Error conditions log to paniclog rather than rejectlog.
249 Debug lines prefixed by "DCC: " to remove any ambiguity.
251 TF/03 Avoid unnecessary rebuilds of lookup-related code.
253 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
254 Bug spotted by Jeremy Harris; was flawed since initial commit.
255 Would have resulted in OCSP responses post-SNI triggering an Exim
256 NULL dereference and crash.
258 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
260 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
261 Bug detection, analysis and fix by Samuel Thibault.
262 Bugzilla 1331, Debian bug #698092.
264 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
266 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
267 Server implementation by Todd Lyons, client by JH.
268 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
269 config variable "prdr_enable" controls whether the server
270 advertises the facility. If the client requests PRDR a new
271 acl_data_smtp_prdr ACL is called once for each recipient, after
272 the body content is received and before the acl_smtp_data ACL.
273 The client is controlled by bolth of: a hosts_try_prdr option
274 on the smtp transport, and the server advertisement.
275 Default client logging of deliveries and rejections involving
276 PRDR are flagged with the string "PRDR".
278 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
279 fclose(). Diagnosis by Todd Lyons.
281 PP/17 Update configure.default to handle IPv6 localhost better.
282 Patch by Alain Williams (plus minor tweaks).
285 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
286 This is now consistent with GnuTLS, and is now documented: the
287 previous undocumented portable approach to treating the option as
288 unset was to force an expansion failure. That still works, and
289 an empty string is now equivalent.
291 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
292 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
293 not performing validation itself.
295 PP/20 Added force_command boolean option to pipe transport.
296 Patch from Nick Koston, of cPanel Inc.
298 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
301 TF/04 Added udpsend ACL modifer and hexquote expansion operator
303 PP/21 Fix eximon continuous updating with timestamped log-files.
304 Broken in a format-string cleanup in 4.80, missed when I repaired the
305 other false fix of the same issue.
306 Report and fix from Heiko Schlichting.
309 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
310 Report from Prashanth Katuri.
312 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
313 It's SecureTransport, so affects any MacOS clients which use the
314 system-integrated TLS libraries, including email clients.
316 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
317 using a MIME ACL for non-SMTP local injection.
318 Report and assistance in diagnosis by Warren Baker.
320 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
322 JH/16 Fix comparisons for 64b. Bugzilla 1385.
324 TL/09 Add expansion variable $authenticated_fail_id to keep track of
325 last id that failed so it may be referenced in subsequent ACL's.
327 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
330 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
331 ldap library initialization, allowing self-signed CA's to be
332 used. Also properly sets require_cert option later in code by
333 using NULL (global ldap config) instead of ldap handle (per
334 session). Bug diagnosis and testing by alxgomz.
336 TL/12 Enhanced documentation in the ratelimit.pl script provided in
337 the src/util/ subdirectory.
339 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
340 renamed to Transport Post Delivery Action by Jeremy Harris, as
343 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
344 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
345 redis_servers = needs to be configured which will be used by the redis
346 lookup. Patch from Warren Baker, of The Packet Hub.
348 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
350 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
351 hostname or reverse DNS when processing a host list. Used suggestions
352 from multiple comments on this bug.
354 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
356 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
357 Missed a few lines, added it to make the runtest require no keyboard
360 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
361 contains upper case chars. Make router use caseful_local_part.
363 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
364 support when GnuTLS has been built with p11-kit.
370 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
372 This, or similar/improved, will also be change PP/11 of 4.82.
378 PP/01 Handle short writes when writing local log-files.
379 In practice, only affects FreeBSD (8 onwards).
380 Bugzilla 1053, with thanks to Dmitry Isaikin.
382 NM/01 Bugzilla 949 - Documentation tweak
384 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
387 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
389 PP/02 Implemented gsasl authenticator.
391 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
393 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
394 `pkg-config foo` for cflags/libs.
396 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
397 with rest of GSASL and with heimdal_gssapi.
399 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
400 `pkg-config foo` for cflags/libs for the TLS implementation.
402 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
403 properties get this fed in as external SSF. A number of robustness
404 and debugging improvements to the cyrus_sasl authenticator.
406 PP/08 cyrus_sasl server now expands the server_realm option.
408 PP/09 Bugzilla 1214 - Log authentication information in reject log.
409 Patch by Jeremy Harris.
411 PP/10 Added dbmjz lookup type.
413 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
415 PP/12 MAIL args handles TAB as well as SP, for better interop with
416 non-compliant senders.
417 Analysis and variant patch by Todd Lyons.
419 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
420 Bug report from Lars Müller <lars@samba.org> (via SUSE),
421 Patch from Dirk Mueller <dmueller@suse.com>
423 PP/13 tls_peerdn now print-escaped for spool files.
424 Observed some $tls_peerdn in wild which contained \n, which resulted
425 in spool file corruption.
427 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
428 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
429 or write after TLS renegotiation, which otherwise led to messages
432 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
433 as a tracking header (ie: a signed header comes before the signature).
434 Patch from Wolfgang Breyha.
436 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
437 comma-sep list; embedded commas doubled.
439 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
441 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
443 Report and patch from Dmitry Banschikov.
445 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
446 Removed SSL_clear() after SSL_new() which led to protocol negotiation
447 failures. We appear to now support TLS1.1+ with Exim.
449 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
450 lets Exim select keys and certificates based upon TLS SNI from client.
451 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
452 before an outbound SMTP session. New log_selector, +tls_sni.
454 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
455 NULL dereference. Report and patch from Alun Jones.
457 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
458 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
459 Not seeing resolver debug output on NetBSD, but suspect this is a
460 resolver implementation change.
462 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
463 Left warnings. Added "eximon gdb" invocation mode.
465 PP/21 Defaulting "accept_8bitmime" to true, not false.
467 PP/22 Added -bw for inetd wait mode support.
469 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
470 locate the relevant includes and libraries. Made this the default.
472 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
473 Bugzilla 1246, report and most of solution from Tomasz Kusy.
475 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
476 This may cause build issues on older platforms.
478 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
479 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
480 gnutls_require_mac & gnutls_require_protocols (no longer supported).
481 Added SNI support via GnuTLS too.
482 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
484 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
486 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
487 Applied second patch from Janne, implementing suggestion to default
488 multiple-strings-in-record handling to match SPF spec.
490 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
492 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
493 read-only, out of scope).
494 Patch from Wolfgang Breyha, report from Stuart Northfield.
496 PP/29 Fix three issues highlighted by clang analyser static analysis.
497 Only crash-plausible issue would require the Cambridge-specific
498 iplookup router and a misconfiguration.
499 Report from Marcin Mirosław.
501 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
503 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
504 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
505 As part of this, removing so much warning spew let me fix some minor
506 real issues in debug logging.
508 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
509 assignment on my part. Fixed.
511 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
512 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
513 Janne Snabb (who went above and beyond: thank you).
515 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
516 string otherwise requires a connection and a bunch more work and it's
517 relatively easy to get wrong. Should also expose TLS library linkage
520 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
521 64-bit ${eval} (JH/03).
523 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
524 GNU libc to support some of the 64-bit stuff, should not lead to
525 conflicts. Defined before os.h is pulled in, so if a given platform
526 needs to override this, it can.
528 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
529 protection layer was required, which is not implemented.
530 Bugzilla 1254, patch from Wolfgang Breyha.
532 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
533 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
534 tls_dhparam take prime identifiers. Also unbreak combination of
535 OpenSSL+DH_params+TLSSNI.
537 PP/39 Disable SSLv2 by default in OpenSSL support.
543 PP/01 Solaris build fix for Oracle's LDAP libraries.
544 Bugzilla 1109, patch from Stephen Usher.
546 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
548 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
551 TF/02 Fix a couple more cases where we did not log the error message
552 when unlink() failed. See also change 4.74-TF/03.
554 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
555 lock up or crash if it happened to be inside a call to libc when it
556 got a SIGUSR1 from exiwhat.
558 The SIGUSR1 handler appends the current process status to the process
559 log which is later printed by exiwhat. It used to use the general
560 purpose logging code to do this, but several functions it calls are
561 not safe for signals.
563 The new output code in the SIGUSR1 handler is specific to the process
564 log, and simple enough that it's easy to inspect for signal safety.
565 Removing some special cases also simplifies the general logging code.
566 Removing the spurious timestamps from the process log simplifies
569 TF/04 Improved ratelimit ACL condition.
571 The /noupdate option has been deprecated in favour of /readonly which
572 has clearer semantics. The /leaky, /strict, and /readonly update modes
573 are mutually exclusive. The update mode is no longer included in the
574 database key; it just determines when the database is updated. (This
575 means that when you upgrde Exim will forget old rate measurements.)
577 Exim now checks that the per_* options are used with an update mode that
578 makes sense for the current ACL. For example, when Exim is processing a
579 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
580 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
581 must specify per_mail/readonly. If you omit the update mode it defaults to
582 /leaky where that makes sense (as before) or /readonly where required.
584 The /noupdate option is now undocumented but still supported for
585 backwards compatibility. It is equivalent to /readonly except that in
586 ACLs where /readonly is required you may specify /leaky/noupdate or
587 /strict/noupdate which are treated the same as /readonly.
589 A useful new feature is the /count= option. This is a generalization
590 of the per_byte option, so that you can measure the throughput of other
591 aggregate values. For example, the per_byte option is now equivalent
592 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
594 The per_rcpt option has been generalized using the /count= mechanism
595 (though it's more complicated than the per_byte equivalence). When it is
596 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
597 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
598 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
599 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
600 Note that using per_rcpt with a non-readonly update mode in more than
601 one ACL will cause the recipients to be double-counted. (The per_mail
602 and per_byte options don't have this problem.)
604 The handling of very low rates has changed slightly. If the computed rate
605 is less than the event's count (usually one) then this event is the first
606 after a long gap. In this case the rate is set to the same as this event's
607 count, so that the first message of a spam run is counted properly.
609 The major new feature is a mechanism for counting the rate of unique
610 events. The new per_addr option counts the number of different
611 recipients that someone has sent messages to in the last time period. It
612 behaves like per_rcpt if all the recipient addresses are different, but
613 duplicate recipient addresses do not increase the measured rate. Like
614 the /count= option this is a general mechanism, so the per_addr option
615 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
616 example, measure the rate that a client uses different sender addresses
617 with the options per_mail/unique=$sender_address. There are further
618 details in the main documentation.
620 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
622 TF/06 Removed a few PCRE remnants.
624 TF/07 Automatically extract Exim's version number from tags in the git
625 repository when doing development or release builds.
627 PP/02 Raise smtp_cmd_buffer_size to 16kB.
628 Bugzilla 879. Patch from Paul Fisher.
630 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
631 Heavily based on revision 40f9a89a from Simon Arlott's tree.
634 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
636 PP/05 Variable $av_failed, true if the AV scanner deferred.
637 Bugzilla 1078. Patch from John Horne.
639 PP/06 Stop make process more reliably on build failure.
640 Bugzilla 1087. Patch from Heiko Schlittermann.
642 PP/07 Make maildir_use_size_file an _expandable_ boolean.
643 Bugzilla 1089. Patch from Heiko Schlittermann.
645 PP/08 Handle ${run} returning more data than OS pipe buffer size.
646 Bugzilla 1131. Patch from Holger Weiß.
648 PP/09 Handle IPv6 addresses with SPF.
649 Bugzilla 860. Patch from Wolfgang Breyha.
651 PP/10 GnuTLS: support TLS 1.2 & 1.1.
653 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
656 PP/11 match_* no longer expand right-hand-side by default.
657 New compile-time build option, EXPAND_LISTMATCH_RHS.
658 New expansion conditions, "inlist", "inlisti".
660 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
662 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
664 PP/14 fix log_write() format string regression from TF/03.
665 Bugzilla 1152. Patch from Dmitry Isaikin.
671 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
673 PP/02 Harmonised TLS library version reporting; only show if debugging.
674 Layout now matches that introduced for other libraries in 4.74 PP/03.
676 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
678 PP/04 New "dns_use_edns0" global option.
680 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
683 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
684 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
686 TK/01 Updated PolarSSL code to 0.14.2.
687 Bugzilla 1097. Patch from Andreas Metzler.
689 PP/07 Catch divide-by-zero in ${eval:...}.
692 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
695 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
696 format-string attack -- SECURITY: remote arbitrary code execution.
698 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
699 time unintentionally subject to list matching rules, letting the header
700 cause arbitrary Exim lookups (of items which can occur in lists, *not*
701 arbitrary string expansion). This allowed for information disclosure.
703 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
704 INT_MIN/-1 -- value coerced to INT_MAX.
710 NM/01 Workround for PCRE version dependency in version reporting
713 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
714 This fixes portability to compilers other than gcc, notably
715 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
717 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
718 makefiles for portability to HP-UX and POSIX correctness.
720 PP/01 Permit LOOKUP_foo enabling on the make command-line.
721 Also via indented variable definition in the Makefile.
722 (Debugging by Oliver Heesakkers).
724 PP/02 Restore caching of spamd results with expanded spamd_address.
725 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
727 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
728 Improves build reliability. Fix from: Frank Elsner
730 NM/02 Fix wide character breakage in the rfc2047 coding
731 Fixes bug 1064. Patch from Andrey N. Oktyabrski
733 NM/03 Allow underscore in dnslist lookups
734 Fixes bug 1026. Patch from Graeme Fowler
736 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
737 Code patches from Adam Ciarcinski of NetBSD.
739 NM/04 Fixed exiqgrep to cope with mailq missing size issue
742 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
743 is logged, to avoid truncation. Patch from John Horne.
745 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
746 Patch from Jakob Hirsch.
748 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
749 SQL string expansion failure details.
750 Patch from Andrey Oktyabrski.
752 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
753 Patch from Simon Arlott.
755 PP/09 New lookups functionality failed to compile on old gcc which rejects
756 extern declarations in function scope.
757 Patch from Oliver Fleischmann
759 PP/10 Use sig_atomic_t for flags set from signal handlers.
760 Check getgroups() return and improve debugging.
761 Fixed developed for diagnosis in bug 927 (which turned out to be
764 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
765 Patch from Mark Zealey.
767 PP/12 Bugzilla 1056: Improved spamd server selection.
768 Patch from Mark Zealey.
770 PP/13 Bugzilla 1086: Deal with maildir quota file races.
771 Based on patch from Heiko Schlittermann.
773 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
774 Patch from Uwe Doering, sign-off by Michael Haardt.
776 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
777 variable declaration deep within a block. Bug and patch from
780 PP/15 lookups-Makefile IRIX compatibilty coercion.
782 PP/16 Make DISABLE_DKIM build knob functional.
784 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
785 Patch by Simon Arlott
787 TF/03 Fix valgrind.h portability to C89 compilers that do not support
788 variable argument macros. Our copy now differs from upstream.
794 TF/01 Failure to get a lock on a hints database can have serious
795 consequences so log it to the panic log.
797 TF/02 Log LMTP confirmation messages in the same way as SMTP,
798 controlled using the smtp_confirmation log selector.
800 TF/03 Include the error message when we fail to unlink a spool file.
802 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
803 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
804 for maintaining out-of-tree patches for some time.
806 PP/01 Bugzilla 139: Documentation and portability issues.
807 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
808 Handle per-OS dynamic-module compilation flags.
810 PP/02 Let /dev/null have normal permissions.
811 The 4.73 fixes were a little too stringent and complained about the
812 permissions on /dev/null. Exempt it from some checks.
813 Reported by Andreas M. Kirchwitz.
815 PP/03 Report version information for many libraries, including
816 Exim version information for dynamically loaded libraries. Created
817 version.h, now support a version extension string for distributors
818 who patch heavily. Dynamic module ABI change.
820 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
821 privilege escalation vulnerability whereby the Exim run-time user
822 can cause root to append content of the attacker's choosing to
825 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
828 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
829 If dropping privileges for untrusted macros, we disabled normal logging
830 on the basis that it would fail; for the Exim run-time user, this is not
831 the case, and it resulted in successful deliveries going unlogged.
832 Fixed. Reported by Andreas Metzler.
838 PP/01 Date: & Message-Id: revert to normally being appended to a message,
839 only prepend for the Resent-* case. Fixes regression introduced in
840 Exim 4.70 by NM/22 for Bugzilla 607.
842 PP/02 Include check_rfc2047_length in configure.default because we're seeing
843 increasing numbers of administrators be bitten by this.
845 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
847 PP/03 Bugzilla 994: added openssl_options main configuration option.
849 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
851 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
853 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
855 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
856 without a peer certificate, leading to a segfault because of an
857 assumption that peers always have certificates. Be a little more
858 paranoid. Problem reported by Martin Tscholak.
860 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
861 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
862 NB: ClamAV planning to remove STREAM in "middle of 2010".
863 CL also introduces -bmalware, various -d+acl logging additions and
864 more caution in buffer sizes.
866 PP/09 Implemented reverse_ip expansion operator.
868 PP/10 Bugzilla 937: provide a "debug" ACL control.
870 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
872 PP/12 Bugzilla 973: Implement --version.
874 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
876 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
878 PP/15 Bugzilla 816: support multiple condition rules on Routers.
880 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
881 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
882 ignore trailing whitespace.
884 JJ/02 prevent non-panic DKIM error from being sent to paniclog
886 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
889 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
890 Notification from Dr Andrew Aitchison.
892 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
893 ExtendedDetectionInfo response format.
894 Notification from John Horne.
896 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
899 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
900 XSL and documented dependency on system catalogs, with examples of how
903 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
906 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
907 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
908 configuration file which is writeable by the Exim user or group.
910 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
911 of configuration files to cover files specified with the -C option if
912 they are going to be used with root privileges, not just the default
915 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
916 option (effectively making it always true).
918 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
919 files to be used while preserving root privileges.
921 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
922 that rogue child processes cannot use them.
924 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
925 run-time user, instead of root.
927 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
928 Exim run-time user without dropping privileges.
930 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
931 result string, instead of calling string_vformat() twice with the same
934 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
935 for other users. Others should always drop root privileges if they use
936 -C on the command line, even for a whitelisted configure file.
938 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
940 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
946 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
947 $data_path, and $header_path variables; fixed documentation bugs and
950 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
951 exipick to access non-standard spools, including the "frozen" queue
954 NM/01 Bugzilla 965: Support mysql stored procedures.
955 Patch from Alain Williams
957 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
959 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
960 Patch from Andreas Metzler
962 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
963 Patch from Kirill Miazine
965 NM/05 Bugzilla 671: Added umask to procmail example.
967 JJ/03 installed exipick 20100323.0, fixing doc bug
969 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
970 directory. Notification and patch from Dan Rosenberg.
972 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
974 TK/02 Improve log output when DKIM signing operation fails.
976 MH/01 Treat the transport option dkim_domain as a colon separated
977 list, not as a single string, and sign the message with each element,
978 omitting multiple occurences of the same signer.
980 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
981 Bugzilla 985, 986. Patch by Simon Arlott
983 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
984 Patch by Simon Arlott
986 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
987 MBX locking. Notification from Dan Rosenberg.
993 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
995 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
997 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
999 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1001 NM/04 Bugzilla 915: Flag broken perl installation during build.
1007 TK/01 Added patch by Johannes Berg that expands the main option
1008 "spamd_address" if it starts with a dollar sign.
1010 TK/02 Write list of recipients to X-Envelope-Sender header when building
1011 the mbox-format spool file for content scanning (suggested by Jakob
1014 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1015 (http://www.dcc-servers.net/) support via dccifd. Activated by
1016 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1018 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1019 by Mark Daniel Reidel <mr@df.eu>.
1021 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1022 When building exim an external PCRE library is now needed -
1023 PCRE is a system library on the majority of modern systems.
1024 See entry on PCRE_LIBS in EDITME file.
1026 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1027 conversation. Added nologin parameter to request.
1028 Patch contributed by Kirill Miazine.
1030 TF/01 Do not log submission mode rewrites if they do not change the address.
1032 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1034 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1035 log files in place. Contributed by Roberto Lima.
1037 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1039 TF/03 Bugzilla 615: When checking the local_parts router precondition
1040 after a local_part_suffix or local_part_prefix option, Exim now
1041 does not use the address's named list lookup cache, since this
1042 contains cached lookups for the whole local part.
1044 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1045 Robert Millan. Documentation is in experimental-spec.txt.
1047 TF/04 Bugzilla 668: Fix parallel build (make -j).
1049 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1051 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1052 Patch provided by Jan Srzednicki.
1054 TF/05 Leading white space used to be stripped from $spam_report which
1055 wrecked the formatting. Now it is preserved.
1057 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1058 that they are available at delivery time.
1060 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1062 TF/08 TLS error reporting now respects the incoming_interface and
1063 incoming_port log selectors.
1065 TF/09 Produce a more useful error message if an SMTP transport's hosts
1066 setting expands to an empty string.
1068 NM/06 Bugzilla 744: EXPN did not work under TLS.
1069 Patch provided by Phil Pennock.
1071 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1072 Patch provided by Richard Godbee.
1074 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1075 acl_smtp_notquit, added index entry.
1077 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1078 Patch provided by Eugene Bujak.
1080 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1081 accept(). Patch provided by Maxim Dounin.
1083 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1084 Patch provided by Phil Pennock.
1086 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1088 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1089 Patch provided by Brad "anomie" Jorsch.
1091 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1092 Patch provided by Dean Brooks.
1094 TK/05 Add native DKIM support (does not depend on external libraries).
1096 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1097 Patch provided by Graeme Fowler.
1099 NM/16 Bugzilla 851: Documentation example syntax fix.
1101 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1103 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1106 NM/19 Bugzilla 745: TLS version reporting.
1107 Patch provided by Phil Pennock.
1109 NM/20 Bugzilla 167: bool: condition support.
1110 Patch provided by Phil Pennock.
1112 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1113 clients. Patch provided by Phil Pennock.
1115 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1116 Patch provided by Brad "anomie" Jorsch.
1118 NM/23 Bugzilla 687: Fix misparses in eximstats.
1119 Patch provided by Heiko Schlittermann.
1121 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1122 Patch provided by Heiko Schlittermann.
1124 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1125 plus update to original patch.
1127 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1129 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1130 Patch provided by David Brownlee.
1132 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1134 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1136 NM/30 Bugzilla 888: TLS documentation bugfixes.
1138 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1140 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1141 Unlike the original bugzilla I have changed all shell scripts in src tree.
1143 NM/33 Bugzilla 898: Transport filter timeout fix.
1144 Patch by Todd Rinaldo.
1146 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1147 Patch by Serge Demonchaux.
