1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router and the smtp transport (applying to the forward
91 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
92 of ldap servers used for a specific lookup. Patch provided by Heiko
95 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 New variable $lookup_dnssec_authenticated for observability.
98 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
99 Patch submitted by Lars Timman.
101 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
103 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
104 Requires trusted mode and valid format message id, aborts otherwise.
105 Patch contributed by Heiko Schlichting.
107 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
108 certextract with support for various fields. Bug 1358.
110 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
111 is requested by default, modifiable by smtp transport option
114 JH/22 Expansion operators ${md5:string} and ${sha1::string} can now
115 operate on certificate variables to give certificate fingerprints
116 Also new ${sha256:cert_variable}.
118 JH/23 The PRDR feature is moved from being Experimental into the mainline.
124 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
126 PP/02 Make -n do something, by making it not do something.
127 When combined with -bP, the name of an option is not output.
129 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
132 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
133 $sender_host_name and config options to manage this, and basic check
136 PP/05 DSCP support for outbound connections and control modifier for inbound.
138 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
139 (Only plugin which currently uses this is kerberos4, which nobody should
140 be using, but we should make it available and other future plugins might
141 conceivably use it, even though it would break NAT; stuff *should* be
142 using channel bindings instead).
144 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
145 name; added for Sendmail compatibility; requires admin caller.
146 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
147 just ignore it); requires trusted caller.
148 Also parse but ignore: -Ac -Am -X<logfile>
151 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
153 TL/02 Add +smtp_confirmation as a default logging option.
155 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
156 Patch by Magnus Holmgren from 2007-02-20.
158 TL/04 Bugzilla 1281 - Spec typo.
159 Bugzilla 1283 - Spec typo.
160 Bugzilla 1290 - Spec grammar fixes.
162 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
164 TL/06 Add Experimental DMARC support using libopendmarc libraries.
166 TL/07 Fix an out of order global option causing a segfault. Reported to dev
167 mailing list by by Dmitry Isaikin.
169 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
171 JH/02 Support "G" suffix to numbers in ${if comparisons.
173 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
175 NM/01 Bugzilla 1197 - Spec typo
176 Bugzilla 1196 - Spec examples corrections
178 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
180 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
181 gnutls_enable_pkcs11, but renamed to more accurately indicate its
184 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
185 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
187 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
188 "acl {{name}{arg}...}", and optional args on acl condition
191 JH/05 Permit multiple router/transport headers_add/remove lines.
193 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
195 JH/07 Avoid using a waiting database for a single-message-only transport.
196 Performance patch from Paul Fisher. Bugzilla 1262.
198 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
201 JH/09 Add $headers_added variable, with content from use of ACL modifier
202 add_header (but not yet added to the message). Bugzilla 199.
204 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
205 Pulled from Bugzilla 817 by Wolfgang Breyha.
207 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
209 (nb: this is the same fix as in Exim 4.80.1)
211 JH/11 Add A= logging on delivery lines, and a client_set_id option on
214 JH/12 Add optional authenticated_sender logging to A= and a log_selector
217 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
219 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
220 advertise SMTP AUTH mechanism to us, instead of a generic
221 protocol violation error. Also, make Exim more robust to bad
222 data from the Dovecot auth socket.
224 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
226 When a queue runner is handling a message, Exim first routes the
227 recipient addresses, during which it prunes them based on the retry
228 hints database. After that it attempts to deliver the message to
229 any remaining recipients. It then updates the hints database using
232 So if a recipient address works intermittently, it can get repeatedly
233 deferred at routing time. The retry hints record remains fresh so the
234 address never reaches the final cutoff time.
236 This is a fairly common occurrence when a user is bumping up against
237 their storage quota. Exim had some logic in its local delivery code
238 to deal with this. However it did not apply to per-recipient defers
239 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
241 This change adds a proper retry rule check during routing so that the
242 final cutoff time is checked against the message's age. We only do
243 this check if there is an address retry record and there is not a
244 domain retry record; this implies that previous attempts to handle
245 the address had the retry_use_local_parts option turned on. We use
246 this as an approximation for the destination being like a local
247 delivery, as in LMTP.
249 I suspect this new check makes the old local delivery cutoff check
250 redundant, but I have not verified this so I left the code in place.
252 TF/02 Correct gecos expansion when From: is a prefix of the username.
254 Test 0254 submits a message to Exim with the header
258 When I ran the test suite under the user fanf2, Exim expanded
259 the header to contain my full name, whereas it should have added
260 a Resent-Sender: header. It erroneously treats any prefix of the
261 username as equal to the username.
263 This change corrects that bug.
265 GF/01 DCC debug and logging tidyup
266 Error conditions log to paniclog rather than rejectlog.
267 Debug lines prefixed by "DCC: " to remove any ambiguity.
269 TF/03 Avoid unnecessary rebuilds of lookup-related code.
271 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
272 Bug spotted by Jeremy Harris; was flawed since initial commit.
273 Would have resulted in OCSP responses post-SNI triggering an Exim
274 NULL dereference and crash.
276 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
278 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
279 Bug detection, analysis and fix by Samuel Thibault.
280 Bugzilla 1331, Debian bug #698092.
282 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
284 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
285 Server implementation by Todd Lyons, client by JH.
286 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
287 config variable "prdr_enable" controls whether the server
288 advertises the facility. If the client requests PRDR a new
289 acl_data_smtp_prdr ACL is called once for each recipient, after
290 the body content is received and before the acl_smtp_data ACL.
291 The client is controlled by bolth of: a hosts_try_prdr option
292 on the smtp transport, and the server advertisement.
293 Default client logging of deliveries and rejections involving
294 PRDR are flagged with the string "PRDR".
296 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
297 fclose(). Diagnosis by Todd Lyons.
299 PP/17 Update configure.default to handle IPv6 localhost better.
300 Patch by Alain Williams (plus minor tweaks).
303 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
304 This is now consistent with GnuTLS, and is now documented: the
305 previous undocumented portable approach to treating the option as
306 unset was to force an expansion failure. That still works, and
307 an empty string is now equivalent.
309 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
310 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
311 not performing validation itself.
313 PP/20 Added force_command boolean option to pipe transport.
314 Patch from Nick Koston, of cPanel Inc.
316 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
319 TF/04 Added udpsend ACL modifer and hexquote expansion operator
321 PP/21 Fix eximon continuous updating with timestamped log-files.
322 Broken in a format-string cleanup in 4.80, missed when I repaired the
323 other false fix of the same issue.
324 Report and fix from Heiko Schlichting.
327 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
328 Report from Prashanth Katuri.
330 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
331 It's SecureTransport, so affects any MacOS clients which use the
332 system-integrated TLS libraries, including email clients.
334 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
335 using a MIME ACL for non-SMTP local injection.
336 Report and assistance in diagnosis by Warren Baker.
338 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
340 JH/16 Fix comparisons for 64b. Bugzilla 1385.
342 TL/09 Add expansion variable $authenticated_fail_id to keep track of
343 last id that failed so it may be referenced in subsequent ACL's.
345 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
348 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
349 ldap library initialization, allowing self-signed CA's to be
350 used. Also properly sets require_cert option later in code by
351 using NULL (global ldap config) instead of ldap handle (per
352 session). Bug diagnosis and testing by alxgomz.
354 TL/12 Enhanced documentation in the ratelimit.pl script provided in
355 the src/util/ subdirectory.
357 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
358 renamed to Transport Post Delivery Action by Jeremy Harris, as
361 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
362 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
363 redis_servers = needs to be configured which will be used by the redis
364 lookup. Patch from Warren Baker, of The Packet Hub.
366 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
368 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
369 hostname or reverse DNS when processing a host list. Used suggestions
370 from multiple comments on this bug.
372 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
374 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
375 Missed a few lines, added it to make the runtest require no keyboard
378 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
379 contains upper case chars. Make router use caseful_local_part.
381 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
382 support when GnuTLS has been built with p11-kit.
388 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
390 This, or similar/improved, will also be change PP/11 of 4.82.
396 PP/01 Handle short writes when writing local log-files.
397 In practice, only affects FreeBSD (8 onwards).
398 Bugzilla 1053, with thanks to Dmitry Isaikin.
400 NM/01 Bugzilla 949 - Documentation tweak
402 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
405 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
407 PP/02 Implemented gsasl authenticator.
409 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
411 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
412 `pkg-config foo` for cflags/libs.
414 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
415 with rest of GSASL and with heimdal_gssapi.
417 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
418 `pkg-config foo` for cflags/libs for the TLS implementation.
420 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
421 properties get this fed in as external SSF. A number of robustness
422 and debugging improvements to the cyrus_sasl authenticator.
424 PP/08 cyrus_sasl server now expands the server_realm option.
426 PP/09 Bugzilla 1214 - Log authentication information in reject log.
427 Patch by Jeremy Harris.
429 PP/10 Added dbmjz lookup type.
431 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
433 PP/12 MAIL args handles TAB as well as SP, for better interop with
434 non-compliant senders.
435 Analysis and variant patch by Todd Lyons.
437 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
438 Bug report from Lars Müller <lars@samba.org> (via SUSE),
439 Patch from Dirk Mueller <dmueller@suse.com>
441 PP/13 tls_peerdn now print-escaped for spool files.
442 Observed some $tls_peerdn in wild which contained \n, which resulted
443 in spool file corruption.
445 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
446 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
447 or write after TLS renegotiation, which otherwise led to messages
450 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
451 as a tracking header (ie: a signed header comes before the signature).
452 Patch from Wolfgang Breyha.
454 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
455 comma-sep list; embedded commas doubled.
457 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
459 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
461 Report and patch from Dmitry Banschikov.
463 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
464 Removed SSL_clear() after SSL_new() which led to protocol negotiation
465 failures. We appear to now support TLS1.1+ with Exim.
467 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
468 lets Exim select keys and certificates based upon TLS SNI from client.
469 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
470 before an outbound SMTP session. New log_selector, +tls_sni.
472 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
473 NULL dereference. Report and patch from Alun Jones.
475 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
476 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
477 Not seeing resolver debug output on NetBSD, but suspect this is a
478 resolver implementation change.
480 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
481 Left warnings. Added "eximon gdb" invocation mode.
483 PP/21 Defaulting "accept_8bitmime" to true, not false.
485 PP/22 Added -bw for inetd wait mode support.
487 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
488 locate the relevant includes and libraries. Made this the default.
490 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
491 Bugzilla 1246, report and most of solution from Tomasz Kusy.
493 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
494 This may cause build issues on older platforms.
496 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
497 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
498 gnutls_require_mac & gnutls_require_protocols (no longer supported).
499 Added SNI support via GnuTLS too.
500 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
502 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
504 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
505 Applied second patch from Janne, implementing suggestion to default
506 multiple-strings-in-record handling to match SPF spec.
508 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
510 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
511 read-only, out of scope).
512 Patch from Wolfgang Breyha, report from Stuart Northfield.
514 PP/29 Fix three issues highlighted by clang analyser static analysis.
515 Only crash-plausible issue would require the Cambridge-specific
516 iplookup router and a misconfiguration.
517 Report from Marcin Mirosław.
519 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
521 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
522 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
523 As part of this, removing so much warning spew let me fix some minor
524 real issues in debug logging.
526 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
527 assignment on my part. Fixed.
529 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
530 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
531 Janne Snabb (who went above and beyond: thank you).
533 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
534 string otherwise requires a connection and a bunch more work and it's
535 relatively easy to get wrong. Should also expose TLS library linkage
538 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
539 64-bit ${eval} (JH/03).
541 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
542 GNU libc to support some of the 64-bit stuff, should not lead to
543 conflicts. Defined before os.h is pulled in, so if a given platform
544 needs to override this, it can.
546 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
547 protection layer was required, which is not implemented.
548 Bugzilla 1254, patch from Wolfgang Breyha.
550 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
551 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
552 tls_dhparam take prime identifiers. Also unbreak combination of
553 OpenSSL+DH_params+TLSSNI.
555 PP/39 Disable SSLv2 by default in OpenSSL support.
561 PP/01 Solaris build fix for Oracle's LDAP libraries.
562 Bugzilla 1109, patch from Stephen Usher.
564 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
566 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
569 TF/02 Fix a couple more cases where we did not log the error message
570 when unlink() failed. See also change 4.74-TF/03.
572 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
573 lock up or crash if it happened to be inside a call to libc when it
574 got a SIGUSR1 from exiwhat.
576 The SIGUSR1 handler appends the current process status to the process
577 log which is later printed by exiwhat. It used to use the general
578 purpose logging code to do this, but several functions it calls are
579 not safe for signals.
581 The new output code in the SIGUSR1 handler is specific to the process
582 log, and simple enough that it's easy to inspect for signal safety.
583 Removing some special cases also simplifies the general logging code.
584 Removing the spurious timestamps from the process log simplifies
587 TF/04 Improved ratelimit ACL condition.
589 The /noupdate option has been deprecated in favour of /readonly which
590 has clearer semantics. The /leaky, /strict, and /readonly update modes
591 are mutually exclusive. The update mode is no longer included in the
592 database key; it just determines when the database is updated. (This
593 means that when you upgrde Exim will forget old rate measurements.)
595 Exim now checks that the per_* options are used with an update mode that
596 makes sense for the current ACL. For example, when Exim is processing a
597 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
598 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
599 must specify per_mail/readonly. If you omit the update mode it defaults to
600 /leaky where that makes sense (as before) or /readonly where required.
602 The /noupdate option is now undocumented but still supported for
603 backwards compatibility. It is equivalent to /readonly except that in
604 ACLs where /readonly is required you may specify /leaky/noupdate or
605 /strict/noupdate which are treated the same as /readonly.
607 A useful new feature is the /count= option. This is a generalization
608 of the per_byte option, so that you can measure the throughput of other
609 aggregate values. For example, the per_byte option is now equivalent
610 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
612 The per_rcpt option has been generalized using the /count= mechanism
613 (though it's more complicated than the per_byte equivalence). When it is
614 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
615 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
616 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
617 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
618 Note that using per_rcpt with a non-readonly update mode in more than
619 one ACL will cause the recipients to be double-counted. (The per_mail
620 and per_byte options don't have this problem.)
622 The handling of very low rates has changed slightly. If the computed rate
623 is less than the event's count (usually one) then this event is the first
624 after a long gap. In this case the rate is set to the same as this event's
625 count, so that the first message of a spam run is counted properly.
627 The major new feature is a mechanism for counting the rate of unique
628 events. The new per_addr option counts the number of different
629 recipients that someone has sent messages to in the last time period. It
630 behaves like per_rcpt if all the recipient addresses are different, but
631 duplicate recipient addresses do not increase the measured rate. Like
632 the /count= option this is a general mechanism, so the per_addr option
633 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
634 example, measure the rate that a client uses different sender addresses
635 with the options per_mail/unique=$sender_address. There are further
636 details in the main documentation.
638 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
640 TF/06 Removed a few PCRE remnants.
642 TF/07 Automatically extract Exim's version number from tags in the git
643 repository when doing development or release builds.
645 PP/02 Raise smtp_cmd_buffer_size to 16kB.
646 Bugzilla 879. Patch from Paul Fisher.
648 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
649 Heavily based on revision 40f9a89a from Simon Arlott's tree.
652 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
654 PP/05 Variable $av_failed, true if the AV scanner deferred.
655 Bugzilla 1078. Patch from John Horne.
657 PP/06 Stop make process more reliably on build failure.
658 Bugzilla 1087. Patch from Heiko Schlittermann.
660 PP/07 Make maildir_use_size_file an _expandable_ boolean.
661 Bugzilla 1089. Patch from Heiko Schlittermann.
663 PP/08 Handle ${run} returning more data than OS pipe buffer size.
664 Bugzilla 1131. Patch from Holger Weiß.
666 PP/09 Handle IPv6 addresses with SPF.
667 Bugzilla 860. Patch from Wolfgang Breyha.
669 PP/10 GnuTLS: support TLS 1.2 & 1.1.
671 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
674 PP/11 match_* no longer expand right-hand-side by default.
675 New compile-time build option, EXPAND_LISTMATCH_RHS.
676 New expansion conditions, "inlist", "inlisti".
678 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
680 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
682 PP/14 fix log_write() format string regression from TF/03.
683 Bugzilla 1152. Patch from Dmitry Isaikin.
689 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
691 PP/02 Harmonised TLS library version reporting; only show if debugging.
692 Layout now matches that introduced for other libraries in 4.74 PP/03.
694 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
696 PP/04 New "dns_use_edns0" global option.
698 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
701 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
702 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
704 TK/01 Updated PolarSSL code to 0.14.2.
705 Bugzilla 1097. Patch from Andreas Metzler.
707 PP/07 Catch divide-by-zero in ${eval:...}.
710 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
713 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
714 format-string attack -- SECURITY: remote arbitrary code execution.
716 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
717 time unintentionally subject to list matching rules, letting the header
718 cause arbitrary Exim lookups (of items which can occur in lists, *not*
719 arbitrary string expansion). This allowed for information disclosure.
721 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
722 INT_MIN/-1 -- value coerced to INT_MAX.
728 NM/01 Workround for PCRE version dependency in version reporting
731 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
732 This fixes portability to compilers other than gcc, notably
733 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
735 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
736 makefiles for portability to HP-UX and POSIX correctness.
738 PP/01 Permit LOOKUP_foo enabling on the make command-line.
739 Also via indented variable definition in the Makefile.
740 (Debugging by Oliver Heesakkers).
742 PP/02 Restore caching of spamd results with expanded spamd_address.
743 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
745 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
746 Improves build reliability. Fix from: Frank Elsner
748 NM/02 Fix wide character breakage in the rfc2047 coding
749 Fixes bug 1064. Patch from Andrey N. Oktyabrski
751 NM/03 Allow underscore in dnslist lookups
752 Fixes bug 1026. Patch from Graeme Fowler
754 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
755 Code patches from Adam Ciarcinski of NetBSD.
757 NM/04 Fixed exiqgrep to cope with mailq missing size issue
760 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
761 is logged, to avoid truncation. Patch from John Horne.
763 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
764 Patch from Jakob Hirsch.
766 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
767 SQL string expansion failure details.
768 Patch from Andrey Oktyabrski.
770 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
771 Patch from Simon Arlott.
773 PP/09 New lookups functionality failed to compile on old gcc which rejects
774 extern declarations in function scope.
775 Patch from Oliver Fleischmann
777 PP/10 Use sig_atomic_t for flags set from signal handlers.
778 Check getgroups() return and improve debugging.
779 Fixed developed for diagnosis in bug 927 (which turned out to be
782 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
783 Patch from Mark Zealey.
785 PP/12 Bugzilla 1056: Improved spamd server selection.
786 Patch from Mark Zealey.
788 PP/13 Bugzilla 1086: Deal with maildir quota file races.
789 Based on patch from Heiko Schlittermann.
791 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
792 Patch from Uwe Doering, sign-off by Michael Haardt.
794 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
795 variable declaration deep within a block. Bug and patch from
798 PP/15 lookups-Makefile IRIX compatibilty coercion.
800 PP/16 Make DISABLE_DKIM build knob functional.
802 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
803 Patch by Simon Arlott
805 TF/03 Fix valgrind.h portability to C89 compilers that do not support
806 variable argument macros. Our copy now differs from upstream.
812 TF/01 Failure to get a lock on a hints database can have serious
813 consequences so log it to the panic log.
815 TF/02 Log LMTP confirmation messages in the same way as SMTP,
816 controlled using the smtp_confirmation log selector.
818 TF/03 Include the error message when we fail to unlink a spool file.
820 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
821 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
822 for maintaining out-of-tree patches for some time.
824 PP/01 Bugzilla 139: Documentation and portability issues.
825 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
826 Handle per-OS dynamic-module compilation flags.
828 PP/02 Let /dev/null have normal permissions.
829 The 4.73 fixes were a little too stringent and complained about the
830 permissions on /dev/null. Exempt it from some checks.
831 Reported by Andreas M. Kirchwitz.
833 PP/03 Report version information for many libraries, including
834 Exim version information for dynamically loaded libraries. Created
835 version.h, now support a version extension string for distributors
836 who patch heavily. Dynamic module ABI change.
838 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
839 privilege escalation vulnerability whereby the Exim run-time user
840 can cause root to append content of the attacker's choosing to
843 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
846 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
847 If dropping privileges for untrusted macros, we disabled normal logging
848 on the basis that it would fail; for the Exim run-time user, this is not
849 the case, and it resulted in successful deliveries going unlogged.
850 Fixed. Reported by Andreas Metzler.
856 PP/01 Date: & Message-Id: revert to normally being appended to a message,
857 only prepend for the Resent-* case. Fixes regression introduced in
858 Exim 4.70 by NM/22 for Bugzilla 607.
860 PP/02 Include check_rfc2047_length in configure.default because we're seeing
861 increasing numbers of administrators be bitten by this.
863 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
865 PP/03 Bugzilla 994: added openssl_options main configuration option.
867 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
869 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
871 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
873 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
874 without a peer certificate, leading to a segfault because of an
875 assumption that peers always have certificates. Be a little more
876 paranoid. Problem reported by Martin Tscholak.
878 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
879 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
880 NB: ClamAV planning to remove STREAM in "middle of 2010".
881 CL also introduces -bmalware, various -d+acl logging additions and
882 more caution in buffer sizes.
884 PP/09 Implemented reverse_ip expansion operator.
886 PP/10 Bugzilla 937: provide a "debug" ACL control.
888 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
890 PP/12 Bugzilla 973: Implement --version.
892 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
894 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
896 PP/15 Bugzilla 816: support multiple condition rules on Routers.
898 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
899 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
900 ignore trailing whitespace.
902 JJ/02 prevent non-panic DKIM error from being sent to paniclog
904 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
907 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
908 Notification from Dr Andrew Aitchison.
910 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
911 ExtendedDetectionInfo response format.
912 Notification from John Horne.
914 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
917 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
918 XSL and documented dependency on system catalogs, with examples of how
921 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
924 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
925 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
926 configuration file which is writeable by the Exim user or group.
928 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
929 of configuration files to cover files specified with the -C option if
930 they are going to be used with root privileges, not just the default
933 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
934 option (effectively making it always true).
936 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
937 files to be used while preserving root privileges.
939 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
940 that rogue child processes cannot use them.
942 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
943 run-time user, instead of root.
945 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
946 Exim run-time user without dropping privileges.
948 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
949 result string, instead of calling string_vformat() twice with the same
952 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
953 for other users. Others should always drop root privileges if they use
954 -C on the command line, even for a whitelisted configure file.
956 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
958 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
964 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
965 $data_path, and $header_path variables; fixed documentation bugs and
968 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
969 exipick to access non-standard spools, including the "frozen" queue
972 NM/01 Bugzilla 965: Support mysql stored procedures.
973 Patch from Alain Williams
975 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
977 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
978 Patch from Andreas Metzler
980 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
981 Patch from Kirill Miazine
983 NM/05 Bugzilla 671: Added umask to procmail example.
985 JJ/03 installed exipick 20100323.0, fixing doc bug
987 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
988 directory. Notification and patch from Dan Rosenberg.
990 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
992 TK/02 Improve log output when DKIM signing operation fails.
994 MH/01 Treat the transport option dkim_domain as a colon separated
995 list, not as a single string, and sign the message with each element,
996 omitting multiple occurences of the same signer.
998 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
999 Bugzilla 985, 986. Patch by Simon Arlott
1001 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1002 Patch by Simon Arlott
1004 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1005 MBX locking. Notification from Dan Rosenberg.
1011 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1013 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1015 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1017 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1019 NM/04 Bugzilla 915: Flag broken perl installation during build.
1025 TK/01 Added patch by Johannes Berg that expands the main option
1026 "spamd_address" if it starts with a dollar sign.
1028 TK/02 Write list of recipients to X-Envelope-Sender header when building
1029 the mbox-format spool file for content scanning (suggested by Jakob
1032 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1033 (http://www.dcc-servers.net/) support via dccifd. Activated by
1034 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1036 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1037 by Mark Daniel Reidel <mr@df.eu>.
1039 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1040 When building exim an external PCRE library is now needed -
1041 PCRE is a system library on the majority of modern systems.
1042 See entry on PCRE_LIBS in EDITME file.
1044 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1045 conversation. Added nologin parameter to request.
1046 Patch contributed by Kirill Miazine.
1048 TF/01 Do not log submission mode rewrites if they do not change the address.
1050 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1052 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1053 log files in place. Contributed by Roberto Lima.
1055 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1057 TF/03 Bugzilla 615: When checking the local_parts router precondition
1058 after a local_part_suffix or local_part_prefix option, Exim now
1059 does not use the address's named list lookup cache, since this
1060 contains cached lookups for the whole local part.
1062 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1063 Robert Millan. Documentation is in experimental-spec.txt.
1065 TF/04 Bugzilla 668: Fix parallel build (make -j).
1067 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1069 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1070 Patch provided by Jan Srzednicki.
1072 TF/05 Leading white space used to be stripped from $spam_report which
1073 wrecked the formatting. Now it is preserved.
1075 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1076 that they are available at delivery time.
1078 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1080 TF/08 TLS error reporting now respects the incoming_interface and
1081 incoming_port log selectors.
1083 TF/09 Produce a more useful error message if an SMTP transport's hosts
1084 setting expands to an empty string.
1086 NM/06 Bugzilla 744: EXPN did not work under TLS.
1087 Patch provided by Phil Pennock.
1089 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1090 Patch provided by Richard Godbee.
1092 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1093 acl_smtp_notquit, added index entry.
1095 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1096 Patch provided by Eugene Bujak.
1098 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1099 accept(). Patch provided by Maxim Dounin.
1101 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1102 Patch provided by Phil Pennock.
1104 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1106 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1107 Patch provided by Brad "anomie" Jorsch.
1109 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1110 Patch provided by Dean Brooks.
1112 TK/05 Add native DKIM support (does not depend on external libraries).
1114 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1115 Patch provided by Graeme Fowler.
1117 NM/16 Bugzilla 851: Documentation example syntax fix.
1119 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1121 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1124 NM/19 Bugzilla 745: TLS version reporting.
1125 Patch provided by Phil Pennock.
1127 NM/20 Bugzilla 167: bool: condition support.
1128 Patch provided by Phil Pennock.
1130 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1131 clients. Patch provided by Phil Pennock.
1133 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1134 Patch provided by Brad "anomie" Jorsch.
1136 NM/23 Bugzilla 687: Fix misparses in eximstats.
1137 Patch provided by Heiko Schlittermann.
1139 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1140 Patch provided by Heiko Schlittermann.
1142 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1143 plus update to original patch.
1145 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1147 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1148 Patch provided by David Brownlee.
1150 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1152 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1154 NM/30 Bugzilla 888: TLS documentation bugfixes.
1156 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1158 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1159 Unlike the original bugzilla I have changed all shell scripts in src tree.
1161 NM/33 Bugzilla 898: Transport filter timeout fix.
1162 Patch by Todd Rinaldo.
1164 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1165 Patch by Serge Demonchaux.
1167 NM/35 Bugzilla 39: Base64 decode bug fixes.
1168 Patch by Jakob Hirsch.
1170 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1172 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1174 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1176 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1182 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1183 ALT-N's libdkim that I have put here:
1184 http://duncanthrax.net/exim-experimental/
1186 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1187 were called 'true' and it seems that C99 defines that as a reserved
1188 keyword to be used with 'bool' variable types. That means you could
1189 not include C99-style headers which use bools without triggering
1190 build errors in sieve.c.
1192 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1193 as mailq or other aliases. Changed the --help handling significantly
1194 to do whats expected. exim_usage() emits usage/help information.
1196 SC/01 Added the -bylocaldomain option to eximstats.
1198 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1200 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1202 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1208 PH/01 Another patch from the Sieve maintainer.
1210 PH/02 When an IPv6 address is converted to a string for single-key lookup
1211 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1212 dots are used instead of colons so that keys in lsearch files need not
1213 contain colons. This was done some time before quoting was made available
1214 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1215 (notated using the quote facility) so as to distinguish them from IPv4
1216 keys. This meant that lookups for IP addresses in host lists did not work
1217 for iplsearch lookups.
1219 This has been fixed by arranging for IPv6 addresses to be expressed with
1220 colons if the lookup type is iplsearch. This is not incompatible, because
1221 previously such lookups could never work.
1223 The situation is now rather anomolous, since one *can* have colons in
1224 ordinary lsearch keys. However, making the change in all cases is
1225 incompatible and would probably break a number of configurations.
1227 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1230 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1231 conversion specification without a maximum field width, thereby enabling
1232 a rogue spamd server to cause a buffer overflow. While nobody in their
1233 right mind would setup Exim to query an untrusted spamd server, an
1234 attacker that gains access to a server running spamd could potentially
1235 exploit this vulnerability to run arbitrary code as the Exim user.
1237 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1238 $primary_hostname instead of what libspf2 thinks the hosts name is.
1240 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1241 a directory entry by the name of the lookup key. Previously, if a
1242 symlink pointed to a non-existing file or a file in a directory that
1243 Exim lacked permissions to read, a lookup for a key matching that
1244 symlink would fail. Now it is enough that a matching directory entry
1245 exists, symlink or not. (Bugzilla 503.)
1247 PH/03 The body_linecount and body_zerocount variables are now exported in the
1250 PH/04 Added the $dnslist_matched variable.
1252 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1253 This means they are set thereafter only if the connection becomes
1256 PH/06 Added the client_condition to authenticators so that some can be skipped
1257 by clients under certain conditions.
1259 PH/07 The error message for a badly-placed control=no_multiline_responses left
1260 "_responses" off the end of the name.
1262 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1264 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1265 (without spaces) instead of just copying the configuration text.
1267 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1269 PH/11 Added $max_received_linelength.
1271 PH/12 Added +ignore_defer and +include_defer to host lists.
1273 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1274 way in which PCRE > 7.0 is built.
1276 PH/14 Implemented queue_only_load_latch.
1278 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1279 MAIL command. The effect was to mangle the value on 64-bit systems.
1281 PH/16 Another patch from the Sieve maintainer.
1283 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1285 PH/18 If a system quota error occurred while trying to create the file for
1286 a maildir delivery, the message "Mailbox is full" was not appended to the
1287 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1288 only to a quota excession during the actual writing of the file.
1290 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1291 characters?) which causes problems in log lines. The DN values are now
1292 passed through string_printing() before being added to log lines.
1294 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1295 and InterBase are left for another time.)
1297 PH/21 Added message_body_newlines option.
1299 PH/22 Guard against possible overflow in moan_check_errorcopy().
1301 PH/23 POSIX allows open() to be a macro; guard against that.
1303 PH/24 If the recipient of an error message contained an @ in the local part
1304 (suitably quoted, of course), incorrect values were put in $domain and
1305 $local_part during the evaluation of errors_copy.
1311 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1312 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1315 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1316 issue a MAIL command.
1318 PH/02 In an ACL statement such as
1320 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1322 if a client was not listed at all, or was listed with a value other than
1323 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1324 the condition was not true (as it should be), so access was not denied.
1325 The bug was that the ! inversion was incorrectly passed on to the second
1326 item. This has been fixed.
1328 PH/03 Added additional dnslists conditions == and =& which are different from
1329 = and & when the dns lookup returns more than one IP address.
1331 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1332 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1334 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1335 FSYNC, which compiles an option called disable_fsync that allows for
1336 bypassing fsync(). The documentation is heavily laced with warnings.
1338 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1340 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1341 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1342 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1343 including adding "make clean"; (3) Added -fPIC when compiling the test
1344 dynamically loaded module, to get rid of a warning.
1346 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1347 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1348 = 0s. The bug is otherwise harmless.
1350 PH/07 There was a bug in the dovecot authenticator such that the value of
1351 $auth1 could be overwritten, and so not correctly preserved, after a
1352 successful authentication. This usually meant that the value preserved by
1353 the server_setid option was incorrect.
1355 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1357 PH/09 Installed PCRE release 7.0.
1359 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1360 run for batched SMTP input. It is now run at the start of every message
1361 in the batch. While fixing this I discovered that the process information
1362 (output by running exiwhat) was not always getting set for -bs and -bS
1363 input. This is fixed, and it now also says "batched" for BSMTP.
1365 PH/11 Added control=no_pipelining.
1367 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1368 patch, slightly modified), and move the expansion of helo_data till after
1369 the connection is made in the smtp transport (so it can use these
1372 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1374 PH/14 Added log_selector = +pid.
1376 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1378 PH/16 Add ${if forany and ${if forall.
1380 PH/17 Added dsn_from option to vary the From: line in DSNs.
1382 PH/18 Flush SMTP output before performing a callout, unless control =
1383 no_callout_flush is set.
1385 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1386 was true (the default) a successful delivery failed to delete the retry
1387 item, thus causing premature timeout of the address. The bug is now
1390 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1392 PH/21 Long custom messages for fakedefer and fakereject are now split up
1393 into multiline reponses in the same way that messages for "deny" and
1394 other ACL rejections are.
1396 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1397 with slight modification.
1399 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1400 draft, changing the syntax and factoring some duplicate code".
1402 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1403 for deliveries of the second and subsequent messages over the same SMTP
1406 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1407 ${reduce, with only minor "tidies".
1409 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1411 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1412 expansion side effects.
1414 PH/27 When a message times out after an over-quota error from an Exim-imposed
1415 quota, the bounce message says "mailbox is full". This message was not
1416 being given when it was a system quota that was exceeded. It now should
1419 MH/03 Made $recipients available in local_scan(). local_scan() already has
1420 better access to the recipient list through recipients_list[], but
1421 $recipients can be useful in postmaster-provided expansion strings.
1423 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1424 in the case of a MAIL command with additional options following the
1425 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1426 were accidentally chopped off.
1428 PH/29 SMTP synchronization checks are implemented when a command is read -
1429 there is a check that no more input is waiting when there shouldn't be
1430 any. However, for some commands, a delay in an ACL can mean that it is
1431 some time before the response is written. In this time, more input might
1432 arrive, invalidly. So now there are extra checks after an ACL has run for
1433 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1434 pipelining has not been advertised.
1436 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1438 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1439 returned in the SMTP response when smtp_return_error_details was false.
1440 This has been fixed.
1442 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1443 instead of the C I/O that was originally supplied, because problems were
1444 reported on Solaris.
1446 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1447 Exim which did not show up earlier: it was assuming that a call to
1448 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1449 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1450 was a macro that became an assignment, so it seemed to work. This has
1451 changed to a proper function call with a void return, hence the compile
1452 error. Exim's code has been fixed.
1454 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1457 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1459 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1461 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1462 because it turns out to be too big or there is a timeout) there is no
1463 "Completed" line in the log. When some messages of this type were
1464 selected by exigrep, they were listed as "not completed". Others were
1465 picked up by some special patterns. I have improved the selection
1466 criteria to be more general.
1468 PH/38 The host_find_failed option in the manualroute router can now be set
1469 to "ignore", to completely ignore a host whose IP address cannot be
1470 found. If all hosts are ignored, the behaviour is controlled by the new
1471 host_all_ignored option.
1473 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1474 homing or because of multiple MX records with /mx) generated more than
1475 one IP address, and the following item turned out to be the local host,
1476 all the secondary addresses of the first item were incorrectly removed
1477 from the list, along with the local host and any following hosts (which
1478 is what is supposed to happen).
1480 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1481 whoever called Exim into the -H file. In the case of the daemon it was
1482 behaving confusingly. When first started, it used values for whoever
1483 started the daemon, but after a SIGHUP it used the Exim user (because it
1484 calls itself on a restart). I have changed the code so that it now always
1487 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1488 message are rejected with the same error (e.g. no authentication or bad
1489 sender address), and a DATA command is nevertheless sent (as can happen
1490 with PIPELINING or a stupid MUA), the error message that was given to the
1491 RCPT commands is included in the rejection of the DATA command. This is
1492 intended to be helpful for MUAs that show only the final error to their
1495 PH/42 Another patch from the Sieve maintainer.
1497 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1498 Eximstats - Fixed some broken HTML links and added missing column headers
1500 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1501 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1503 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1505 PH/43 Yet another patch from the Sieve maintainer.
1507 PH/44 I found a way to check for a TCP/IP connection going away before sending
1508 the response to the final '.' that terminates a message, but only in the
1509 case where the client has not sent further data following the '.'
1510 (unfortunately, this is allowed). However, in many cases there won't be
1511 any further data because there won't be any more messages to send. A call
1512 to select() can be used: if it shows that the input is "ready", there is
1513 either input waiting, or the socket has been closed. An attempt to read
1514 the next input character can distinguish the two cases. Previously, Exim
1515 would have sent an OK response which the client would never have see.
1516 This could lead to message repetition. This fix should cure that, at
1517 least in a lot of common cases.
1519 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1520 advertised in response to EHLO.
1526 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1527 fixed by 4.65/MH/01 (is this a record?) are fixed:
1529 (i) An empty string was always treated as zero by the numeric comparison
1530 operators. This behaviour has been restored.
1532 (ii) It is documented that the numeric comparison operators always treat
1533 their arguments as decimal numbers. This was broken in that numbers
1534 starting with 0 were being interpreted as octal.
1536 While fixing these problems I realized that there was another issue that
1537 hadn't been noticed. Values of message_size_limit (both the global option
1538 and the transport option) were treated as octal if they started with 0.
1539 The documentation was vague. These values are now always treated as
1540 decimal, and I will make that clear in the documentation.
1546 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1547 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1550 MH/01 Don't check that the operands of numeric comparison operators are
1551 integers when their expansion is in "skipping" mode (fixes bug
1552 introduced by 4.64-PH/07).
1554 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1555 child addresses, Exim now panics and dies. Previously, because the count
1556 is held in a short int, deliveries were likely to be lost. As such a
1557 large number of recipients for a single message is ridiculous
1558 (performance will be very, very poor), I have chosen to impose a limit
1559 rather than extend the field.
1565 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1566 leftover -K file (the existence of which was triggered by #402).
1567 While we were at it, introduced process PID as part of the -K
1568 filename. This should rule out race conditions when creating
1571 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1572 processing considerably. Previous code took too long for large mails,
1573 triggering a timeout which in turn triggers #401.
1575 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1576 in the DK code in transports.c. sendfile() is not really portable,
1577 hence the _LINUX specificness.
1579 TF/01 In the add_headers option to the mail command in an Exim filter,
1580 there was a bug that Exim would claim a syntax error in any
1581 header after the first one which had an odd number of characters
1584 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1585 callout verification, Exim cached a "reject" for the entire domain. This
1586 is correct for most verifications, but it is not correct for a recipient
1587 verification with use_sender or use_postmaster set, because in that case
1588 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1589 case of MAIL FROM:<> rejection from other early rejections (e.g.
1590 rejection of HELO). When verifying a recipient using a non-null MAIL
1591 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1592 Whatever the result of the callout, the value of the domain cache is
1593 left unchanged (for any other kind of callout, getting as far as trying
1594 RCPT means that the domain itself is ok).
1596 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1599 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1600 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1601 session, the daemon ignores ECONNECT errors and logs others; it now
1602 ignores EPIPE as well.
1604 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1605 (quoted-printable decoding).
1607 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1608 later the small subsequent patch to fix an introduced bug.
1610 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1612 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1614 PH/08 An error is now given if message_size_limit is specified negative.
1616 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1617 to be given (somewhat) arbitrary names.
1619 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1622 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1623 miscellaneous code fixes
1625 PH/10 Added the log_reject_target ACL modifier to specify where to log
1628 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1629 hostname. This is wrong, because it relates to the incoming message (and
1630 probably the interface on which it is arriving) and not to the outgoing
1631 callout (which could be using a different interface). This has been
1632 changed to use the value of the helo_data option from the smtp transport
1633 instead - this is what is used when a message is actually being sent. If
1634 there is no remote transport (possible with a router that sets up host
1635 addresses), $smtp_active_hostname is used.
1637 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1638 tweaks were necessary in order to get it to work (see also 21 below):
1639 (a) The code assumed that strncpy() returns a negative number on buffer
1640 overflow, which isn't the case. Replaced with Exim's string_format()
1642 (b) There were several signed/unsigned issues. I just did the minimum
1643 hacking in of casts. There is scope for a larger refactoring.
1644 (c) The code used strcasecmp() which is not a standard C function.
1645 Replaced with Exim's strcmpic() function.
1646 (d) The code set only $1; it now sets $auth1 as well.
1647 (e) A simple test gave the error "authentication client didn't specify
1648 service in request". It would seem that Dovecot has changed its
1649 interface. Fortunately there's a specification; I followed it and
1650 changed what the client sends and it appears to be working now.
1652 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1655 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1656 address A is aliased to B and C, where B exists and C does not. Without
1657 -v the output is "A verified" because verification stops after a
1658 successful redirection if more than one address is generated. However,
1659 with -v the child addresses are also verified. Exim was outputting "A
1660 failed to verify" and then showing the successful verification for C,
1661 with its parentage. It now outputs "B failed to verify", showing B's
1662 parentage before showing the successful verification of C.
1664 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1665 look up a TXT record in a specific list after matching in a combined
1668 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1669 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1670 they consult the DNS. I had assumed they would set it the way they
1671 wanted; and indeed my experiments on Linux seem to show that in some
1672 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1673 To be on the safe side, however, I have now made the interface to
1674 host_find_byname() similar to host_find_bydns(), with an argument
1675 containing the DNS resolver options. The host_find_byname() function now
1676 sets these options at its start, just as host_find_bydns() does. The smtp
1677 transport options dns_qualify_single and dns_search_parents are passed to
1678 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1679 of host_find_byname() use the default settings of RES_DEFNAMES
1680 (qualify_single) but not RES_DNSRCH (search_parents).
1682 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1683 spool_read_header() do less string testing, by means of a preliminary
1684 switch on the second character of optional "-foo" lines. (This is
1685 overdue, caused by the large number of possibilities that now exist.
1686 Originally there were few.) While I was there, I also converted the
1687 str(n)cmp tests so they don't re-test the leading "-" and the first
1688 character, in the hope this might squeeze out yet more improvement.
1690 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1691 flag allowing group syntax was set by the header_syntax check but not
1692 turned off, possible causing trouble later; (2) The flag was not being
1693 set at all for the header_verify test, causing "group"-style headers to
1694 be rejected. I have now set it in this case, and also caused header_
1695 verify to ignore an empty address taken from a group. While doing this, I
1696 came across some other cases where the code for allowing group syntax
1697 while scanning a header line wasn't quite right (mostly, not resetting
1698 the flag correctly in the right place). These bugs could have caused
1699 trouble for malformed header lines. I hope it is now all correct.
1701 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1702 with the "reply" argument non-NULL. The code, however (which originally
1703 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1704 but it didn't always do it. This confused somebody who was copying the
1705 code for some other use. I have removed all the tests.
1707 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1708 feature that was used to support insecure browsers during the U.S. crypto
1709 embargo. It requires special client support, and Exim is probably the
1710 only MTA that supported it -- and would never use it because real RSA is
1711 always available. This code has been removed, because it had the bad
1712 effect of slowing Exim down by computing (never used) parameters for the
1713 RSA_EXPORT functionality.
1715 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1716 authenticator to fail if there's a tab character in the incoming data
1717 (there should never be unless someone is messing about, as it's supposed
1718 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1719 if the connection is using TLS or if the remote IP is the same as the
1720 local IP, and the "valid-client-cert option" if a client certificate has
1723 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1724 authenticators. This can be used for authorization after authentication
1725 succeeds. (In the case of plaintext, it servers for both authentication
1728 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1729 if any retry times were supplied.
1731 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1732 connection, where there is no client IP address to check. In this
1733 situation, the verify now always succeeds.
1735 PH/25 Applied John Jetmore's -Mset patch.
1737 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1739 PH/27 In a string expansion for a processed (not raw) header when multiple
1740 headers of the same name were present, leading whitespace was being
1741 removed from all of them, but trailing whitespace was being removed only
1742 from the last one. Now trailing whitespace is removed from each header
1743 before concatenation. Completely empty headers in a concatenation (as
1744 before) are ignored.
1746 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1747 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1749 PH/29 [Removed. This was a change that I later backed out, and forgot to
1750 correct the ChangeLog entry (that I had efficiently created) before
1751 committing the later change.]