1149 NM/35 Bugzilla 39: Base64 decode bug fixes.
1150 Patch by Jakob Hirsch.
1152 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1154 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1156 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1158 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1164 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1165 ALT-N's libdkim that I have put here:
1166 http://duncanthrax.net/exim-experimental/
1168 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1169 were called 'true' and it seems that C99 defines that as a reserved
1170 keyword to be used with 'bool' variable types. That means you could
1171 not include C99-style headers which use bools without triggering
1172 build errors in sieve.c.
1174 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1175 as mailq or other aliases. Changed the --help handling significantly
1176 to do whats expected. exim_usage() emits usage/help information.
1178 SC/01 Added the -bylocaldomain option to eximstats.
1180 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1182 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1184 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1190 PH/01 Another patch from the Sieve maintainer.
1192 PH/02 When an IPv6 address is converted to a string for single-key lookup
1193 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1194 dots are used instead of colons so that keys in lsearch files need not
1195 contain colons. This was done some time before quoting was made available
1196 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1197 (notated using the quote facility) so as to distinguish them from IPv4
1198 keys. This meant that lookups for IP addresses in host lists did not work
1199 for iplsearch lookups.
1201 This has been fixed by arranging for IPv6 addresses to be expressed with
1202 colons if the lookup type is iplsearch. This is not incompatible, because
1203 previously such lookups could never work.
1205 The situation is now rather anomolous, since one *can* have colons in
1206 ordinary lsearch keys. However, making the change in all cases is
1207 incompatible and would probably break a number of configurations.
1209 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1212 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1213 conversion specification without a maximum field width, thereby enabling
1214 a rogue spamd server to cause a buffer overflow. While nobody in their
1215 right mind would setup Exim to query an untrusted spamd server, an
1216 attacker that gains access to a server running spamd could potentially
1217 exploit this vulnerability to run arbitrary code as the Exim user.
1219 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1220 $primary_hostname instead of what libspf2 thinks the hosts name is.
1222 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1223 a directory entry by the name of the lookup key. Previously, if a
1224 symlink pointed to a non-existing file or a file in a directory that
1225 Exim lacked permissions to read, a lookup for a key matching that
1226 symlink would fail. Now it is enough that a matching directory entry
1227 exists, symlink or not. (Bugzilla 503.)
1229 PH/03 The body_linecount and body_zerocount variables are now exported in the
1232 PH/04 Added the $dnslist_matched variable.
1234 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1235 This means they are set thereafter only if the connection becomes
1238 PH/06 Added the client_condition to authenticators so that some can be skipped
1239 by clients under certain conditions.
1241 PH/07 The error message for a badly-placed control=no_multiline_responses left
1242 "_responses" off the end of the name.
1244 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1246 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1247 (without spaces) instead of just copying the configuration text.
1249 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1251 PH/11 Added $max_received_linelength.
1253 PH/12 Added +ignore_defer and +include_defer to host lists.
1255 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1256 way in which PCRE > 7.0 is built.
1258 PH/14 Implemented queue_only_load_latch.
1260 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1261 MAIL command. The effect was to mangle the value on 64-bit systems.
1263 PH/16 Another patch from the Sieve maintainer.
1265 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1267 PH/18 If a system quota error occurred while trying to create the file for
1268 a maildir delivery, the message "Mailbox is full" was not appended to the
1269 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1270 only to a quota excession during the actual writing of the file.
1272 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1273 characters?) which causes problems in log lines. The DN values are now
1274 passed through string_printing() before being added to log lines.
1276 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1277 and InterBase are left for another time.)
1279 PH/21 Added message_body_newlines option.
1281 PH/22 Guard against possible overflow in moan_check_errorcopy().
1283 PH/23 POSIX allows open() to be a macro; guard against that.
1285 PH/24 If the recipient of an error message contained an @ in the local part
1286 (suitably quoted, of course), incorrect values were put in $domain and
1287 $local_part during the evaluation of errors_copy.
1293 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1294 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1297 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1298 issue a MAIL command.
1300 PH/02 In an ACL statement such as
1302 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1304 if a client was not listed at all, or was listed with a value other than
1305 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1306 the condition was not true (as it should be), so access was not denied.
1307 The bug was that the ! inversion was incorrectly passed on to the second
1308 item. This has been fixed.
1310 PH/03 Added additional dnslists conditions == and =& which are different from
1311 = and & when the dns lookup returns more than one IP address.
1313 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1314 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1316 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1317 FSYNC, which compiles an option called disable_fsync that allows for
1318 bypassing fsync(). The documentation is heavily laced with warnings.
1320 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1322 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1323 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1324 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1325 including adding "make clean"; (3) Added -fPIC when compiling the test
1326 dynamically loaded module, to get rid of a warning.
1328 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1329 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1330 = 0s. The bug is otherwise harmless.
1332 PH/07 There was a bug in the dovecot authenticator such that the value of
1333 $auth1 could be overwritten, and so not correctly preserved, after a
1334 successful authentication. This usually meant that the value preserved by
1335 the server_setid option was incorrect.
1337 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1339 PH/09 Installed PCRE release 7.0.
1341 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1342 run for batched SMTP input. It is now run at the start of every message
1343 in the batch. While fixing this I discovered that the process information
1344 (output by running exiwhat) was not always getting set for -bs and -bS
1345 input. This is fixed, and it now also says "batched" for BSMTP.
1347 PH/11 Added control=no_pipelining.
1349 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1350 patch, slightly modified), and move the expansion of helo_data till after
1351 the connection is made in the smtp transport (so it can use these
1354 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1356 PH/14 Added log_selector = +pid.
1358 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1360 PH/16 Add ${if forany and ${if forall.
1362 PH/17 Added dsn_from option to vary the From: line in DSNs.
1364 PH/18 Flush SMTP output before performing a callout, unless control =
1365 no_callout_flush is set.
1367 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1368 was true (the default) a successful delivery failed to delete the retry
1369 item, thus causing premature timeout of the address. The bug is now
1372 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1374 PH/21 Long custom messages for fakedefer and fakereject are now split up
1375 into multiline reponses in the same way that messages for "deny" and
1376 other ACL rejections are.
1378 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1379 with slight modification.
1381 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1382 draft, changing the syntax and factoring some duplicate code".
1384 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1385 for deliveries of the second and subsequent messages over the same SMTP
1388 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1389 ${reduce, with only minor "tidies".
1391 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1393 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1394 expansion side effects.
1396 PH/27 When a message times out after an over-quota error from an Exim-imposed
1397 quota, the bounce message says "mailbox is full". This message was not
1398 being given when it was a system quota that was exceeded. It now should
1401 MH/03 Made $recipients available in local_scan(). local_scan() already has
1402 better access to the recipient list through recipients_list[], but
1403 $recipients can be useful in postmaster-provided expansion strings.
1405 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1406 in the case of a MAIL command with additional options following the
1407 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1408 were accidentally chopped off.
1410 PH/29 SMTP synchronization checks are implemented when a command is read -
1411 there is a check that no more input is waiting when there shouldn't be
1412 any. However, for some commands, a delay in an ACL can mean that it is
1413 some time before the response is written. In this time, more input might
1414 arrive, invalidly. So now there are extra checks after an ACL has run for
1415 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1416 pipelining has not been advertised.
1418 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1420 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1421 returned in the SMTP response when smtp_return_error_details was false.
1422 This has been fixed.
1424 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1425 instead of the C I/O that was originally supplied, because problems were
1426 reported on Solaris.
1428 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1429 Exim which did not show up earlier: it was assuming that a call to
1430 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1431 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1432 was a macro that became an assignment, so it seemed to work. This has
1433 changed to a proper function call with a void return, hence the compile
1434 error. Exim's code has been fixed.
1436 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1439 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1441 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1443 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1444 because it turns out to be too big or there is a timeout) there is no
1445 "Completed" line in the log. When some messages of this type were
1446 selected by exigrep, they were listed as "not completed". Others were
1447 picked up by some special patterns. I have improved the selection
1448 criteria to be more general.
1450 PH/38 The host_find_failed option in the manualroute router can now be set
1451 to "ignore", to completely ignore a host whose IP address cannot be
1452 found. If all hosts are ignored, the behaviour is controlled by the new
1453 host_all_ignored option.
1455 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1456 homing or because of multiple MX records with /mx) generated more than
1457 one IP address, and the following item turned out to be the local host,
1458 all the secondary addresses of the first item were incorrectly removed
1459 from the list, along with the local host and any following hosts (which
1460 is what is supposed to happen).
1462 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1463 whoever called Exim into the -H file. In the case of the daemon it was
1464 behaving confusingly. When first started, it used values for whoever
1465 started the daemon, but after a SIGHUP it used the Exim user (because it
1466 calls itself on a restart). I have changed the code so that it now always
1469 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1470 message are rejected with the same error (e.g. no authentication or bad
1471 sender address), and a DATA command is nevertheless sent (as can happen
1472 with PIPELINING or a stupid MUA), the error message that was given to the
1473 RCPT commands is included in the rejection of the DATA command. This is
1474 intended to be helpful for MUAs that show only the final error to their
1477 PH/42 Another patch from the Sieve maintainer.
1479 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1480 Eximstats - Fixed some broken HTML links and added missing column headers
1482 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1483 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1485 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1487 PH/43 Yet another patch from the Sieve maintainer.
1489 PH/44 I found a way to check for a TCP/IP connection going away before sending
1490 the response to the final '.' that terminates a message, but only in the
1491 case where the client has not sent further data following the '.'
1492 (unfortunately, this is allowed). However, in many cases there won't be
1493 any further data because there won't be any more messages to send. A call
1494 to select() can be used: if it shows that the input is "ready", there is
1495 either input waiting, or the socket has been closed. An attempt to read
1496 the next input character can distinguish the two cases. Previously, Exim
1497 would have sent an OK response which the client would never have see.
1498 This could lead to message repetition. This fix should cure that, at
1499 least in a lot of common cases.
1501 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1502 advertised in response to EHLO.
1508 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1509 fixed by 4.65/MH/01 (is this a record?) are fixed:
1511 (i) An empty string was always treated as zero by the numeric comparison
1512 operators. This behaviour has been restored.
1514 (ii) It is documented that the numeric comparison operators always treat
1515 their arguments as decimal numbers. This was broken in that numbers
1516 starting with 0 were being interpreted as octal.
1518 While fixing these problems I realized that there was another issue that
1519 hadn't been noticed. Values of message_size_limit (both the global option
1520 and the transport option) were treated as octal if they started with 0.
1521 The documentation was vague. These values are now always treated as
1522 decimal, and I will make that clear in the documentation.
1528 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1529 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1532 MH/01 Don't check that the operands of numeric comparison operators are
1533 integers when their expansion is in "skipping" mode (fixes bug
1534 introduced by 4.64-PH/07).
1536 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1537 child addresses, Exim now panics and dies. Previously, because the count
1538 is held in a short int, deliveries were likely to be lost. As such a
1539 large number of recipients for a single message is ridiculous
1540 (performance will be very, very poor), I have chosen to impose a limit
1541 rather than extend the field.
1547 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1548 leftover -K file (the existence of which was triggered by #402).
1549 While we were at it, introduced process PID as part of the -K
1550 filename. This should rule out race conditions when creating
1553 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1554 processing considerably. Previous code took too long for large mails,
1555 triggering a timeout which in turn triggers #401.
1557 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1558 in the DK code in transports.c. sendfile() is not really portable,
1559 hence the _LINUX specificness.
1561 TF/01 In the add_headers option to the mail command in an Exim filter,
1562 there was a bug that Exim would claim a syntax error in any
1563 header after the first one which had an odd number of characters
1566 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1567 callout verification, Exim cached a "reject" for the entire domain. This
1568 is correct for most verifications, but it is not correct for a recipient
1569 verification with use_sender or use_postmaster set, because in that case
1570 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1571 case of MAIL FROM:<> rejection from other early rejections (e.g.
1572 rejection of HELO). When verifying a recipient using a non-null MAIL
1573 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1574 Whatever the result of the callout, the value of the domain cache is
1575 left unchanged (for any other kind of callout, getting as far as trying
1576 RCPT means that the domain itself is ok).
1578 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1581 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1582 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1583 session, the daemon ignores ECONNECT errors and logs others; it now
1584 ignores EPIPE as well.
1586 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1587 (quoted-printable decoding).
1589 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1590 later the small subsequent patch to fix an introduced bug.
1592 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1594 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1596 PH/08 An error is now given if message_size_limit is specified negative.
1598 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1599 to be given (somewhat) arbitrary names.
1601 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1604 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1605 miscellaneous code fixes
1607 PH/10 Added the log_reject_target ACL modifier to specify where to log
1610 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1611 hostname. This is wrong, because it relates to the incoming message (and
1612 probably the interface on which it is arriving) and not to the outgoing
1613 callout (which could be using a different interface). This has been
1614 changed to use the value of the helo_data option from the smtp transport
1615 instead - this is what is used when a message is actually being sent. If
1616 there is no remote transport (possible with a router that sets up host
1617 addresses), $smtp_active_hostname is used.
1619 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1620 tweaks were necessary in order to get it to work (see also 21 below):
1621 (a) The code assumed that strncpy() returns a negative number on buffer
1622 overflow, which isn't the case. Replaced with Exim's string_format()
1624 (b) There were several signed/unsigned issues. I just did the minimum
1625 hacking in of casts. There is scope for a larger refactoring.
1626 (c) The code used strcasecmp() which is not a standard C function.
1627 Replaced with Exim's strcmpic() function.
1628 (d) The code set only $1; it now sets $auth1 as well.
1629 (e) A simple test gave the error "authentication client didn't specify
1630 service in request". It would seem that Dovecot has changed its
1631 interface. Fortunately there's a specification; I followed it and
1632 changed what the client sends and it appears to be working now.
1634 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1637 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1638 address A is aliased to B and C, where B exists and C does not. Without
1639 -v the output is "A verified" because verification stops after a
1640 successful redirection if more than one address is generated. However,
1641 with -v the child addresses are also verified. Exim was outputting "A
1642 failed to verify" and then showing the successful verification for C,
1643 with its parentage. It now outputs "B failed to verify", showing B's
1644 parentage before showing the successful verification of C.
1646 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1647 look up a TXT record in a specific list after matching in a combined
1650 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1651 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1652 they consult the DNS. I had assumed they would set it the way they
1653 wanted; and indeed my experiments on Linux seem to show that in some
1654 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1655 To be on the safe side, however, I have now made the interface to
1656 host_find_byname() similar to host_find_bydns(), with an argument
1657 containing the DNS resolver options. The host_find_byname() function now
1658 sets these options at its start, just as host_find_bydns() does. The smtp
1659 transport options dns_qualify_single and dns_search_parents are passed to
1660 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1661 of host_find_byname() use the default settings of RES_DEFNAMES
1662 (qualify_single) but not RES_DNSRCH (search_parents).
1664 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1665 spool_read_header() do less string testing, by means of a preliminary
1666 switch on the second character of optional "-foo" lines. (This is
1667 overdue, caused by the large number of possibilities that now exist.
1668 Originally there were few.) While I was there, I also converted the
1669 str(n)cmp tests so they don't re-test the leading "-" and the first
1670 character, in the hope this might squeeze out yet more improvement.
1672 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1673 flag allowing group syntax was set by the header_syntax check but not
1674 turned off, possible causing trouble later; (2) The flag was not being
1675 set at all for the header_verify test, causing "group"-style headers to
1676 be rejected. I have now set it in this case, and also caused header_
1677 verify to ignore an empty address taken from a group. While doing this, I
1678 came across some other cases where the code for allowing group syntax
1679 while scanning a header line wasn't quite right (mostly, not resetting
1680 the flag correctly in the right place). These bugs could have caused
1681 trouble for malformed header lines. I hope it is now all correct.
1683 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1684 with the "reply" argument non-NULL. The code, however (which originally
1685 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1686 but it didn't always do it. This confused somebody who was copying the
1687 code for some other use. I have removed all the tests.
1689 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1690 feature that was used to support insecure browsers during the U.S. crypto
1691 embargo. It requires special client support, and Exim is probably the
1692 only MTA that supported it -- and would never use it because real RSA is
1693 always available. This code has been removed, because it had the bad
1694 effect of slowing Exim down by computing (never used) parameters for the
1695 RSA_EXPORT functionality.
1697 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1698 authenticator to fail if there's a tab character in the incoming data
1699 (there should never be unless someone is messing about, as it's supposed
1700 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1701 if the connection is using TLS or if the remote IP is the same as the
1702 local IP, and the "valid-client-cert option" if a client certificate has
1705 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1706 authenticators. This can be used for authorization after authentication
1707 succeeds. (In the case of plaintext, it servers for both authentication
1710 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1711 if any retry times were supplied.
1713 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1714 connection, where there is no client IP address to check. In this
1715 situation, the verify now always succeeds.
1717 PH/25 Applied John Jetmore's -Mset patch.
1719 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1721 PH/27 In a string expansion for a processed (not raw) header when multiple
1722 headers of the same name were present, leading whitespace was being
1723 removed from all of them, but trailing whitespace was being removed only
1724 from the last one. Now trailing whitespace is removed from each header
1725 before concatenation. Completely empty headers in a concatenation (as
1726 before) are ignored.
1728 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1729 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1731 PH/29 [Removed. This was a change that I later backed out, and forgot to
1732 correct the ChangeLog entry (that I had efficiently created) before
1733 committing the later change.]
1735 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1736 address errors (4xx response to RCPT) over the same connection as other
1737 messages routed to the same hosts. Such deliveries are always "forced",
1738 so retry times are not inspected. This resulted in far too many retries
1739 for the affected addresses. The effect occurred only when there were more
1740 hosts than the hosts_max_try setting in the smtp transport when it had
1741 the 4xx errors. Those hosts that it had tried were not added to the list
1742 of hosts for which the message was waiting, so if all were tried, there
1743 was no problem. Two fixes have been applied:
1745 (i) If there are any address or message errors in an SMTP delivery, none
1746 of the hosts (tried or untried) are now added to the list of hosts
1747 for which the message is waiting, so the message should not be a
1748 candidate for sending over the same connection that was used for a
1749 successful delivery of some other message. This seems entirely
1750 reasonable: after all the message is NOT "waiting for some host".
1751 This is so "obvious" that I'm not sure why it wasn't done
1752 previously. Hope I haven't missed anything, but it can't do any
1753 harm, as the worst effect is to miss an optimization.
1755 (ii) If, despite (i), such a delivery is accidentally attempted, the
1756 routing retry time is respected, so at least it doesn't keep
1757 hammering the server.
1759 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1760 in ${readsocket because some servers need this prod.
1762 PH/32 Added some extra debug output when updating a wait-xxx database.
1764 PH/33 The hint "could be header name not terminated by colon", which has been
1765 given for certain expansion errors for a long time, was not being given
1766 for the ${if def:h_colon_omitted{... case.
1768 PH/34 The spec says: "With one important exception, whenever a domain list is
1769 being scanned, $domain contains the subject domain." There was at least
1770 one case where this was not true.
1772 PH/35 The error "getsockname() failed: connection reset by peer" was being
1773 written to the panic log as well as the main log, but it isn't really
1774 panic-worthy as it just means the connection died rather early on. I have
1775 removed the panic log writing for the ECONNRESET error when getsockname()
1778 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1779 runs only) independently of the message's sender address. This meant
1780 that, if the 4xx error was in fact related to the sender, a different
1781 message to the same recipient with a different sender could confuse
1782 things. In particualar, this can happen when sending to a greylisting
1783 server, but other circumstances could also provoke similar problems.
1784 I have changed the default so that the retry time for these errors is now
1785 based a combination of the sender and recipient addresses. This change
1786 can be overridden by setting address_retry_include_sender=false in the
1789 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1790 remote server are returned as part of bounce messages. This was not
1791 happening for LMTP over a pipe (the lmtp transport), but now it is the
1792 same for both kinds of LMTP.
1794 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1795 in header lines that were in fact CNAMEs. This is no longer the case.
1797 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1798 and queue runs started by the daemon processed all messages. This has
1799 been fixed so that -R and -S can now usefully be given with -q<time>.
1801 PH/40 Import PCRE release 6.7 (fixes some bugs).
1803 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1805 PH/42 Give an error if -q is specified more than once.
1807 PH/43 Renamed the variables $interface_address and $interface_port as
1808 $received_ip_address and $received_port, to make it clear that these
1809 values apply to message reception, and not to the outgoing interface when
1810 a message is delivered. (The old names remain recognized, of course.)
1812 PH/44 There was no timeout on the connect() call when using a Unix domain
1813 socket in the ${readsocket expansion. There now is.
1815 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1816 be meaningful with "accept".
1818 SC/01 Eximstats V1.43
1819 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1821 SC/02 Eximstats V1.44
1822 Use a glob alias rather than an array ref in the generated
1823 parser. This improves both readability and performance.
1825 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1826 Collect SpamAssassin and rejection statistics.
1827 Don't display local sender or destination tables unless
1828 there is data to show.
1829 Added average volumes into the top table text output.
1831 SC/04 Eximstats V1.46
1832 Collect data on the number of addresses (recipients)
1833 as well as the number of messages.
1835 SC/05 Eximstats V1.47
1836 Added 'Message too big' to the list of mail rejection
1837 reasons (thanks to Marco Gaiarin).
1839 SC/06 Eximstats V1.48
1840 Mainlog lines which have GMT offsets and are too short to
1841 have a flag are now skipped.
1843 SC/07 Eximstats V1.49 (Alain Williams)
1844 Added the -emptyok flag.
1846 SC/08 Eximstats V1.50
1847 Fixes for obtaining the IP address from reject messages.
1849 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1850 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1851 whitesspace changes from 4.64-PH/27
1853 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1856 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1857 are found, allow negative numbers in numeric criteria)
1859 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1861 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1862 to match changes made in 4.64-PH/43
1864 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1866 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1867 "domain" is actually an IP address, and force a failure. This locks out
1868 those revolvers/nameservers that support "A-for-A" lookups, in
1869 contravention of the specifications.
1871 PH/48 When a host name was looked up from an IP address, and the subsequent
1872 forward lookup of the name timed out, the host name was left in
1873 $sender_host_name, contrary to the specification.
1875 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1876 restricted to single-key lookups, Exim was not diagnosing an error if
1877 * or *@ was used with a query-style lookup.
1879 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1881 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1882 long ago, but noone interested enough thought of it. Let's just say that
1883 the "1.1" means that there are some new functions that weren't there at
1884 some point in the past.
1886 PH/51 Error processing for expansion failure of helo_data from an smtp
1887 transport during callout processing was broken.