1753 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1754 address errors (4xx response to RCPT) over the same connection as other
1755 messages routed to the same hosts. Such deliveries are always "forced",
1756 so retry times are not inspected. This resulted in far too many retries
1757 for the affected addresses. The effect occurred only when there were more
1758 hosts than the hosts_max_try setting in the smtp transport when it had
1759 the 4xx errors. Those hosts that it had tried were not added to the list
1760 of hosts for which the message was waiting, so if all were tried, there
1761 was no problem. Two fixes have been applied:
1763 (i) If there are any address or message errors in an SMTP delivery, none
1764 of the hosts (tried or untried) are now added to the list of hosts
1765 for which the message is waiting, so the message should not be a
1766 candidate for sending over the same connection that was used for a
1767 successful delivery of some other message. This seems entirely
1768 reasonable: after all the message is NOT "waiting for some host".
1769 This is so "obvious" that I'm not sure why it wasn't done
1770 previously. Hope I haven't missed anything, but it can't do any
1771 harm, as the worst effect is to miss an optimization.
1773 (ii) If, despite (i), such a delivery is accidentally attempted, the
1774 routing retry time is respected, so at least it doesn't keep
1775 hammering the server.
1777 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1778 in ${readsocket because some servers need this prod.
1780 PH/32 Added some extra debug output when updating a wait-xxx database.
1782 PH/33 The hint "could be header name not terminated by colon", which has been
1783 given for certain expansion errors for a long time, was not being given
1784 for the ${if def:h_colon_omitted{... case.
1786 PH/34 The spec says: "With one important exception, whenever a domain list is
1787 being scanned, $domain contains the subject domain." There was at least
1788 one case where this was not true.
1790 PH/35 The error "getsockname() failed: connection reset by peer" was being
1791 written to the panic log as well as the main log, but it isn't really
1792 panic-worthy as it just means the connection died rather early on. I have
1793 removed the panic log writing for the ECONNRESET error when getsockname()
1796 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1797 runs only) independently of the message's sender address. This meant
1798 that, if the 4xx error was in fact related to the sender, a different
1799 message to the same recipient with a different sender could confuse
1800 things. In particualar, this can happen when sending to a greylisting
1801 server, but other circumstances could also provoke similar problems.
1802 I have changed the default so that the retry time for these errors is now
1803 based a combination of the sender and recipient addresses. This change
1804 can be overridden by setting address_retry_include_sender=false in the
1807 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1808 remote server are returned as part of bounce messages. This was not
1809 happening for LMTP over a pipe (the lmtp transport), but now it is the
1810 same for both kinds of LMTP.
1812 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1813 in header lines that were in fact CNAMEs. This is no longer the case.
1815 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1816 and queue runs started by the daemon processed all messages. This has
1817 been fixed so that -R and -S can now usefully be given with -q<time>.
1819 PH/40 Import PCRE release 6.7 (fixes some bugs).
1821 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1823 PH/42 Give an error if -q is specified more than once.
1825 PH/43 Renamed the variables $interface_address and $interface_port as
1826 $received_ip_address and $received_port, to make it clear that these
1827 values apply to message reception, and not to the outgoing interface when
1828 a message is delivered. (The old names remain recognized, of course.)
1830 PH/44 There was no timeout on the connect() call when using a Unix domain
1831 socket in the ${readsocket expansion. There now is.
1833 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1834 be meaningful with "accept".
1836 SC/01 Eximstats V1.43
1837 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1839 SC/02 Eximstats V1.44
1840 Use a glob alias rather than an array ref in the generated
1841 parser. This improves both readability and performance.
1843 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1844 Collect SpamAssassin and rejection statistics.
1845 Don't display local sender or destination tables unless
1846 there is data to show.
1847 Added average volumes into the top table text output.
1849 SC/04 Eximstats V1.46
1850 Collect data on the number of addresses (recipients)
1851 as well as the number of messages.
1853 SC/05 Eximstats V1.47
1854 Added 'Message too big' to the list of mail rejection
1855 reasons (thanks to Marco Gaiarin).
1857 SC/06 Eximstats V1.48
1858 Mainlog lines which have GMT offsets and are too short to
1859 have a flag are now skipped.
1861 SC/07 Eximstats V1.49 (Alain Williams)
1862 Added the -emptyok flag.
1864 SC/08 Eximstats V1.50
1865 Fixes for obtaining the IP address from reject messages.
1867 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1868 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1869 whitesspace changes from 4.64-PH/27
1871 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1874 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1875 are found, allow negative numbers in numeric criteria)
1877 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1879 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1880 to match changes made in 4.64-PH/43
1882 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1884 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1885 "domain" is actually an IP address, and force a failure. This locks out
1886 those revolvers/nameservers that support "A-for-A" lookups, in
1887 contravention of the specifications.
1889 PH/48 When a host name was looked up from an IP address, and the subsequent
1890 forward lookup of the name timed out, the host name was left in
1891 $sender_host_name, contrary to the specification.
1893 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1894 restricted to single-key lookups, Exim was not diagnosing an error if
1895 * or *@ was used with a query-style lookup.
1897 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1899 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1900 long ago, but noone interested enough thought of it. Let's just say that
1901 the "1.1" means that there are some new functions that weren't there at
1902 some point in the past.
1904 PH/51 Error processing for expansion failure of helo_data from an smtp
1905 transport during callout processing was broken.
1907 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1908 tested/used via the -bh/-bhc/-bs options.
1910 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1911 bug, fixed in subsequent PCRE releases).
1913 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1914 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1916 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1922 SC/01 Use a glob alias rather than an array ref in eximstats generated
1923 parser. This improves both readability and performance.
1925 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1926 Don't display local sender or destination tables in eximstats unless
1927 there is data to show.
1928 Added average volumes into the eximstats top table text output.
1930 SC/03 Collect data on the number of addresses (recipients) as well
1931 as the number of messages in eximstats.
1933 TF/01 Correct an error in the documentation for the redirect router. Exim
1934 does (usually) call initgroups() when daemonizing.
1936 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1937 with consistent privilege compared to when running as a daemon.
1939 TF/03 Note in the spec that $authenticated_id is not set for local
1940 submissions from trusted users.
1942 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1943 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1945 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1946 by adding some example configuration directives to the default
1947 configuration file. A little bit of work is required to uncomment the
1948 directives and define how usernames and passwords are checked, but
1949 there is now a framework to start from.
1951 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1952 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1953 without this. I don't know how relevant this is to other LDAP libraries.
1955 PH/02 Add the verb name to the "unknown ACL verb" error.
1957 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1959 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1961 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1962 directory not expanded when it should be if an expanded home directory
1963 was set for the address (which is overridden by the transport).
1965 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1968 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1969 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1970 because it is too late at that time, and has no effect.
1972 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1973 security issue with \' (bugzilla #107). I could not use the
1974 PQescapeStringConn() function, because it needs a PGconn value as one of
1977 PH/08 When testing addresses using -bt, indicate those final addresses that
1978 are duplicates that would not cause an additional delivery. At least one
1979 person was confused, thinking that -bt output corresponded to deliveries.
1980 (Suppressing duplicates isn't a good idea as you lose the information
1981 about possibly different redirections that led to the duplicates.)
1983 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1984 systems where poll() doesn't work, in particular OS X.
1986 PH/10 Added more information to debugging output for retry time not reached.
1988 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1989 operations in malware.c.
1991 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1994 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1995 syslog_duplication set false, log lines that would normally be written
1996 both the the main log and to the reject log were not written to syslog at
1999 PH/14 In the default configuration, change the use of "message" in ACL warn
2000 statements to "add_header".
2002 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2003 not followed by a command (e.g. "seen endif").
2005 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2006 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2009 PH/17 Added extra conditions to the default value of delay_warning_condition
2013 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2014 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2015 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2018 The Auto-Submitted: and various List- headers are standardised, whereas I
2019 don't think Precedence: ever was.
2021 PH/18 Refactored debugging code in route_finduser() to show more information,
2022 in particular, the error code if getpwnam() issues one.
2024 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2025 This is apparently needed in addition to the PH/07 change above to avoid
2026 any possible encoding problems.
2028 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2029 but not after initializing Perl.
2031 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2032 output them only if debugging. By default they are written stderr,
2033 apparently, which is not desirable.
2035 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2038 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2041 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2043 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2044 authenticated or an ident call has been made. Suppress the default
2045 values for $authenticated_id and $authenticated_sender (but permit -oMai
2046 and -oMas) when testing with -bh.
2048 PH/24 Re-jigged the order of the tests in the default configuration so that the
2049 tests for valid domains and recipients precede the DNS black list and CSA
2050 tests, on the grounds that those ones are more expensive.
2052 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2053 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2054 This bug exists in every version of Exim that I still have, right back to
2057 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2058 However, an attempt to turn on case-sensitivity in a regex key by
2059 including (?-i) didn't work because the subject string was already
2060 lowercased, and the effects were non-intuitive. It turns out that a
2061 one-line patch can be used to allow (?-i) to work as expected.
2067 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2068 other effects) broke the use of negated acl sub-conditions.
2070 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2073 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2074 "Deny" causes Exim to reject the incoming connection with a 554 error.
2075 Unfortunately, if there is a major crisis, such as a disk failure,
2076 tcp-wrappers gives "deny", whereas what one would like would be some
2077 kind of temporary error. A kludge has been added to help with this.
2078 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2079 554 error is used if errno is still zero or contains ENOENT (which occurs
2080 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2083 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2085 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2086 errors. Otherwise a message that provokes a temporary error (when other
2087 messages do not) can cause a whole host to time out.
2089 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2090 addresses were routed directly to files or pipes from a redirect router.
2091 File deliveries just didn't batch; pipe deliveries might have suffered
2094 PH/06 A failure to get a lock for a hints database would erroneously always say
2095 "Failed to get write lock", even when it was really a read lock.
2097 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2098 of 0600. This has been changed to use the value of the lockfile_mode
2099 option (which defaults to 0600).
2101 PH/08 Applied small patch from the Sieve maintainer.
2103 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2104 folder from quota calculations, a direct delivery into this folder messed
2105 up the contents of the maildirsize file. This was because the regex was
2106 used only to exclude .Trash (or whatever) when the size of the mailbox
2107 was calculated. There was no check that a delivery was happening into an
2108 excluded directory. This bug has been fixed by ignoring all quota
2109 processing for deliveries into excluded directories.
2111 PH/10 Added the maildirfolder_create_regex option to appendfile.
2117 PH/01 The code for finding all the local interface addresses on a FreeBSD
2118 system running IPv6 was broken. This may well have applied to all BSD
2119 systems, as well as to others that have similar system calls. The broken
2120 code found IPv4 interfaces correctly, but gave incorrect values for the
2121 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2122 that it would not match correctly against @[] and not recognize the IPv6
2125 PH/02 The ipliteral router was not recognizing addresses of the form user@
2126 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2128 PH/03 Added disable_ipv6.
2130 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2131 decoded form, because it is most often used to construct To: headers
2132 lines in autoreplies, and the decoded form may well be syntactically
2133 invalid. However, $reply_address has leading white space removed, and all
2134 newlines turned into spaces so that the autoreply transport does not
2137 PH/05 If group was specified without a user on a router, and no group or user
2138 was specified on a transport, the group from the router was ignored.
2140 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2141 for visible compile-time settings that can be used to change these
2142 numbers, for those that want even more. Backwards compatibility with old
2143 spool files has been maintained. However, going back to a previous Exim
2144 release will lost any variables that are in spool files.
2146 PH/07 Two small changes when running in the test harness: increase delay when
2147 passing a TCP/IP connection to a new process, in case the original
2148 process has to generate a bounce, and remove special handling of
2149 127.0.0.2 (sic), which is no longer necessary.
2151 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2152 be the same on different OS.
2154 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2157 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2158 whether --show-vars was specified or not
2160 JJ/02 exipick: Added support for new ACL variable spool format introduced
2163 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2164 syntactically invalid From: or Reply-to: line, and a filter used this to
2165 generate an autoreply, and therefore failed to obtain an address for the
2166 autoreply, Exim could try to deliver to a non-existent relative file
2167 name, causing unrelated and misleading errors. What now happens is that
2168 it logs this as a hard delivery error, but does not attempt to create a
2171 PH/11 The exinext utility has a -C option for testing purposes, but although
2172 the given file was scanned by exinext itself; it wasn't being passed on
2173 when Exim was called.
2175 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2176 an end-of-file indication when reading a command response.
2178 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2179 compiled. In many other places in Exim, IPv6 addresses are always
2180 recognized, so I have changed this. It also means that IPv4 domain
2181 literals of the form [IPV4:n.n.n.n] are now always recognized.
2183 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2184 used if the router is not running as root, for example, when verifying at
2185 ACL time, or when using -bh. The debugging output from this situation was
2186 non-existent - all you got was a failure to exec. I have made two
2189 (a) Failures to set uid/gid, the current directory, or a process leader
2190 in a subprocess such as that created by queryprogram now generate
2191 suitable debugging ouput when -d is set.
2193 (b) The queryprogram router detects when it is not running as root,
2194 outputs suitable debugging information if -d is set, and then runs
2195 the subprocess without attempting to change uid/gid.
2197 PH/15 Minor change to Makefile for building test_host (undocumented testing
2200 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2201 additional section of a DNS packet that returns MX or SRV records.
2202 Instead, it always explicitly searches for A/AAAA records. This avoids
2203 major problems that occur when a DNS server includes only records of one
2204 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2205 fixed another bug: if SRV records were looked up and the corresponding
2206 address records were *not* found in the additional section, the port
2207 values from the SRV records were lost.
2209 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2210 using the correct key (the original address) when searching the retry
2211 rules in order to find which one to use for generating the retry hint.
2213 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2214 adding the default one. Similarly, if it contains a Reply-To: header, the
2215 errors_reply_to option, if set, is not used.
2217 PH/19 When calculating a retry time, Exim used to measure the "time since
2218 failure" by looking at the "first failed" field in the retry record. Now
2219 it does not use this if it is later than than the arrival time of the
2220 message. Instead it uses the arrival time. This makes for better
2221 behaviour in cases where some deliveries succeed, thus re-setting the
2222 "first failed" field. An example is a quota failure for a huge message
2223 when small messages continue to be delivered. Without this change, the
2224 "time since failure" will always be short, possible causing more frequent
2225 delivery attempts for the huge message than are intended.
2226 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2228 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2229 $1, $2, $3) because the numerical variables can be reset during some
2230 expansion items (e.g. "match"), thereby losing the authentication data.
2232 PH/21 Make -bV show the size of off_t variables so that the test suite can
2233 decide whether to run tests for quotas > 2G.
2235 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2236 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2237 filecount value is greater than 2G or if a quota value is greater than 2G
2238 on a system where the size of off_t is not greater than 4, a panic error
2241 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2242 never match. The debug and -bh output now contains an explicit error
2243 message indicating a malformed IPv4 address or mask.
2245 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2246 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2247 PH/23 above applies.
2249 PH/25 Do not write to syslog when running in the test harness. The only
2250 occasion when this arises is a failure to open the main or panic logs
2251 (for which there is an explicit test).
2253 PH/26 Added the /no_tell option to "control=freeze".
2255 PH/27 If a host name lookup failed very early in a connection, for example, if
2256 the IP address matched host_lookup and the reverse lookup yielded a name
2257 that did not have a forward lookup, an error message of the form "no IP
2258 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2259 could be logged. Now it outputs the IP address instead of "NULL".
2261 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2262 allows the sender and the authenticated sender to be set when
2263 submitting a message from within Exim. Since child_open_exim() is
2264 documented for local_scan(), the new function should be too.
2266 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2267 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2268 results in an empty string is now treated as unset.
2270 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2272 PH/31 Added sender_verify_fail logging option.
2274 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2275 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2276 tidied the source and removed it altogether.
2278 PH/33 When a queue run was abandoned because the load average was too high, a
2279 log line was always written; now it is written only if the queue_run log
2280 selector is set. In addition, the log line for abandonment now contains
2281 information about the queue run such as the pid. This is always present
2282 in "start" and "stop" lines but was omitted from the "abandon" line.
2284 PH/34 Omit spaces between a header name and the colon in the error message that
2285 is given when verify = headers_syntax fails (if there are lots of them,
2286 the message gets confusing).
2288 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2289 names, as there are now some PTR records that contain slashes. This check
2290 is only to protect against broken name servers that fall over on strange
2291 characters, so the fact that it applies to all lookups doesn't matter.
2293 PH/36 Now that the new test suite is complete, we can remove some of the
2294 special code in Exim that was needed for the old test suite. For example,
2295 sorting DNS records because real resolvers return them in an arbitrary
2296 order. The new test suite's fake resolver always returns records in the
2299 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2300 bounces) except when queue_only is set, to avoid logging races between
2301 the different processes.
2303 PH/38 Panic-die if .include specifies a non-absolute path.
2305 PH/39 A tweak to the "H" retry rule from its user.
2307 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2308 a label. They prevented compilation on older perls.
2310 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2311 a warning to be raised on newish perls.
2313 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2314 on queue. Changes to match documented behaviour of showing count of
2315 messages matching specified criteria.
2317 PH/40 Changed the default ident timeout from 30s to 5s.
2319 PH/41 Added support for the use of login_cap features, on those BSD systems
2320 that have them, for controlling the resources used by pipe deliveries.
2322 PH/42 The content-scanning code uses fopen() to create files in which to put
2323 message data. Previously it was not paying any attention to the mode of
2324 the files. Exim runs with umask(0) because the rest of the code creates
2325 files with open(), and sets the required mode explicitly. Thus, these
2326 files were ending up world-writeable. This was not a big issue, because,
2327 being within the spool directory, they were not world-accessible. I have
2328 created a function called modefopen, which takes an additional mode
2329 argument. It sets umask(777), creates the file, chmods it to the required
2330 mode, then resets the umask. All the relevant calls to fopen() in the
2331 content scanning code have been changed to use this function.
2333 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2334 to 24 hours. This avoids potential overflow problems when processing G
2335 and H retry rules. I suspect nobody ever tinkers with this value.
2337 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2339 PH/45 When the plaintext authenticator is running as a client, the server's
2340 challenges are checked to ensure they are valid base64 strings. By
2341 default, the authentication attempt is cancelled if an invalid string is
2342 received. Setting client_ignore_invalid_base64 true ignores these errors.
2343 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2344 they are received. Thus, the responses can be made to depend on the
2345 challenges. If an invalid string is ignored, an empty string is placed in
2348 PH/46 Messages that are created by the autoreply transport now contains a
2349 References: header, in accordance with RFCs 2822 and 3834.
2351 PH/47 Added authenticated_sender_force to the smtp transport.
2353 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2355 PH/49 Installed latest patch from the Sieve maintainer.
2357 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2358 was also set, the appendfile transport was unnecessarily scanning a
2359 directory of message files (e.g. for maildir delivery) to find the count
2360 of files (along with the size), even though it did not need this
2361 information. It now does the scan only if it needs to find either the
2362 size of the count of files.
2364 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2366 PH/52 Two bugs concerned with error handling when the smtp transport is
2369 (i) Exim was not creating retry information for temporary errors given
2370 for individual recipients after the DATA command when the smtp transport
2371 was used in LMTP mode. This meant that they could be retried too
2372 frequently, and not timed out correctly.
2374 (ii) Exim was setting the flag that allows error details to be returned
2375 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2376 recipients that were returned after the DATA command.
2378 PH/53 This is related to PH/52, but is more general: for any failing address,
2379 when detailed error information was permitted to be returned to the
2380 sender, but the error was temporary, then after the final timeout, only
2381 "retry timeout exceeded" was returned. Now it returns the full error as
2382 well as "retry timeout exceeded".
2384 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2385 do this, and (what is worse) MTAs that accept it.
2387 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2388 will now be deprecated.
2390 PH/56 New os.c-cygwin from the Cygwin maintainer.
2392 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2393 formats (previously only available in exim formats via -bpr, -bpru,
2394 and -bpra. Now also available in native and exiqgrep formats)
2396 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2397 with very large, slow to parse queues
2399 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2401 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2403 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2404 responses to authentication challenges, though it was showing the
2405 challenges; (ii) I've removed the CR characters from the debug output for
2408 PH/58 Allow for the insertion of a newline as well as a space when a string
2409 is turned into more than one encoded-word during RFC 2047 encoding. The
2410 Sieve code now uses this.
2412 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2413 data_4xx, lost_connection, tls_required.
2415 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2416 message was being sent as an SMTP response.
2418 PH/61 Add -l and -k options to exicyclog.
2420 PH/62 When verifying, if an address was redirected to one new address, so that
2421 verification continued, and the new address failed or deferred after
2422 having set something in $address_data, the value of $address_data was not
2423 passed back to the ACL. This was different to the case when no
2424 redirection occurred. The value is now passed back in both cases.
2426 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2427 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2428 use login_cap.h, so on its own it isn't the distinguishing feature. The
2429 new name refers directly to the setclassresources() function.
2431 PH/65 Added configuration files for NetBSD3.
2433 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2435 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2436 is preferred over IPv4.
2438 PH/68 The bounce_return_message and bounce_return_body options were not being
2439 honoured for bounces generated during the reception of non-SMTP messages.
2440 In particular, this applied to messages rejected by the ACL. This bug has
2441 been fixed. However, if bounce_return_message is true and bounce_return_
2442 body is false, the headers that are returned for a non-SMTP message
2443 include only those that have been read before the error was detected.
2444 (In the case of an ACL rejection, they have all been read.)
2446 PH/69 The HTML version of the specification is now built in a directory called
2447 spec_html instead of spec.html, because the latter looks like a path with
2448 a MIME-type, and this confuses some software.
2450 PH/70 Catch two compiler warnings in sieve.c.
2452 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2453 function verify_get_ident() calls ip_connect() to connect a socket, but
2454 if the "connect()" function timed out, ip_connect() used to close the
2455 socket. However, verify_get_ident() also closes the socket later, and in
2456 between Exim writes to the log, which may get opened at this point. When
2457 the socket was closed in ip_connect(), the log could get the same file
2458 descriptor number as the socket. This naturally causes chaos. The fix is
2459 not to close the socket in ip_connect(); the socket should be closed by
2460 the function that creates it. There was only one place in the code where
2461 this was missing, in the iplookup router, which I don't think anybody now
2462 uses, but I've fixed it anyway.
2464 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2465 well as to direct DNS lookups. Otherwise the handling of names in host
2466 lists is inconsistent and therefore confusing.
2472 PH/01 Two changes to the default runtime configuration:
2474 (1) Move the checks for relay_from_hosts and authenticated clients from
2475 after to before the (commented out) DNS black list checks.
2477 (2) Add control=submission to the relay_from_hosts and authenticated
2478 clients checks, on the grounds that messages accepted by these
2479 statements are most likely to be submissions.
2481 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2483 (1) Generate an error if the third argument for the ${prvs expansion is
2486 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2489 (3) Reset the variables that are obtained from the first argument of
2490 ${prvscheck and used in the second argument before leaving the code,
2491 because their memory is reclaimed, so using them afterwards may do
2494 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2495 one (it's much easier than Tom thought :-).
2497 (5) Because of (4), we can now allow for the use of $prvscheck_result
2498 inside the third argument.
2500 PH/03 For some reason, the default setting of PATH when running a command from
2501 a pipe transport was just "/usr/bin". I have changed it to
2504 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2505 anything to be listed in the output from -bV.
2507 PH/05 When a filter generated an autoreply, the entire To: header line was
2508 quoted in the delivery log line, like this:
2510 => >A.N.Other <ano@some.domain> <original@ddress> ...
2512 This has been changed so that it extracts the operative address. There
2513 may be more than one such address. If so, they are comma-separated, like
2516 => >ano@some.domain,ona@other.domain <original@ddress> ...
2518 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2519 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2520 not being looked up in the rDNS to get a host name, Exim was showing the
2521 IP address twice in Received: lines, even though the IP addresses were
2522 identical. For example:
2524 Received: from [1.2.3.4] (helo=[1.2.3.4])
2526 However, if the real host name was known, it was omitting the HELO data
2527 if it matched the actual IP address. This has been tidied up so that it
2528 doesn't show the same IP address twice.
2530 PH/07 When both +timestamp and +memory debugging was on, the value given by
2531 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2532 called by the debug printing, thereby overwriting the timestamp buffer.
2533 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2535 PH/08 When the original message was included in an autoreply transport, it
2536 always said "this is a copy of the message, including all the headers",
2537 even if body_only or headers_only was set. It now gives an appropriate
2540 PH/09 Applied a patch from the Sieve maintainer which:
2542 o fixes some comments
2543 o adds the (disabled) notify extension core
2544 o adds some debug output for the result of if/elsif tests
2545 o points to the current vacation draft in the documentation
2546 and documents the missing references header update
2550 o fixes a bug in processing the envelope test (when testing
2551 multiple envelope elements, the last element determinted the
2554 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2555 Electronic Mail") by including:
2557 Auto-submitted: auto-generated
2559 in the messages that it generates (bounce messages and others, such as
2560 warnings). In the case of bounce messages for non-SMTP mesages, there was
2561 also a typo: it was using "Auto_submitted" (underscore instead of
2562 hyphen). Since every message generated by Exim is necessarily in response
2563 to another message, thes have all been changed to:
2565 Auto-Submitted: auto-replied
2567 in accordance with these statements in the RFC:
2569 The auto-replied keyword:
2571 - SHOULD be used on messages sent in direct response to another
2572 message by an automatic process,
2574 - MUST NOT be used on manually-generated messages,
2576 - MAY be used on Delivery Status Notifications (DSNs) and Message
2577 Disposition Notifications (MDNs),
2579 - MUST NOT be used on messages generated by automatic or periodic
2580 processes, except for messages which are automatic responses to
2583 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2584 to the default Received: header definition.
2586 PH/12 Added log selector acl_warn_skipped (default on).
2588 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2589 variables because (a) they are in the wrong storage pool and (b) even if
2590 they were copied, it wouldn't work properly because of the caching.
2592 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2593 checking when decoding. Apparently there are clients that generate
2594 overlong encoded strings. Why am I not surprised?
2596 PH/15 If the first argument of "${if match_address" was not empty, but did not
2597 contain an "@" character, Exim crashed. Now it writes a panic log message
2598 and treats the condition as false.
2600 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2602 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2603 extension "envelope-auth". The code is finished and in agreement with
2604 other implementations, but there is no documentation so far and in fact,
2605 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2606 not changing the active code.
2608 Print executed "if" and "elsif" statements when debugging is used. This
2609 helps a great deal to understand what a filter does.
2611 Document more things not specified clearly in RFC3028. I had all this
2612 sorted out, when out of a sudden new issues came to my mind. Oops."
2614 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2617 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2618 canonical form (as documented). However, after a host name lookup from
2619 the IP address, check_host() was doing a simple string comparison with
2620 addresses acquired from the DNS when checking that the found name did
2621 have the original IP as one of its addresses. Since any found IPv6
2622 addresses are likely to be in abbreviated form, the comparison could
2623 fail. Luckily, there already exists a function for doing the comparison
2624 by converting both addresses to binary, so now that is used instead of
2625 the text comparison.
2627 PH/20 There was another similar case to PH/19, when a complete host name was
2628 given in a host list; looking up its IP address could give an abbreviated
2629 form, whereas the current host's name might or might not be abbreviated.
2630 The same fix has been applied.
2636 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2637 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2640 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2641 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2643 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2645 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2646 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2647 Also, for "Auto-Submitted", treat anything other than "no" as
2648 non-personal, in accordance with RFC 3834. (Previously it treated
2649 anything starting "auto-" as non-personal.)
2651 TF/01 The control=submission/name=... option had a problem with syntax
2652 errors if the name included a slash character. The /name= option
2653 now slurps the rest of the string, so it can include any characters
2654 but it must come last in the list of options (after /sender_retain
2657 PH/05 Some modifications to the interface to the fake nameserver for the new
2665 TK/01 Added the "success_on_redirect" address verification option. See
2666 NewStuff for rationale and an example.
2668 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2670 PH/02 Patch to exigrep to allow it to work on syslog lines.
2672 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2673 fread() to skip over the body file's header line, because in Cygwin the
2674 header line is locked and is inaccessible.
2676 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2677 co-exist for some time) to make it clear that it is the Exim ID that is
2678 referenced, not the Message-ID: header line.
2680 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2681 string_format() function, because snprintf() does not exist on all
2684 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2685 ${stat: expansion item.
2687 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2688 protocol synchronization error", to keep the pedants happy.
2690 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2691 well as for IRIX systems, when gcc is being used. See the host.c source
2694 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2696 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2699 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2700 transport and to the smtp transport in LMTP mode.
2702 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2704 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2705 run a filter in a subprocess. This could lead to confusion in subsequent
2706 lookups in the parent process. There should also be a search_tidyup() at
2707 the end of the subprocess.
2709 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2710 only if the host matched helo_try_verify_hosts, which caused the
2711 verification to occur when the EHLO/HELO command was issued. The ACL just
2712 tested the remembered result. Now, if a previous verification attempt has
2713 not happened, "verify = helo" does it there and then.
2715 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2717 TK/03 Fix log output including CR from clamd.
2719 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2720 exist provoked a memory error which could cause a segfault.
2722 PH/15 Installed PCRE 6.2
2724 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2726 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2727 of the problem. Specifically, suggested +O2 rather than +O1 for the
2730 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2732 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2733 record, the port was not correctly logged when the outgoing_port log
2734 selector was set (it logged the transort's default port).
2736 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2737 fallback_hosts, and "hosts" in the smtp transport.
2739 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2740 host errors such as "Connection refused".
2742 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2743 authentication with radiusclient 0.4.9:
2745 - Error returned from rc_read_config was caught wrongly
2746 - Username/password not passed on to radius server due to wrong length.
2748 The presumption is that some radiusclient API changes for 4.51/PH/17
2749 were not taken care of correctly. The code is still untested by me (my
2750 Linux distribution still has 0.3.2 of radiusclient), but it was
2751 contributed by a Radius user.
2753 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2754 expanding the "port" option of the smtp transport.
2756 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2757 while reading a MIME header. Thanks to Tom Hughes for a patch.
2759 PH/24 Include config.h inside local_scan.h so that configuration settings are
2762 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2763 that in an ACL for RCPT (for example), you can examine exactly what was
2766 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2767 commands, but it was not correctly comparing the address with the actual
2768 client host address. Thus, it would show the EHLO address in Received:
2769 header lines when this was not necessary.
2771 PH/27 Added the % operator to ${eval:}.
2773 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2774 it should be ignoring failures (because with -C, for example, it has lost
2775 privilege). It wasn't ignoring creation failures other than "already
2778 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2781 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2782 because an input line was too long, either on its own, or by virtue of
2783 too many continuations, the temporary file was not being removed, and the
2784 return code was incorrect.
2786 PH/31 Missing "BOOL" in function definition in filtertest.c.
2788 PH/32 Applied Sieve patches from the maintainer.
2790 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2792 PH/33 Added "verify = not_blind".
2794 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2795 Local/Makefile (with some defaults set). These are used in built scripts
2796 such as exicyclog, but they have never been used in the exim_install
2797 script (though there are many overriding facilities there). I have
2798 arranged that the exim_install script now takes note of these two
2801 PH/35 Installed configuration files for Dragonfly.
2803 PH/36 When a locally submitted message by a trusted user did not contain a
2804 From: header, and the sender address was obtained from -f or from an SMTP
2805 MAIL command, and the trusted user did not use -F to supply a sender
2806 name, $originator_name was incorrectly used when constructing a From:
2807 header. Furthermore, $originator_name was used for submission mode
2808 messages from external hosts without From: headers in a similar way,
2809 which is clearly wrong.
2811 PH/37 Added control=suppress_local_fixups.
2813 PH/38 When log_selector = +received_sender was set, and the addition of the
2814 sender made the log line's construction buffer exactly full, or one byte
2815 less than full, an overflow happened when the terminating "\n" was
2818 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2819 when the result of a list match is failure because a DNS lookup failed.
2821 PH/40 RM_COMMAND is now used in the building process.
2823 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2824 the "build-* directories that it finds.
2826 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2827 address in a domain literal was a prefix of an interface address.
2829 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2830 when verifying a sender address, unless rewrite_headers is false.
2832 PH/44 Wrote a long comment about why errors_to addresses are verified as
2833 recipients, not senders.
2835 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2836 the ratelimit ACL was added.
2838 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2840 PH/46 Added extra information about PostgreSQL errors to the error string.
2842 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2843 avoiding the need to install special zones in a real server. This is
2844 backwards compatible; if it can't find the fake resolver, it drops back.
2845 Thus, both old and new test suites can be run.
2847 TF/02 Added util/ratelimit.pl
2849 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2850 clock is set back in time.
2852 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2853 Candler <B.Candler@pobox.com>.
2855 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2856 OK for addresses that are the result of redirections.
2858 PH/48 A number of further additions for the benefit of the new test suite,
2859 including a fake gethostbyname() that interfaces to the fake DNS resolver
2862 TF/06 The fix for widen_domains has also been applied to qualify_single and
2863 search_parents which are the other dnslookup options that can cause
2866 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2869 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2871 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2872 service" if it encountered an MX record with an empty target hostname.
2873 The message is now "an MX or SRV record indicated no SMTP service".
2875 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2876 if the DNS of the sending site is misconfigured. This is quite a
2877 common situation. This change restores the behaviour of treating a
2878 helo verification defer as a failure.
2880 PH/51 If self=fail was set on a router, the bounce message did not include the
2881 actual error message.
2887 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2889 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2890 command itself ended in error, the underlying message about the transport
2891 filter timeout was being overwritten with the pipe command error. Now the
2892 underlying error message should be appended to the second error message.
2894 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2896 PH/02 Reduce the amount of output that "make" produces by default. Full output
2897 can still be requested.
2899 PH/03 The warning log line about a condition test deferring for a "warn" verb
2900 was being output only once per connection, rather than after each
2901 occurrence (because it was using the same function as for successful
2902 "warn" verbs). This seems wrong, so I have changed it.
2904 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2905 it should not have, which might have caused a crash in the right
2906 circumstances, but probably never did.
2908 PH/04 Installed a modified version of Tony Finch's patch to make submission
2909 mode fix the return path as well as the Sender: header line, and to
2910 add a /name= option so that you can make the user's friendly name appear
2913 TF/03 Added the control = fakedefer ACL modifier.
2915 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2916 Mark Lowes for thorough testing.
2918 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2920 TK/03 Merged latest SRS patch from Miles Wilton.
2922 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2923 with the definition in sysexits.h (which is #included earlier).
2924 Fortunately, Exim does not actually use EX_OK. The code used to try to
2925 preserve the sysexits.h value, by assumimg that macro definitions were
2926 scanned for macro replacements. I have been disabused of this notion,
2927 so now the code just undefines EX_OK before #including unistd.h.
2929 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2930 in the smtp transport. When a block could not be written in a single
2931 write() function, the timeout was being re-applied to each part-write.
2932 This seems wrong - if the receiver was accepting one byte at a time it
2933 would take for ever. The timeout is now adjusted when this happens. It
2934 doesn't have to be particularly precise.
2936 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2937 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2939 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2940 without a domain if the check to <postmaster@domain> fails.
2942 SC/01 Eximstats: added -xls and the ability to specify output files
2943 (patch written by Frank Heydlauf).
2945 SC/02 Eximstats: use FileHandles for outputing results.
2947 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2949 SC/04 Eximstats: fixed display of large numbers with -nvr option
2951 SC/05 Eximstats: fixed merging of reports with empty tables.
2953 SC/06 Eximstats: added the -include_original_destination flag
2955 SC/07 Eximstats: removed tabs and trailing whitespace.
2957 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2959 TK/06 MBOX spool code: Add real "From " MBOX separator line
2960 so the .eml file is really in mbox format (even though
2961 most programs do not really care). Patch from Alex Miller.
2963 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2964 The latter is generated from $received_to and is only set if the
2965 message has one envelope recipient. SA can use these headers,
2966 obviously out-of-the-box. Patch from Alex Miller.
2968 PH/08 The ${def test on a variable was returning false if the variable's
2969 value was "0", contrary to what the specification has always said!
2970 The result should be true unless the variable is empty.
2972 PH/09 The syntax error of a character other than { following "${if
2973 def:variable_name" (after optional whitespace) was not being diagnosed.
2974 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2975 accidental colon was present, for example, could give incorrect results.
2977 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2978 result is used (not including appendfile, where other changes are about
2981 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2982 This involved changing a lot of size variables from int to off_t. It
2983 should work with maildirs and everything.
2985 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2986 spamd dying while we are connected to it.
2988 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2991 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2992 function for BDB 4.3.
2994 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2996 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2997 the sources for PCRE have been re-arranged and more files are now
3000 PH/15 The code I had for printing potentially long long variables in PH/11
3001 above was not the best (it lost precision). The length of off_t variables
3002 is now inspected at build time, and an appropriate printing format (%ld
3003 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3004 to be "long long int" or "long int". This is needed for the internal
3005 formatting function string_vformat().
3007 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3008 the configuration file to be ":syslog", then the script "guesses" where
3009 the logs files are, rather than using the compiled in default. In our
3010 case the guess is not the same as the compiled default, so the script
3011 suddenly stopped working when I started to use syslog. The patch checks
3012 to see if log_file_path is "". If so, it attempts to read it from exim
3013 with no configuration file to get the compiled in version, before it
3014 falls back to the previous guessing code."
3016 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3017 implementing BATV in an Exim configuration. See NewStuff for the gory
3020 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3021 Makefile that are specific to HP-UX.
3023 PH/18 If the "use_postmaster" option was set for a recipient callout together
3024 with the "random" option, the postmaster address was used as the MAIL
3025 FROM address for the random test, but not for the subsequent recipient
3026 test. It is now used for both.
3028 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3029 patch removes a few documentation additions to RFC 3028, because the
3030 latest draft now contains them. It adds the new en;ascii-case comparator
3031 and a new error check for 8bit text in MIME parts. Comparator and
3032 require names are now matched exactly. I enabled the subaddress
3033 extension, but it is not well tested yet (read: it works for me)."
3035 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3036 rework some of the code of TK/09 above to avoid the hardwired use of
3037 "%lld" and "long long". Replaced the call to snprintf() with a call to
3040 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3041 records point to non-existent hosts", "retry timeout exceeded", and
3042 "retry time not reached for any host after a long failure period".
3044 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3045 experimental DomainKeys support:
3047 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3048 (2) On an error such as an illegally used "control", the wrong name for
3049 the control was given.
3051 These problems did NOT occur unless DomainKeys support was compiled.
3053 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3055 PH/24 Added ${if match_ip condition.
3057 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3058 databases so that it will be absolutely obvious if a crash occurs in the
3059 DB library. This is a regular occurrence (often caused by mis-matched
3062 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3063 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3064 picked up on a user's system that detects such things. There doesn't seem
3065 to be a gcc warning option for this - only an attribute that has to be
3066 put on the function's prototype. It seems that in Fedora Core 4 they have
3067 set this on a number of new functions. No doubt there will be more in due
3070 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3071 specify a transport. However, if an address that was verified by such a
3072 router was the subject of a callout, Exim crashed because it tried to
3073 read the rcpt_include_affixes from the non-existent transport. Now it
3074 just assumes that the setting of that option is false. This bug was
3075 introduced by 4.51/PH/31.
3077 PH/28 Changed -d+all to exclude +memory, because that information is very
3078 rarely of interest, but it makes the output a lot bigger. People tend to
3079 do -d+all out of habit.
3081 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3082 code in os-type was giving problems when libc.so lives in lib64, like on
3085 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3086 aren't the modern standard, and it seems that some systems' include files
3087 don't always have them. Exim was already checking for some of the newer
3088 ones like T_AAAA, and defining it itself. I've added checks for all the
3089 record types that Exim uses.
3091 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3092 not automatically generating a new one, as it is supposed to. This
3093 prevented TLS from working. If the file did exist, but contained invalid
3094 data, a new version was generated, as expected. It was only the case of a
3095 non-existent file that was broken.
3097 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3098 with a change in libdomainkeys > 0.64.
3100 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3101 from DNS. If the selector record carries the flag, it now has
3102 precedence over the domain-wide flag.
3104 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3106 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3107 the use of a port name that isn't defined in /etc/services) occurred, the
3108 message was deferred as in a normal delivery, and thus remained on the
3109 spool, instead of being failed because of the mua_wrapper setting. This
3110 is now fixed, and I tidied up some of the mua_wrapper messages at the
3113 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3114 the messages in a hash of arrays rather than using individual hashes.
3115 This is a bit cleaner and results in dramatic memory savings, albeit
3116 at a slight CPU cost.
3118 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3119 as requested by Marc Sherman.
3121 SC/10 Eximstats: added histograms for user specified patterns as requested
3124 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3126 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3127 fopen() in the content-scanning modules that did not already have it.
3133 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3134 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3136 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3138 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3140 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3141 to test Sieve filters that use "vacation".
3143 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3144 that changes the way the GnuTLS parameters are stored in the cache file.
3145 The new format can be generated externally. For backward compatibility,
3146 if the data in the cache doesn't make sense, Exim assumes it has read an
3147 old-format file, and it generates new data and writes a new file. This
3148 means that you can't go back to an older release without removing the
3151 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3152 work if there are any delivery delays because "one_time" forces the
3153 parent to be marked "delivered", so its unseen clone is never tried
3154 again. For this reason, Exim now forbids the simultaneous setting of
3157 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3158 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3159 entry if you want to know the details. The fix, however, neglected to
3160 consider the case when local delivery batching is involved. The test for
3161 "previously delivered" was not happening when checking to see if an
3162 address could be batched with a previous (undelivered) one; under
3163 certain circumstances this could lead to multiple deliveries to the same
3166 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3167 in its include files, and this causes problems building Exim.
3169 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3170 header_syntax) but Exim was just ignoring anything given after a slash.
3171 In particular, this caused confusion with an attempt to use "verify =
3172 reverse_host_lookup/defer_ok". An error is now given when options are
3173 supplied for verify items that do not have them. (Maybe reverse_host_
3174 lookup should have a defer_ok option, but that's a different point.)
3176 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3177 defined by RFC 821) to 2048, because there were problems with some AUTH
3178 commands, and RFC 1869 says the size should be increased for extended
3179 SMTP commands that take arguments.
3181 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3184 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3185 "unknown" error; now it says that the functionality isn't in the binary.
3187 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3188 an address' error message when a string expansion fails (syntax or
3189 whatever). Otherwise the password may appear in the log. Following change
3190 PH/42 below, there is no longer a chance of it appearing in a bounce
3193 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3195 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3196 of its addresses was ever tried. (Bugzilla bug #2.)
3198 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3199 the result incorrectly in the debug output. (It correctly added a newline
3200 to what was transported.)
3202 TF/01 Added $received_time.
3204 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3205 commented out examples of how to interface to a virus scanner and to
3206 SpamAssassin. Also added commented examples of av_scanner and
3207 spamd_address settings.
3209 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3210 and controls are allowed in which ACLs. There were a couple of minor
3211 errors. Some of the entries in the conditions table (which is a table of
3212 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3213 negation of where the condition IS allowed.
3215 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3217 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3218 header file does not have a version number, so I've had to invent a new
3219 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3220 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3221 radiusclient), but it was contributed by a Radius user.
3223 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3224 files or directories, for OpenSSL.
3226 PH/19 When an Exim process that is running as root has to create an Exim log
3227 file, it does so in a subprocess that runs as exim:exim so as to get the
3228 ownership right at creation (otherwise, other Exim processes might see
3229 the file with the wrong ownership). There was no test for failure of this
3230 fork() call, which would lead to the process getting stuck as it waited
3231 for a non-existent subprocess. Forks do occasionally fail when resources
3232 run out. I reviewed all the other calls to fork(); they all seem to check
3235 PH/20 When checking for unexpected SMTP input at connect time (before writing
3236 the banner), Exim was not dealing correctly with a non-positive return
3237 from the read() function. If the client had disconnected by this time,
3238 the result was a log entry for a synchronization error with an empty
3239 string after "input=" when read() returned zero. If read() returned -1
3240 (an event I could not check), uninitialized data bytes were printed.
3241 There were reports of junk text (parts of files, etc) appearing after
3244 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3246 PH/22 Added support for macro redefinition, and (re)definition in between
3247 driver and ACL definitions.
3249 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3250 forgetting to use the resulting value; it was using the unexpanded value.