1889 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1890 tested/used via the -bh/-bhc/-bs options.
1892 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1893 bug, fixed in subsequent PCRE releases).
1895 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1896 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1898 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1904 SC/01 Use a glob alias rather than an array ref in eximstats generated
1905 parser. This improves both readability and performance.
1907 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1908 Don't display local sender or destination tables in eximstats unless
1909 there is data to show.
1910 Added average volumes into the eximstats top table text output.
1912 SC/03 Collect data on the number of addresses (recipients) as well
1913 as the number of messages in eximstats.
1915 TF/01 Correct an error in the documentation for the redirect router. Exim
1916 does (usually) call initgroups() when daemonizing.
1918 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1919 with consistent privilege compared to when running as a daemon.
1921 TF/03 Note in the spec that $authenticated_id is not set for local
1922 submissions from trusted users.
1924 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1925 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1927 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1928 by adding some example configuration directives to the default
1929 configuration file. A little bit of work is required to uncomment the
1930 directives and define how usernames and passwords are checked, but
1931 there is now a framework to start from.
1933 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1934 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1935 without this. I don't know how relevant this is to other LDAP libraries.
1937 PH/02 Add the verb name to the "unknown ACL verb" error.
1939 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1941 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1943 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1944 directory not expanded when it should be if an expanded home directory
1945 was set for the address (which is overridden by the transport).
1947 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1950 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1951 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1952 because it is too late at that time, and has no effect.
1954 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1955 security issue with \' (bugzilla #107). I could not use the
1956 PQescapeStringConn() function, because it needs a PGconn value as one of
1959 PH/08 When testing addresses using -bt, indicate those final addresses that
1960 are duplicates that would not cause an additional delivery. At least one
1961 person was confused, thinking that -bt output corresponded to deliveries.
1962 (Suppressing duplicates isn't a good idea as you lose the information
1963 about possibly different redirections that led to the duplicates.)
1965 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1966 systems where poll() doesn't work, in particular OS X.
1968 PH/10 Added more information to debugging output for retry time not reached.
1970 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1971 operations in malware.c.
1973 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1976 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1977 syslog_duplication set false, log lines that would normally be written
1978 both the the main log and to the reject log were not written to syslog at
1981 PH/14 In the default configuration, change the use of "message" in ACL warn
1982 statements to "add_header".
1984 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1985 not followed by a command (e.g. "seen endif").
1987 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1988 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1991 PH/17 Added extra conditions to the default value of delay_warning_condition
1995 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1996 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1997 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2000 The Auto-Submitted: and various List- headers are standardised, whereas I
2001 don't think Precedence: ever was.
2003 PH/18 Refactored debugging code in route_finduser() to show more information,
2004 in particular, the error code if getpwnam() issues one.
2006 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2007 This is apparently needed in addition to the PH/07 change above to avoid
2008 any possible encoding problems.
2010 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2011 but not after initializing Perl.
2013 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2014 output them only if debugging. By default they are written stderr,
2015 apparently, which is not desirable.
2017 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2020 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2023 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2025 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2026 authenticated or an ident call has been made. Suppress the default
2027 values for $authenticated_id and $authenticated_sender (but permit -oMai
2028 and -oMas) when testing with -bh.
2030 PH/24 Re-jigged the order of the tests in the default configuration so that the
2031 tests for valid domains and recipients precede the DNS black list and CSA
2032 tests, on the grounds that those ones are more expensive.
2034 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2035 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2036 This bug exists in every version of Exim that I still have, right back to
2039 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2040 However, an attempt to turn on case-sensitivity in a regex key by
2041 including (?-i) didn't work because the subject string was already
2042 lowercased, and the effects were non-intuitive. It turns out that a
2043 one-line patch can be used to allow (?-i) to work as expected.
2049 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2050 other effects) broke the use of negated acl sub-conditions.
2052 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2055 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2056 "Deny" causes Exim to reject the incoming connection with a 554 error.
2057 Unfortunately, if there is a major crisis, such as a disk failure,
2058 tcp-wrappers gives "deny", whereas what one would like would be some
2059 kind of temporary error. A kludge has been added to help with this.
2060 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2061 554 error is used if errno is still zero or contains ENOENT (which occurs
2062 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2065 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2067 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2068 errors. Otherwise a message that provokes a temporary error (when other
2069 messages do not) can cause a whole host to time out.
2071 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2072 addresses were routed directly to files or pipes from a redirect router.
2073 File deliveries just didn't batch; pipe deliveries might have suffered
2076 PH/06 A failure to get a lock for a hints database would erroneously always say
2077 "Failed to get write lock", even when it was really a read lock.
2079 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2080 of 0600. This has been changed to use the value of the lockfile_mode
2081 option (which defaults to 0600).
2083 PH/08 Applied small patch from the Sieve maintainer.
2085 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2086 folder from quota calculations, a direct delivery into this folder messed
2087 up the contents of the maildirsize file. This was because the regex was
2088 used only to exclude .Trash (or whatever) when the size of the mailbox
2089 was calculated. There was no check that a delivery was happening into an
2090 excluded directory. This bug has been fixed by ignoring all quota
2091 processing for deliveries into excluded directories.
2093 PH/10 Added the maildirfolder_create_regex option to appendfile.
2099 PH/01 The code for finding all the local interface addresses on a FreeBSD
2100 system running IPv6 was broken. This may well have applied to all BSD
2101 systems, as well as to others that have similar system calls. The broken
2102 code found IPv4 interfaces correctly, but gave incorrect values for the
2103 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2104 that it would not match correctly against @[] and not recognize the IPv6
2107 PH/02 The ipliteral router was not recognizing addresses of the form user@
2108 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2110 PH/03 Added disable_ipv6.
2112 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2113 decoded form, because it is most often used to construct To: headers
2114 lines in autoreplies, and the decoded form may well be syntactically
2115 invalid. However, $reply_address has leading white space removed, and all
2116 newlines turned into spaces so that the autoreply transport does not
2119 PH/05 If group was specified without a user on a router, and no group or user
2120 was specified on a transport, the group from the router was ignored.
2122 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2123 for visible compile-time settings that can be used to change these
2124 numbers, for those that want even more. Backwards compatibility with old
2125 spool files has been maintained. However, going back to a previous Exim
2126 release will lost any variables that are in spool files.
2128 PH/07 Two small changes when running in the test harness: increase delay when
2129 passing a TCP/IP connection to a new process, in case the original
2130 process has to generate a bounce, and remove special handling of
2131 127.0.0.2 (sic), which is no longer necessary.
2133 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2134 be the same on different OS.
2136 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2139 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2140 whether --show-vars was specified or not
2142 JJ/02 exipick: Added support for new ACL variable spool format introduced
2145 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2146 syntactically invalid From: or Reply-to: line, and a filter used this to
2147 generate an autoreply, and therefore failed to obtain an address for the
2148 autoreply, Exim could try to deliver to a non-existent relative file
2149 name, causing unrelated and misleading errors. What now happens is that
2150 it logs this as a hard delivery error, but does not attempt to create a
2153 PH/11 The exinext utility has a -C option for testing purposes, but although
2154 the given file was scanned by exinext itself; it wasn't being passed on
2155 when Exim was called.
2157 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2158 an end-of-file indication when reading a command response.
2160 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2161 compiled. In many other places in Exim, IPv6 addresses are always
2162 recognized, so I have changed this. It also means that IPv4 domain
2163 literals of the form [IPV4:n.n.n.n] are now always recognized.
2165 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2166 used if the router is not running as root, for example, when verifying at
2167 ACL time, or when using -bh. The debugging output from this situation was
2168 non-existent - all you got was a failure to exec. I have made two
2171 (a) Failures to set uid/gid, the current directory, or a process leader
2172 in a subprocess such as that created by queryprogram now generate
2173 suitable debugging ouput when -d is set.
2175 (b) The queryprogram router detects when it is not running as root,
2176 outputs suitable debugging information if -d is set, and then runs
2177 the subprocess without attempting to change uid/gid.
2179 PH/15 Minor change to Makefile for building test_host (undocumented testing
2182 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2183 additional section of a DNS packet that returns MX or SRV records.
2184 Instead, it always explicitly searches for A/AAAA records. This avoids
2185 major problems that occur when a DNS server includes only records of one
2186 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2187 fixed another bug: if SRV records were looked up and the corresponding
2188 address records were *not* found in the additional section, the port
2189 values from the SRV records were lost.
2191 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2192 using the correct key (the original address) when searching the retry
2193 rules in order to find which one to use for generating the retry hint.
2195 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2196 adding the default one. Similarly, if it contains a Reply-To: header, the
2197 errors_reply_to option, if set, is not used.
2199 PH/19 When calculating a retry time, Exim used to measure the "time since
2200 failure" by looking at the "first failed" field in the retry record. Now
2201 it does not use this if it is later than than the arrival time of the
2202 message. Instead it uses the arrival time. This makes for better
2203 behaviour in cases where some deliveries succeed, thus re-setting the
2204 "first failed" field. An example is a quota failure for a huge message
2205 when small messages continue to be delivered. Without this change, the
2206 "time since failure" will always be short, possible causing more frequent
2207 delivery attempts for the huge message than are intended.
2208 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2210 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2211 $1, $2, $3) because the numerical variables can be reset during some
2212 expansion items (e.g. "match"), thereby losing the authentication data.
2214 PH/21 Make -bV show the size of off_t variables so that the test suite can
2215 decide whether to run tests for quotas > 2G.
2217 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2218 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2219 filecount value is greater than 2G or if a quota value is greater than 2G
2220 on a system where the size of off_t is not greater than 4, a panic error
2223 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2224 never match. The debug and -bh output now contains an explicit error
2225 message indicating a malformed IPv4 address or mask.
2227 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2228 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2229 PH/23 above applies.
2231 PH/25 Do not write to syslog when running in the test harness. The only
2232 occasion when this arises is a failure to open the main or panic logs
2233 (for which there is an explicit test).
2235 PH/26 Added the /no_tell option to "control=freeze".
2237 PH/27 If a host name lookup failed very early in a connection, for example, if
2238 the IP address matched host_lookup and the reverse lookup yielded a name
2239 that did not have a forward lookup, an error message of the form "no IP
2240 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2241 could be logged. Now it outputs the IP address instead of "NULL".
2243 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2244 allows the sender and the authenticated sender to be set when
2245 submitting a message from within Exim. Since child_open_exim() is
2246 documented for local_scan(), the new function should be too.
2248 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2249 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2250 results in an empty string is now treated as unset.
2252 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2254 PH/31 Added sender_verify_fail logging option.
2256 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2257 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2258 tidied the source and removed it altogether.
2260 PH/33 When a queue run was abandoned because the load average was too high, a
2261 log line was always written; now it is written only if the queue_run log
2262 selector is set. In addition, the log line for abandonment now contains
2263 information about the queue run such as the pid. This is always present
2264 in "start" and "stop" lines but was omitted from the "abandon" line.
2266 PH/34 Omit spaces between a header name and the colon in the error message that
2267 is given when verify = headers_syntax fails (if there are lots of them,
2268 the message gets confusing).
2270 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2271 names, as there are now some PTR records that contain slashes. This check
2272 is only to protect against broken name servers that fall over on strange
2273 characters, so the fact that it applies to all lookups doesn't matter.
2275 PH/36 Now that the new test suite is complete, we can remove some of the
2276 special code in Exim that was needed for the old test suite. For example,
2277 sorting DNS records because real resolvers return them in an arbitrary
2278 order. The new test suite's fake resolver always returns records in the
2281 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2282 bounces) except when queue_only is set, to avoid logging races between
2283 the different processes.
2285 PH/38 Panic-die if .include specifies a non-absolute path.
2287 PH/39 A tweak to the "H" retry rule from its user.
2289 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2290 a label. They prevented compilation on older perls.
2292 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2293 a warning to be raised on newish perls.
2295 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2296 on queue. Changes to match documented behaviour of showing count of
2297 messages matching specified criteria.
2299 PH/40 Changed the default ident timeout from 30s to 5s.
2301 PH/41 Added support for the use of login_cap features, on those BSD systems
2302 that have them, for controlling the resources used by pipe deliveries.
2304 PH/42 The content-scanning code uses fopen() to create files in which to put
2305 message data. Previously it was not paying any attention to the mode of
2306 the files. Exim runs with umask(0) because the rest of the code creates
2307 files with open(), and sets the required mode explicitly. Thus, these
2308 files were ending up world-writeable. This was not a big issue, because,
2309 being within the spool directory, they were not world-accessible. I have
2310 created a function called modefopen, which takes an additional mode
2311 argument. It sets umask(777), creates the file, chmods it to the required
2312 mode, then resets the umask. All the relevant calls to fopen() in the
2313 content scanning code have been changed to use this function.
2315 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2316 to 24 hours. This avoids potential overflow problems when processing G
2317 and H retry rules. I suspect nobody ever tinkers with this value.
2319 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2321 PH/45 When the plaintext authenticator is running as a client, the server's
2322 challenges are checked to ensure they are valid base64 strings. By
2323 default, the authentication attempt is cancelled if an invalid string is
2324 received. Setting client_ignore_invalid_base64 true ignores these errors.
2325 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2326 they are received. Thus, the responses can be made to depend on the
2327 challenges. If an invalid string is ignored, an empty string is placed in
2330 PH/46 Messages that are created by the autoreply transport now contains a
2331 References: header, in accordance with RFCs 2822 and 3834.
2333 PH/47 Added authenticated_sender_force to the smtp transport.
2335 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2337 PH/49 Installed latest patch from the Sieve maintainer.
2339 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2340 was also set, the appendfile transport was unnecessarily scanning a
2341 directory of message files (e.g. for maildir delivery) to find the count
2342 of files (along with the size), even though it did not need this
2343 information. It now does the scan only if it needs to find either the
2344 size of the count of files.
2346 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2348 PH/52 Two bugs concerned with error handling when the smtp transport is
2351 (i) Exim was not creating retry information for temporary errors given
2352 for individual recipients after the DATA command when the smtp transport
2353 was used in LMTP mode. This meant that they could be retried too
2354 frequently, and not timed out correctly.
2356 (ii) Exim was setting the flag that allows error details to be returned
2357 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2358 recipients that were returned after the DATA command.
2360 PH/53 This is related to PH/52, but is more general: for any failing address,
2361 when detailed error information was permitted to be returned to the
2362 sender, but the error was temporary, then after the final timeout, only
2363 "retry timeout exceeded" was returned. Now it returns the full error as
2364 well as "retry timeout exceeded".
2366 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2367 do this, and (what is worse) MTAs that accept it.
2369 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2370 will now be deprecated.
2372 PH/56 New os.c-cygwin from the Cygwin maintainer.
2374 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2375 formats (previously only available in exim formats via -bpr, -bpru,
2376 and -bpra. Now also available in native and exiqgrep formats)
2378 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2379 with very large, slow to parse queues
2381 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2383 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2385 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2386 responses to authentication challenges, though it was showing the
2387 challenges; (ii) I've removed the CR characters from the debug output for
2390 PH/58 Allow for the insertion of a newline as well as a space when a string
2391 is turned into more than one encoded-word during RFC 2047 encoding. The
2392 Sieve code now uses this.
2394 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2395 data_4xx, lost_connection, tls_required.
2397 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2398 message was being sent as an SMTP response.
2400 PH/61 Add -l and -k options to exicyclog.
2402 PH/62 When verifying, if an address was redirected to one new address, so that
2403 verification continued, and the new address failed or deferred after
2404 having set something in $address_data, the value of $address_data was not
2405 passed back to the ACL. This was different to the case when no
2406 redirection occurred. The value is now passed back in both cases.
2408 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2409 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2410 use login_cap.h, so on its own it isn't the distinguishing feature. The
2411 new name refers directly to the setclassresources() function.
2413 PH/65 Added configuration files for NetBSD3.
2415 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2417 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2418 is preferred over IPv4.
2420 PH/68 The bounce_return_message and bounce_return_body options were not being
2421 honoured for bounces generated during the reception of non-SMTP messages.
2422 In particular, this applied to messages rejected by the ACL. This bug has
2423 been fixed. However, if bounce_return_message is true and bounce_return_
2424 body is false, the headers that are returned for a non-SMTP message
2425 include only those that have been read before the error was detected.
2426 (In the case of an ACL rejection, they have all been read.)
2428 PH/69 The HTML version of the specification is now built in a directory called
2429 spec_html instead of spec.html, because the latter looks like a path with
2430 a MIME-type, and this confuses some software.
2432 PH/70 Catch two compiler warnings in sieve.c.
2434 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2435 function verify_get_ident() calls ip_connect() to connect a socket, but
2436 if the "connect()" function timed out, ip_connect() used to close the
2437 socket. However, verify_get_ident() also closes the socket later, and in
2438 between Exim writes to the log, which may get opened at this point. When
2439 the socket was closed in ip_connect(), the log could get the same file
2440 descriptor number as the socket. This naturally causes chaos. The fix is
2441 not to close the socket in ip_connect(); the socket should be closed by
2442 the function that creates it. There was only one place in the code where
2443 this was missing, in the iplookup router, which I don't think anybody now
2444 uses, but I've fixed it anyway.
2446 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2447 well as to direct DNS lookups. Otherwise the handling of names in host
2448 lists is inconsistent and therefore confusing.
2454 PH/01 Two changes to the default runtime configuration:
2456 (1) Move the checks for relay_from_hosts and authenticated clients from
2457 after to before the (commented out) DNS black list checks.
2459 (2) Add control=submission to the relay_from_hosts and authenticated
2460 clients checks, on the grounds that messages accepted by these
2461 statements are most likely to be submissions.
2463 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2465 (1) Generate an error if the third argument for the ${prvs expansion is
2468 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2471 (3) Reset the variables that are obtained from the first argument of
2472 ${prvscheck and used in the second argument before leaving the code,
2473 because their memory is reclaimed, so using them afterwards may do
2476 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2477 one (it's much easier than Tom thought :-).
2479 (5) Because of (4), we can now allow for the use of $prvscheck_result
2480 inside the third argument.
2482 PH/03 For some reason, the default setting of PATH when running a command from
2483 a pipe transport was just "/usr/bin". I have changed it to
2486 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2487 anything to be listed in the output from -bV.
2489 PH/05 When a filter generated an autoreply, the entire To: header line was
2490 quoted in the delivery log line, like this:
2492 => >A.N.Other <ano@some.domain> <original@ddress> ...
2494 This has been changed so that it extracts the operative address. There
2495 may be more than one such address. If so, they are comma-separated, like
2498 => >ano@some.domain,ona@other.domain <original@ddress> ...
2500 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2501 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2502 not being looked up in the rDNS to get a host name, Exim was showing the
2503 IP address twice in Received: lines, even though the IP addresses were
2504 identical. For example:
2506 Received: from [1.2.3.4] (helo=[1.2.3.4])
2508 However, if the real host name was known, it was omitting the HELO data
2509 if it matched the actual IP address. This has been tidied up so that it
2510 doesn't show the same IP address twice.
2512 PH/07 When both +timestamp and +memory debugging was on, the value given by
2513 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2514 called by the debug printing, thereby overwriting the timestamp buffer.
2515 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2517 PH/08 When the original message was included in an autoreply transport, it
2518 always said "this is a copy of the message, including all the headers",
2519 even if body_only or headers_only was set. It now gives an appropriate
2522 PH/09 Applied a patch from the Sieve maintainer which:
2524 o fixes some comments
2525 o adds the (disabled) notify extension core
2526 o adds some debug output for the result of if/elsif tests
2527 o points to the current vacation draft in the documentation
2528 and documents the missing references header update
2532 o fixes a bug in processing the envelope test (when testing
2533 multiple envelope elements, the last element determinted the
2536 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2537 Electronic Mail") by including:
2539 Auto-submitted: auto-generated
2541 in the messages that it generates (bounce messages and others, such as
2542 warnings). In the case of bounce messages for non-SMTP mesages, there was
2543 also a typo: it was using "Auto_submitted" (underscore instead of
2544 hyphen). Since every message generated by Exim is necessarily in response
2545 to another message, thes have all been changed to:
2547 Auto-Submitted: auto-replied
2549 in accordance with these statements in the RFC:
2551 The auto-replied keyword:
2553 - SHOULD be used on messages sent in direct response to another
2554 message by an automatic process,
2556 - MUST NOT be used on manually-generated messages,
2558 - MAY be used on Delivery Status Notifications (DSNs) and Message
2559 Disposition Notifications (MDNs),
2561 - MUST NOT be used on messages generated by automatic or periodic
2562 processes, except for messages which are automatic responses to
2565 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2566 to the default Received: header definition.
2568 PH/12 Added log selector acl_warn_skipped (default on).
2570 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2571 variables because (a) they are in the wrong storage pool and (b) even if
2572 they were copied, it wouldn't work properly because of the caching.
2574 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2575 checking when decoding. Apparently there are clients that generate
2576 overlong encoded strings. Why am I not surprised?
2578 PH/15 If the first argument of "${if match_address" was not empty, but did not
2579 contain an "@" character, Exim crashed. Now it writes a panic log message
2580 and treats the condition as false.
2582 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2584 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2585 extension "envelope-auth". The code is finished and in agreement with
2586 other implementations, but there is no documentation so far and in fact,
2587 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2588 not changing the active code.
2590 Print executed "if" and "elsif" statements when debugging is used. This
2591 helps a great deal to understand what a filter does.
2593 Document more things not specified clearly in RFC3028. I had all this
2594 sorted out, when out of a sudden new issues came to my mind. Oops."
2596 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2599 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2600 canonical form (as documented). However, after a host name lookup from
2601 the IP address, check_host() was doing a simple string comparison with
2602 addresses acquired from the DNS when checking that the found name did
2603 have the original IP as one of its addresses. Since any found IPv6
2604 addresses are likely to be in abbreviated form, the comparison could
2605 fail. Luckily, there already exists a function for doing the comparison
2606 by converting both addresses to binary, so now that is used instead of
2607 the text comparison.
2609 PH/20 There was another similar case to PH/19, when a complete host name was
2610 given in a host list; looking up its IP address could give an abbreviated
2611 form, whereas the current host's name might or might not be abbreviated.
2612 The same fix has been applied.
2618 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2619 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2622 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2623 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2625 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2627 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2628 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2629 Also, for "Auto-Submitted", treat anything other than "no" as
2630 non-personal, in accordance with RFC 3834. (Previously it treated
2631 anything starting "auto-" as non-personal.)