3252 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3253 hadn't been configured. The fix is from Juergen Kreileder, who
3254 understands it better than I do:
3256 "Here's what I see happening with three configured cyrus_sasl
3257 authenticators configured (plain, login, cram-md5):
3259 On startup auth_cyrus_sasl_init() gets called for each of these.
3260 This means three calls to sasl_listmech() without a specified mech_list.
3261 => SASL tests which mechs of all available mechs actually work
3262 => three warnings about OTP not working
3263 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3265 With the patch, sasl_listmech() also gets called three times. But now
3266 SASL's mech_list option is set to the server_mech specified in the the
3267 authenticator. Or in other words, the answer from sasl_listmech()
3268 gets limited to just the mech you're testing for (which is different
3270 => the return list contains just 'plain' or 'login', 'cram-md5' or
3271 nothing depending on the value of ob->server_mech.
3273 I've just tested the patch: Authentication still works fine,
3274 unavailable mechs specified in the exim configuration are still
3275 caught, and the auth.log warnings about OTP are gone."
3277 PH/25 When debugging is enabled, the contents of the command line are added
3278 to the debugging output, even when log_selector=+arguments is not
3281 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3282 answer is "GNU", and only if the return is "GNU/something" is the answer
3285 PH/27 $acl_verify_message is now set immediately after the failure of a
3286 verification in an ACL, and so is available in subsequent modifiers. In
3287 particular, the message can be preserved by coding like this:
3289 warn !verify = sender
3290 set acl_m0 = $acl_verify_message
3292 Previously, $acl_verify_message was set only while expanding "message"
3293 and "log_message" when a very denied access.
3295 PH/28 Modified OS/os.c-Linux with
3297 -#ifndef OS_LOAD_AVERAGE
3298 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3300 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3301 nomenclature these days.)
3303 PH/29 Installed patch from the Sieve maintainer that adds the options
3304 sieve_useraddress and sieve_subaddress to the redirect router.
3306 PH/30 In these circumstances:
3307 . Two addresses routed to the same list of hosts;
3308 . First host does not offer TLS;
3309 . First host accepts first address;
3310 . First host gives temporary error to second address;
3311 . Second host offers TLS and a TLS session is established;
3312 . Second host accepts second address.
3313 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3314 and peerdn, if requested) that were in fact used only for the second
3317 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3318 attention to any local part prefix or suffix that was matched by the
3319 router that accepted the address. It now behaves in the same way as it
3320 does for delivery: the affixes are removed from the local part unless
3321 rcpt_include_affixes is set on the transport.
3323 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3324 timeout during the DATA phase of an incoming message.
3326 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3327 applied a patch sanctioned by the Sieve maintainer.
3329 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3330 the uid or gid is negative. A case of a negative gid caused this to be
3331 noticed. The fix allows for either to be negative.
3333 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3334 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3337 PH/36 The change PH/12 above was broken. Fixed it.
3339 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3340 the grounds that routing the same address twice would always produce the
3341 same answer. This might have been true once, but it is certainly no
3342 longer true now. Routing a child address may depend on the previous
3343 routing that produced that child. Some complicated redirection strategies
3344 went wrong when messages had multiple recipients, and made Exim's
3345 behaviour dependent on the order in which the addresses were given.
3347 I have moved the duplicate checking until after the routing is complete.
3348 Exim scans the addresses that are assigned to local and remote
3349 transports, and removes any duplicates. This means that more work will be
3350 done, as duplicates will always all be routed, but duplicates are
3351 presumably rare, so I don't expect this is of any significance.
3353 For deliveries to pipes, files, and autoreplies, the duplicate checking
3354 still happens during the routing process, since they are not going to be
3357 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3358 It corrects a timeout issue with spamd. This is Ian's comment: "The
3359 background is that sometimes spamd either never reads data from a
3360 connection it has accepted, or it never writes response data. The exiscan
3361 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3362 blindly assumes that writes won't block so it may never time out."
3364 PH/39 Allow G after quota size as well as K and M.
3366 PH/40 The value set for $authenticated_id in an authenticator may not contain
3367 binary zeroes or newlines because the value is written to log lines and
3368 to spool files. There was no check on this. Now the value is run through
3369 the string_printing() function so that such characters are converted to
3370 printable escape sequences.
3372 PH/41 $message_linecount is a new variable that contains the total number of
3373 lines in the message. Compare $body_linecount, which is the count for the
3376 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3377 bounce and delay warning messages, except in certain special cases, which
3380 (a) An SMTP error message from a remote host;
3381 (b) A message specified in a :fail: redirection;
3382 (c) A message specified in a "fail" command in a system filter;
3383 (d) A message specified in a FAIL return from the queryprogram router;
3384 (e) A message specified by the cannot_route_message router option.
3386 In these cases only, Exim does include the error details in bounce and
3387 warning messages. There are also a few cases where bland messages such
3388 as "unrouteable address" or "local delivery error" are given.
3390 PH/43 $value is now also set for the "else" part of a ${run expansion.
3392 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3393 being worked on, but at least Exim now implements the latest version to
3396 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3397 process to complete was treated as a delivery failure, a timeout while
3398 writing the message to the pipe was logged, but erroneously treated as a
3399 successful delivery. Such timeouts include transport filter timeouts. For
3400 consistency with the overall process timeout, these timeouts are now
3401 treated as errors, giving rise to delivery failures by default. However,
3402 there is now a new Boolean option for the pipe transport called
3403 timeout_defer, which, if set TRUE, converts the failures into defers for
3404 both kinds of timeout. A transport filter timeout is now identified in
3407 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3408 systems where "make" and "gmake" are different, calling "gmake" at top
3409 level broke things. I've arranged for the value of $(MAKE) to be passed
3410 from the Makefile to this script so that it can call the same version of
3414 A note about Exim versions 4.44 and 4.50
3415 ----------------------------------------
3417 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3418 changes of various kinds. As a consequence, a big documentation update was
3419 needed. This delayed the release for rather longer than seemed good, especially
3420 in the light of a couple of (minor) security issues. Therefore, the changes
3421 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3422 release. So 4.44 and 4.50 are in effect two different branches that both start
3425 I have left the 4.50 change log unchanged; it contains all the changes since
3426 4.43. The change log for 4.44 is below; many of its items are identical to
3427 those for 4.50. This seems to be the most sensible way to preserve the
3428 historical information.
3434 1. Minor wording change to the doc/README.SIEVE file.
3436 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3437 computation of the current number of files was incorrect.
3439 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3440 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3443 4. Give more explanation in the error message when the command for a transport
3444 filter fails to execute.
3446 5. There are several places where Exim runs a non-Exim command in a
3447 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3448 was being done only for the command run by the queryprogram router. It is
3449 now done for all such subprocesses. The other cases are: ${run, transport
3450 filters, and the commands run by the lmtp and pipe transports.
3452 6. Added CONFIGURE_GROUP build-time option.
3454 7. Some older OS have a limit of 256 on the maximum number of file
3455 descriptors. Exim was using setrlimit() to set 1000 as a large value
3456 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3457 systems. I've change it so that if it can't get 1000, it tries for 256.
3459 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3460 was an oversight, and furthermore, ever since the addition of extra
3461 controls (e.g. 4.43/32), the checks on when to allow different forms of
3462 "control" were broken. There should now be diagnostics for all cases when a
3463 control that does not make sense is encountered.
3465 9. Added the /retain_sender option to "control=submission".
3467 10. $recipients is now available in the predata ACL (oversight).
3469 11. Tidy the search cache before the fork to do a delivery from a message
3470 received from the command line. Otherwise the child will trigger a lookup
3471 failure and thereby defer the delivery if it tries to use (for example) a
3472 cached ldap connection that the parent has called unbind on.
3474 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3475 of $address_data from the recipient verification was clobbered by the
3476 sender verification.
3478 13. The value of address_data from a sender verification is now available in
3479 $sender_address_data in subsequent conditions in the ACL statement.
3481 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3483 15. Added a new option "connect=<time>" to callout options, to set a different
3486 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3487 was its contents. (It was OK if the option was not defined at all.)
3489 17. A "Completed" log line is now written for messages that are removed from
3490 the spool by the -Mrm option.
3492 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3493 information about exactly what failed.
3495 19. Added -dd to debug only the daemon process.
3497 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3498 handles timeouts, both on the server side and network timeouts. Renamed the
3499 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3501 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3502 It is now set to "smtps".
3504 22. $host_address is now set to the target address during the checking of
3505 ignore_target_hosts.
3507 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3508 being passed; this would have caused $sender_host_name to have been used if
3509 matching the list had actually called for a host name (not very likely,
3510 since this list is usually IP addresses). A host name is now passed as
3513 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3514 code that specifies a non-restarting handler (typically sigaction() in
3515 modern systems) in an attempt to fix a rare and obscure crash bug.
3517 25. Narrowed the window for a race in the daemon that could cause it to ignore
3518 SIGCHLD signals. This is not a major problem, because they are used only to
3519 wake it up if nothing else does.
3521 26. A malformed maildirsize file could cause Exim to calculate negative values
3522 for the mailbox size or file count. Odd effects could occur as a result.
3523 The maildirsize information is now recalculated if the size or filecount
3526 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3527 support for a long time. Removed HAVE_SYS_VFS_H.
3529 28. Installed the latest version of exipick from John Jetmore.
3531 29. In an address list, if the pattern was not a regular expression, an empty
3532 subject address (from a bounce message) matched only if the pattern was an
3533 empty string. Non-empty patterns were not even tested. This was the wrong
3534 because it is perfectly reasonable to use an empty address as part of a
3535 database query. An empty address is now tested by patterns that are
3536 lookups. However, all the other forms of pattern expect the subject to
3537 contain a local part and a domain, and therefore, for them, an empty
3538 address still always fails if the pattern is not itself empty.
3540 30. Exim went into a mad DNS loop when attempting to do a callout where the
3541 host was specified on an smtp transport, and looking it up yielded more
3542 than one IP address.
3544 31. Re-factored the code for checking spool and log partition space into a
3545 function that finds that data and another that does the check. The former
3546 is then used to implement four new variables: $spool_space, $log_space,
3547 $spool_inodes, and $log_inodes.
3549 32. The RFC2047 encoding function was originally intended for short strings
3550 such as real names; it was not keeping to the 75-character limit for
3551 encoded words that the RFC imposes. It now respects the limit, and
3552 generates multiple encoded words if necessary. To be on the safe side, I
3553 have increased the buffer size for the ${rfc2047: expansion operator from
3556 33. It is now permitted to omit both strings after an "if" condition; if the
3557 condition is true, the result is "true". As before, when the second string
3558 is omitted, a false condition yields an empty string. This makes it less
3559 cumbersome to write custom ACL and router conditions.
3561 34. Failure to deliver a bounce message always caused it to be frozen, even if
3562 there was an errors_to setting on the router. The errors_to setting is now
3565 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3566 canonical form (fully expanded) before being placed in
3567 $sender_host_address.
3569 36. The table in the code that translates DNS record types into text (T_A to
3570 "A" for instance) was missing entries for NS and CNAME. It is just possible
3571 that this could have caused confusion if both these types were looked up
3572 for the same domain, because the text type is used as part of Exim's
3573 per-process caching. But the chance of anyone hitting this buglet seems
3576 37. The dnsdb lookup has been extended in a number of ways.
3578 (1) There is a new type, "zns", which walks up the domain tree until it
3579 finds some nameserver records. It should be used with care.
3581 (2) There is a new type, "mxh", which is like "mx" except that it returns
3582 just the host names, not the priorities.
3584 (3) It is now possible to give a list of domains (or IP addresses) to be
3585 looked up. The behaviour when one of the lookups defers can be
3586 controlled by a keyword.
3588 (4) It is now possible to specify the separator character for use when
3589 multiple records are returned.
3591 38. The dnslists ACL condition has been extended: it is now possible to supply
3592 a list of IP addresses and/or domains to be looked up in a particular DNS
3595 39. Added log_selector=+queue_time_overall.
3597 40. When running the queue in the test harness, wait just a tad after forking a
3598 delivery process, to get repeatability of debugging output.
3600 41. Include certificate and key file names in error message when GnuTLS fails
3601 to set them up, because the GnuTLS error message doesn't include the name
3602 of the failing file when there is a problem reading it.
3604 42. Allow both -bf and -bF in the same test run.
3606 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3608 44. The "Exiscan patch" is now merged into the mainline Exim source.
3610 45. Sometimes the final signoff response after QUIT could fail to get
3611 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3612 < 0 before doing a fflush(). This bug looks as though it goes back to the
3613 introduction of TLS in release 3.20, but "sometimes" must have been rare
3614 because the tests only now provoked it.
3616 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3617 (this can affect the format of dates).
3619 47. exim_tidydb, when checking for the continued existence of a message for
3620 which it has found a message-specific retry record, was not finding
3621 messages that were in split spool directories. Consequently, it was
3622 deleting retry records that should have stayed in existence.
3624 48. Steve fixed some bugs in eximstats.
3626 49. The SPA authentication driver was not abandoning authentication and moving
3627 on to the next authenticator when an expansion was forced to fail,
3628 contradicting the general specification for all authenticators. Instead it
3629 was generating a temporary error. It now behaves as specified.
3631 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3632 (the order specifies the preference for clients). The order is now AES256,
3633 AES128, 3DES, ARCFOUR128.
3635 51. Small patch to Sieve code - explicitly set From: when generating an
3638 52. Exim crashed if a remote delivery caused a very long error message to be
3639 recorded - for instance if somebody sent an entire SpamAssassin report back
3640 as a large number of 550 error lines. This bug was coincidentally fixed by
3641 increasing the size of one of Exim's internal buffers (big_buffer) that
3642 happened as part of the Exiscan merge. However, to be on the safe side, I
3643 have made the code more robust (and fixed the comments that describe what
3646 53. Now that there can be additional text after "Completed" in log lines (if
3647 the queue_time_overall log selector is set), a one-byte patch to exigrep
3648 was needed to allow it to recognize "Completed" as not the last thing in
3651 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3652 patch that reportedly fixes this has been added. I am not expert enough to
3653 create a test for it. This is what the patch creator wrote:
3655 "I found a little strange behaviour of ldap code when working with
3656 Windows 2003 AD Domain, where users was placed in more than one
3657 Organization Units. When I tried to give exim partial DN, the exit code
3658 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3659 But simultaneously result of request was absolutely normal ldap result,
3660 so I produce this patch..."
3662 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3663 REFERENCE, so I have modified the code to exclude the patch when that macro
3666 55. Some experimental protocols are using DNS PTR records for new purposes. The
3667 keys for these records are domain names, not reversed IP addresses. The
3668 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3669 leaves it alone. Component reversal etc. now happens only for IP addresses.
3672 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3674 57. Double the size of the debug message buffer (to 2048) so that more of very
3675 long debug lines gets shown.
3677 58. The exicyclog utility now does better if the number of log files to keep
3678 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3680 59. Two changes related to the smtp_active_hostname option:
3682 (1) $smtp_active_hostname is now available as a variable.
3683 (2) The default for smtp_banner uses $smtp_active_hostname instead
3684 of $primary_hostname.
3686 60. The host_aton() function is supposed to be passed a string that is known
3687 to be a valid IP address. However, in the case of IPv6 addresses, it was
3688 not checking this. This is a hostage to fortune. Exim now panics and dies
3689 if the condition is not met. A case was found where this could be provoked
3690 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3691 components; fortuitously, this particular loophole had already been fixed
3692 by change 4.50/55 above.
3694 If there are any other similar loopholes, the new check in host_aton()
3695 itself should stop them being exploited. The report I received stated that
3696 data on the command line could provoke the exploit when Exim was running as
3697 exim, but did not say which command line option was involved. All I could
3698 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3699 running as the user.
3702 61. There was a buffer overflow vulnerability in the SPA authentication code
3703 (which came originally from the Samba project). I have added a test to the
3704 spa_base64_to_bits() function which I hope fixes it.
3707 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3708 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3710 63. The daemon start-up calls getloadavg() while still root for those OS that
3711 need the first call to be done as root, but it missed one case: when
3712 deliver_queue_load_max is set with deliver_drop_privilege. This is
3713 necessary for the benefit of the queue runner, because there is no re-exec
3714 when deliver_drop_privilege is set.
3716 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3717 This has been fixed.
3719 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3720 was in use, was not putting the data itself into the right store pool;
3721 consequently, it could be overwritten for a subsequent message in the same
3722 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3725 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3727 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3728 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3729 the function were treating the return as a boolean value, which happened to
3730 work because 0=false and not-0=true, but is not correct code.
3732 68. The host_aton() function was not handling scoped IPv6 addresses (those
3733 with, for example, "%eth0" on the end) correctly.
3735 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3736 negated items (that is, ~something) in unsigned ints. Some compilers
3737 apparently mutter when there is no cast.
3739 70. If an address verification called from an ACL failed, and did not produce a
3740 user-specific message (i.e. there was only a "system" message), nothing was
3741 put in $acl_verify_message. In this situation, it now puts the system
3744 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3745 session; change 4.31/43 added the unwanted input to the log line - except
3746 that it did not do this in the start of session case. It now does.
3748 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3749 This is wrong and can cause the other end to generate a synchronization
3750 error if it is another Exim or anything else that does the synchronization
3751 check. A QUIT command is no longer sent after a timeout.
3753 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3754 during host lookups.
3756 74. The defer_ok option of callout verification was not working if it was used
3757 when verifying addresses in header lines, that is, for this case:
3759 verify = header_sender/callout=defer_ok
3761 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3762 those file descriptors could be used for SMTP connections. If anything
3763 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3764 could be sent to the SMTP client, causing chaos. The daemon now opens
3765 stdin, stdout, and stderr to /dev/null when it puts itself into the
3768 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3769 log by default. The user can override this with suitable Perl magic.
3771 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3772 the log message when discard triggers, was not working for the DATA ACL or
3773 for the non-SMTP ACL.
3775 78. Error message wording change in sieve.c.
3777 79. If smtp_accept_max_per_host was set, the number of connections could be
3778 restricted to fewer than expected, because the daemon was trying to set up
3779 a new connection before checking whether the processes handling previous
3780 connections had finished. The check for completed processes is now done
3781 earlier. On busy systems, this bug wouldn't be noticed because something
3782 else would have woken the daemon, and it would have reaped the completed
3785 80. If a message was submitted locally by a user whose login name contained one
3786 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3787 It caused a spool format error. I have fixed the spool reading code. A
3788 related problem was that the "from" clause in the Received: line became
3789 illegal because of the space(s). It is now covered by ${quote_local_part.
3791 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3794 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3796 83. Patch from Sieve maintainer to fix unterminated string problem in
3797 "vacation" handling.
3799 84. Some minor changes to the Linux configuration files to help with other
3800 OS variants using glibc.
3802 85. One more patch for Sieve to update vacation handling to latest spec.
3805 ----------------------------------------------------
3806 See the note above about the 4.44 and 4.50 releases.
3807 ----------------------------------------------------
3813 1. Change 4.43/35 introduced a bug that caused file counts to be
3814 incorrectly computed when quota_filecount was set in an appendfile
3817 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3818 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3821 3. Give more explanation in the error message when the command for a transport
3822 filter fails to execute.
3824 4. There are several places where Exim runs a non-Exim command in a
3825 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3826 was being done only for the command run by the queryprogram router. It is
3827 now done for all such subprocesses. The other cases are: ${run, transport
3828 filters, and the commands run by the lmtp and pipe transports.
3830 5. Some older OS have a limit of 256 on the maximum number of file
3831 descriptors. Exim was using setrlimit() to set 1000 as a large value
3832 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3833 systems. I've change it so that if it can't get 1000, it tries for 256.
3835 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3836 was an oversight, and furthermore, ever since the addition of extra
3837 controls (e.g. 4.43/32), the checks on when to allow different forms of
3838 "control" were broken. There should now be diagnostics for all cases when a
3839 control that does not make sense is encountered.
3841 7. $recipients is now available in the predata ACL (oversight).
3843 8. Tidy the search cache before the fork to do a delivery from a message
3844 received from the command line. Otherwise the child will trigger a lookup
3845 failure and thereby defer the delivery if it tries to use (for example) a
3846 cached ldap connection that the parent has called unbind on.
3848 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3849 of $address_data from the recipient verification was clobbered by the
3850 sender verification.
3852 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3853 was its contents. (It was OK if the option was not defined at all.)
3855 11. A "Completed" log line is now written for messages that are removed from
3856 the spool by the -Mrm option.
3858 12. $host_address is now set to the target address during the checking of
3859 ignore_target_hosts.
3861 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3862 being passed; this would have caused $sender_host_name to have been used if
3863 matching the list had actually called for a host name (not very likely,
3864 since this list is usually IP addresses). A host name is now passed as
3867 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3868 code that specifies a non-restarting handler (typically sigaction() in
3869 modern systems) in an attempt to fix a rare and obscure crash bug.
3871 15. Narrowed the window for a race in the daemon that could cause it to ignore
3872 SIGCHLD signals. This is not a major problem, because they are used only to
3873 wake it up if nothing else does.
3875 16. A malformed maildirsize file could cause Exim to calculate negative values
3876 for the mailbox size or file count. Odd effects could occur as a result.
3877 The maildirsize information is now recalculated if the size or filecount
3880 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3881 support for a long time. Removed HAVE_SYS_VFS_H.
3883 18. Updated exipick to current release from John Jetmore.
3885 19. Allow an empty sender to be matched against a lookup in an address list.
3886 Previously the only cases considered were a regular expression, or an
3889 20. Exim went into a mad DNS lookup loop when doing a callout where the
3890 host was specified on the transport, if the DNS lookup yielded more than
3893 21. The RFC2047 encoding function was originally intended for short strings
3894 such as real names; it was not keeping to the 75-character limit for
3895 encoded words that the RFC imposes. It now respects the limit, and
3896 generates multiple encoded words if necessary. To be on the safe side, I
3897 have increased the buffer size for the ${rfc2047: expansion operator from
3900 22. Failure to deliver a bounce message always caused it to be frozen, even if
3901 there was an errors_to setting on the router. The errors_to setting is now
3904 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3905 canonical form (fully expanded) before being placed in
3906 $sender_host_address.
3908 24. Updated eximstats to version 1.33
3910 25. Include certificate and key file names in error message when GnuTLS fails
3911 to set them up, because the GnuTLS error message doesn't include the name
3912 of the failing file when there is a problem reading it.
3914 26. Expand error message when OpenSSL has problems setting up cert/key files.
3917 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3918 (this can affect the format of dates).