2633 TF/01 The control=submission/name=... option had a problem with syntax
2634 errors if the name included a slash character. The /name= option
2635 now slurps the rest of the string, so it can include any characters
2636 but it must come last in the list of options (after /sender_retain
2639 PH/05 Some modifications to the interface to the fake nameserver for the new
2647 TK/01 Added the "success_on_redirect" address verification option. See
2648 NewStuff for rationale and an example.
2650 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2652 PH/02 Patch to exigrep to allow it to work on syslog lines.
2654 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2655 fread() to skip over the body file's header line, because in Cygwin the
2656 header line is locked and is inaccessible.
2658 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2659 co-exist for some time) to make it clear that it is the Exim ID that is
2660 referenced, not the Message-ID: header line.
2662 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2663 string_format() function, because snprintf() does not exist on all
2666 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2667 ${stat: expansion item.
2669 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2670 protocol synchronization error", to keep the pedants happy.
2672 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2673 well as for IRIX systems, when gcc is being used. See the host.c source
2676 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2678 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2681 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2682 transport and to the smtp transport in LMTP mode.
2684 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2686 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2687 run a filter in a subprocess. This could lead to confusion in subsequent
2688 lookups in the parent process. There should also be a search_tidyup() at
2689 the end of the subprocess.
2691 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2692 only if the host matched helo_try_verify_hosts, which caused the
2693 verification to occur when the EHLO/HELO command was issued. The ACL just
2694 tested the remembered result. Now, if a previous verification attempt has
2695 not happened, "verify = helo" does it there and then.
2697 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2699 TK/03 Fix log output including CR from clamd.
2701 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2702 exist provoked a memory error which could cause a segfault.
2704 PH/15 Installed PCRE 6.2
2706 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2708 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2709 of the problem. Specifically, suggested +O2 rather than +O1 for the
2712 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2714 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2715 record, the port was not correctly logged when the outgoing_port log
2716 selector was set (it logged the transort's default port).
2718 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2719 fallback_hosts, and "hosts" in the smtp transport.
2721 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2722 host errors such as "Connection refused".
2724 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2725 authentication with radiusclient 0.4.9:
2727 - Error returned from rc_read_config was caught wrongly
2728 - Username/password not passed on to radius server due to wrong length.
2730 The presumption is that some radiusclient API changes for 4.51/PH/17
2731 were not taken care of correctly. The code is still untested by me (my
2732 Linux distribution still has 0.3.2 of radiusclient), but it was
2733 contributed by a Radius user.
2735 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2736 expanding the "port" option of the smtp transport.
2738 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2739 while reading a MIME header. Thanks to Tom Hughes for a patch.
2741 PH/24 Include config.h inside local_scan.h so that configuration settings are
2744 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2745 that in an ACL for RCPT (for example), you can examine exactly what was
2748 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2749 commands, but it was not correctly comparing the address with the actual
2750 client host address. Thus, it would show the EHLO address in Received:
2751 header lines when this was not necessary.
2753 PH/27 Added the % operator to ${eval:}.
2755 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2756 it should be ignoring failures (because with -C, for example, it has lost
2757 privilege). It wasn't ignoring creation failures other than "already
2760 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2763 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2764 because an input line was too long, either on its own, or by virtue of
2765 too many continuations, the temporary file was not being removed, and the
2766 return code was incorrect.
2768 PH/31 Missing "BOOL" in function definition in filtertest.c.
2770 PH/32 Applied Sieve patches from the maintainer.
2772 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2774 PH/33 Added "verify = not_blind".
2776 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2777 Local/Makefile (with some defaults set). These are used in built scripts
2778 such as exicyclog, but they have never been used in the exim_install
2779 script (though there are many overriding facilities there). I have
2780 arranged that the exim_install script now takes note of these two
2783 PH/35 Installed configuration files for Dragonfly.
2785 PH/36 When a locally submitted message by a trusted user did not contain a
2786 From: header, and the sender address was obtained from -f or from an SMTP
2787 MAIL command, and the trusted user did not use -F to supply a sender
2788 name, $originator_name was incorrectly used when constructing a From:
2789 header. Furthermore, $originator_name was used for submission mode
2790 messages from external hosts without From: headers in a similar way,
2791 which is clearly wrong.
2793 PH/37 Added control=suppress_local_fixups.
2795 PH/38 When log_selector = +received_sender was set, and the addition of the
2796 sender made the log line's construction buffer exactly full, or one byte
2797 less than full, an overflow happened when the terminating "\n" was
2800 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2801 when the result of a list match is failure because a DNS lookup failed.
2803 PH/40 RM_COMMAND is now used in the building process.
2805 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2806 the "build-* directories that it finds.
2808 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2809 address in a domain literal was a prefix of an interface address.
2811 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2812 when verifying a sender address, unless rewrite_headers is false.
2814 PH/44 Wrote a long comment about why errors_to addresses are verified as
2815 recipients, not senders.
2817 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2818 the ratelimit ACL was added.
2820 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2822 PH/46 Added extra information about PostgreSQL errors to the error string.
2824 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2825 avoiding the need to install special zones in a real server. This is
2826 backwards compatible; if it can't find the fake resolver, it drops back.
2827 Thus, both old and new test suites can be run.
2829 TF/02 Added util/ratelimit.pl
2831 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2832 clock is set back in time.
2834 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2835 Candler <B.Candler@pobox.com>.
2837 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2838 OK for addresses that are the result of redirections.
2840 PH/48 A number of further additions for the benefit of the new test suite,
2841 including a fake gethostbyname() that interfaces to the fake DNS resolver
2844 TF/06 The fix for widen_domains has also been applied to qualify_single and
2845 search_parents which are the other dnslookup options that can cause
2848 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2851 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2853 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2854 service" if it encountered an MX record with an empty target hostname.
2855 The message is now "an MX or SRV record indicated no SMTP service".
2857 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2858 if the DNS of the sending site is misconfigured. This is quite a
2859 common situation. This change restores the behaviour of treating a
2860 helo verification defer as a failure.
2862 PH/51 If self=fail was set on a router, the bounce message did not include the
2863 actual error message.
2869 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2871 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2872 command itself ended in error, the underlying message about the transport
2873 filter timeout was being overwritten with the pipe command error. Now the
2874 underlying error message should be appended to the second error message.
2876 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2878 PH/02 Reduce the amount of output that "make" produces by default. Full output
2879 can still be requested.
2881 PH/03 The warning log line about a condition test deferring for a "warn" verb
2882 was being output only once per connection, rather than after each
2883 occurrence (because it was using the same function as for successful
2884 "warn" verbs). This seems wrong, so I have changed it.
2886 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2887 it should not have, which might have caused a crash in the right
2888 circumstances, but probably never did.
2890 PH/04 Installed a modified version of Tony Finch's patch to make submission
2891 mode fix the return path as well as the Sender: header line, and to
2892 add a /name= option so that you can make the user's friendly name appear
2895 TF/03 Added the control = fakedefer ACL modifier.
2897 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2898 Mark Lowes for thorough testing.
2900 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2902 TK/03 Merged latest SRS patch from Miles Wilton.
2904 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2905 with the definition in sysexits.h (which is #included earlier).
2906 Fortunately, Exim does not actually use EX_OK. The code used to try to
2907 preserve the sysexits.h value, by assumimg that macro definitions were
2908 scanned for macro replacements. I have been disabused of this notion,
2909 so now the code just undefines EX_OK before #including unistd.h.
2911 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2912 in the smtp transport. When a block could not be written in a single
2913 write() function, the timeout was being re-applied to each part-write.
2914 This seems wrong - if the receiver was accepting one byte at a time it
2915 would take for ever. The timeout is now adjusted when this happens. It
2916 doesn't have to be particularly precise.
2918 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2919 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2921 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2922 without a domain if the check to <postmaster@domain> fails.
2924 SC/01 Eximstats: added -xls and the ability to specify output files
2925 (patch written by Frank Heydlauf).
2927 SC/02 Eximstats: use FileHandles for outputing results.
2929 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2931 SC/04 Eximstats: fixed display of large numbers with -nvr option
2933 SC/05 Eximstats: fixed merging of reports with empty tables.
2935 SC/06 Eximstats: added the -include_original_destination flag
2937 SC/07 Eximstats: removed tabs and trailing whitespace.
2939 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2941 TK/06 MBOX spool code: Add real "From " MBOX separator line
2942 so the .eml file is really in mbox format (even though
2943 most programs do not really care). Patch from Alex Miller.
2945 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2946 The latter is generated from $received_to and is only set if the
2947 message has one envelope recipient. SA can use these headers,
2948 obviously out-of-the-box. Patch from Alex Miller.
2950 PH/08 The ${def test on a variable was returning false if the variable's
2951 value was "0", contrary to what the specification has always said!
2952 The result should be true unless the variable is empty.
2954 PH/09 The syntax error of a character other than { following "${if
2955 def:variable_name" (after optional whitespace) was not being diagnosed.
2956 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2957 accidental colon was present, for example, could give incorrect results.
2959 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2960 result is used (not including appendfile, where other changes are about
2963 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2964 This involved changing a lot of size variables from int to off_t. It
2965 should work with maildirs and everything.
2967 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2968 spamd dying while we are connected to it.
2970 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2973 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2974 function for BDB 4.3.
2976 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2978 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2979 the sources for PCRE have been re-arranged and more files are now
2982 PH/15 The code I had for printing potentially long long variables in PH/11
2983 above was not the best (it lost precision). The length of off_t variables
2984 is now inspected at build time, and an appropriate printing format (%ld
2985 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2986 to be "long long int" or "long int". This is needed for the internal
2987 formatting function string_vformat().
2989 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2990 the configuration file to be ":syslog", then the script "guesses" where
2991 the logs files are, rather than using the compiled in default. In our
2992 case the guess is not the same as the compiled default, so the script
2993 suddenly stopped working when I started to use syslog. The patch checks
2994 to see if log_file_path is "". If so, it attempts to read it from exim
2995 with no configuration file to get the compiled in version, before it
2996 falls back to the previous guessing code."
2998 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2999 implementing BATV in an Exim configuration. See NewStuff for the gory
3002 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3003 Makefile that are specific to HP-UX.
3005 PH/18 If the "use_postmaster" option was set for a recipient callout together
3006 with the "random" option, the postmaster address was used as the MAIL
3007 FROM address for the random test, but not for the subsequent recipient
3008 test. It is now used for both.
3010 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3011 patch removes a few documentation additions to RFC 3028, because the
3012 latest draft now contains them. It adds the new en;ascii-case comparator
3013 and a new error check for 8bit text in MIME parts. Comparator and
3014 require names are now matched exactly. I enabled the subaddress
3015 extension, but it is not well tested yet (read: it works for me)."
3017 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3018 rework some of the code of TK/09 above to avoid the hardwired use of
3019 "%lld" and "long long". Replaced the call to snprintf() with a call to
3022 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3023 records point to non-existent hosts", "retry timeout exceeded", and
3024 "retry time not reached for any host after a long failure period".
3026 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3027 experimental DomainKeys support:
3029 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3030 (2) On an error such as an illegally used "control", the wrong name for
3031 the control was given.
3033 These problems did NOT occur unless DomainKeys support was compiled.
3035 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3037 PH/24 Added ${if match_ip condition.
3039 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3040 databases so that it will be absolutely obvious if a crash occurs in the
3041 DB library. This is a regular occurrence (often caused by mis-matched
3044 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3045 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3046 picked up on a user's system that detects such things. There doesn't seem
3047 to be a gcc warning option for this - only an attribute that has to be
3048 put on the function's prototype. It seems that in Fedora Core 4 they have
3049 set this on a number of new functions. No doubt there will be more in due
3052 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3053 specify a transport. However, if an address that was verified by such a
3054 router was the subject of a callout, Exim crashed because it tried to
3055 read the rcpt_include_affixes from the non-existent transport. Now it
3056 just assumes that the setting of that option is false. This bug was
3057 introduced by 4.51/PH/31.
3059 PH/28 Changed -d+all to exclude +memory, because that information is very
3060 rarely of interest, but it makes the output a lot bigger. People tend to
3061 do -d+all out of habit.
3063 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3064 code in os-type was giving problems when libc.so lives in lib64, like on
3067 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3068 aren't the modern standard, and it seems that some systems' include files
3069 don't always have them. Exim was already checking for some of the newer
3070 ones like T_AAAA, and defining it itself. I've added checks for all the
3071 record types that Exim uses.
3073 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3074 not automatically generating a new one, as it is supposed to. This
3075 prevented TLS from working. If the file did exist, but contained invalid
3076 data, a new version was generated, as expected. It was only the case of a
3077 non-existent file that was broken.
3079 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3080 with a change in libdomainkeys > 0.64.
3082 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3083 from DNS. If the selector record carries the flag, it now has
3084 precedence over the domain-wide flag.
3086 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3088 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3089 the use of a port name that isn't defined in /etc/services) occurred, the
3090 message was deferred as in a normal delivery, and thus remained on the
3091 spool, instead of being failed because of the mua_wrapper setting. This
3092 is now fixed, and I tidied up some of the mua_wrapper messages at the
3095 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3096 the messages in a hash of arrays rather than using individual hashes.
3097 This is a bit cleaner and results in dramatic memory savings, albeit
3098 at a slight CPU cost.
3100 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3101 as requested by Marc Sherman.
3103 SC/10 Eximstats: added histograms for user specified patterns as requested
3106 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3108 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3109 fopen() in the content-scanning modules that did not already have it.
3115 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3116 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3118 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3120 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3122 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3123 to test Sieve filters that use "vacation".
3125 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3126 that changes the way the GnuTLS parameters are stored in the cache file.
3127 The new format can be generated externally. For backward compatibility,
3128 if the data in the cache doesn't make sense, Exim assumes it has read an
3129 old-format file, and it generates new data and writes a new file. This
3130 means that you can't go back to an older release without removing the
3133 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3134 work if there are any delivery delays because "one_time" forces the
3135 parent to be marked "delivered", so its unseen clone is never tried
3136 again. For this reason, Exim now forbids the simultaneous setting of
3139 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3140 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3141 entry if you want to know the details. The fix, however, neglected to
3142 consider the case when local delivery batching is involved. The test for
3143 "previously delivered" was not happening when checking to see if an
3144 address could be batched with a previous (undelivered) one; under
3145 certain circumstances this could lead to multiple deliveries to the same
3148 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3149 in its include files, and this causes problems building Exim.
3151 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3152 header_syntax) but Exim was just ignoring anything given after a slash.
3153 In particular, this caused confusion with an attempt to use "verify =
3154 reverse_host_lookup/defer_ok". An error is now given when options are
3155 supplied for verify items that do not have them. (Maybe reverse_host_
3156 lookup should have a defer_ok option, but that's a different point.)
3158 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3159 defined by RFC 821) to 2048, because there were problems with some AUTH
3160 commands, and RFC 1869 says the size should be increased for extended
3161 SMTP commands that take arguments.
3163 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3166 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3167 "unknown" error; now it says that the functionality isn't in the binary.
3169 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3170 an address' error message when a string expansion fails (syntax or
3171 whatever). Otherwise the password may appear in the log. Following change
3172 PH/42 below, there is no longer a chance of it appearing in a bounce
3175 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3177 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3178 of its addresses was ever tried. (Bugzilla bug #2.)
3180 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3181 the result incorrectly in the debug output. (It correctly added a newline
3182 to what was transported.)
3184 TF/01 Added $received_time.
3186 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3187 commented out examples of how to interface to a virus scanner and to
3188 SpamAssassin. Also added commented examples of av_scanner and
3189 spamd_address settings.
3191 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3192 and controls are allowed in which ACLs. There were a couple of minor
3193 errors. Some of the entries in the conditions table (which is a table of
3194 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3195 negation of where the condition IS allowed.
3197 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3199 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3200 header file does not have a version number, so I've had to invent a new
3201 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3202 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3203 radiusclient), but it was contributed by a Radius user.
3205 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3206 files or directories, for OpenSSL.
3208 PH/19 When an Exim process that is running as root has to create an Exim log
3209 file, it does so in a subprocess that runs as exim:exim so as to get the
3210 ownership right at creation (otherwise, other Exim processes might see
3211 the file with the wrong ownership). There was no test for failure of this
3212 fork() call, which would lead to the process getting stuck as it waited
3213 for a non-existent subprocess. Forks do occasionally fail when resources
3214 run out. I reviewed all the other calls to fork(); they all seem to check
3217 PH/20 When checking for unexpected SMTP input at connect time (before writing
3218 the banner), Exim was not dealing correctly with a non-positive return
3219 from the read() function. If the client had disconnected by this time,
3220 the result was a log entry for a synchronization error with an empty
3221 string after "input=" when read() returned zero. If read() returned -1
3222 (an event I could not check), uninitialized data bytes were printed.
3223 There were reports of junk text (parts of files, etc) appearing after
3226 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3228 PH/22 Added support for macro redefinition, and (re)definition in between
3229 driver and ACL definitions.
3231 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3232 forgetting to use the resulting value; it was using the unexpanded value.
3234 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3235 hadn't been configured. The fix is from Juergen Kreileder, who
3236 understands it better than I do:
3238 "Here's what I see happening with three configured cyrus_sasl
3239 authenticators configured (plain, login, cram-md5):
3241 On startup auth_cyrus_sasl_init() gets called for each of these.
3242 This means three calls to sasl_listmech() without a specified mech_list.
3243 => SASL tests which mechs of all available mechs actually work
3244 => three warnings about OTP not working
3245 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3247 With the patch, sasl_listmech() also gets called three times. But now
3248 SASL's mech_list option is set to the server_mech specified in the the
3249 authenticator. Or in other words, the answer from sasl_listmech()
3250 gets limited to just the mech you're testing for (which is different
3252 => the return list contains just 'plain' or 'login', 'cram-md5' or
3253 nothing depending on the value of ob->server_mech.
3255 I've just tested the patch: Authentication still works fine,
3256 unavailable mechs specified in the exim configuration are still
3257 caught, and the auth.log warnings about OTP are gone."
3259 PH/25 When debugging is enabled, the contents of the command line are added
3260 to the debugging output, even when log_selector=+arguments is not
3263 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3264 answer is "GNU", and only if the return is "GNU/something" is the answer
3267 PH/27 $acl_verify_message is now set immediately after the failure of a
3268 verification in an ACL, and so is available in subsequent modifiers. In
3269 particular, the message can be preserved by coding like this:
3271 warn !verify = sender
3272 set acl_m0 = $acl_verify_message
3274 Previously, $acl_verify_message was set only while expanding "message"
3275 and "log_message" when a very denied access.
3277 PH/28 Modified OS/os.c-Linux with
3279 -#ifndef OS_LOAD_AVERAGE
3280 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3282 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3283 nomenclature these days.)
3285 PH/29 Installed patch from the Sieve maintainer that adds the options
3286 sieve_useraddress and sieve_subaddress to the redirect router.
3288 PH/30 In these circumstances:
3289 . Two addresses routed to the same list of hosts;
3290 . First host does not offer TLS;
3291 . First host accepts first address;
3292 . First host gives temporary error to second address;
3293 . Second host offers TLS and a TLS session is established;
3294 . Second host accepts second address.
3295 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3296 and peerdn, if requested) that were in fact used only for the second
3299 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3300 attention to any local part prefix or suffix that was matched by the
3301 router that accepted the address. It now behaves in the same way as it
3302 does for delivery: the affixes are removed from the local part unless
3303 rcpt_include_affixes is set on the transport.
3305 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3306 timeout during the DATA phase of an incoming message.
3308 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3309 applied a patch sanctioned by the Sieve maintainer.
3311 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3312 the uid or gid is negative. A case of a negative gid caused this to be
3313 noticed. The fix allows for either to be negative.
3315 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3316 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3319 PH/36 The change PH/12 above was broken. Fixed it.
3321 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3322 the grounds that routing the same address twice would always produce the
3323 same answer. This might have been true once, but it is certainly no
3324 longer true now. Routing a child address may depend on the previous
3325 routing that produced that child. Some complicated redirection strategies
3326 went wrong when messages had multiple recipients, and made Exim's
3327 behaviour dependent on the order in which the addresses were given.
3329 I have moved the duplicate checking until after the routing is complete.
3330 Exim scans the addresses that are assigned to local and remote
3331 transports, and removes any duplicates. This means that more work will be
3332 done, as duplicates will always all be routed, but duplicates are
3333 presumably rare, so I don't expect this is of any significance.
3335 For deliveries to pipes, files, and autoreplies, the duplicate checking
3336 still happens during the routing process, since they are not going to be
3339 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3340 It corrects a timeout issue with spamd. This is Ian's comment: "The
3341 background is that sometimes spamd either never reads data from a
3342 connection it has accepted, or it never writes response data. The exiscan
3343 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3344 blindly assumes that writes won't block so it may never time out."
3346 PH/39 Allow G after quota size as well as K and M.
3348 PH/40 The value set for $authenticated_id in an authenticator may not contain
3349 binary zeroes or newlines because the value is written to log lines and
3350 to spool files. There was no check on this. Now the value is run through
3351 the string_printing() function so that such characters are converted to
3352 printable escape sequences.
3354 PH/41 $message_linecount is a new variable that contains the total number of
3355 lines in the message. Compare $body_linecount, which is the count for the
3358 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3359 bounce and delay warning messages, except in certain special cases, which
3362 (a) An SMTP error message from a remote host;
3363 (b) A message specified in a :fail: redirection;
3364 (c) A message specified in a "fail" command in a system filter;
3365 (d) A message specified in a FAIL return from the queryprogram router;
3366 (e) A message specified by the cannot_route_message router option.
3368 In these cases only, Exim does include the error details in bounce and
3369 warning messages. There are also a few cases where bland messages such
3370 as "unrouteable address" or "local delivery error" are given.
3372 PH/43 $value is now also set for the "else" part of a ${run expansion.
3374 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3375 being worked on, but at least Exim now implements the latest version to
3378 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3379 process to complete was treated as a delivery failure, a timeout while
3380 writing the message to the pipe was logged, but erroneously treated as a
3381 successful delivery. Such timeouts include transport filter timeouts. For
3382 consistency with the overall process timeout, these timeouts are now
3383 treated as errors, giving rise to delivery failures by default. However,
3384 there is now a new Boolean option for the pipe transport called
3385 timeout_defer, which, if set TRUE, converts the failures into defers for
3386 both kinds of timeout. A transport filter timeout is now identified in
3389 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3390 systems where "make" and "gmake" are different, calling "gmake" at top
3391 level broke things. I've arranged for the value of $(MAKE) to be passed
3392 from the Makefile to this script so that it can call the same version of
3396 A note about Exim versions 4.44 and 4.50
3397 ----------------------------------------
3399 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3400 changes of various kinds. As a consequence, a big documentation update was
3401 needed. This delayed the release for rather longer than seemed good, especially
3402 in the light of a couple of (minor) security issues. Therefore, the changes
3403 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3404 release. So 4.44 and 4.50 are in effect two different branches that both start
3407 I have left the 4.50 change log unchanged; it contains all the changes since
3408 4.43. The change log for 4.44 is below; many of its items are identical to
3409 those for 4.50. This seems to be the most sensible way to preserve the
3410 historical information.
3416 1. Minor wording change to the doc/README.SIEVE file.
3418 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3419 computation of the current number of files was incorrect.
3421 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3422 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3425 4. Give more explanation in the error message when the command for a transport
3426 filter fails to execute.
3428 5. There are several places where Exim runs a non-Exim command in a
3429 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3430 was being done only for the command run by the queryprogram router. It is
3431 now done for all such subprocesses. The other cases are: ${run, transport
3432 filters, and the commands run by the lmtp and pipe transports.
3434 6. Added CONFIGURE_GROUP build-time option.
3436 7. Some older OS have a limit of 256 on the maximum number of file
3437 descriptors. Exim was using setrlimit() to set 1000 as a large value
3438 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3439 systems. I've change it so that if it can't get 1000, it tries for 256.
3441 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3442 was an oversight, and furthermore, ever since the addition of extra
3443 controls (e.g. 4.43/32), the checks on when to allow different forms of
3444 "control" were broken. There should now be diagnostics for all cases when a
3445 control that does not make sense is encountered.
3447 9. Added the /retain_sender option to "control=submission".
3449 10. $recipients is now available in the predata ACL (oversight).
3451 11. Tidy the search cache before the fork to do a delivery from a message
3452 received from the command line. Otherwise the child will trigger a lookup
3453 failure and thereby defer the delivery if it tries to use (for example) a
3454 cached ldap connection that the parent has called unbind on.
3456 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3457 of $address_data from the recipient verification was clobbered by the
3458 sender verification.
3460 13. The value of address_data from a sender verification is now available in
3461 $sender_address_data in subsequent conditions in the ACL statement.
3463 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3465 15. Added a new option "connect=<time>" to callout options, to set a different
3468 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3469 was its contents. (It was OK if the option was not defined at all.)
3471 17. A "Completed" log line is now written for messages that are removed from
3472 the spool by the -Mrm option.
3474 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3475 information about exactly what failed.
3477 19. Added -dd to debug only the daemon process.
3479 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3480 handles timeouts, both on the server side and network timeouts. Renamed the
3481 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3483 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3484 It is now set to "smtps".
3486 22. $host_address is now set to the target address during the checking of
3487 ignore_target_hosts.
3489 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3490 being passed; this would have caused $sender_host_name to have been used if
3491 matching the list had actually called for a host name (not very likely,
3492 since this list is usually IP addresses). A host name is now passed as
3495 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3496 code that specifies a non-restarting handler (typically sigaction() in
3497 modern systems) in an attempt to fix a rare and obscure crash bug.
3499 25. Narrowed the window for a race in the daemon that could cause it to ignore
3500 SIGCHLD signals. This is not a major problem, because they are used only to
3501 wake it up if nothing else does.
3503 26. A malformed maildirsize file could cause Exim to calculate negative values
3504 for the mailbox size or file count. Odd effects could occur as a result.
3505 The maildirsize information is now recalculated if the size or filecount
3508 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3509 support for a long time. Removed HAVE_SYS_VFS_H.
3511 28. Installed the latest version of exipick from John Jetmore.
3513 29. In an address list, if the pattern was not a regular expression, an empty
3514 subject address (from a bounce message) matched only if the pattern was an
3515 empty string. Non-empty patterns were not even tested. This was the wrong
3516 because it is perfectly reasonable to use an empty address as part of a
3517 database query. An empty address is now tested by patterns that are
3518 lookups. However, all the other forms of pattern expect the subject to
3519 contain a local part and a domain, and therefore, for them, an empty
3520 address still always fails if the pattern is not itself empty.
3522 30. Exim went into a mad DNS loop when attempting to do a callout where the
3523 host was specified on an smtp transport, and looking it up yielded more
3524 than one IP address.
3526 31. Re-factored the code for checking spool and log partition space into a
3527 function that finds that data and another that does the check. The former
3528 is then used to implement four new variables: $spool_space, $log_space,
3529 $spool_inodes, and $log_inodes.
3531 32. The RFC2047 encoding function was originally intended for short strings
3532 such as real names; it was not keeping to the 75-character limit for
3533 encoded words that the RFC imposes. It now respects the limit, and
3534 generates multiple encoded words if necessary. To be on the safe side, I
3535 have increased the buffer size for the ${rfc2047: expansion operator from
3538 33. It is now permitted to omit both strings after an "if" condition; if the
3539 condition is true, the result is "true". As before, when the second string
3540 is omitted, a false condition yields an empty string. This makes it less
3541 cumbersome to write custom ACL and router conditions.
3543 34. Failure to deliver a bounce message always caused it to be frozen, even if
3544 there was an errors_to setting on the router. The errors_to setting is now
3547 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3548 canonical form (fully expanded) before being placed in
3549 $sender_host_address.
3551 36. The table in the code that translates DNS record types into text (T_A to
3552 "A" for instance) was missing entries for NS and CNAME. It is just possible
3553 that this could have caused confusion if both these types were looked up
3554 for the same domain, because the text type is used as part of Exim's
3555 per-process caching. But the chance of anyone hitting this buglet seems
3558 37. The dnsdb lookup has been extended in a number of ways.
3560 (1) There is a new type, "zns", which walks up the domain tree until it
3561 finds some nameserver records. It should be used with care.
3563 (2) There is a new type, "mxh", which is like "mx" except that it returns
3564 just the host names, not the priorities.
3566 (3) It is now possible to give a list of domains (or IP addresses) to be
3567 looked up. The behaviour when one of the lookups defers can be
3568 controlled by a keyword.
3570 (4) It is now possible to specify the separator character for use when
3571 multiple records are returned.
3573 38. The dnslists ACL condition has been extended: it is now possible to supply
3574 a list of IP addresses and/or domains to be looked up in a particular DNS
3577 39. Added log_selector=+queue_time_overall.
3579 40. When running the queue in the test harness, wait just a tad after forking a
3580 delivery process, to get repeatability of debugging output.
3582 41. Include certificate and key file names in error message when GnuTLS fails
3583 to set them up, because the GnuTLS error message doesn't include the name
3584 of the failing file when there is a problem reading it.
3586 42. Allow both -bf and -bF in the same test run.
3588 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3590 44. The "Exiscan patch" is now merged into the mainline Exim source.
3592 45. Sometimes the final signoff response after QUIT could fail to get
3593 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3594 < 0 before doing a fflush(). This bug looks as though it goes back to the
3595 introduction of TLS in release 3.20, but "sometimes" must have been rare
3596 because the tests only now provoked it.
3598 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3599 (this can affect the format of dates).
3601 47. exim_tidydb, when checking for the continued existence of a message for
3602 which it has found a message-specific retry record, was not finding
3603 messages that were in split spool directories. Consequently, it was
3604 deleting retry records that should have stayed in existence.
3606 48. Steve fixed some bugs in eximstats.
3608 49. The SPA authentication driver was not abandoning authentication and moving
3609 on to the next authenticator when an expansion was forced to fail,
3610 contradicting the general specification for all authenticators. Instead it
3611 was generating a temporary error. It now behaves as specified.
3613 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3614 (the order specifies the preference for clients). The order is now AES256,
3615 AES128, 3DES, ARCFOUR128.
3617 51. Small patch to Sieve code - explicitly set From: when generating an
3620 52. Exim crashed if a remote delivery caused a very long error message to be
3621 recorded - for instance if somebody sent an entire SpamAssassin report back
3622 as a large number of 550 error lines. This bug was coincidentally fixed by
3623 increasing the size of one of Exim's internal buffers (big_buffer) that
3624 happened as part of the Exiscan merge. However, to be on the safe side, I
3625 have made the code more robust (and fixed the comments that describe what
3628 53. Now that there can be additional text after "Completed" in log lines (if
3629 the queue_time_overall log selector is set), a one-byte patch to exigrep
3630 was needed to allow it to recognize "Completed" as not the last thing in
3633 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3634 patch that reportedly fixes this has been added. I am not expert enough to
3635 create a test for it. This is what the patch creator wrote:
3637 "I found a little strange behaviour of ldap code when working with
3638 Windows 2003 AD Domain, where users was placed in more than one
3639 Organization Units. When I tried to give exim partial DN, the exit code
3640 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3641 But simultaneously result of request was absolutely normal ldap result,
3642 so I produce this patch..."
3644 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3645 REFERENCE, so I have modified the code to exclude the patch when that macro
3648 55. Some experimental protocols are using DNS PTR records for new purposes. The
3649 keys for these records are domain names, not reversed IP addresses. The
3650 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3651 leaves it alone. Component reversal etc. now happens only for IP addresses.
3654 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3656 57. Double the size of the debug message buffer (to 2048) so that more of very
3657 long debug lines gets shown.
3659 58. The exicyclog utility now does better if the number of log files to keep
3660 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3662 59. Two changes related to the smtp_active_hostname option:
3664 (1) $smtp_active_hostname is now available as a variable.
3665 (2) The default for smtp_banner uses $smtp_active_hostname instead
3666 of $primary_hostname.
3668 60. The host_aton() function is supposed to be passed a string that is known
3669 to be a valid IP address. However, in the case of IPv6 addresses, it was
3670 not checking this. This is a hostage to fortune. Exim now panics and dies
3671 if the condition is not met. A case was found where this could be provoked
3672 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3673 components; fortuitously, this particular loophole had already been fixed
3674 by change 4.50/55 above.
3676 If there are any other similar loopholes, the new check in host_aton()
3677 itself should stop them being exploited. The report I received stated that
3678 data on the command line could provoke the exploit when Exim was running as
3679 exim, but did not say which command line option was involved. All I could
3680 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3681 running as the user.
3684 61. There was a buffer overflow vulnerability in the SPA authentication code
3685 (which came originally from the Samba project). I have added a test to the
3686 spa_base64_to_bits() function which I hope fixes it.
3689 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3690 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3692 63. The daemon start-up calls getloadavg() while still root for those OS that
3693 need the first call to be done as root, but it missed one case: when
3694 deliver_queue_load_max is set with deliver_drop_privilege. This is
3695 necessary for the benefit of the queue runner, because there is no re-exec
3696 when deliver_drop_privilege is set.
3698 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3699 This has been fixed.
3701 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3702 was in use, was not putting the data itself into the right store pool;
3703 consequently, it could be overwritten for a subsequent message in the same
3704 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3707 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3709 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3710 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3711 the function were treating the return as a boolean value, which happened to
3712 work because 0=false and not-0=true, but is not correct code.
3714 68. The host_aton() function was not handling scoped IPv6 addresses (those
3715 with, for example, "%eth0" on the end) correctly.
3717 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3718 negated items (that is, ~something) in unsigned ints. Some compilers
3719 apparently mutter when there is no cast.
3721 70. If an address verification called from an ACL failed, and did not produce a
3722 user-specific message (i.e. there was only a "system" message), nothing was
3723 put in $acl_verify_message. In this situation, it now puts the system
3726 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3727 session; change 4.31/43 added the unwanted input to the log line - except
3728 that it did not do this in the start of session case. It now does.
3730 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3731 This is wrong and can cause the other end to generate a synchronization
3732 error if it is another Exim or anything else that does the synchronization
3733 check. A QUIT command is no longer sent after a timeout.
3735 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3736 during host lookups.
3738 74. The defer_ok option of callout verification was not working if it was used
3739 when verifying addresses in header lines, that is, for this case:
3741 verify = header_sender/callout=defer_ok
3743 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3744 those file descriptors could be used for SMTP connections. If anything
3745 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3746 could be sent to the SMTP client, causing chaos. The daemon now opens
3747 stdin, stdout, and stderr to /dev/null when it puts itself into the
3750 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3751 log by default. The user can override this with suitable Perl magic.
3753 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3754 the log message when discard triggers, was not working for the DATA ACL or
3755 for the non-SMTP ACL.
3757 78. Error message wording change in sieve.c.
3759 79. If smtp_accept_max_per_host was set, the number of connections could be
3760 restricted to fewer than expected, because the daemon was trying to set up
3761 a new connection before checking whether the processes handling previous
3762 connections had finished. The check for completed processes is now done
3763 earlier. On busy systems, this bug wouldn't be noticed because something
3764 else would have woken the daemon, and it would have reaped the completed
3767 80. If a message was submitted locally by a user whose login name contained one
3768 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3769 It caused a spool format error. I have fixed the spool reading code. A
3770 related problem was that the "from" clause in the Received: line became
3771 illegal because of the space(s). It is now covered by ${quote_local_part.
3773 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3776 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3778 83. Patch from Sieve maintainer to fix unterminated string problem in
3779 "vacation" handling.
3781 84. Some minor changes to the Linux configuration files to help with other
3782 OS variants using glibc.
3784 85. One more patch for Sieve to update vacation handling to latest spec.
3787 ----------------------------------------------------
3788 See the note above about the 4.44 and 4.50 releases.
3789 ----------------------------------------------------
3795 1. Change 4.43/35 introduced a bug that caused file counts to be
3796 incorrectly computed when quota_filecount was set in an appendfile
3799 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3800 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3803 3. Give more explanation in the error message when the command for a transport
3804 filter fails to execute.
3806 4. There are several places where Exim runs a non-Exim command in a
3807 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3808 was being done only for the command run by the queryprogram router. It is
3809 now done for all such subprocesses. The other cases are: ${run, transport
3810 filters, and the commands run by the lmtp and pipe transports.
3812 5. Some older OS have a limit of 256 on the maximum number of file
3813 descriptors. Exim was using setrlimit() to set 1000 as a large value
3814 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3815 systems. I've change it so that if it can't get 1000, it tries for 256.
3817 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3818 was an oversight, and furthermore, ever since the addition of extra
3819 controls (e.g. 4.43/32), the checks on when to allow different forms of
3820 "control" were broken. There should now be diagnostics for all cases when a
3821 control that does not make sense is encountered.
3823 7. $recipients is now available in the predata ACL (oversight).
3825 8. Tidy the search cache before the fork to do a delivery from a message
3826 received from the command line. Otherwise the child will trigger a lookup
3827 failure and thereby defer the delivery if it tries to use (for example) a
3828 cached ldap connection that the parent has called unbind on.
3830 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3831 of $address_data from the recipient verification was clobbered by the
3832 sender verification.
3834 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3835 was its contents. (It was OK if the option was not defined at all.)
3837 11. A "Completed" log line is now written for messages that are removed from
3838 the spool by the -Mrm option.
3840 12. $host_address is now set to the target address during the checking of
3841 ignore_target_hosts.
3843 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3844 being passed; this would have caused $sender_host_name to have been used if
3845 matching the list had actually called for a host name (not very likely,
3846 since this list is usually IP addresses). A host name is now passed as
3849 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3850 code that specifies a non-restarting handler (typically sigaction() in
3851 modern systems) in an attempt to fix a rare and obscure crash bug.
3853 15. Narrowed the window for a race in the daemon that could cause it to ignore
3854 SIGCHLD signals. This is not a major problem, because they are used only to
3855 wake it up if nothing else does.
3857 16. A malformed maildirsize file could cause Exim to calculate negative values
3858 for the mailbox size or file count. Odd effects could occur as a result.
3859 The maildirsize information is now recalculated if the size or filecount
3862 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3863 support for a long time. Removed HAVE_SYS_VFS_H.
3865 18. Updated exipick to current release from John Jetmore.
3867 19. Allow an empty sender to be matched against a lookup in an address list.
3868 Previously the only cases considered were a regular expression, or an
3871 20. Exim went into a mad DNS lookup loop when doing a callout where the
3872 host was specified on the transport, if the DNS lookup yielded more than
3875 21. The RFC2047 encoding function was originally intended for short strings
3876 such as real names; it was not keeping to the 75-character limit for
3877 encoded words that the RFC imposes. It now respects the limit, and
3878 generates multiple encoded words if necessary. To be on the safe side, I
3879 have increased the buffer size for the ${rfc2047: expansion operator from
3882 22. Failure to deliver a bounce message always caused it to be frozen, even if
3883 there was an errors_to setting on the router. The errors_to setting is now
3886 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3887 canonical form (fully expanded) before being placed in
3888 $sender_host_address.
3890 24. Updated eximstats to version 1.33
3892 25. Include certificate and key file names in error message when GnuTLS fails
3893 to set them up, because the GnuTLS error message doesn't include the name
3894 of the failing file when there is a problem reading it.
3896 26. Expand error message when OpenSSL has problems setting up cert/key files.
3899 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3900 (this can affect the format of dates).
3902 28. exim_tidydb, when checking for the continued existence of a message for
3903 which it has found a message-specific retry record, was not finding
3904 messages that were in split spool directories. Consequently, it was
3905 deleting retry records that should have stayed in existence.
3907 29. eximstats updated to version 1.35
3908 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3909 1.35 - bugfix such that pie charts by volume are generated correctly
3911 30. The SPA authentication driver was not abandoning authentication and moving
3912 on to the next authenticator when an expansion was forced to fail,
3913 contradicting the general specification for all authenticators. Instead it
3914 was generating a temporary error. It now behaves as specified.
3916 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3917 (the order specifies the preference for clients). The order is now AES256,
3918 AES128, 3DES, ARCFOUR128.
3920 31. Small patch to Sieve code - explicitly set From: when generating an
3923 32. Exim crashed if a remote delivery caused a very long error message to be
3924 recorded - for instance if somebody sent an entire SpamAssassin report back
3925 as a large number of 550 error lines. This bug was coincidentally fixed by
3926 increasing the size of one of Exim's internal buffers (big_buffer) that
3927 happened as part of the Exiscan merge. However, to be on the safe side, I
3928 have made the code more robust (and fixed the comments that describe what
3931 33. Some experimental protocols are using DNS PTR records for new purposes. The
3932 keys for these records are domain names, not reversed IP addresses. The
3933 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3934 leaves it alone. Component reversal etc. now happens only for IP addresses.
3937 34. The host_aton() function is supposed to be passed a string that is known
3938 to be a valid IP address. However, in the case of IPv6 addresses, it was
3939 not checking this. This is a hostage to fortune. Exim now panics and dies
3940 if the condition is not met. A case was found where this could be provoked
3941 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3942 components; fortuitously, this particular loophole had already been fixed
3943 by change 4.50/55 or 4.44/33 above.
3945 If there are any other similar loopholes, the new check in host_aton()
3946 itself should stop them being exploited. The report I received stated that
3947 data on the command line could provoke the exploit when Exim was running as
3948 exim, but did not say which command line option was involved. All I could
3949 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3950 running as the user.
3953 35. There was a buffer overflow vulnerability in the SPA authentication code
3954 (which came originally from the Samba project). I have added a test to the
3955 spa_base64_to_bits() function which I hope fixes it.
3958 36. The daemon start-up calls getloadavg() while still root for those OS that
3959 need the first call to be done as root, but it missed one case: when
3960 deliver_queue_load_max is set with deliver_drop_privilege. This is
3961 necessary for the benefit of the queue runner, because there is no re-exec
3962 when deliver_drop_privilege is set.
3964 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3965 was in use, was not putting the data itself into the right store pool;
3966 consequently, it could be overwritten for a subsequent message in the same
3967 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3970 38. Sometimes the final signoff response after QUIT could fail to get
3971 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3972 < 0 before doing a fflush(). This bug looks as though it goes back to the
3973 introduction of TLS in release 3.20, but "sometimes" must have been rare
3974 because the tests only now provoked it.
3980 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3981 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3982 Now it returns an integer. A number of calls were still expecting a T/F
3983 return. Fortuitously, in all cases, the tests worked in OK situations,
3984 which is the norm. However, things would have gone wrong on any write
3985 failures on the smtp file descriptor. This function is used when sending
3986 messages over SMTP and also when doing verify callouts.
3988 2. When Exim is called to do synchronous delivery of a locally submitted
3989 message (the -odf or -odi options), it no longer closes stderr before doing
3992 3. Implemented the mua_wrapper option.
3994 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3996 5. Implemented the functions header_remove(), header_testname(),
3997 header_add_at_position(), and receive_remove_recipient(), and exported them
4000 6. If an ACL "warn" statement specified the addition of headers, Exim already
4001 inserted X-ACL-Warn: at the start if there was no header name. However, it
4002 was not making this test for the second and subsequent header lines if
4003 there were newlines in the string. This meant that an invalid header could
4004 be inserted if Exim was badly configured.
4006 7. Allow an ACL "warn" statement to add header lines at the start or after all
4007 the Received: headers, as well as at the end.
4009 8. Added the rcpt_4xx retry error code.
4011 9. Added postmaster_mailfrom=xxx to callout verification option.
4013 10. Added mailfrom=xxxx to the callout verification option, for verify=
4016 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4017 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4019 12. Inserted some casts to stop certain compilers warning when using pointer
4020 differences as field lengths or precisions in printf-type calls (mostly
4021 affecting debugging statements).
4023 13. Added optional readline() support for -be (dynamically loaded).
4025 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4026 same clock tick as a message's arrival, so that its received time was the
4027 same as the "first fail" time on the retry record, and that message
4028 remained on the queue past the ultimate address timeout, every queue runner
4029 would try a delivery (because it was past the ultimate address timeout) but
4030 after another failure, the ultimate address timeout, which should have then
4031 bounced the address, did not kick in. This was a "< instead of <=" error;
4032 in most cases the first failure would have been in the next clock tick
4033 after the received time, and all would be well.
4035 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4036 being recognized when the domain list was tested by the match_domain
4037 condition in an expansion string.
4039 16. Added the ${str2b64: operator.
4041 17. Exim was always calling setrlimit() to set a large limit for the number of
4042 processes, without checking whether the existing limit was already
4043 adequate. (It did check for the limit on file descriptors.) Furthermore,
4044 errors from getrlimit() and setrlimit() were being ignored. Now they are
4045 logged to the main and panic logs, but Exim does carry on, to try to do its
4046 job under whatever limits there are.
4048 18. Imported PCRE 5.0.
4050 19. Trivial typo in log message " temporarily refused connection" (the leading
4053 20. If the log selector return_path_on_delivery was set and an address was
4054 redirected to /dev/null, the delivery process crashed because it assumed
4055 that a return path would always be set for a "successful" delivery. In this
4056 case, the whole delivery is bypassed as an optimization, and therefore no
4059 21. Internal re-arrangement: the function for sending a challenge and reading
4060 a response while authentication was assuming a zero-terminated challenge
4061 string. It's now changed to take a pointer and a length, to allow for
4062 binary data in such strings.