3920 28. exim_tidydb, when checking for the continued existence of a message for
3921 which it has found a message-specific retry record, was not finding
3922 messages that were in split spool directories. Consequently, it was
3923 deleting retry records that should have stayed in existence.
3925 29. eximstats updated to version 1.35
3926 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3927 1.35 - bugfix such that pie charts by volume are generated correctly
3929 30. The SPA authentication driver was not abandoning authentication and moving
3930 on to the next authenticator when an expansion was forced to fail,
3931 contradicting the general specification for all authenticators. Instead it
3932 was generating a temporary error. It now behaves as specified.
3934 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3935 (the order specifies the preference for clients). The order is now AES256,
3936 AES128, 3DES, ARCFOUR128.
3938 31. Small patch to Sieve code - explicitly set From: when generating an
3941 32. Exim crashed if a remote delivery caused a very long error message to be
3942 recorded - for instance if somebody sent an entire SpamAssassin report back
3943 as a large number of 550 error lines. This bug was coincidentally fixed by
3944 increasing the size of one of Exim's internal buffers (big_buffer) that
3945 happened as part of the Exiscan merge. However, to be on the safe side, I
3946 have made the code more robust (and fixed the comments that describe what
3949 33. Some experimental protocols are using DNS PTR records for new purposes. The
3950 keys for these records are domain names, not reversed IP addresses. The
3951 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3952 leaves it alone. Component reversal etc. now happens only for IP addresses.
3955 34. The host_aton() function is supposed to be passed a string that is known
3956 to be a valid IP address. However, in the case of IPv6 addresses, it was
3957 not checking this. This is a hostage to fortune. Exim now panics and dies
3958 if the condition is not met. A case was found where this could be provoked
3959 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3960 components; fortuitously, this particular loophole had already been fixed
3961 by change 4.50/55 or 4.44/33 above.
3963 If there are any other similar loopholes, the new check in host_aton()
3964 itself should stop them being exploited. The report I received stated that
3965 data on the command line could provoke the exploit when Exim was running as
3966 exim, but did not say which command line option was involved. All I could
3967 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3968 running as the user.
3971 35. There was a buffer overflow vulnerability in the SPA authentication code
3972 (which came originally from the Samba project). I have added a test to the
3973 spa_base64_to_bits() function which I hope fixes it.
3976 36. The daemon start-up calls getloadavg() while still root for those OS that
3977 need the first call to be done as root, but it missed one case: when
3978 deliver_queue_load_max is set with deliver_drop_privilege. This is
3979 necessary for the benefit of the queue runner, because there is no re-exec
3980 when deliver_drop_privilege is set.
3982 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3983 was in use, was not putting the data itself into the right store pool;
3984 consequently, it could be overwritten for a subsequent message in the same
3985 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3988 38. Sometimes the final signoff response after QUIT could fail to get
3989 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3990 < 0 before doing a fflush(). This bug looks as though it goes back to the
3991 introduction of TLS in release 3.20, but "sometimes" must have been rare
3992 because the tests only now provoked it.
3998 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3999 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4000 Now it returns an integer. A number of calls were still expecting a T/F
4001 return. Fortuitously, in all cases, the tests worked in OK situations,
4002 which is the norm. However, things would have gone wrong on any write
4003 failures on the smtp file descriptor. This function is used when sending
4004 messages over SMTP and also when doing verify callouts.
4006 2. When Exim is called to do synchronous delivery of a locally submitted
4007 message (the -odf or -odi options), it no longer closes stderr before doing
4010 3. Implemented the mua_wrapper option.
4012 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4014 5. Implemented the functions header_remove(), header_testname(),
4015 header_add_at_position(), and receive_remove_recipient(), and exported them
4018 6. If an ACL "warn" statement specified the addition of headers, Exim already
4019 inserted X-ACL-Warn: at the start if there was no header name. However, it
4020 was not making this test for the second and subsequent header lines if
4021 there were newlines in the string. This meant that an invalid header could
4022 be inserted if Exim was badly configured.
4024 7. Allow an ACL "warn" statement to add header lines at the start or after all
4025 the Received: headers, as well as at the end.
4027 8. Added the rcpt_4xx retry error code.
4029 9. Added postmaster_mailfrom=xxx to callout verification option.
4031 10. Added mailfrom=xxxx to the callout verification option, for verify=
4034 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4035 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4037 12. Inserted some casts to stop certain compilers warning when using pointer
4038 differences as field lengths or precisions in printf-type calls (mostly
4039 affecting debugging statements).
4041 13. Added optional readline() support for -be (dynamically loaded).
4043 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4044 same clock tick as a message's arrival, so that its received time was the
4045 same as the "first fail" time on the retry record, and that message
4046 remained on the queue past the ultimate address timeout, every queue runner
4047 would try a delivery (because it was past the ultimate address timeout) but
4048 after another failure, the ultimate address timeout, which should have then
4049 bounced the address, did not kick in. This was a "< instead of <=" error;
4050 in most cases the first failure would have been in the next clock tick
4051 after the received time, and all would be well.
4053 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4054 being recognized when the domain list was tested by the match_domain
4055 condition in an expansion string.
4057 16. Added the ${str2b64: operator.
4059 17. Exim was always calling setrlimit() to set a large limit for the number of
4060 processes, without checking whether the existing limit was already
4061 adequate. (It did check for the limit on file descriptors.) Furthermore,
4062 errors from getrlimit() and setrlimit() were being ignored. Now they are
4063 logged to the main and panic logs, but Exim does carry on, to try to do its
4064 job under whatever limits there are.
4066 18. Imported PCRE 5.0.
4068 19. Trivial typo in log message " temporarily refused connection" (the leading
4071 20. If the log selector return_path_on_delivery was set and an address was
4072 redirected to /dev/null, the delivery process crashed because it assumed
4073 that a return path would always be set for a "successful" delivery. In this
4074 case, the whole delivery is bypassed as an optimization, and therefore no
4077 21. Internal re-arrangement: the function for sending a challenge and reading
4078 a response while authentication was assuming a zero-terminated challenge
4079 string. It's now changed to take a pointer and a length, to allow for
4080 binary data in such strings.
4082 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4084 23. Exim was not respecting finduser_retries when seeking the login of the
4085 uid under which it was called; it was always trying 10 times. (The default
4086 setting of finduser_retries is zero.) Also, it was sleeping after the final
4087 failure, which is pointless.
4089 24. Implemented tls_on_connect_ports.
4091 25. Implemented acl_smtp_predata.
4093 26. If the domain in control=submission is set empty, Exim assumes that the
4094 authenticated id is a complete email address when it generates From: or
4095 Sender: header lines.
4097 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4098 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4099 chown and chgrp in /bin and hostname in /usr/bin.
4101 28. Exim was keeping the "process log" file open after each use, just as it
4102 does for the main log. This opens the possibility of it remaining open for
4103 long periods when the USR1 signal hits a daemon. Occasional processlog
4104 errors were reported, that could have been caused by this. Anyway, it seems
4105 much more sensible not to leave this file open at all, so that is what now
4108 29. The long-running daemon process does not normally write to the log once it
4109 has entered its main loop, and it closes the log before doing so. This is
4110 so that log files can straightforwardly be renamed and moved. However,
4111 there are a couple of unusual error situations where the daemon does write
4112 log entries, and I had neglected to close the log afterwards.
4114 30. The text of an SMTP error response that was received during a remote
4115 delivery was being truncated at 512 bytes. This is too short for some of
4116 the long messages that one sometimes sees. I've increased the limit to
4119 31. It is now possible to make retry rules that apply only when a message has a
4120 specific sender, in particular, an empty sender.
4122 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4123 it possible to be selective about when SMTP synchronization is enforced.
4125 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4127 32. Implemented hosts_connection_nolog.
4129 33. Added an ACL for QUIT.
4131 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4134 35. Added mailbox_size and mailbox_filecount to appendfile.
4136 36. Added control = no_multiline_responses to ACLs.
4138 37. There was a bug in the logic of the code that waits for the clock to tick
4139 in the case where the clock went backwards by a substantial amount such
4140 that the microsecond fraction of "now" was more than the microsecond
4141 fraction of "then" (but the whole seconds number was less).
4143 38. Added support for the libradius Radius client library this is found on
4144 FreeBSD (previously only the radiusclient library was supported).
4150 1. When certain lookups returned multiple values in the form name=value, the
4151 quoting of the values was not always being done properly. Specifically:
4152 (a) If the value started with a double quote, but contained no whitespace,
4154 (b) If the value contained whitespace other than a space character (i.e.
4155 tabs or newlines or carriage returns) it was not quoted.
4156 This fix has been applied to the mysql and pgsql lookups by writing a
4157 separate quoting function and calling it from the lookup code. The fix
4158 should probably also be applied to nisplus, ibase and oracle lookups, but
4159 since I cannot test any of those, I have not disturbed their existing code.
4161 2. A hit in the callout cache for a specific address caused a log line with no
4162 reason for rejecting RCPT. Now it says "Previous (cached) callout
4163 verification failure".
4165 3. There was an off-by-one bug in the queryprogram router. An over-long
4166 return line was truncated at 256 instead of 255 characters, thereby
4167 overflowing its buffer with the terminating zero. As well as fixing this, I
4168 have increased the buffer size to 1024 (and made a note to document this).
4170 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4171 when Exim is waiting for an SMTP response from a remote server, Exim
4172 restarts its select() call on the socket, thereby resetting its timeout.
4173 This is not a problem when such interrupts are rare. Somebody set up a cron
4174 job to run exiwhat every 2 minutes, which is less than the normal select()
4175 timeout (5 or 10 minutes). This meant that the select() timeout never
4176 kicked in because it was always reset. I have fixed this by comparing the
4177 time when an interrupt arrives with the time at the start of the first call
4178 to select(). If more time than the timeout has elapsed, the interrupt is
4179 treated as a timeout.
4181 5. Some internal re-factoring in preparation for the addition of Sieve
4182 extensions (by MH). In particular, the "personal" test is moved to a
4183 separate function, and given an option for scanning Cc: and Bcc: (which is
4184 not set for Exim filters).
4186 6. When Exim created an email address using the login of the caller as the
4187 local part (e.g. when creating a From: or Sender: header line), it was not
4188 quoting the local part when it contained special characters such as @.
4190 7. Installed new OpenBSD configuration files.
4192 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4193 try to make them clearer.
4195 9. Callout options, other than the timeout value, were being ignored when
4196 verifying sender addresses in header lines. For example, when using
4198 verify = header_sender/callout=no_cache
4200 the cache was (incorrectly) being used.
4202 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4203 only the Cygwin environment.
4205 11. When return_path_on_delivery was set as a log selector, if different remote
4206 addresses in the same message used different return paths and parallel
4207 remote delivery occurred, the wrong values would sometimes be logged.
4208 (Whenever a remote delivery process finished, the return path value from
4209 the most recently started remote delivery process was logged.)
4211 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4212 lines when AUTH and/or TLS are in use. This is the "received protocol"
4213 field. Exim used to use "asmtp" for authenticated SMTP, without any
4214 indication (in the protocol name) for TLS use. Now it follows the RFC and
4215 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4216 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4217 names appear in log lines as well as in Received: header lines.
4219 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4220 extensions, and comparison tests, and to fix some bugs.
4222 14. Changes to the "personal" filter test:
4224 (1) The test was buggy in that it was just doing the equivalent of
4225 "contains" tests on header lines. For example, if a user's address was
4226 anne@some.where, the "personal" test would incorrectly be true for
4228 To: susanne@some.where
4230 This test is now done by extracting each address from the header in turn,
4231 and checking the entire address. Other tests that are part of "personal"
4232 are now done using regular expressions (for example, to check local parts
4233 of addresses in From: header lines).
4235 (2) The list of non-personal local parts in From: addresses has been
4236 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4237 taken from the Sieve specification recommendations.
4239 (3) If the message contains any header line starting with "List-" it is
4240 treated as non-personal.
4242 (4) The test for "circular" in the Subject: header line has been removed
4243 because it now seems ill-conceived.
4245 15. Minor typos in src/EDITME comments corrected.
4247 16. Installed latest exipick from John Jetmore.
4249 17. If headers_add on a router specified a text string that was too long for
4250 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4251 of string_sprintf() is now avoided.
4253 18. $message_body_size was not set (it was always zero) when running the DATA
4254 ACL and the local_scan() function.
4256 19. For the "mail" command in an Exim filter, no default was being set for
4257 the once_repeat time, causing a random time value to be used if "once" was
4258 specified. (If the value happened to be <= 0, no repeat happened.) The
4259 default is now 0s, meaning "never repeat". The "vacation" command was OK
4260 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4261 (I found it when inspecting the code).
4263 20. There is now an overall timeout for performing a callout verification. It
4264 defaults to 4 times the callout timeout, which applies to individual SMTP
4265 commands during the callout. The overall timeout applies when there is more
4266 than one host that can be tried. The timeout is checked before trying the
4267 next host. This prevents very long delays if there are a large number of
4268 hosts and all are timing out (e.g. when the network connections are timing
4269 out). The value of the overall timeout can be changed by specifying an
4270 additional sub-option for "callout", called "maxwait". For example:
4272 verify = sender/callout=5s,maxwait=20s
4274 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4275 to the end before writing, but this should make it even safer).
4277 22. Exim was forgetting that it had advertised PIPELINING for the second and
4278 subsequent messages on an SMTP connection. It was also not resetting its
4279 memory on STARTTLS and an internal HELO.
4281 23. When Exim logs an SMTP synchronization error within a session, it now
4282 records whether PIPELINING has been advertised or not.
4284 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4285 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4286 rather than long int.
4288 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4290 26. Added the never_mail option to autoreply.
4296 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4297 crash if the getsockname() call failed; this can happen if a connection is
4298 closed very soon after it is established. The problem was simply in the
4299 order in which certain operations were done, causing Exim to try to write
4300 to the SMTP stream before it had set up the file descriptor. The bug has
4301 been fixed by making things happen in the correct order.
4307 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4308 before the connection was closed, thus losing the rejection response.
4310 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4311 some early Solaris releases, but causes trouble in current releases where
4312 socklen_t is defined.
4314 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4317 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4320 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4321 that releasing the top of it at the end releases what was used for sub-
4322 expansions (unless the block got too big). However, discard this block if
4323 the first thing is a variable or header, so that we can use its block when
4324 it is dynamic (useful for very large $message_headers, for example).
4326 6. Lookups now cache *every* query, not just the most recent. A new, separate
4327 store pool is used for this. It can be recovered when all lookup caches are
4328 flushed. Lookups now release memory at the end of their result strings.
4329 This has involved some general refactoring of the lookup sources.
4331 7. Some code has been added to the store_xxx() functions to reduce the amount
4332 of flapping under certain conditions.
4334 8. log_incoming_interface used to affect only the <= reception log lines. Now
4335 it causes the local interface and port to be added to several more SMTP log
4336 lines, for example "SMTP connection from", and rejection lines.
4338 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4340 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4342 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4343 could be overwritten at the end of the current message (or the start of a
4344 new message if it was set in a HELO ACL). The value is now preserved for
4345 the duration of the SMTP connection.
4347 12. If a transport had a headers_rewrite setting, and a matching header line
4348 contained an unqualified address, that address was qualified, even if it
4349 did not match any rewriting rules. The underlying bug was that the values
4350 of the flags that permit the existence of unqualified sender and recipient
4351 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4352 non-local messages, and by -bnq for local messages) were not being
4353 preserved with the message after it was received.
4355 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4356 "next input=" as part of the text comprising the host identity instead of
4357 the correct text. The code was using the same buffer for two different
4358 strings. However, depending on which order the printing function evaluated
4359 its arguments, the bug did not always show up. Under Linux, for example, my
4360 test suite worked just fine.
4362 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4363 doesn't work with some older versions of Perl. It has been changed to "my",
4364 which in any case is probably the better facility to use.
4366 15. A really picky compiler found some instances of statements for creating
4367 error messages that either had too many or two few arguments for the format
4370 16. The size of the buffer for calls to the DNS resolver has been increased
4371 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4372 for addresses that have a lot of PTR records. This alleviates a problem; it
4373 does not fully solve it.
4375 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4376 buffer now truncates the list and logs the incident, which is the same
4377 action as happens when Exim is looking up a host name and its aliases.
4378 Previously in this situation something unpredictable would happen;
4379 sometimes it was "internal error: store_reset failed".
4381 18. If a server dropped the connection unexpectedly when an Exim client was
4382 using GnuTLS and trying to read a response, the client delivery process
4383 crashed while trying to generate an error log message.
4385 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4386 string, for example:
4388 warn message = H1: something\nH2: something
4390 the text was added as a single header line from Exim's point of view
4391 though it ended up OK in the delivered message. However, searching for the
4392 second and subsequent header lines using $h_h2: did not work. This has been
4393 fixed. Similarly, if a system filter added multiple headers in this way,
4394 the routers could not see them.
4396 20. Expanded the error message when iplsearch is called with an invalid key to
4397 suggest using net-iplsearch in a host list.
4399 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4400 ACLs are no longer actually imposed (and a message to that effect is
4403 22. If a "gecos" field in a passwd entry contained escaped characters, in
4404 particular, if it contained a \" sequence, Exim got it wrong when building
4405 a From: or a Sender: header from that name. A second bug also caused
4406 incorrect handling when an unquoted " was present following a character
4407 that needed quoting.
4409 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4410 was not being matched caselessly.
4412 24. Arranged for all hyphens in the exim.8 source to be escaped with
4415 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4416 properly. Recipient callout cache records were still being keyed to include
4417 the sender, even when use_sender was set false. This led to far more
4418 callouts that were necessary. The sender is no longer included in the key
4419 when use_sender is false.
4421 26. Added "control = submission" modifier to ACLs.
4423 27. Added the ${base62d: operator to decode base 62 numbers.
4425 28. dnsdb lookups can now access SRV records.
4427 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4428 the configuration file.
4430 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4431 (-v) mode. This makes the output for a verbose queue run more intelligible.
4433 31. Added a use_postmaster feature to recipient callouts.
4435 32. Added the $body_zerocount variable, containing the number of binary zero
4436 bytes in the message body.
4438 33. The time of last modification of the "new" subdirectory is now used as the
4439 "mailbox time last read" when there is a quota error for a maildir
4442 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4444 35. Added +ignore_unknown as a special item in host lists.
4446 36. Code for decoding IPv6 addresses in host lists is now included, even if
4447 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4448 address was recognized as an IP address, but was then not correctly decoded
4449 into binary, causing unexpected and incorrect effects when compared with
4456 1. Very minor rewording of debugging text in manualroute to say "list of
4457 hosts" instead of "hostlist".
4459 2. If verify=header_syntax was set, and a header line with an unqualified
4460 address (no domain) and a large number of spaces between the end of the
4461 name and the colon was received, the reception process suffered a buffer
4462 overflow, and (when I tested it) crashed. This was caused by some obsolete
4463 code that should have been removed. The fix is to remove it!
4465 3. When running in the test harness, delay a bit after writing a bounce
4466 message to get a bit more predictability in the log output.
4468 4. Added a call to search_tidyup() just before forking a reception process. In
4469 theory, someone could use a lookup in the expansion of smtp_accept_max_
4470 per_host which, without the tidyup, could leave open a database connection.
4472 5. Added the variables $recipient_data and $sender_data which get set from a
4473 lookup success in an ACL "recipients" or "senders" condition, or a router
4474 "senders" option, similar to $domain_data and $local_part_data.
4476 6. Moved the writing of debug_print from before to after the "senders" test
4479 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4480 problems for message scanning, either using a data ACL, or using
4481 local_scan() because the Received: header was not generated till after they
4482 were called (in order to set the time as the time of reception completion).
4483 I have revised the way this works. The header is now generated after the
4484 body is received, but before the ACL or local_scan() are called. After they
4485 are run, the timestamp in the header is updated.
4491 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4492 before starting a queue runner without re-exec. This happened only when
4493 deliver_drop_privilege was set or when the Exim user was set to root. The
4494 effect of the bug was that timeouts during subsequent deliveries caused
4495 crashes instead of being properly handled. The handler is now left at its
4496 default (and expected) setting.
4498 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4499 message, again when deliver_drop_privilege is set or Exim is run as root.
4500 The bug described in (1) was not present in this case, but the tidying up
4501 of the other signals was missing. I have made the two cases consistent.
4503 3. The ignore_target_hosts setting on a manualroute router was being ignored
4504 for hosts that were looked up using the /MX notation.
4506 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4509 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4510 operated on the sender address. After changing the $sender_address to <>
4511 for the sender address verify, Exim was re-instated it as the original
4512 (before rewriting) address, but remembering that it had rewritten it, so it
4513 wasn't rewriting it again. This bug also had the effect of breaking the
4514 sender address verification caching when the sender address was rewritten.
4516 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4517 This has been changed so that if the ip literal address matches
4518 ignore_target_hosts, the router declines.
4520 7. Added expansion conditions match_domain, match_address, and match_local_
4521 part (NOT match_host).
4523 8. The placeholder for the Received: header didn't have a length field set.
4525 9. Added code to Exim itself and to exim_lock to test for a specific race
4526 condition that could lead to file corruption when using MBX delivery. The
4527 issue is with the lockfile that is created in /tmp. If this file is removed
4528 after a process has opened it but before that process has acquired a lock,
4529 there is the potential for a second process to recreate the file and also
4530 acquire a lock. This could lead to two Exim processes writing to the file
4531 at the same time. The added code performs the same test as UW imapd; it
4532 checks after acquiring the lock that its file descriptor still refers to
4533 the same named file.
4535 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4536 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4539 11. Added the smtp_active_hostname option. If used, this will typically be made
4540 to depend on the incoming interface address. Because $interface_address is
4541 not set up until the daemon has forked a reception process, error responses
4542 that can happen earlier (such as "too many connections") no longer contain
4545 12. If an expansion in a condition on a "warn" statement fails because a lookup
4546 defers, the "warn" statement is abandoned, and the next ACL statement is
4547 processed. Previously this caused the whole ACL to be aborted.
4549 13. Added the iplsearch lookup type.
4551 14. Added ident_timeout as a log selector.
4553 15. Added tls_certificate_verified as a log selector.
4555 16. Added a global option tls_require_ciphers (compare the smtp transport
4556 option of the same name). This controls incoming TLS connections.
4558 17. I finally figured out how to make tls_require_ciphers do a similar thing
4559 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4560 before starting the TLS session.
4562 18. Tabs are now shown as \t in -bP output.
4564 19. If the log selector return_path_on_delivery was set, Exim crashed when
4565 bouncing a message because it had too many Received: header lines.
4567 20. If two routers both had headers_remove settings, and the first one included
4568 a superfluous trailing colon, the final name in the first list and the
4569 first name in the second list were incorrectly joined into one item (with a
4570 colon in the middle).
4576 1. Added -C and -D options to the exinext utility, mainly to make it easier
4577 to include in the automated testing, but these could be helpful when
4578 multiple configurations are in use.
4580 2. The exinext utility was not formatting the output nicely when there was
4581 an alternate port involved in the retry record key, nor when there was a
4582 message id as well (for retries that were specific to a specific message
4583 and a specific host). It was also confused by IPv6 addresses, because of
4584 the additional colons they contain. I have fixed the IPv4 problem, and
4585 patched it up to do a reasonable job for IPv6.
4587 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4588 delivery, the log line now contains "pipelined" if PIPELINING was used.
4590 4. An SMTP transport process used to panic and die if the bind() call to set
4591 an explicit outgoing interface failed. This has been changed; it is now
4592 treated in the same way as a connect() failure.
4594 5. A reference to $sender_host_name in the part of a conditional expansion
4595 that was being skipped was still causing a DNS lookup. This no longer
4598 6. The def: expansion condition was not recognizing references to header lines
4599 that used bh_ and bheader_.
4601 7. Added the _cache feature to named lists.
4603 8. The code for checking quota_filecount in the appendfile transport was
4604 allowing one more file than it should have been.
4606 9. For compatibility with Sendmail, the command line option
4614 and sets the incoming protocol and host name (for trusted callers). The
4615 host name and its colon can be omitted when only the protocol is to be set.
4616 Note the Exim already has two private options, -pd and -ps, that refer to
4617 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4618 "s", but I don't think that's a major issue.
4620 10. A number of refactoring changes to the code, none of which should affect
4623 (a) The number of logging options was getting close to filling up the
4624 32-bit word that was used as a bit map. I have split them into two classes:
4625 those that are passed in the argument to log_write(), and those that are
4626 only ever tested independently outside of that function. These are now in
4627 separate 32-bit words, so there is plenty of room for expansion again.
4628 There is no change in the user interface or the logging behaviour.
4630 (b) When building, for example, log lines, the code previously used a
4631 macro that called string_cat() twice, in order to add two strings. This is
4632 not really sufficiently general. Furthermore, there was one instance where
4633 it was actually wrong because one of the argument was used twice, and in
4634 one call a function was used. (As it happened, calling the function twice
4635 did not affect the overall behaviour.) The macro has been replaced by a
4636 function that can join an arbitrary number of extra strings onto a growing
4639 (c) The code for expansion conditions now uses a table and a binary chop
4640 instead of a serial search (which was left over from when there were very
4641 few conditions). Also, it now recognizes conditions like "pam" even when
4642 the relevant support is not compiled in: a suitably worded error message is
4643 given if an attempt is made to use such a condition.
4645 11. Added ${time_interval:xxxxx}.
4647 12. A bug was causing one of the ddress fields not to be passed back correctly
4648 from remote delivery subprocesses. The field in question was not being
4649 subsequently used, so this caused to problems in practice.
4651 13. Added new log selectors queue_time and deliver_time.
4653 14. Might have fixed a bug in maildirsizefile handling that threw up
4654 "unexpected character" debug warnings, and recalculated the data
4655 unnecessarily. In any case, I expanded the warning message to give more
4658 15. Added the message "Restricted characters in address" to the statements in
4659 the default ACL that block characters like @ and % in local parts.
4661 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4662 Three changes have been made:
4664 (a) There was a serious bug; a negative response to MAIL caused the whole
4665 recipient domain to be cached as invalid, thereby blocking all messages
4666 to all local parts at the same domain, from all senders. This bug has
4667 been fixed. The domain is no longer cached after a negative response to
4668 MAIL if the sender used is not empty.
4670 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4673 (c) A new callout option, "use_sender" has been added for people who want
4674 the modified behaviour.
4680 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4683 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4684 indeed breaks things for older releases.
4686 3. Added additional logging to the case where there is a problem reading data
4687 from a filter that is running in a subprocess using a pipe, in order to
4688 try to track down a specific problem.
4690 4. Testing facility fudge: when running in the test harness and attempting
4691 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4692 getting "No route to host". Convert this to a timeout.
4694 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4697 6. Some OS don't have socklen_t but use size_t instead. This affects the
4698 fifth argument of getsockopt() amongst other things. This is now
4699 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4700 can be set for individual OS. I have set it for SunOS5, OSF1, and
4701 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4702 some earlier ones do not.
4704 7. Change 4.30/15 was not doing the test caselessly.
4706 8. The standard form for an IPv6 address literal was being rejected by address
4707 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4708 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4709 this, as well as the form without the "IPv6" on the front (but only when
4710 address literals are enabled, of course).
4712 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4714 10. Exim crashed if a message with an empty sender address specified by -f
4715 encountered a router with an errors_to setting. This could be provoked only
4716 by a command such as
4720 where an empty string was supplied; "<>" did not hit this bug.
4722 11. Installed PCRE release 4.5.
4724 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4725 remained set. It is now erased.
4727 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4728 times from message ids (which are base 36 rather than the normal 62).
4730 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4731 were being counted as actual protocol errors, and logged if the log
4732 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4733 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4734 and DATA following a set of rejected RCPTs do not count as protocol errors.
4735 In other words, Exim assumes they were pipelined, though this may not
4736 actually be the case. Of course, in all cases the client gets an
4737 appropriate error code.
4739 15. If a lookup fails in an ACL condition, a message about the failure may
4740 be available; it is used if testing the ACL cannot continue, because most
4741 such messages specify what the cause of the deferral is. However, some
4742 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4743 that caused an old message to be retained and used if a later statement
4744 caused a defer, replacing the real cause of the deferral.
4746 16. If an IP address had so many PTR records that the DNS lookup buffer
4747 was not large enough to hold them, Exim could crash while trying to process
4748 the truncated data. It now detects and logs this case.
4750 17. Further to 4.21/58, another change has been made: if (and only if) the
4751 first line of a message (the first header line) ends with CRLF, a bare LF
4752 in a subsequent header line has a space inserted after it, so as not to
4753 terminate the header.
4755 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4756 unnecessarily, used atoi() instead of strtol(), and didn't check the
4757 termination when getting file sizes from file names by regex.
4759 19. Completely re-implemented the support for maildirsize files, in the light
4760 of a number of problems with the previous contributed implementation
4761 (4.30/29). In particular:
4763 . If the quota is zero, the maildirsize file is maintained, but no quota is
4766 . If the maildir directory does not exist, it is created before any attempt
4767 to write a maildirsize file.
4769 . The quota value in the file is just a cache; if the quota is changed in
4770 the transport, the new value overrides.
4772 . A regular expression is available for excluding directories from the
4775 20. The autoreply transport checks the characters in options that define the
4776 message's headers; it allows continued headers, but it was checking with
4777 isspace() after an embedded newline instead of explicitly looking for a
4780 21. If all the "regular" hosts to which an address was routed had passed their
4781 expiry times, and had not reached their retry times, the address was
4782 bounced, even if fallback hosts were defined. Now Exim should go on to try
4785 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4786 equivalent code in the SMTP transport. Some hosts send humungous responses
4787 to HELO/EHLO, more than 1024 it seems.
4789 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4790 gives compiler warnings in some environments. I've now done it "properly",
4793 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4794 (because of problems with the built-in one) was declared to return uschar *
4795 instead of char *, causing compiler failure.
4797 25. Fixed a file descriptor leak when processing alias/forward files.
4799 26. Fixed a minor format string issue in dbfn.c.
4801 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4803 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4804 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4807 29. When a sender address is verified, it is cached, to save repeating the test
4808 when there is more than one recipient in a message. However, when the
4809 verification involves a callout, it is possible for different callout
4810 options to be set for different recipients. It is too complicated to keep
4811 track of this in the cache, so now Exim always runs a verification when a
4812 callout is required, relying on the callout cache for the optimization.
4813 The overhead is duplication of the address routing, but this should not be
4816 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4817 to be verified with callout=postmaster, and the main callout worked but the
4818 postmaster check failed, the verification correctly failed. However, if a
4819 subsequent RCPT command asked for sender verification *without* the
4820 postmaster check, incorrect caching caused this verification also to fail,
4823 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4824 it was not caching the DNS options (qualify_single, search_parents) that
4825 were used when the lookup failed. A subsequent lookup with different
4826 options therefore always gave the same answer, though there were cases
4827 where it should not have. (Example: a "domains = !$mx_any" option on a
4828 dnslookup router: the "domains" option is always processed without any
4829 widening, but the router might have qualify_single set.) Now Exim uses the
4830 cached value only when the same options are set.
4832 32. Added John Jetmore's "exipick" utility to the distribution.
4834 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4835 than a timeout (e.g. a certificate is required, and is not provided), an
4836 Exim server now closes the connection immediately. Previously it waited for
4837 the client to close - but if the client is SSL, it seems that they each
4838 wait for each other, leading to a delay before one of them times out.
4840 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4841 maintained 0.8.x compatibility because I don't think many are using it, and
4842 it is clearly obsolete.
4844 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4847 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4848 client certificate was expired. A simple patch fixes this, though I don't
4849 understand the full logic of why the verify callback is called multiple
4852 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4853 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4854 which causes problems with some clients (such as the Certicom SSL Plus
4855 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4856 disables the coutermeasure allowing Eudora to connect."
4858 38. Exim was not checking that a write() to a log file succeeded. This could
4859 lead to Bad Things if a log got too big, in particular if it hit a file
4860 size limit. Exim now panics and dies if it cannot write to a log file, just
4861 as it does if it cannot open a log file.
4863 39. Modified OS/Makefile-Linux so that it now contains
4865 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4867 The two -D definitions ensure that Exim is compiled with large file
4868 support, which makes it possible to handle log files that are bigger than
4871 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4872 instance) a domain was checked against a named list that involved a lookup,
4873 causing $domain_data to be set, then another domain was checked against the
4874 same list, then the first domain was re-checked, the value of $domain_data
4875 after the final check could be wrong. In particular, if the second check
4876 failed, it could be set empty. This bug probably also applied to
4879 41. The strip_trailing_dot option was not being applied to the address given
4880 with the -f command-line option.
4882 42. The code for reading a message's header from the spool was incrementing
4883 $received_count, but never initializing it. This meant that the value was
4884 incorrect (doubled) while delivering a message in the same process in which
4885 it was received. In the most common configuration of Exim, this never
4886 happens - a fresh exec is done - but it can happen when
4887 deliver_drop_privilege is set.
4889 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4890 it now includes up to 150 characters of the unexpected data in the log
4893 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4894 and building data strings. The size of both of these buffers was 10 000
4895 bytes - far larger than anybody would *ever* want, thought I. Needless to
4896 say, somebody hit the limit. I have increased the maximum line length to
4897 20 000 and the maximum data length of concatenated lines to 100 000. I have
4898 also fixed two bugs, because there was no checking on these buffers. Tsk,
4899 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4900 buffer is too small.
4902 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4903 lsearch lookups. Now it does.
4905 46. When parsing a route_list item in a manualroute router, a fixed-length
4906 buffer was used for the list of hosts. I made this 1024 bytes long,
4907 thinking that nobody would ever have a list of hosts that long. Wrong.
4908 Somebody had a whole pile of complicated expansion conditions, and the
4909 string was silently truncated, leading to an expansion error. It turns out
4910 that it is easier to change to an unlimited length (owing to other changes
4911 that have happened since this code was originally written) than to build
4912 structure for giving a limitation error. The length of the item that
4913 expands into the list of hosts is now unlimited.
4915 47. The lsearch lookup could not handle data where the length of text line was
4916 more than 4095 characters. Such lines were truncated, leading to shortened
4917 data being returned. It should now handle lines of any length.
4919 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4920 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4923 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4924 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4925 no longer happens in comment lines. A list of replacements is now placed
4926 at the head of all of the source files, except those whose only change is
4927 to replace PERL_COMMAND in the very first #! line.
4929 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4930 messages on the queue, with a bottom-up merge sort, using code contributed
4931 by Michael Haardt. This should make operations like -bp somewhat faster on
4932 large queues. It won't affect queue runners, except when queue_run_in_order
4935 51. Installed eximstats 1.31 in the distribution.
4937 52. Added support for SRV lookups to the dnslookup router.
4939 53. If an ACL referred to $message_body or $message_body_end, the value was not
4940 reset for any messages that followed in the same SMTP session.
4942 54. The store-handling optimization for building very long strings was not
4943 differentiating between the different store pools. I don't think this
4944 actually made any difference in practice, but I've tidied it.
4946 55. While running the routers to verify a sender address, $sender_address
4947 was still set to the sender address. This is wrong, because when routing to
4948 send a bounce to the sender, it would be empty. Therefore, I have changed
4949 it so that, while verifying a sender address, $sender_address is set to <>.
4950 (There is no change to what happens when verifying a recipient address.)
4952 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4953 target A or AAAA records (if not already returned) without resetting the
4954 qualify_single or search_parents options of the DNS resolver. These are
4955 inappropriate in this case because the targets of MX and SRV records must
4956 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4957 target that, when qualified, matched something in the local domain. These
4958 two options are now turned off when doing these lookups.
4960 57. It seems that at least some releases of Reiserfs (which does not have the
4961 concept of a fixed number of inodes) returns zero and not -1 for the
4962 number of available inodes. This interacted badly with check_spool_inodes,
4963 which assumed that -1 was the "no such thing" setting. What I have done is
4964 to check that the total number of inodes is greater than zero before doing
4965 the test of how many are available.
4967 58. When a "warn" ACL statement has a log_message modifier, the message is
4968 remembered, and not repeated. This is to avoid a lot of repetition when a
4969 message has many recipients that cause the same warning to be written.
4970 Howewer, Exim was preserving the list of already written lines for an
4971 entire SMTP session, which doesn't seem right. The memory is now reset if a
4972 new message is started.
4974 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4975 debugging output unless log_rewrite was also set.
4977 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4978 of (int)(handle) when we know that handle contains (void *)(-1).
4980 61. The Exim daemon panic-logs an error return when it closes the incoming
4981 connection. However "connection reset by peer" seems to be common, and
4982 isn't really an error worthy of noting specially, so that particular error
4985 62. When Exim is trying to find all the local interfaces, it used to panic and
4986 die if the ioctl to get the interface flags failed. However, it seems that
4987 on at least one OS (Solaris 9) it is possible to have an interface that is
4988 included in the list of interfaces, but for which you get a failure error
4989 for this call. This happens when the interface is not "plumbed" into a
4990 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4991 failure of the "get flags" call assumes that the interface is down.
4993 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4994 makes life easier for people who are doing arithmetic on fields extracted
4995 from dates, where you often get leading zeros that should not be
4996 interpreted as octal.
4998 64. Added qualify_domain to the redirect router, to override the global
5001 65. If a pathologically long header line contained very many addresses (the
5002 report of this problem mentioned 10 000) and each of them was rewritten,
5003 Exim could use up a very large amount of memory. (It kept on making new
5004 copies of the header line as it rewrote, and never released the old ones.)
5005 At the expense of a bit more processing, the header rewriting function has
5006 been changed so that it no longer eats memory in this way.
5008 66. The generation of the Received: header has been moved from the time that a
5009 message starts to be received, to the time that it finishes. The timestamp
5010 in the Received: header should now be very close to that of the <= log
5011 line. There are two side-effects of this change:
5013 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5014 logged header lines no longer include the local Received: line, because
5015 it has not yet been created. The same applies to a copy of the message
5016 that is returned to a non-SMTP sender when a message is rejected.
5018 (b) When a filter file is tested using -bf, no additional Received: header
5019 is added to the test message. After some thought, I decided that this
5022 This change does not affect the value of $received_for. It is still set
5023 after address rewriting, but before local_scan() is called.
5025 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5027 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5028 gave an unhelpful panic error message, and a defer error. I have managed to
5029 change this behaviour so that it now rejects any supplied certificate,
5030 which seems right, as the list of acceptable certificates is empty.
5032 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5033 gave an unhelpful defer error. I have not managed to make this reject any
5034 supplied certificates, but the error message it gives is "no certificate
5035 supplied", which is not helpful.
5037 70. exigrep's output now also includes lines that are not associated with any
5038 message, but which match the given pattern. Implemented by a patch from
5039 Martin Sluka, which also tidied up the Perl a bit.
5041 71. Recipient callout verification, like sender verification, was using <> in
5042 the MAIL FROM command. This isn't really the right thing, since the actual
5043 sender may affect whether the remote host accepts the recipient or not. I
5044 have changed it to use the actual sender in the callout; this means that
5045 the cache record is now keyed on a recipient/sender pair, not just the
5046 recipient address. There doesn't seem to be a real danger of callout loops,
5047 since a callout by the remote host to check the sender would use <>.
5048 [SEE ABOVE: changed after hitting problems.]
5050 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5051 temporary errors. However, in the case of such a code being given after
5052 the end of a data transmission (i.e. after ".") Exim was failing to write
5053 a retry record for the message. (Yes, there was some broken host that was
5054 actually sending 8xx at this point.)
5056 73. An unknown lookup type in a host list could cause Exim to panic-die when
5057 the list was checked. (An example that provoked this was putting <; in the
5058 middle of a list instead of at the start.) If this happened during a DATA
5059 ACL check, a -D file could be left lying around. This kind of configuration
5060 error no longer causes Exim to die; instead it causes a defer errror. The
5061 incident is still logged to the main and panic logs.
5063 74. Buglet left over from Exim 3 conversion. The message "too many messages
5064 in one connection" was written to the rejectlog but not the mainlog, except
5065 when address rewriting (yes!) was being logged.
5067 75. Added write_rejectlog option.
5069 76. When a system filter was run not as root (that is, when system_filter_user
5070 was set), the values of the $n variables were not being returned to the
5071 main process; thus, they were not subsequently available in the $sn
5074 77. Added +return_path_on_delivery log selector.
5076 78. A connection timeout was being treated differently from recipients deferred
5077 when testing hosts_max_try with a message that was older than the host's
5078 retry timeout. (The host should not be counted, thus allowing all hosts to
5079 be tried at least once before bouncing.) This may have been the cause of an
5080 occasionally reported bug whereby a message would remain on the queue
5081 longer than the retry timeout, but would be bounced if a delivery was
5082 forced. I say "may" because I never totally pinned down the problem;
5083 setting up timeout/retry tests is difficult. See also the next item.
5085 79. The ultimate address timeout was not being applied to errors that involved
5086 a combination of host plus message (for example, a timeout on a MAIL
5087 command). When an address resolved to a number of possible hosts, and they
5088 were not all tried for each delivery (e.g. because of hosts_max_try), a
5089 message could remain on the queue longer than the retry timeout.
5091 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5094 81. Fixed an obscure SMTP outgoing bug which required at least the following
5095 conditions: (a) there was another message waiting for the same server;
5096 (b) the server returned 5xx to all RCPT commands in the first message so
5097 that the message was not completed; (c) the server dropped the connection
5098 or gave a negative response to the RSET that Exim sends to abort the
5099 transaction. The observed case was a dropped connection after DATA that had
5100 been sent in pipelining mode. That is, the server had advertised PIPELINING
5101 but was not implementing it correctly. The effect of the bug was incorrect
5102 behaviour, such as trying another host, and this could lead to a crash.
5108 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5109 and daemon.c were passed as pointers to ints; they should have been
5110 pointers to socklen_t variables (which are typically unsigned ints).
5112 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5115 3. Fixed a really odd bug that affected only the testing scheme; patching a
5116 certain fixed string in the binary changed the value of another string that
5117 happened to be identical to the end of the original first string.
5119 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5120 name", it returns that address as the IP address. On some operating
5121 systems (e.g. Solaris), it also passes back the IP address string as the
5122 "host name". However, on others (e.g. Linux), it passes back an empty
5123 string. Exim wasn't checking for this, and was changing the host name to an
5124 empty string, assuming it had been canonicized.
5126 5. Although rare, it is permitted to have more than one PTR record for a given
5127 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5128 all the names associated with an address, because they do in Solaris.
5129 However, it seems that they do not in Linux for data that comes from the
5130 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5131 I found this out when I moved to a new Linux workstation and tried to run
5132 the Exim test suite.
5134 To get round this problem I have changed the code so that it now does its
5135 own call to the DNS to look up PTR records when searching for a host name.
5136 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5137 addresses that are only in /etc/hosts are still found.
5139 This behaviour is, however, controlled by an option called host_lookup_
5140 order, which defaults to "bydns:byaddr". If people want to use the other
5141 order, or indeed, just use one or the other means of lookup, they can
5142 specify it in this variable.
5144 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5145 some operating systems, this comes back from gethostbyaddr() as an empty
5146 string, and this is what Exim used to test for. However, it seems that in
5147 other systems, "." is yielded. Exim now tests for this case too.
5149 7. The values of check_spool_space and check_log_space are now held internally
5150 as a number of kilobytes instead of an absolute number of bytes. If a
5151 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5152 kilobyte. This means that much larger values can be stored.
5154 8. Exim monitor: an attempt to get the action menu when not actually pointing
5155 at a message produces an empty menu entitled "No message selected". This
5156 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5157 no entries in it ("Shell widget menu has zero width and/or height"). So I
5158 have added a single, blank menu entry in this case.
5160 9. Added ${quote_local_part.
5162 10. MIME decoding is now applied to the contents of Subject: header lines when
5165 11. Now that a reference to $sender_host_address automatically causes a reverse
5166 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5167 host lookup before query-style lookups in lists that might use this
5168 variable. This has therefore been abolished, and the "net-" prefix is no
5169 longer necessary for query-style lookups.
5171 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5172 have been a typo for LFLAGS, so it has been changed.
5174 13. The install script calls Exim with "-C /dev/null" in order to find the
5175 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5176 to be output. Howeve, since Exim outputs its version number before the
5177 error, it didn't break the script. It just looked ugly. I fixed this by
5178 always allowing "-C /dev/null" if the caller is root.