4064 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4066 23. Exim was not respecting finduser_retries when seeking the login of the
4067 uid under which it was called; it was always trying 10 times. (The default
4068 setting of finduser_retries is zero.) Also, it was sleeping after the final
4069 failure, which is pointless.
4071 24. Implemented tls_on_connect_ports.
4073 25. Implemented acl_smtp_predata.
4075 26. If the domain in control=submission is set empty, Exim assumes that the
4076 authenticated id is a complete email address when it generates From: or
4077 Sender: header lines.
4079 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4080 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4081 chown and chgrp in /bin and hostname in /usr/bin.
4083 28. Exim was keeping the "process log" file open after each use, just as it
4084 does for the main log. This opens the possibility of it remaining open for
4085 long periods when the USR1 signal hits a daemon. Occasional processlog
4086 errors were reported, that could have been caused by this. Anyway, it seems
4087 much more sensible not to leave this file open at all, so that is what now
4090 29. The long-running daemon process does not normally write to the log once it
4091 has entered its main loop, and it closes the log before doing so. This is
4092 so that log files can straightforwardly be renamed and moved. However,
4093 there are a couple of unusual error situations where the daemon does write
4094 log entries, and I had neglected to close the log afterwards.
4096 30. The text of an SMTP error response that was received during a remote
4097 delivery was being truncated at 512 bytes. This is too short for some of
4098 the long messages that one sometimes sees. I've increased the limit to
4101 31. It is now possible to make retry rules that apply only when a message has a
4102 specific sender, in particular, an empty sender.
4104 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4105 it possible to be selective about when SMTP synchronization is enforced.
4107 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4109 32. Implemented hosts_connection_nolog.
4111 33. Added an ACL for QUIT.
4113 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4116 35. Added mailbox_size and mailbox_filecount to appendfile.
4118 36. Added control = no_multiline_responses to ACLs.
4120 37. There was a bug in the logic of the code that waits for the clock to tick
4121 in the case where the clock went backwards by a substantial amount such
4122 that the microsecond fraction of "now" was more than the microsecond
4123 fraction of "then" (but the whole seconds number was less).
4125 38. Added support for the libradius Radius client library this is found on
4126 FreeBSD (previously only the radiusclient library was supported).
4132 1. When certain lookups returned multiple values in the form name=value, the
4133 quoting of the values was not always being done properly. Specifically:
4134 (a) If the value started with a double quote, but contained no whitespace,
4136 (b) If the value contained whitespace other than a space character (i.e.
4137 tabs or newlines or carriage returns) it was not quoted.
4138 This fix has been applied to the mysql and pgsql lookups by writing a
4139 separate quoting function and calling it from the lookup code. The fix
4140 should probably also be applied to nisplus, ibase and oracle lookups, but
4141 since I cannot test any of those, I have not disturbed their existing code.
4143 2. A hit in the callout cache for a specific address caused a log line with no
4144 reason for rejecting RCPT. Now it says "Previous (cached) callout
4145 verification failure".
4147 3. There was an off-by-one bug in the queryprogram router. An over-long
4148 return line was truncated at 256 instead of 255 characters, thereby
4149 overflowing its buffer with the terminating zero. As well as fixing this, I
4150 have increased the buffer size to 1024 (and made a note to document this).
4152 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4153 when Exim is waiting for an SMTP response from a remote server, Exim
4154 restarts its select() call on the socket, thereby resetting its timeout.
4155 This is not a problem when such interrupts are rare. Somebody set up a cron
4156 job to run exiwhat every 2 minutes, which is less than the normal select()
4157 timeout (5 or 10 minutes). This meant that the select() timeout never
4158 kicked in because it was always reset. I have fixed this by comparing the
4159 time when an interrupt arrives with the time at the start of the first call
4160 to select(). If more time than the timeout has elapsed, the interrupt is
4161 treated as a timeout.
4163 5. Some internal re-factoring in preparation for the addition of Sieve
4164 extensions (by MH). In particular, the "personal" test is moved to a
4165 separate function, and given an option for scanning Cc: and Bcc: (which is
4166 not set for Exim filters).
4168 6. When Exim created an email address using the login of the caller as the
4169 local part (e.g. when creating a From: or Sender: header line), it was not
4170 quoting the local part when it contained special characters such as @.
4172 7. Installed new OpenBSD configuration files.
4174 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4175 try to make them clearer.
4177 9. Callout options, other than the timeout value, were being ignored when
4178 verifying sender addresses in header lines. For example, when using
4180 verify = header_sender/callout=no_cache
4182 the cache was (incorrectly) being used.
4184 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4185 only the Cygwin environment.
4187 11. When return_path_on_delivery was set as a log selector, if different remote
4188 addresses in the same message used different return paths and parallel
4189 remote delivery occurred, the wrong values would sometimes be logged.
4190 (Whenever a remote delivery process finished, the return path value from
4191 the most recently started remote delivery process was logged.)
4193 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4194 lines when AUTH and/or TLS are in use. This is the "received protocol"
4195 field. Exim used to use "asmtp" for authenticated SMTP, without any
4196 indication (in the protocol name) for TLS use. Now it follows the RFC and
4197 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4198 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4199 names appear in log lines as well as in Received: header lines.
4201 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4202 extensions, and comparison tests, and to fix some bugs.
4204 14. Changes to the "personal" filter test:
4206 (1) The test was buggy in that it was just doing the equivalent of
4207 "contains" tests on header lines. For example, if a user's address was
4208 anne@some.where, the "personal" test would incorrectly be true for
4210 To: susanne@some.where
4212 This test is now done by extracting each address from the header in turn,
4213 and checking the entire address. Other tests that are part of "personal"
4214 are now done using regular expressions (for example, to check local parts
4215 of addresses in From: header lines).
4217 (2) The list of non-personal local parts in From: addresses has been
4218 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4219 taken from the Sieve specification recommendations.
4221 (3) If the message contains any header line starting with "List-" it is
4222 treated as non-personal.
4224 (4) The test for "circular" in the Subject: header line has been removed
4225 because it now seems ill-conceived.
4227 15. Minor typos in src/EDITME comments corrected.
4229 16. Installed latest exipick from John Jetmore.
4231 17. If headers_add on a router specified a text string that was too long for
4232 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4233 of string_sprintf() is now avoided.
4235 18. $message_body_size was not set (it was always zero) when running the DATA
4236 ACL and the local_scan() function.
4238 19. For the "mail" command in an Exim filter, no default was being set for
4239 the once_repeat time, causing a random time value to be used if "once" was
4240 specified. (If the value happened to be <= 0, no repeat happened.) The
4241 default is now 0s, meaning "never repeat". The "vacation" command was OK
4242 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4243 (I found it when inspecting the code).
4245 20. There is now an overall timeout for performing a callout verification. It
4246 defaults to 4 times the callout timeout, which applies to individual SMTP
4247 commands during the callout. The overall timeout applies when there is more
4248 than one host that can be tried. The timeout is checked before trying the
4249 next host. This prevents very long delays if there are a large number of
4250 hosts and all are timing out (e.g. when the network connections are timing
4251 out). The value of the overall timeout can be changed by specifying an
4252 additional sub-option for "callout", called "maxwait". For example:
4254 verify = sender/callout=5s,maxwait=20s
4256 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4257 to the end before writing, but this should make it even safer).
4259 22. Exim was forgetting that it had advertised PIPELINING for the second and
4260 subsequent messages on an SMTP connection. It was also not resetting its
4261 memory on STARTTLS and an internal HELO.
4263 23. When Exim logs an SMTP synchronization error within a session, it now
4264 records whether PIPELINING has been advertised or not.
4266 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4267 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4268 rather than long int.
4270 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4272 26. Added the never_mail option to autoreply.
4278 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4279 crash if the getsockname() call failed; this can happen if a connection is
4280 closed very soon after it is established. The problem was simply in the
4281 order in which certain operations were done, causing Exim to try to write
4282 to the SMTP stream before it had set up the file descriptor. The bug has
4283 been fixed by making things happen in the correct order.
4289 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4290 before the connection was closed, thus losing the rejection response.
4292 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4293 some early Solaris releases, but causes trouble in current releases where
4294 socklen_t is defined.
4296 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4299 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4302 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4303 that releasing the top of it at the end releases what was used for sub-
4304 expansions (unless the block got too big). However, discard this block if
4305 the first thing is a variable or header, so that we can use its block when
4306 it is dynamic (useful for very large $message_headers, for example).
4308 6. Lookups now cache *every* query, not just the most recent. A new, separate
4309 store pool is used for this. It can be recovered when all lookup caches are
4310 flushed. Lookups now release memory at the end of their result strings.
4311 This has involved some general refactoring of the lookup sources.
4313 7. Some code has been added to the store_xxx() functions to reduce the amount
4314 of flapping under certain conditions.
4316 8. log_incoming_interface used to affect only the <= reception log lines. Now
4317 it causes the local interface and port to be added to several more SMTP log
4318 lines, for example "SMTP connection from", and rejection lines.
4320 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4322 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4324 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4325 could be overwritten at the end of the current message (or the start of a
4326 new message if it was set in a HELO ACL). The value is now preserved for
4327 the duration of the SMTP connection.
4329 12. If a transport had a headers_rewrite setting, and a matching header line
4330 contained an unqualified address, that address was qualified, even if it
4331 did not match any rewriting rules. The underlying bug was that the values
4332 of the flags that permit the existence of unqualified sender and recipient
4333 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4334 non-local messages, and by -bnq for local messages) were not being
4335 preserved with the message after it was received.
4337 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4338 "next input=" as part of the text comprising the host identity instead of
4339 the correct text. The code was using the same buffer for two different
4340 strings. However, depending on which order the printing function evaluated
4341 its arguments, the bug did not always show up. Under Linux, for example, my
4342 test suite worked just fine.
4344 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4345 doesn't work with some older versions of Perl. It has been changed to "my",
4346 which in any case is probably the better facility to use.
4348 15. A really picky compiler found some instances of statements for creating
4349 error messages that either had too many or two few arguments for the format
4352 16. The size of the buffer for calls to the DNS resolver has been increased
4353 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4354 for addresses that have a lot of PTR records. This alleviates a problem; it
4355 does not fully solve it.
4357 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4358 buffer now truncates the list and logs the incident, which is the same
4359 action as happens when Exim is looking up a host name and its aliases.
4360 Previously in this situation something unpredictable would happen;
4361 sometimes it was "internal error: store_reset failed".
4363 18. If a server dropped the connection unexpectedly when an Exim client was
4364 using GnuTLS and trying to read a response, the client delivery process
4365 crashed while trying to generate an error log message.
4367 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4368 string, for example:
4370 warn message = H1: something\nH2: something
4372 the text was added as a single header line from Exim's point of view
4373 though it ended up OK in the delivered message. However, searching for the
4374 second and subsequent header lines using $h_h2: did not work. This has been
4375 fixed. Similarly, if a system filter added multiple headers in this way,
4376 the routers could not see them.
4378 20. Expanded the error message when iplsearch is called with an invalid key to
4379 suggest using net-iplsearch in a host list.
4381 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4382 ACLs are no longer actually imposed (and a message to that effect is
4385 22. If a "gecos" field in a passwd entry contained escaped characters, in
4386 particular, if it contained a \" sequence, Exim got it wrong when building
4387 a From: or a Sender: header from that name. A second bug also caused
4388 incorrect handling when an unquoted " was present following a character
4389 that needed quoting.
4391 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4392 was not being matched caselessly.
4394 24. Arranged for all hyphens in the exim.8 source to be escaped with
4397 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4398 properly. Recipient callout cache records were still being keyed to include
4399 the sender, even when use_sender was set false. This led to far more
4400 callouts that were necessary. The sender is no longer included in the key
4401 when use_sender is false.
4403 26. Added "control = submission" modifier to ACLs.
4405 27. Added the ${base62d: operator to decode base 62 numbers.
4407 28. dnsdb lookups can now access SRV records.
4409 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4410 the configuration file.
4412 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4413 (-v) mode. This makes the output for a verbose queue run more intelligible.
4415 31. Added a use_postmaster feature to recipient callouts.
4417 32. Added the $body_zerocount variable, containing the number of binary zero
4418 bytes in the message body.
4420 33. The time of last modification of the "new" subdirectory is now used as the
4421 "mailbox time last read" when there is a quota error for a maildir
4424 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4426 35. Added +ignore_unknown as a special item in host lists.
4428 36. Code for decoding IPv6 addresses in host lists is now included, even if
4429 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4430 address was recognized as an IP address, but was then not correctly decoded
4431 into binary, causing unexpected and incorrect effects when compared with
4438 1. Very minor rewording of debugging text in manualroute to say "list of
4439 hosts" instead of "hostlist".
4441 2. If verify=header_syntax was set, and a header line with an unqualified
4442 address (no domain) and a large number of spaces between the end of the
4443 name and the colon was received, the reception process suffered a buffer
4444 overflow, and (when I tested it) crashed. This was caused by some obsolete
4445 code that should have been removed. The fix is to remove it!
4447 3. When running in the test harness, delay a bit after writing a bounce
4448 message to get a bit more predictability in the log output.
4450 4. Added a call to search_tidyup() just before forking a reception process. In
4451 theory, someone could use a lookup in the expansion of smtp_accept_max_
4452 per_host which, without the tidyup, could leave open a database connection.
4454 5. Added the variables $recipient_data and $sender_data which get set from a
4455 lookup success in an ACL "recipients" or "senders" condition, or a router
4456 "senders" option, similar to $domain_data and $local_part_data.
4458 6. Moved the writing of debug_print from before to after the "senders" test
4461 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4462 problems for message scanning, either using a data ACL, or using
4463 local_scan() because the Received: header was not generated till after they
4464 were called (in order to set the time as the time of reception completion).
4465 I have revised the way this works. The header is now generated after the
4466 body is received, but before the ACL or local_scan() are called. After they
4467 are run, the timestamp in the header is updated.
4473 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4474 before starting a queue runner without re-exec. This happened only when
4475 deliver_drop_privilege was set or when the Exim user was set to root. The
4476 effect of the bug was that timeouts during subsequent deliveries caused
4477 crashes instead of being properly handled. The handler is now left at its
4478 default (and expected) setting.
4480 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4481 message, again when deliver_drop_privilege is set or Exim is run as root.
4482 The bug described in (1) was not present in this case, but the tidying up
4483 of the other signals was missing. I have made the two cases consistent.
4485 3. The ignore_target_hosts setting on a manualroute router was being ignored
4486 for hosts that were looked up using the /MX notation.
4488 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4491 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4492 operated on the sender address. After changing the $sender_address to <>
4493 for the sender address verify, Exim was re-instated it as the original
4494 (before rewriting) address, but remembering that it had rewritten it, so it
4495 wasn't rewriting it again. This bug also had the effect of breaking the
4496 sender address verification caching when the sender address was rewritten.
4498 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4499 This has been changed so that if the ip literal address matches
4500 ignore_target_hosts, the router declines.
4502 7. Added expansion conditions match_domain, match_address, and match_local_
4503 part (NOT match_host).
4505 8. The placeholder for the Received: header didn't have a length field set.
4507 9. Added code to Exim itself and to exim_lock to test for a specific race
4508 condition that could lead to file corruption when using MBX delivery. The
4509 issue is with the lockfile that is created in /tmp. If this file is removed
4510 after a process has opened it but before that process has acquired a lock,
4511 there is the potential for a second process to recreate the file and also
4512 acquire a lock. This could lead to two Exim processes writing to the file
4513 at the same time. The added code performs the same test as UW imapd; it
4514 checks after acquiring the lock that its file descriptor still refers to
4515 the same named file.
4517 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4518 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4521 11. Added the smtp_active_hostname option. If used, this will typically be made
4522 to depend on the incoming interface address. Because $interface_address is
4523 not set up until the daemon has forked a reception process, error responses
4524 that can happen earlier (such as "too many connections") no longer contain
4527 12. If an expansion in a condition on a "warn" statement fails because a lookup
4528 defers, the "warn" statement is abandoned, and the next ACL statement is
4529 processed. Previously this caused the whole ACL to be aborted.
4531 13. Added the iplsearch lookup type.
4533 14. Added ident_timeout as a log selector.
4535 15. Added tls_certificate_verified as a log selector.
4537 16. Added a global option tls_require_ciphers (compare the smtp transport
4538 option of the same name). This controls incoming TLS connections.
4540 17. I finally figured out how to make tls_require_ciphers do a similar thing
4541 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4542 before starting the TLS session.
4544 18. Tabs are now shown as \t in -bP output.
4546 19. If the log selector return_path_on_delivery was set, Exim crashed when
4547 bouncing a message because it had too many Received: header lines.
4549 20. If two routers both had headers_remove settings, and the first one included
4550 a superfluous trailing colon, the final name in the first list and the
4551 first name in the second list were incorrectly joined into one item (with a
4552 colon in the middle).
4558 1. Added -C and -D options to the exinext utility, mainly to make it easier
4559 to include in the automated testing, but these could be helpful when
4560 multiple configurations are in use.
4562 2. The exinext utility was not formatting the output nicely when there was
4563 an alternate port involved in the retry record key, nor when there was a
4564 message id as well (for retries that were specific to a specific message
4565 and a specific host). It was also confused by IPv6 addresses, because of
4566 the additional colons they contain. I have fixed the IPv4 problem, and
4567 patched it up to do a reasonable job for IPv6.
4569 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4570 delivery, the log line now contains "pipelined" if PIPELINING was used.
4572 4. An SMTP transport process used to panic and die if the bind() call to set
4573 an explicit outgoing interface failed. This has been changed; it is now
4574 treated in the same way as a connect() failure.
4576 5. A reference to $sender_host_name in the part of a conditional expansion
4577 that was being skipped was still causing a DNS lookup. This no longer
4580 6. The def: expansion condition was not recognizing references to header lines
4581 that used bh_ and bheader_.
4583 7. Added the _cache feature to named lists.
4585 8. The code for checking quota_filecount in the appendfile transport was
4586 allowing one more file than it should have been.
4588 9. For compatibility with Sendmail, the command line option
4596 and sets the incoming protocol and host name (for trusted callers). The
4597 host name and its colon can be omitted when only the protocol is to be set.
4598 Note the Exim already has two private options, -pd and -ps, that refer to
4599 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4600 "s", but I don't think that's a major issue.
4602 10. A number of refactoring changes to the code, none of which should affect
4605 (a) The number of logging options was getting close to filling up the
4606 32-bit word that was used as a bit map. I have split them into two classes:
4607 those that are passed in the argument to log_write(), and those that are
4608 only ever tested independently outside of that function. These are now in
4609 separate 32-bit words, so there is plenty of room for expansion again.
4610 There is no change in the user interface or the logging behaviour.
4612 (b) When building, for example, log lines, the code previously used a
4613 macro that called string_cat() twice, in order to add two strings. This is
4614 not really sufficiently general. Furthermore, there was one instance where
4615 it was actually wrong because one of the argument was used twice, and in
4616 one call a function was used. (As it happened, calling the function twice
4617 did not affect the overall behaviour.) The macro has been replaced by a
4618 function that can join an arbitrary number of extra strings onto a growing
4621 (c) The code for expansion conditions now uses a table and a binary chop
4622 instead of a serial search (which was left over from when there were very
4623 few conditions). Also, it now recognizes conditions like "pam" even when
4624 the relevant support is not compiled in: a suitably worded error message is
4625 given if an attempt is made to use such a condition.
4627 11. Added ${time_interval:xxxxx}.
4629 12. A bug was causing one of the ddress fields not to be passed back correctly
4630 from remote delivery subprocesses. The field in question was not being
4631 subsequently used, so this caused to problems in practice.
4633 13. Added new log selectors queue_time and deliver_time.
4635 14. Might have fixed a bug in maildirsizefile handling that threw up
4636 "unexpected character" debug warnings, and recalculated the data
4637 unnecessarily. In any case, I expanded the warning message to give more
4640 15. Added the message "Restricted characters in address" to the statements in
4641 the default ACL that block characters like @ and % in local parts.
4643 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4644 Three changes have been made:
4646 (a) There was a serious bug; a negative response to MAIL caused the whole
4647 recipient domain to be cached as invalid, thereby blocking all messages
4648 to all local parts at the same domain, from all senders. This bug has
4649 been fixed. The domain is no longer cached after a negative response to
4650 MAIL if the sender used is not empty.
4652 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4655 (c) A new callout option, "use_sender" has been added for people who want
4656 the modified behaviour.
4662 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4665 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4666 indeed breaks things for older releases.
4668 3. Added additional logging to the case where there is a problem reading data
4669 from a filter that is running in a subprocess using a pipe, in order to
4670 try to track down a specific problem.
4672 4. Testing facility fudge: when running in the test harness and attempting
4673 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4674 getting "No route to host". Convert this to a timeout.
4676 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4679 6. Some OS don't have socklen_t but use size_t instead. This affects the
4680 fifth argument of getsockopt() amongst other things. This is now
4681 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4682 can be set for individual OS. I have set it for SunOS5, OSF1, and
4683 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4684 some earlier ones do not.
4686 7. Change 4.30/15 was not doing the test caselessly.
4688 8. The standard form for an IPv6 address literal was being rejected by address
4689 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4690 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4691 this, as well as the form without the "IPv6" on the front (but only when
4692 address literals are enabled, of course).
4694 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4696 10. Exim crashed if a message with an empty sender address specified by -f
4697 encountered a router with an errors_to setting. This could be provoked only
4698 by a command such as
4702 where an empty string was supplied; "<>" did not hit this bug.
4704 11. Installed PCRE release 4.5.
4706 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4707 remained set. It is now erased.
4709 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4710 times from message ids (which are base 36 rather than the normal 62).
4712 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4713 were being counted as actual protocol errors, and logged if the log
4714 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4715 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4716 and DATA following a set of rejected RCPTs do not count as protocol errors.
4717 In other words, Exim assumes they were pipelined, though this may not
4718 actually be the case. Of course, in all cases the client gets an
4719 appropriate error code.
4721 15. If a lookup fails in an ACL condition, a message about the failure may
4722 be available; it is used if testing the ACL cannot continue, because most
4723 such messages specify what the cause of the deferral is. However, some
4724 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4725 that caused an old message to be retained and used if a later statement
4726 caused a defer, replacing the real cause of the deferral.
4728 16. If an IP address had so many PTR records that the DNS lookup buffer
4729 was not large enough to hold them, Exim could crash while trying to process
4730 the truncated data. It now detects and logs this case.
4732 17. Further to 4.21/58, another change has been made: if (and only if) the
4733 first line of a message (the first header line) ends with CRLF, a bare LF
4734 in a subsequent header line has a space inserted after it, so as not to
4735 terminate the header.
4737 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4738 unnecessarily, used atoi() instead of strtol(), and didn't check the
4739 termination when getting file sizes from file names by regex.
4741 19. Completely re-implemented the support for maildirsize files, in the light
4742 of a number of problems with the previous contributed implementation
4743 (4.30/29). In particular:
4745 . If the quota is zero, the maildirsize file is maintained, but no quota is
4748 . If the maildir directory does not exist, it is created before any attempt
4749 to write a maildirsize file.
4751 . The quota value in the file is just a cache; if the quota is changed in
4752 the transport, the new value overrides.
4754 . A regular expression is available for excluding directories from the
4757 20. The autoreply transport checks the characters in options that define the
4758 message's headers; it allows continued headers, but it was checking with
4759 isspace() after an embedded newline instead of explicitly looking for a
4762 21. If all the "regular" hosts to which an address was routed had passed their
4763 expiry times, and had not reached their retry times, the address was
4764 bounced, even if fallback hosts were defined. Now Exim should go on to try
4767 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4768 equivalent code in the SMTP transport. Some hosts send humungous responses
4769 to HELO/EHLO, more than 1024 it seems.
4771 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4772 gives compiler warnings in some environments. I've now done it "properly",
4775 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4776 (because of problems with the built-in one) was declared to return uschar *
4777 instead of char *, causing compiler failure.
4779 25. Fixed a file descriptor leak when processing alias/forward files.
4781 26. Fixed a minor format string issue in dbfn.c.
4783 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4785 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4786 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4789 29. When a sender address is verified, it is cached, to save repeating the test
4790 when there is more than one recipient in a message. However, when the
4791 verification involves a callout, it is possible for different callout
4792 options to be set for different recipients. It is too complicated to keep
4793 track of this in the cache, so now Exim always runs a verification when a
4794 callout is required, relying on the callout cache for the optimization.
4795 The overhead is duplication of the address routing, but this should not be
4798 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4799 to be verified with callout=postmaster, and the main callout worked but the
4800 postmaster check failed, the verification correctly failed. However, if a
4801 subsequent RCPT command asked for sender verification *without* the
4802 postmaster check, incorrect caching caused this verification also to fail,
4805 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4806 it was not caching the DNS options (qualify_single, search_parents) that
4807 were used when the lookup failed. A subsequent lookup with different
4808 options therefore always gave the same answer, though there were cases
4809 where it should not have. (Example: a "domains = !$mx_any" option on a
4810 dnslookup router: the "domains" option is always processed without any
4811 widening, but the router might have qualify_single set.) Now Exim uses the
4812 cached value only when the same options are set.
4814 32. Added John Jetmore's "exipick" utility to the distribution.
4816 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4817 than a timeout (e.g. a certificate is required, and is not provided), an
4818 Exim server now closes the connection immediately. Previously it waited for
4819 the client to close - but if the client is SSL, it seems that they each
4820 wait for each other, leading to a delay before one of them times out.
4822 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4823 maintained 0.8.x compatibility because I don't think many are using it, and
4824 it is clearly obsolete.
4826 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4829 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4830 client certificate was expired. A simple patch fixes this, though I don't
4831 understand the full logic of why the verify callback is called multiple
4834 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4835 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4836 which causes problems with some clients (such as the Certicom SSL Plus
4837 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4838 disables the coutermeasure allowing Eudora to connect."
4840 38. Exim was not checking that a write() to a log file succeeded. This could
4841 lead to Bad Things if a log got too big, in particular if it hit a file
4842 size limit. Exim now panics and dies if it cannot write to a log file, just
4843 as it does if it cannot open a log file.
4845 39. Modified OS/Makefile-Linux so that it now contains
4847 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4849 The two -D definitions ensure that Exim is compiled with large file
4850 support, which makes it possible to handle log files that are bigger than
4853 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4854 instance) a domain was checked against a named list that involved a lookup,
4855 causing $domain_data to be set, then another domain was checked against the
4856 same list, then the first domain was re-checked, the value of $domain_data
4857 after the final check could be wrong. In particular, if the second check
4858 failed, it could be set empty. This bug probably also applied to
4861 41. The strip_trailing_dot option was not being applied to the address given
4862 with the -f command-line option.
4864 42. The code for reading a message's header from the spool was incrementing
4865 $received_count, but never initializing it. This meant that the value was
4866 incorrect (doubled) while delivering a message in the same process in which
4867 it was received. In the most common configuration of Exim, this never
4868 happens - a fresh exec is done - but it can happen when
4869 deliver_drop_privilege is set.
4871 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4872 it now includes up to 150 characters of the unexpected data in the log
4875 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4876 and building data strings. The size of both of these buffers was 10 000
4877 bytes - far larger than anybody would *ever* want, thought I. Needless to
4878 say, somebody hit the limit. I have increased the maximum line length to
4879 20 000 and the maximum data length of concatenated lines to 100 000. I have
4880 also fixed two bugs, because there was no checking on these buffers. Tsk,
4881 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4882 buffer is too small.
4884 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4885 lsearch lookups. Now it does.
4887 46. When parsing a route_list item in a manualroute router, a fixed-length
4888 buffer was used for the list of hosts. I made this 1024 bytes long,
4889 thinking that nobody would ever have a list of hosts that long. Wrong.
4890 Somebody had a whole pile of complicated expansion conditions, and the
4891 string was silently truncated, leading to an expansion error. It turns out
4892 that it is easier to change to an unlimited length (owing to other changes
4893 that have happened since this code was originally written) than to build
4894 structure for giving a limitation error. The length of the item that
4895 expands into the list of hosts is now unlimited.
4897 47. The lsearch lookup could not handle data where the length of text line was
4898 more than 4095 characters. Such lines were truncated, leading to shortened
4899 data being returned. It should now handle lines of any length.
4901 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4902 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4905 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4906 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4907 no longer happens in comment lines. A list of replacements is now placed
4908 at the head of all of the source files, except those whose only change is
4909 to replace PERL_COMMAND in the very first #! line.
4911 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4912 messages on the queue, with a bottom-up merge sort, using code contributed
4913 by Michael Haardt. This should make operations like -bp somewhat faster on
4914 large queues. It won't affect queue runners, except when queue_run_in_order
4917 51. Installed eximstats 1.31 in the distribution.
4919 52. Added support for SRV lookups to the dnslookup router.
4921 53. If an ACL referred to $message_body or $message_body_end, the value was not
4922 reset for any messages that followed in the same SMTP session.
4924 54. The store-handling optimization for building very long strings was not
4925 differentiating between the different store pools. I don't think this
4926 actually made any difference in practice, but I've tidied it.
4928 55. While running the routers to verify a sender address, $sender_address
4929 was still set to the sender address. This is wrong, because when routing to
4930 send a bounce to the sender, it would be empty. Therefore, I have changed
4931 it so that, while verifying a sender address, $sender_address is set to <>.
4932 (There is no change to what happens when verifying a recipient address.)
4934 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4935 target A or AAAA records (if not already returned) without resetting the
4936 qualify_single or search_parents options of the DNS resolver. These are
4937 inappropriate in this case because the targets of MX and SRV records must
4938 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4939 target that, when qualified, matched something in the local domain. These
4940 two options are now turned off when doing these lookups.
4942 57. It seems that at least some releases of Reiserfs (which does not have the
4943 concept of a fixed number of inodes) returns zero and not -1 for the
4944 number of available inodes. This interacted badly with check_spool_inodes,
4945 which assumed that -1 was the "no such thing" setting. What I have done is
4946 to check that the total number of inodes is greater than zero before doing
4947 the test of how many are available.
4949 58. When a "warn" ACL statement has a log_message modifier, the message is
4950 remembered, and not repeated. This is to avoid a lot of repetition when a
4951 message has many recipients that cause the same warning to be written.
4952 Howewer, Exim was preserving the list of already written lines for an
4953 entire SMTP session, which doesn't seem right. The memory is now reset if a
4954 new message is started.
4956 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4957 debugging output unless log_rewrite was also set.
4959 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4960 of (int)(handle) when we know that handle contains (void *)(-1).
4962 61. The Exim daemon panic-logs an error return when it closes the incoming
4963 connection. However "connection reset by peer" seems to be common, and
4964 isn't really an error worthy of noting specially, so that particular error
4967 62. When Exim is trying to find all the local interfaces, it used to panic and
4968 die if the ioctl to get the interface flags failed. However, it seems that
4969 on at least one OS (Solaris 9) it is possible to have an interface that is
4970 included in the list of interfaces, but for which you get a failure error
4971 for this call. This happens when the interface is not "plumbed" into a
4972 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4973 failure of the "get flags" call assumes that the interface is down.
4975 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4976 makes life easier for people who are doing arithmetic on fields extracted
4977 from dates, where you often get leading zeros that should not be
4978 interpreted as octal.
4980 64. Added qualify_domain to the redirect router, to override the global
4983 65. If a pathologically long header line contained very many addresses (the
4984 report of this problem mentioned 10 000) and each of them was rewritten,
4985 Exim could use up a very large amount of memory. (It kept on making new
4986 copies of the header line as it rewrote, and never released the old ones.)
4987 At the expense of a bit more processing, the header rewriting function has
4988 been changed so that it no longer eats memory in this way.
4990 66. The generation of the Received: header has been moved from the time that a
4991 message starts to be received, to the time that it finishes. The timestamp
4992 in the Received: header should now be very close to that of the <= log
4993 line. There are two side-effects of this change:
4995 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4996 logged header lines no longer include the local Received: line, because
4997 it has not yet been created. The same applies to a copy of the message
4998 that is returned to a non-SMTP sender when a message is rejected.
5000 (b) When a filter file is tested using -bf, no additional Received: header
5001 is added to the test message. After some thought, I decided that this
5004 This change does not affect the value of $received_for. It is still set
5005 after address rewriting, but before local_scan() is called.
5007 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5009 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5010 gave an unhelpful panic error message, and a defer error. I have managed to
5011 change this behaviour so that it now rejects any supplied certificate,
5012 which seems right, as the list of acceptable certificates is empty.
5014 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5015 gave an unhelpful defer error. I have not managed to make this reject any
5016 supplied certificates, but the error message it gives is "no certificate
5017 supplied", which is not helpful.
5019 70. exigrep's output now also includes lines that are not associated with any
5020 message, but which match the given pattern. Implemented by a patch from
5021 Martin Sluka, which also tidied up the Perl a bit.
5023 71. Recipient callout verification, like sender verification, was using <> in
5024 the MAIL FROM command. This isn't really the right thing, since the actual
5025 sender may affect whether the remote host accepts the recipient or not. I
5026 have changed it to use the actual sender in the callout; this means that
5027 the cache record is now keyed on a recipient/sender pair, not just the
5028 recipient address. There doesn't seem to be a real danger of callout loops,
5029 since a callout by the remote host to check the sender would use <>.
5030 [SEE ABOVE: changed after hitting problems.]
5032 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5033 temporary errors. However, in the case of such a code being given after
5034 the end of a data transmission (i.e. after ".") Exim was failing to write
5035 a retry record for the message. (Yes, there was some broken host that was
5036 actually sending 8xx at this point.)
5038 73. An unknown lookup type in a host list could cause Exim to panic-die when
5039 the list was checked. (An example that provoked this was putting <; in the
5040 middle of a list instead of at the start.) If this happened during a DATA
5041 ACL check, a -D file could be left lying around. This kind of configuration
5042 error no longer causes Exim to die; instead it causes a defer errror. The
5043 incident is still logged to the main and panic logs.
5045 74. Buglet left over from Exim 3 conversion. The message "too many messages
5046 in one connection" was written to the rejectlog but not the mainlog, except
5047 when address rewriting (yes!) was being logged.
5049 75. Added write_rejectlog option.
5051 76. When a system filter was run not as root (that is, when system_filter_user
5052 was set), the values of the $n variables were not being returned to the
5053 main process; thus, they were not subsequently available in the $sn
5056 77. Added +return_path_on_delivery log selector.
5058 78. A connection timeout was being treated differently from recipients deferred
5059 when testing hosts_max_try with a message that was older than the host's
5060 retry timeout. (The host should not be counted, thus allowing all hosts to
5061 be tried at least once before bouncing.) This may have been the cause of an
5062 occasionally reported bug whereby a message would remain on the queue
5063 longer than the retry timeout, but would be bounced if a delivery was
5064 forced. I say "may" because I never totally pinned down the problem;
5065 setting up timeout/retry tests is difficult. See also the next item.
5067 79. The ultimate address timeout was not being applied to errors that involved
5068 a combination of host plus message (for example, a timeout on a MAIL
5069 command). When an address resolved to a number of possible hosts, and they
5070 were not all tried for each delivery (e.g. because of hosts_max_try), a
5071 message could remain on the queue longer than the retry timeout.
5073 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5076 81. Fixed an obscure SMTP outgoing bug which required at least the following
5077 conditions: (a) there was another message waiting for the same server;
5078 (b) the server returned 5xx to all RCPT commands in the first message so
5079 that the message was not completed; (c) the server dropped the connection
5080 or gave a negative response to the RSET that Exim sends to abort the
5081 transaction. The observed case was a dropped connection after DATA that had
5082 been sent in pipelining mode. That is, the server had advertised PIPELINING
5083 but was not implementing it correctly. The effect of the bug was incorrect
5084 behaviour, such as trying another host, and this could lead to a crash.
5090 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5091 and daemon.c were passed as pointers to ints; they should have been
5092 pointers to socklen_t variables (which are typically unsigned ints).
5094 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5097 3. Fixed a really odd bug that affected only the testing scheme; patching a
5098 certain fixed string in the binary changed the value of another string that
5099 happened to be identical to the end of the original first string.
5101 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5102 name", it returns that address as the IP address. On some operating
5103 systems (e.g. Solaris), it also passes back the IP address string as the
5104 "host name". However, on others (e.g. Linux), it passes back an empty
5105 string. Exim wasn't checking for this, and was changing the host name to an
5106 empty string, assuming it had been canonicized.
5108 5. Although rare, it is permitted to have more than one PTR record for a given
5109 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5110 all the names associated with an address, because they do in Solaris.
5111 However, it seems that they do not in Linux for data that comes from the
5112 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5113 I found this out when I moved to a new Linux workstation and tried to run
5114 the Exim test suite.
5116 To get round this problem I have changed the code so that it now does its
5117 own call to the DNS to look up PTR records when searching for a host name.
5118 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5119 addresses that are only in /etc/hosts are still found.
5121 This behaviour is, however, controlled by an option called host_lookup_
5122 order, which defaults to "bydns:byaddr". If people want to use the other
5123 order, or indeed, just use one or the other means of lookup, they can
5124 specify it in this variable.
5126 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5127 some operating systems, this comes back from gethostbyaddr() as an empty
5128 string, and this is what Exim used to test for. However, it seems that in
5129 other systems, "." is yielded. Exim now tests for this case too.
5131 7. The values of check_spool_space and check_log_space are now held internally
5132 as a number of kilobytes instead of an absolute number of bytes. If a
5133 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5134 kilobyte. This means that much larger values can be stored.
5136 8. Exim monitor: an attempt to get the action menu when not actually pointing
5137 at a message produces an empty menu entitled "No message selected". This
5138 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5139 no entries in it ("Shell widget menu has zero width and/or height"). So I
5140 have added a single, blank menu entry in this case.
5142 9. Added ${quote_local_part.
5144 10. MIME decoding is now applied to the contents of Subject: header lines when
5147 11. Now that a reference to $sender_host_address automatically causes a reverse
5148 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5149 host lookup before query-style lookups in lists that might use this
5150 variable. This has therefore been abolished, and the "net-" prefix is no
5151 longer necessary for query-style lookups.
5153 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5154 have been a typo for LFLAGS, so it has been changed.
5156 13. The install script calls Exim with "-C /dev/null" in order to find the
5157 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5158 to be output. Howeve, since Exim outputs its version number before the
5159 error, it didn't break the script. It just looked ugly. I fixed this by
5160 always allowing "-C /dev/null" if the caller is root.
5162 14. Ignore overlarge ACL variable number when reading spool file - insurance
5163 against a later release with more variables having written the file.
5165 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5166 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5167 this, as well as the form without the "IPv6" on the front.
5169 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5170 OS/Makefile-Darwin file.
5172 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5173 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5175 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5176 variable. However, some OS use ints for inodes. I've added cast to long int
5177 to get rid of the compiler warning.
5179 19. I had forgotten to lock out "/../" in configuration file names when
5180 ALT_CONFIG_PREFIX was set.
5182 20. Routers used for verification do not need to specify transports. However,
5183 if such a router generated a host list, and callout was configured, Exim
5184 crashed, because it could not find a port number from the (non-existent)
5185 transport. It now assumes port 25 in this circumstance.
5187 21. Added the -t option to exigrep.
5189 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5190 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5191 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5192 them is set, LOOKUP_LSEARCH is forced.
5194 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5196 24. Added sender and host information to the "rejected by local_scan()" log
5197 line; previously there was no indication of these.
5199 25. Added .include_if_exists.
5201 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5202 Linux. It turns out that not all file systems support this. Apparently some
5203 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5204 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5205 not supported on the file descriptor, is now ignored when Exim is trying to
5206 sync a directory. This applies only to Linux.
5208 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5210 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5211 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5214 29. Added support for maildirsize files from supplied patch (modified a bit).
5216 30. The use of :fail: followed by an empty string could lead Exim to respond to
5217 sender verification failures with (e.g.):
5219 550 Verification failed for <xxx>
5220 550 Sender verify failed
5222 where the first response line was missing the '-' that indicates it is not
5223 the final line of the response.
5225 31. The loop for finding the name of the user that called Exim had a hardwired
5226 limit of 10; it now uses the value of finduser_retries, which is used for
5227 all other user lookups.
5229 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5232 33. Exim was neglecting to zero errno before one call of strtol() when
5233 expanding a string and expecting an integer value. On some systems this
5234 resulted in spurious "integer overflow" errors. Also, it was casting the
5235 result into an int without checking.
5237 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5238 did not work. The code looks as if it has *never* worked, though it appears
5239 to have been documented since at least releast 1.62. I have made it work.
5241 35. The "timeout_DNS" error in retry rules, also documented since at least
5242 1.62, also never worked. As it isn't clear exactly what this means, and
5243 clearly it isn't a major issue, I have abolished the feature by treating it
5244 as "timeout", and writing a warning to the main and panic logs.
5246 36. The display of retry rules for -brt wasn't always showing the error code
5249 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5250 timeout_connect_A, timeout_connect_MX.
5252 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5253 to the empty sender.
5255 39. The daemon was not analysing the content of -oX till after it had closed
5256 stderr and disconnected from the controlling terminal. This meant that any
5257 syntax errors were only noted on the panic log, and the return code from
5258 the command was 0. By re-arranging the code a little, I've made the
5259 decoding happen first, so such errors now appear on stderr, and the return
5260 code is 1. However, the actual setting up of the sockets still happens in
5261 the disconnected process, so errors there are still only recorded on the
5264 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5265 connections (as happens on some IP stacks) was logged at start up time as
5266 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5267 it from "IPv6 and IPv4", which means that two separate sockets are being
5270 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5271 says whether AF_INET or AF_INET6 was passed as an argument.
5273 42. Exiwhat output was messed up when time zones were included in log
5276 43. Exiwhat now gives more information about the daemon's listening ports,
5277 and whether -tls-on-connect was used.
5279 44. The "port" option of the smtp transport is now expanded.
5281 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5282 silently ignored. Now an error message is written to the main and panic
5285 46. There's a new ACL modifier called "logwrite" which writes to a log file
5286 as soon as it is encountered.
5288 47. Added $local_user_uid and $local_user_gid at routing time.
5290 48. Exim crashed when trying to verify a sender address that was being
5293 49. Exim was recognizing only a space character after ".include". It now also
5294 recognizes a tab character.
5296 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5297 extracting the relevant information from the specification. The man page no
5298 longer contains scrambled data for the -d option, and I've added a section
5299 at the front about calling Exim under different names.
5301 51. Added "extra_headers" argument to the "mail" command in filter files.
5303 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5306 53. Installed eximstats 1.29.
5308 54. Added transport_filter_timeout as a generic transport option.
5310 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5311 Cc: header lines. This was required by RFC 822, but it not required by RFC
5314 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5315 incoming messages that did not have them. Now it does so only if the
5316 message originates locally, that is, if there is no associated remote host
5317 address. When Resent- header lines are present, this applies to the Resent-
5318 lines rather than the non-Resent- lines.
5320 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5321 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5323 58. Messages for configuration errors now include the name of the main
5324 configuration file - useful now that there may be more than one file in a
5325 list (.included file names were always shown).
5327 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5328 for those rare installations that do not start the daemon as root or run it
5329 setuid root. I've cut out the call to initgroups() if the daemon is not
5332 60. The Exim user and group can now be bound into the binary as text strings
5333 that are looked up at the start of Exim's processing.
5335 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5337 62. Added $mailstore_basename variable.
5339 63. Installed patch to sieve.c from Michael Haardt.
5341 64. When Exim failed to open the panic log after failing to open the main log,
5342 the original message it was trying to log was written to stderr and debug
5343 output, but if they were not available (the usual case in production), it
5344 was lost. Now it is written to syslog before the two lines that record the
5345 failures to open the logs.
5347 65. Users' Exim filters run in subprocesses under the user's uid. It is
5348 possible for a "deliver" command or an alias in a "personal" command to
5349 provoke an address rewrite. If logging of address rewriting is configured,
5350 this fails because the process is not running as root or exim. There may be
5351 a better way of dealing with this, but for the moment (because 4.30 needs
5352 to be released), I have disabled address rewrite logging when running a
5353 filter in a non-root, non-exim process.
5359 1. The buildconfig auxiliary program wasn't quoting the value set for
5360 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5361 not defined. This bug was masked in 4.22 by the effect that was fixed in
5364 2. Some messages that were rejected after a message id was allocated were
5365 shown as "incomplete" by exigrep. It no longer does this for messages that
5366 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5368 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5369 have allow_domain_literals set, the ID did not get logged in the <= line.
5370 Domain literals are now always recognized in Message-ID: header lines.
5372 4. The first argument for a ${extract expansion item is the key name or field
5373 number. Leading and trailing spaces in this item were not being ignored,
5374 causing some misleading effects.