5180 14. Ignore overlarge ACL variable number when reading spool file - insurance
5181 against a later release with more variables having written the file.
5183 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5184 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5185 this, as well as the form without the "IPv6" on the front.
5187 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5188 OS/Makefile-Darwin file.
5190 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5191 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5193 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5194 variable. However, some OS use ints for inodes. I've added cast to long int
5195 to get rid of the compiler warning.
5197 19. I had forgotten to lock out "/../" in configuration file names when
5198 ALT_CONFIG_PREFIX was set.
5200 20. Routers used for verification do not need to specify transports. However,
5201 if such a router generated a host list, and callout was configured, Exim
5202 crashed, because it could not find a port number from the (non-existent)
5203 transport. It now assumes port 25 in this circumstance.
5205 21. Added the -t option to exigrep.
5207 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5208 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5209 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5210 them is set, LOOKUP_LSEARCH is forced.
5212 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5214 24. Added sender and host information to the "rejected by local_scan()" log
5215 line; previously there was no indication of these.
5217 25. Added .include_if_exists.
5219 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5220 Linux. It turns out that not all file systems support this. Apparently some
5221 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5222 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5223 not supported on the file descriptor, is now ignored when Exim is trying to
5224 sync a directory. This applies only to Linux.
5226 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5228 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5229 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5232 29. Added support for maildirsize files from supplied patch (modified a bit).
5234 30. The use of :fail: followed by an empty string could lead Exim to respond to
5235 sender verification failures with (e.g.):
5237 550 Verification failed for <xxx>
5238 550 Sender verify failed
5240 where the first response line was missing the '-' that indicates it is not
5241 the final line of the response.
5243 31. The loop for finding the name of the user that called Exim had a hardwired
5244 limit of 10; it now uses the value of finduser_retries, which is used for
5245 all other user lookups.
5247 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5250 33. Exim was neglecting to zero errno before one call of strtol() when
5251 expanding a string and expecting an integer value. On some systems this
5252 resulted in spurious "integer overflow" errors. Also, it was casting the
5253 result into an int without checking.
5255 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5256 did not work. The code looks as if it has *never* worked, though it appears
5257 to have been documented since at least releast 1.62. I have made it work.
5259 35. The "timeout_DNS" error in retry rules, also documented since at least
5260 1.62, also never worked. As it isn't clear exactly what this means, and
5261 clearly it isn't a major issue, I have abolished the feature by treating it
5262 as "timeout", and writing a warning to the main and panic logs.
5264 36. The display of retry rules for -brt wasn't always showing the error code
5267 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5268 timeout_connect_A, timeout_connect_MX.
5270 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5271 to the empty sender.
5273 39. The daemon was not analysing the content of -oX till after it had closed
5274 stderr and disconnected from the controlling terminal. This meant that any
5275 syntax errors were only noted on the panic log, and the return code from
5276 the command was 0. By re-arranging the code a little, I've made the
5277 decoding happen first, so such errors now appear on stderr, and the return
5278 code is 1. However, the actual setting up of the sockets still happens in
5279 the disconnected process, so errors there are still only recorded on the
5282 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5283 connections (as happens on some IP stacks) was logged at start up time as
5284 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5285 it from "IPv6 and IPv4", which means that two separate sockets are being
5288 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5289 says whether AF_INET or AF_INET6 was passed as an argument.
5291 42. Exiwhat output was messed up when time zones were included in log
5294 43. Exiwhat now gives more information about the daemon's listening ports,
5295 and whether -tls-on-connect was used.
5297 44. The "port" option of the smtp transport is now expanded.
5299 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5300 silently ignored. Now an error message is written to the main and panic
5303 46. There's a new ACL modifier called "logwrite" which writes to a log file
5304 as soon as it is encountered.
5306 47. Added $local_user_uid and $local_user_gid at routing time.
5308 48. Exim crashed when trying to verify a sender address that was being
5311 49. Exim was recognizing only a space character after ".include". It now also
5312 recognizes a tab character.
5314 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5315 extracting the relevant information from the specification. The man page no
5316 longer contains scrambled data for the -d option, and I've added a section
5317 at the front about calling Exim under different names.
5319 51. Added "extra_headers" argument to the "mail" command in filter files.
5321 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5324 53. Installed eximstats 1.29.
5326 54. Added transport_filter_timeout as a generic transport option.
5328 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5329 Cc: header lines. This was required by RFC 822, but it not required by RFC
5332 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5333 incoming messages that did not have them. Now it does so only if the
5334 message originates locally, that is, if there is no associated remote host
5335 address. When Resent- header lines are present, this applies to the Resent-
5336 lines rather than the non-Resent- lines.
5338 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5339 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5341 58. Messages for configuration errors now include the name of the main
5342 configuration file - useful now that there may be more than one file in a
5343 list (.included file names were always shown).
5345 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5346 for those rare installations that do not start the daemon as root or run it
5347 setuid root. I've cut out the call to initgroups() if the daemon is not
5350 60. The Exim user and group can now be bound into the binary as text strings
5351 that are looked up at the start of Exim's processing.
5353 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5355 62. Added $mailstore_basename variable.
5357 63. Installed patch to sieve.c from Michael Haardt.
5359 64. When Exim failed to open the panic log after failing to open the main log,
5360 the original message it was trying to log was written to stderr and debug
5361 output, but if they were not available (the usual case in production), it
5362 was lost. Now it is written to syslog before the two lines that record the
5363 failures to open the logs.
5365 65. Users' Exim filters run in subprocesses under the user's uid. It is
5366 possible for a "deliver" command or an alias in a "personal" command to
5367 provoke an address rewrite. If logging of address rewriting is configured,
5368 this fails because the process is not running as root or exim. There may be
5369 a better way of dealing with this, but for the moment (because 4.30 needs
5370 to be released), I have disabled address rewrite logging when running a
5371 filter in a non-root, non-exim process.
5377 1. The buildconfig auxiliary program wasn't quoting the value set for
5378 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5379 not defined. This bug was masked in 4.22 by the effect that was fixed in
5382 2. Some messages that were rejected after a message id was allocated were
5383 shown as "incomplete" by exigrep. It no longer does this for messages that
5384 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5386 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5387 have allow_domain_literals set, the ID did not get logged in the <= line.
5388 Domain literals are now always recognized in Message-ID: header lines.
5390 4. The first argument for a ${extract expansion item is the key name or field
5391 number. Leading and trailing spaces in this item were not being ignored,
5392 causing some misleading effects.
5394 5. When deliver_drop_privilege was set, single queue runner processes started
5395 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5396 same command in the process it spins off) were not dropping privilege.
5398 6. When the daemon running as "exim" started a queue runner, it always
5399 re-executed Exim in the spun-off process. This is a waste of effort when
5400 deliver_drop_privilege is set. The new process now just calls the
5401 queue-runner function directly.
5407 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5410 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5411 ignored. Though the use of -oP was forcing the writing of a pid file, it
5412 was always written to the default place.
5414 3. If the message "no IP address found for host xxxx" is generated during
5415 incoming verification, it is now followed by identification of the incoming
5416 connection (so you can more easily find what provoked it).
5418 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5420 5. Added some features to "harden" Exim a bit more against certain attacks:
5422 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5423 be put in Local/Makefile. This is like the never_users runtime option,
5424 but it cannot be overridden. The default setting is "root".
5426 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5427 prefix string with which any file named in a -C command line option
5430 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5431 is retained for -C and -D only if the caller of Exim is root. Without
5432 it, the exim user may also use -C and -D and retain privilege.
5434 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5435 command line option is disabled.
5437 6. Macro names set by the -D option must start with an upper case letter, just
5438 like macro names defined in the configuration file.
5440 7. Added "dereference=" facility to LDAP.
5442 8. Two instances of the typo "uknown" in the source files are fixed.
5444 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5445 the Configure-Makefile script screwed up while processing it.
5447 10. Incorporated PCRE 4.4.
5449 11. The SMTP synchronization check was not operating right at the start of an
5450 SMTP session. For example, it could not catch a HELO sent before the client
5451 waited for the greeting. There is now a check for outstanding input at the
5452 point when the greeting is written. Because of the duplex, asynchronous
5453 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5454 way, but not yet received, when the check is performed.
5456 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5457 on TCP/IP sockets, because this apparently causes some broken clients to
5460 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5461 unchanged) from the Cygwin maintainer.
5463 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5464 was supported instead of testing for mbx. Effectively a typo.
5466 15. The spa authenticator server code was not checking that the input it
5467 received was valid base64.
5469 16. The debug output line for the "set" modifier in ACLs was not showing the
5470 name of the variable that was being set.
5472 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5474 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5475 lookup on its own. Something else was needed to trigger the lookup. For
5476 example, a match in host_lookup or the need for a host name in a host list.
5477 Now, if $sender_host_name is referenced and the host name has not yet been
5478 looked up, a lookup is performed. If the lookup fails, the variable remains
5479 empty, and $host_lookup_failed is set to "1".
5481 19. Added "eqi" as a case-independent comparison operator.
5483 20. The saslauthd authentication condition could segfault if neither service
5484 nor realm was specified.
5486 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5487 error message that was logged was misleading, and incoming SMTP
5488 connections were dropped. The message is now more accurate, and temporary
5489 errors are given to SMTP connections.
5491 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5492 (except RSET) with a 421 error, until QUIT is received. However, it was
5493 failing to send a response to QUIT.
5495 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5496 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5497 ACL is now run after the helo_try_verify_hosts code.
5499 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5500 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5501 case-independent, so other case variants are also recognized). Apparently
5502 some systems use these upper case variants.
5504 25. If more than two messages were waiting for the same host, and a transport
5505 filter was specified for the transport, Exim sent two messages over the
5506 same TCP/IP connection, and then failed with "socket operation on non-
5507 socket" when it tried to send the third.
5509 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5511 27. The extern definition of crypt16() in expand.c was not being excluded when
5512 the OS had its own crypt16() function.
5514 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5515 as a preferred synonym for return_size_limit, both as an option and as an
5518 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5520 30. Changed the default configuration ACL to relax the local part checking rule
5521 for addresses that are not in any local domains. For these addresses,
5522 slashes and pipe symbols are allowed within local parts, but the sequence
5523 /../ is explicitly forbidden.
5525 31. SPA server authentication was not clearing the challenge buffer before
5528 32. log_message in a "warn" ACL statement was writing to the reject log as
5529 well as to the main log, which contradicts the documentation and doesn't
5530 seem right (because no rejection is happening). So I have stopped it.
5532 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5533 However, I am unable to do any testing of this.
5535 34. Fixed an infelicity in the appendfile transport. When checking directories
5536 for a mailbox, to see if any needed to be created, it was accidentally
5537 using path names with one or more superfluous leading slashes; tracing
5538 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5540 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5541 contents are added to the log line that is written for every discarded
5542 recipient. (Previously a log_message setting was ignored.)
5544 36. The ${quote: operator now quotes the string if it is empty.
5546 37. The install script runs exim in order to find its version number. If for
5547 some reason other than non-existence or emptiness, which it checks, it
5548 could not run './exim', it was installing it with an empty version number,
5549 i.e. as "exim-". This error state is now caught, and the installation is
5552 38. An argument was missing from the function that creates an error message
5553 when Exim fails to connect to the socket for saslauthd authentication.
5554 This could cause Exim to crash, or give a corrupted message.
5556 39. Added isip, isip4, and isip6 to ${if conditions.
5558 40. The ACL variables $acl_xx are now saved with the message, and can be
5559 accessed later in routers, transports, and filters.
5561 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5562 strings in the file are not string-expanded.
5564 42. If a MAIL command specified a SIZE value that was too large to fit into an
5565 int variable, the check against message_size_limit failed. Such values are
5566 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5567 day this will have to be increased, but I don't think I want to be around
5568 when emails are that large.
5575 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5576 iconv() is not standard in FreeBSD.
5578 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5579 IPv6 enabled. The observed symptom was a segmentation fault on return from
5580 the function os_common_find_running_interfaces() in src/os.c.
5582 3. In the check_special_case() function in daemon.c I had used "errno" as an
5583 argument name, which causes warnings on some systems. This was basically a
5584 typo, since it was named "eno" in the comments!
5586 4. The code that waits for the clock to tick (at a resolution of some fraction
5587 of a second) so as to ensure message-id uniqueness was always waiting for
5588 at least one whole tick, when it could have waited for less. [This is
5589 almost certainly not relevant at current processor speeds, where it is
5590 unlikely to ever wait at all. But we try to future-proof.]
5592 5. The function that sleeps for a time interval that includes fractions of a
5593 second contained a race. It did not block SIGALRM between setting the
5594 timer, and suspending (a couple of lines later). If the interval was short
5595 and the sigsuspend() was delayed until after it had expired, the suspension
5596 never ended. On busy systems this could lead to processes getting stuck for
5599 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5600 process, before it forks any delivery processes. The open lookup caching
5601 mechanism meant that the open file or database connection was passed into
5602 the delivery process. The problem was that delivery processes always tidy
5603 up cached lookup data. This could cause a problem for the next delivery
5604 process started by the queue runner, because the external queue runner
5605 process does not know about the closure. So the next delivery process
5606 still has data in the lookup cache. In the case of a file lookup, there was
5607 no problem because closing a file descriptor in a subprocess doesn't affect
5608 the parent. However, if the lookup was caching a connection to a database,
5609 the connection was closed, and the second delivery process was likely to
5610 see errors such as "PGSQL: query failed: server closed the connection
5611 unexpectedly". The problem has been fixed by closing all cached lookups
5612 in a queue runner before running a delivery process.
5614 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5615 seem to have the "const" qualifier which it has on other OS. I've
5618 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5619 *of the same type* (client or server) with the same public name that an
5620 error should be diagnosed.
5622 9. When Exim looked up a host name for an IP address, but failed to find the
5623 original IP address when looking up the host name (a safety check), it
5624 output the message "<ip address> does not match any IP for NULL", which was
5625 confusing, to say the least. The bug was that the host name should have
5626 appeared instead of "NULL".
5628 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5629 user that is built into the binary, and the -C or -D options is used, root
5630 privilege is dropped before the configuration file is read. In addition,
5631 logging is switched to stderr instead of the normal log files. If the
5632 configuration then re-defines the Exim user, the unprivileged environment
5633 is probably not what is expected, so Exim logs a panic warning message (but
5636 However, if deliver_drop_privilege is set, the unprivileged state may well
5637 be exactly what is intended, so the warning has been cut out in that case,
5638 and Exim is allowed to try to write to its normal log files.
5644 1. smtp_return_error_details was not giving details for temporary sender
5645 or receiver verification errors.
5647 2. Diagnose a configuration error if two authenticators have the same public
5650 3. Exim used not to create the message log file for a message until the first
5651 delivery attempt. This could be confusing when incoming messages were held
5652 for policy or load reasons. The message log file is now created at the time
5653 the message is received, and an initial "Received" line is written to it.
5655 4. The automatically generated man page for command line options had a minor
5656 bug that caused no ill effects; however, a more serious problem was that
5657 the procedure for building the man page automatically didn't always
5658 operate. Consequently, release 4.20 contains an out-of-date version. This
5659 shouldn't happen again.
5661 5. When building Exim with embedded Perl support, the script that builds the
5662 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5663 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5665 6. The freeze_tell option was not being used for messages that were frozen on
5666 arrival, either by an ACL or by local_scan().
5668 7. Added the smtp_incomplete_transaction log selector.
5670 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5671 was accepting AUTH without a new EHLO.
5673 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5674 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5677 10. Logging of TCP/IP connections (when configured) now happens in the main
5678 daemon process instead of the child process, so that the TCP/IP connection
5679 count is more accurate (but it can never be perfect).
5681 11. The use of "drop" in a nested ACL was not being handled correctly in the
5682 outer ACL. Now, if condition failure induced by the nested "drop" causes
5683 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5684 or "require"), the connection is dropped.
5686 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5687 that yield "discard" can now be used with an "accept" or a "discard" verb,
5688 but an error is generated for any others (because I can't see a useful way
5689 to define what should happen).
5691 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5692 are now processed in the same way as lines in the Exim configuration file.
5693 In particular, continuation lines are supported.
5695 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5697 15. Added -ti meaning -t -i.
5699 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5700 domains, and warn by logging if others are found.
5702 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5703 returned from GIFCONF when Exim is trying to find the list of interfaces on
5704 a host. The code in os.c has been modified to copy each ifreq to an aligned
5705 structure in all cases.
5707 Also, in some cases, the returned ifreq's were being copied to a 'struct
5708 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5709 means the last couple of bytes of an IPv6 address could be chopped if the
5710 ifreq contained only a normal sockaddr (14 bytes storage).
5712 18. Named domain lists were not supported in the hosts_treat_as_local option.
5713 An entry such as +xxxx was not recognized, and was treated as a literal
5716 19. Ensure that header lines added by a DATA ACL are included in the reject log
5717 if the ACL subsequently rejects the message.
5719 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5720 MD5 (which is deprecated).
5722 21. When testing a filter file using -bf, Exim was writing a message when it
5723 took the sender from a "From " line in the message, but it was not doing so
5724 when it took $return_path from a Return-Path: header line. It now does.
5726 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5727 with a valid header line field name (a series of printing characters
5728 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5730 23. Changed "disc" in the source to "disk" to conform to the documentation and
5731 the book and for uniformity.
5733 24. Ignore Sendmail's -Ooption=value command line item.
5735 25. When execve() failed while trying to run a command in a pipe transport,
5736 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5737 could be confused with a return value of 69 from the command itself. This
5738 has been changed to 127, the value the shell returns if it is asked to run
5739 a non-existent command. The wording for the related log line suggests a
5740 non-existent command as the problem.
5742 26. If received_header_text expands to an empty string, do not add a Received:
5743 header line to the message. (Well, it adds a token one on the spool, but
5744 marks it "old" so that it doesn't get used or transmitted.)
5746 27. Installed eximstats 1.28 (addition of -nt option).
5748 28. There was no check for failure on the call to getsockname() in the daemon
5749 code. This can fail if there is a shortage of resources on the system, with
5750 ENOMEM, for example. A temporary error is now given on failure.
5752 29. Contrary to the C standard, it seems that in some environments, the
5753 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5754 program. Exim now does this explicitly; it affects the formatting of
5755 timestamps using strftime().
5757 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5758 complaints. I've now initialized all the variables, to avoid this.
5760 32. Header lines added by a system filter were not being "seen" during
5761 transport-time rewrites.
5763 33. The info_callback() function passed to OpenSSL is set up with type void
5764 (*)(SSL *, int, int), as described somewhere. However, when calling the
5765 function (actually a macro) that sets it up, the type void(*)() is
5766 expected. I've put in a cast to prevent warnings from picky compilers.
5768 34. If a DNS black list lookup found a CNAME record, but there were no A
5769 records associated with the domain it pointed at, Exim crashed.
5771 35. If a DNS black list lookup returned more than one A record, Exim ignored
5772 all but the first. It now scans all returned addresses if a particular IP
5773 value is being sought. In this situation, the contents of the
5774 $dnslist_value variable are a list of all the addresses, separated by a
5777 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5778 to accept a host name and all its aliases if the forward lookup for any of
5779 them yielded the IP address of the incoming connection. Now it accepts only
5780 those names whose forward lookup yields the correct IP address. Any other
5781 names are discarded. This closes a loophole whereby a rogue DNS
5782 administrator could create reverse DNS records to break through a
5783 wildcarded host restriction in an ACL.
5785 37. If a user filter or a system filter that ran in a subprocess used any of
5786 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5787 the wrong values were passed to the pipe command ($thisaddress had the
5788 value of $0, $0 had the value of $1, etc). This bug was introduced by
5789 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5791 38. Improved the line breaking for long SMTP error messages from ACLs.
5792 Previously, if there was no break point between 40 and 75 characters, Exim
5793 left the rest of the message alone. Two changes have been made: (a) I've
5794 reduced the minimum length to 35 characters; (b) if it can't find a break
5795 point between 35 and 75 characters, it looks ahead and uses the first one
5796 that it finds. This may give the occasional overlong line, but at least the
5797 remaining text gets split now.
5799 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5800 file descriptors might be low, and that setting 1000 would always raise it.
5801 It turns out that in some environments, the limit is already over 1000 and
5802 that lowering it causes trouble. So now Exim takes care not to decrease it.
5804 40. When delivering a message, the value of $return_path is set to $sender_
5805 address at the start of routing (routers may change the value). By an
5806 oversight, this default was not being set up when an address was tested by
5807 -bt or -bv, which affected the outcome if any router or filter referred to
5810 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5811 log only when "message" or "log_message" are set. However, if one of the
5812 conditions was an address verification, or a call to a nested ACL, the
5813 messages generated by the underlying test were being passed through. This
5814 no longer happens. The underlying message is available in $acl_verify_
5815 message for both "message" and "log_message" expansions, so it can be
5816 passed through if needed.
5818 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5819 new expansion $bh_ to give the encoded byte string without charset
5820 translation. Translation happens only if iconv() is available; HAVE_ICONV
5821 indicates this at build time. HEADERS_CHARSET gives the charset to
5822 translate to; headers_charset can change it in the configuration, and
5823 "headers charset" can change it in an individual filter file.
5825 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5826 that creates RFC 2047 encoded "words" labels them as that charset instead
5827 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5828 expansion operator; (ii) when Exim creates a From: line for a local
5829 message; (iii) when a header line is rewritten to include a "phrase" part.
5831 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5832 buggy, causing it to skip the first lines of messages whose message ID
5833 ended in 'D'. This would not have bitten before Exim release 4.14, because
5834 message IDs were unlikely to end in 'D' before then. The effect was to have
5835 incorrect size information for certain domains.
5837 45. #include "config.h" was missing at the start of the crypt16.c module. This
5838 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5841 46. If there was a timeout during a "random" callout check, Exim treated it as
5842 a failure of the random address, and carried on sending RSET and the real
5843 address. If the delay was just some slowness somewhere, the response to the
5844 original RCPT would be taken as a response to RSET and so on, causing
5845 mayhem of various kinds.
5847 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5848 when I implemented it. It didn't allow for the fact that some option values
5849 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5850 the right test for positive values.
5852 48. Domain names in DNS records are case-independent. Exim always looks them up
5853 in lower case. Some resolvers return domain names in exactly the case they
5854 