5376 5. When deliver_drop_privilege was set, single queue runner processes started
5377 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5378 same command in the process it spins off) were not dropping privilege.
5380 6. When the daemon running as "exim" started a queue runner, it always
5381 re-executed Exim in the spun-off process. This is a waste of effort when
5382 deliver_drop_privilege is set. The new process now just calls the
5383 queue-runner function directly.
5389 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5392 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5393 ignored. Though the use of -oP was forcing the writing of a pid file, it
5394 was always written to the default place.
5396 3. If the message "no IP address found for host xxxx" is generated during
5397 incoming verification, it is now followed by identification of the incoming
5398 connection (so you can more easily find what provoked it).
5400 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5402 5. Added some features to "harden" Exim a bit more against certain attacks:
5404 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5405 be put in Local/Makefile. This is like the never_users runtime option,
5406 but it cannot be overridden. The default setting is "root".
5408 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5409 prefix string with which any file named in a -C command line option
5412 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5413 is retained for -C and -D only if the caller of Exim is root. Without
5414 it, the exim user may also use -C and -D and retain privilege.
5416 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5417 command line option is disabled.
5419 6. Macro names set by the -D option must start with an upper case letter, just
5420 like macro names defined in the configuration file.
5422 7. Added "dereference=" facility to LDAP.
5424 8. Two instances of the typo "uknown" in the source files are fixed.
5426 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5427 the Configure-Makefile script screwed up while processing it.
5429 10. Incorporated PCRE 4.4.
5431 11. The SMTP synchronization check was not operating right at the start of an
5432 SMTP session. For example, it could not catch a HELO sent before the client
5433 waited for the greeting. There is now a check for outstanding input at the
5434 point when the greeting is written. Because of the duplex, asynchronous
5435 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5436 way, but not yet received, when the check is performed.
5438 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5439 on TCP/IP sockets, because this apparently causes some broken clients to
5442 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5443 unchanged) from the Cygwin maintainer.
5445 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5446 was supported instead of testing for mbx. Effectively a typo.
5448 15. The spa authenticator server code was not checking that the input it
5449 received was valid base64.
5451 16. The debug output line for the "set" modifier in ACLs was not showing the
5452 name of the variable that was being set.
5454 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5456 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5457 lookup on its own. Something else was needed to trigger the lookup. For
5458 example, a match in host_lookup or the need for a host name in a host list.
5459 Now, if $sender_host_name is referenced and the host name has not yet been
5460 looked up, a lookup is performed. If the lookup fails, the variable remains
5461 empty, and $host_lookup_failed is set to "1".
5463 19. Added "eqi" as a case-independent comparison operator.
5465 20. The saslauthd authentication condition could segfault if neither service
5466 nor realm was specified.
5468 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5469 error message that was logged was misleading, and incoming SMTP
5470 connections were dropped. The message is now more accurate, and temporary
5471 errors are given to SMTP connections.
5473 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5474 (except RSET) with a 421 error, until QUIT is received. However, it was
5475 failing to send a response to QUIT.
5477 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5478 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5479 ACL is now run after the helo_try_verify_hosts code.
5481 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5482 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5483 case-independent, so other case variants are also recognized). Apparently
5484 some systems use these upper case variants.
5486 25. If more than two messages were waiting for the same host, and a transport
5487 filter was specified for the transport, Exim sent two messages over the
5488 same TCP/IP connection, and then failed with "socket operation on non-
5489 socket" when it tried to send the third.
5491 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5493 27. The extern definition of crypt16() in expand.c was not being excluded when
5494 the OS had its own crypt16() function.
5496 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5497 as a preferred synonym for return_size_limit, both as an option and as an
5500 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5502 30. Changed the default configuration ACL to relax the local part checking rule
5503 for addresses that are not in any local domains. For these addresses,
5504 slashes and pipe symbols are allowed within local parts, but the sequence
5505 /../ is explicitly forbidden.
5507 31. SPA server authentication was not clearing the challenge buffer before
5510 32. log_message in a "warn" ACL statement was writing to the reject log as
5511 well as to the main log, which contradicts the documentation and doesn't
5512 seem right (because no rejection is happening). So I have stopped it.
5514 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5515 However, I am unable to do any testing of this.
5517 34. Fixed an infelicity in the appendfile transport. When checking directories
5518 for a mailbox, to see if any needed to be created, it was accidentally
5519 using path names with one or more superfluous leading slashes; tracing
5520 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5522 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5523 contents are added to the log line that is written for every discarded
5524 recipient. (Previously a log_message setting was ignored.)
5526 36. The ${quote: operator now quotes the string if it is empty.
5528 37. The install script runs exim in order to find its version number. If for
5529 some reason other than non-existence or emptiness, which it checks, it
5530 could not run './exim', it was installing it with an empty version number,
5531 i.e. as "exim-". This error state is now caught, and the installation is
5534 38. An argument was missing from the function that creates an error message
5535 when Exim fails to connect to the socket for saslauthd authentication.
5536 This could cause Exim to crash, or give a corrupted message.
5538 39. Added isip, isip4, and isip6 to ${if conditions.
5540 40. The ACL variables $acl_xx are now saved with the message, and can be
5541 accessed later in routers, transports, and filters.
5543 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5544 strings in the file are not string-expanded.
5546 42. If a MAIL command specified a SIZE value that was too large to fit into an
5547 int variable, the check against message_size_limit failed. Such values are
5548 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5549 day this will have to be increased, but I don't think I want to be around
5550 when emails are that large.
5557 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5558 iconv() is not standard in FreeBSD.
5560 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5561 IPv6 enabled. The observed symptom was a segmentation fault on return from
5562 the function os_common_find_running_interfaces() in src/os.c.
5564 3. In the check_special_case() function in daemon.c I had used "errno" as an
5565 argument name, which causes warnings on some systems. This was basically a
5566 typo, since it was named "eno" in the comments!
5568 4. The code that waits for the clock to tick (at a resolution of some fraction
5569 of a second) so as to ensure message-id uniqueness was always waiting for
5570 at least one whole tick, when it could have waited for less. [This is
5571 almost certainly not relevant at current processor speeds, where it is
5572 unlikely to ever wait at all. But we try to future-proof.]
5574 5. The function that sleeps for a time interval that includes fractions of a
5575 second contained a race. It did not block SIGALRM between setting the
5576 timer, and suspending (a couple of lines later). If the interval was short
5577 and the sigsuspend() was delayed until after it had expired, the suspension
5578 never ended. On busy systems this could lead to processes getting stuck for
5581 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5582 process, before it forks any delivery processes. The open lookup caching
5583 mechanism meant that the open file or database connection was passed into
5584 the delivery process. The problem was that delivery processes always tidy
5585 up cached lookup data. This could cause a problem for the next delivery
5586 process started by the queue runner, because the external queue runner
5587 process does not know about the closure. So the next delivery process
5588 still has data in the lookup cache. In the case of a file lookup, there was
5589 no problem because closing a file descriptor in a subprocess doesn't affect
5590 the parent. However, if the lookup was caching a connection to a database,
5591 the connection was closed, and the second delivery process was likely to
5592 see errors such as "PGSQL: query failed: server closed the connection
5593 unexpectedly". The problem has been fixed by closing all cached lookups
5594 in a queue runner before running a delivery process.
5596 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5597 seem to have the "const" qualifier which it has on other OS. I've
5600 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5601 *of the same type* (client or server) with the same public name that an
5602 error should be diagnosed.
5604 9. When Exim looked up a host name for an IP address, but failed to find the
5605 original IP address when looking up the host name (a safety check), it
5606 output the message "<ip address> does not match any IP for NULL", which was
5607 confusing, to say the least. The bug was that the host name should have
5608 appeared instead of "NULL".
5610 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5611 user that is built into the binary, and the -C or -D options is used, root
5612 privilege is dropped before the configuration file is read. In addition,
5613 logging is switched to stderr instead of the normal log files. If the
5614 configuration then re-defines the Exim user, the unprivileged environment
5615 is probably not what is expected, so Exim logs a panic warning message (but
5618 However, if deliver_drop_privilege is set, the unprivileged state may well
5619 be exactly what is intended, so the warning has been cut out in that case,
5620 and Exim is allowed to try to write to its normal log files.
5626 1. smtp_return_error_details was not giving details for temporary sender
5627 or receiver verification errors.
5629 2. Diagnose a configuration error if two authenticators have the same public
5632 3. Exim used not to create the message log file for a message until the first
5633 delivery attempt. This could be confusing when incoming messages were held
5634 for policy or load reasons. The message log file is now created at the time
5635 the message is received, and an initial "Received" line is written to it.
5637 4. The automatically generated man page for command line options had a minor
5638 bug that caused no ill effects; however, a more serious problem was that
5639 the procedure for building the man page automatically didn't always
5640 operate. Consequently, release 4.20 contains an out-of-date version. This
5641 shouldn't happen again.
5643 5. When building Exim with embedded Perl support, the script that builds the
5644 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5645 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5647 6. The freeze_tell option was not being used for messages that were frozen on
5648 arrival, either by an ACL or by local_scan().
5650 7. Added the smtp_incomplete_transaction log selector.
5652 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5653 was accepting AUTH without a new EHLO.
5655 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5656 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5659 10. Logging of TCP/IP connections (when configured) now happens in the main
5660 daemon process instead of the child process, so that the TCP/IP connection
5661 count is more accurate (but it can never be perfect).
5663 11. The use of "drop" in a nested ACL was not being handled correctly in the
5664 outer ACL. Now, if condition failure induced by the nested "drop" causes
5665 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5666 or "require"), the connection is dropped.
5668 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5669 that yield "discard" can now be used with an "accept" or a "discard" verb,
5670 but an error is generated for any others (because I can't see a useful way
5671 to define what should happen).
5673 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5674 are now processed in the same way as lines in the Exim configuration file.
5675 In particular, continuation lines are supported.
5677 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5679 15. Added -ti meaning -t -i.
5681 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5682 domains, and warn by logging if others are found.
5684 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5685 returned from GIFCONF when Exim is trying to find the list of interfaces on
5686 a host. The code in os.c has been modified to copy each ifreq to an aligned
5687 structure in all cases.
5689 Also, in some cases, the returned ifreq's were being copied to a 'struct
5690 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5691 means the last couple of bytes of an IPv6 address could be chopped if the
5692 ifreq contained only a normal sockaddr (14 bytes storage).
5694 18. Named domain lists were not supported in the hosts_treat_as_local option.
5695 An entry such as +xxxx was not recognized, and was treated as a literal
5698 19. Ensure that header lines added by a DATA ACL are included in the reject log
5699 if the ACL subsequently rejects the message.
5701 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5702 MD5 (which is deprecated).
5704 21. When testing a filter file using -bf, Exim was writing a message when it
5705 took the sender from a "From " line in the message, but it was not doing so
5706 when it took $return_path from a Return-Path: header line. It now does.
5708 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5709 with a valid header line field name (a series of printing characters
5710 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5712 23. Changed "disc" in the source to "disk" to conform to the documentation and
5713 the book and for uniformity.
5715 24. Ignore Sendmail's -Ooption=value command line item.
5717 25. When execve() failed while trying to run a command in a pipe transport,
5718 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5719 could be confused with a return value of 69 from the command itself. This
5720 has been changed to 127, the value the shell returns if it is asked to run
5721 a non-existent command. The wording for the related log line suggests a
5722 non-existent command as the problem.
5724 26. If received_header_text expands to an empty string, do not add a Received:
5725 header line to the message. (Well, it adds a token one on the spool, but
5726 marks it "old" so that it doesn't get used or transmitted.)
5728 27. Installed eximstats 1.28 (addition of -nt option).
5730 28. There was no check for failure on the call to getsockname() in the daemon
5731 code. This can fail if there is a shortage of resources on the system, with
5732 ENOMEM, for example. A temporary error is now given on failure.
5734 29. Contrary to the C standard, it seems that in some environments, the
5735 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5736 program. Exim now does this explicitly; it affects the formatting of
5737 timestamps using strftime().
5739 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5740 complaints. I've now initialized all the variables, to avoid this.
5742 32. Header lines added by a system filter were not being "seen" during
5743 transport-time rewrites.
5745 33. The info_callback() function passed to OpenSSL is set up with type void
5746 (*)(SSL *, int, int), as described somewhere. However, when calling the
5747 function (actually a macro) that sets it up, the type void(*)() is
5748 expected. I've put in a cast to prevent warnings from picky compilers.
5750 34. If a DNS black list lookup found a CNAME record, but there were no A
5751 records associated with the domain it pointed at, Exim crashed.
5753 35. If a DNS black list lookup returned more than one A record, Exim ignored
5754 all but the first. It now scans all returned addresses if a particular IP
5755 value is being sought. In this situation, the contents of the
5756 $dnslist_value variable are a list of all the addresses, separated by a
5759 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5760 to accept a host name and all its aliases if the forward lookup for any of
5761 them yielded the IP address of the incoming connection. Now it accepts only
5762 those names whose forward lookup yields the correct IP address. Any other
5763 names are discarded. This closes a loophole whereby a rogue DNS
5764 administrator could create reverse DNS records to break through a
5765 wildcarded host restriction in an ACL.
5767 37. If a user filter or a system filter that ran in a subprocess used any of
5768 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5769 the wrong values were passed to the pipe command ($thisaddress had the
5770 value of $0, $0 had the value of $1, etc). This bug was introduced by
5771 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5773 38. Improved the line breaking for long SMTP error messages from ACLs.
5774 Previously, if there was no break point between 40 and 75 characters, Exim
5775 left the rest of the message alone. Two changes have been made: (a) I've
5776 reduced the minimum length to 35 characters; (b) if it can't find a break
5777 point between 35 and 75 characters, it looks ahead and uses the first one
5778 that it finds. This may give the occasional overlong line, but at least the
5779 remaining text gets split now.
5781 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5782 file descriptors might be low, and that setting 1000 would always raise it.
5783 It turns out that in some environments, the limit is already over 1000 and
5784 that lowering it causes trouble. So now Exim takes care not to decrease it.
5786 40. When delivering a message, the value of $return_path is set to $sender_
5787 address at the start of routing (routers may change the value). By an
5788 oversight, this default was not being set up when an address was tested by
5789 -bt or -bv, which affected the outcome if any router or filter referred to
5792 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5793 log only when "message" or "log_message" are set. However, if one of the
5794 conditions was an address verification, or a call to a nested ACL, the
5795 messages generated by the underlying test were being passed through. This
5796 no longer happens. The underlying message is available in $acl_verify_
5797 message for both "message" and "log_message" expansions, so it can be
5798 passed through if needed.
5800 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5801 new expansion $bh_ to give the encoded byte string without charset
5802 translation. Translation happens only if iconv() is available; HAVE_ICONV
5803 indicates this at build time. HEADERS_CHARSET gives the charset to
5804 translate to; headers_charset can change it in the configuration, and
5805 "headers charset" can change it in an individual filter file.
5807 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5808 that creates RFC 2047 encoded "words" labels them as that charset instead
5809 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5810 expansion operator; (ii) when Exim creates a From: line for a local
5811 message; (iii) when a header line is rewritten to include a "phrase" part.
5813 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5814 buggy, causing it to skip the first lines of messages whose message ID
5815 ended in 'D'. This would not have bitten before Exim release 4.14, because
5816 message IDs were unlikely to end in 'D' before then. The effect was to have
5817 incorrect size information for certain domains.
5819 45. #include "config.h" was missing at the start of the crypt16.c module. This
5820 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5823 46. If there was a timeout during a "random" callout check, Exim treated it as
5824 a failure of the random address, and carried on sending RSET and the real
5825 address. If the delay was just some slowness somewhere, the response to the
5826 original RCPT would be taken as a response to RSET and so on, causing
5827 mayhem of various kinds.
5829 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5830 when I implemented it. It didn't allow for the fact that some option values
5831 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5832 the right test for positive values.
5834 48. Domain names in DNS records are case-independent. Exim always looks them up
5835 in lower case. Some resolvers return domain names in exactly the case they
5836 appear in the zone file, that is, they may contain uppercase letters. Not
5837 all resolvers do this - some return always lower case. Exim was treating a
5838 change of case by a resolver as a change of domain, similar to a widening
5839 of a domain abbreviation. This triggered its re-routing code and so it was
5840 trying to route what was effectively the same domain again. This normally
5841 caused routing to fail (because the router wouldn't handle the domain
5842 twice). Now Exim checks for this case specially, and just changes the
5843 casing of the domain that it ultimately uses when it transmits the message
5846 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5849 50. If a filter generated a file delivery with a non-absolute name (possible if
5850 no home directory exists for the router), the forbid_file option was not
5853 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5854 the existing equality matching.
5856 52. Exim was using ints instead of ino_t variables in some places where it was
5857 dealing with inode numbers.
5859 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5860 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5861 called TMPDIR, and if it finds it is different, it changes its value.
5863 54. The smtp_printf() function is now made available to local_scan() so
5864 additional output lines can be written before returning. There is also an
5865 smtp_fflush() function to enable the detection of a dropped connection.
5866 The variables smtp_input and smtp_batched_input are exported to
5869 55. Changed the default runtime configuration: the message "Unknown user"
5870 has been removed from the ACL, and instead placed on the localuser router,
5871 using the cannot_route_message feature. This means that any verification
5872 failures that generate their own messages won't get overridden. Similarly,
5873 the "Unrouteable address" message that was in the ACL for unverifiable
5874 relay addresses has also been removed.
5876 56. Added hosts_avoid_esmtp to the smtp transport.
5878 57. The exicyclog script was not checking for the esoteric option
5879 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5880 will work only if exicyclog is run under the appropriate euid.
5882 58. Following a discussion on the list, the rules by which Exim recognises line
5883 endings on incoming messages have been changed. The -dropcr and drop_cr
5884 options are now no-ops, retained only for backwards compatibility. The
5885 following line terminators are recognized: LF CRLF CR. However, special
5886 processing applies to CR:
5888 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5889 nor a local message in the state where . is a terminator.
5891 (ii) If a bare CR is encountered in a header line, an extra space is added
5892 after the line terminator so as not to end the header. The reasoning
5893 behind this is that bare CRs in header lines are most likely either
5894 to be mistakes, or people trying to play silly games.
5896 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5897 was being incorrectly given as 18 bytes larger than it should have been.
5898 This is a VOB (very old bug).
5900 60. This may never have affected anything current, but just in case it has:
5901 When the local host is found other than at the start of a list of hosts,
5902 the local host, those with the same MX, and any that follow, are discarded.
5903 When the list in question was part of a longer list of hosts, the following
5904 hosts (not currently being processed) were also being discarded. This no
5905 longer happens. I'm not sure if this situation could ever has previously
5908 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5911 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5912 header. This is something that is recommended in a new Internet Draft, and
5913 is something that is documented as being done by Sendmail. There are two
5914 possible values. For messages generated by the autoreply transport, Exim
5917 Auto-Submitted: auto-replied
5919 whereas for all other generated messages (e.g. bounces) it adds
5921 Auto-Submitted: auto-generated
5923 63. The "personal" condition in filters now includes a test for the
5924 Auto-Submitted: header. If it contains the string "auto-" the message it
5925 not considered personal.
5927 64. Added rcpt_include_affixes as a generic transport option.
5929 65. Added queue_only_override (default true).
5931 66. Added the syslog_duplication option.
5933 67. If what should have been the first header line of a message consisted of
5934 a space followed by a colon, Exim was mis-interpreting it as a header line.
5935 It isn't of course - it is syntactically invalid and should therefore be
5936 treated as the start of the message body. The misbehaviour could have
5937 caused a number of strange effects, including loss of data in subsequent
5938 header lines, and spool format errors.
5940 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5941 client host had authenticated. This control can now be exercised by an ACL
5942 for more flexibility.
5944 69. By default, callouts do not happen when testing with -bh. There is now a
5945 variant, -bhc, which does actually run the callout code, including
5946 consulting and updating the callout cache.
5948 70. Added support for saslauthd authentication, courtesy of Alexander
5951 71. If statvfs() failed on the spool or log directories while checking their
5952 size for availability, Exim confusingly gave the error "space shortage".
5953 Furthermore, in debugging mode it crashed with a floating point exception.
5954 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5955 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5956 serious problem, Exim now writes to the main and panic logs when this
5957 happens, with details of the failure. It then refuses to accept the
5958 incoming message, giving the message "spool directory problem" or "log
5959 directory problem" with a 421 code for SMTP messages.
5961 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5962 0, 1, and 2 exist, because some OS complain for execs without them (see
5963 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5964 descriptors. However, the code omitted to check that the open succeeded,
5965 causing mysterious errors if for some reason the permissions on /dev/null
5966 got screwed. Now Exim writes a message to the main and panic logs, and
5967 bombs out if it can't open /dev/null.
5969 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5970 interact so that it is all more flexible. It is supposed to remain
5971 backwards compatible. Also added extra_local_interfaces.
5973 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5974 to bomb out with an assertion failure - to the client this appears as a
5975 connection drop. This problem occurs in the part of the code that was taken
5976 from the Samba project. Fortunately, the assertion is in a very simple
5977 function, so I have fixed this by reproducing the function inline in the
5978 one place where it is called, and arranging for authentication to fail
5979 instead of killing the process with assert().
5981 75. The SPA client code was not working when the server requested OEM rather
5982 than Unicode encoding.
5984 76. Added code to make require_files with a specific uid setting more usable in
5985 the case where statting the file as root fails - usually a non-root-mounted
5986 NFS file system. When this happens and the failure is EACCES, Exim now
5987 forks a subprocess and does the per-uid checking as the relevant uid.
5989 77. Added process_log_path.
5991 78. If log_file_path was not explicitly set, a setting of check_log_space or
5992 check_log_inodes was ignored.
5994 79. If a space check for the spool or log partitions fails, the incident is now
5995 logged. Of course, in the latter case the data may get lost...
5997 80. Added the %p formatting code to string_format() so that it can be used to
5998 print addresses in debug_print(). Adjusted all the address printing in the
5999 debugging in store.c to use %p rather than %d.
6001 81. There was a concern that a line of code in smtp_in.c could overflow a
6002 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6003 initially expressed, the concern was not well-founded, because trailing
6004 spaces are removed early. However, if the trailing spaces were followed by
6005 a NULL, they did not get removed, so the overflow was possible. Two fixes
6008 (a) I re-wrote the offending code in a cleaner fashion.
6009 (b) If an incoming SMTP command contains a NULL character, it is rejected
6012 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6013 runs initgroups(), so that if the Exim user is in any additional groups,
6014 they will be used during message reception.
6020 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob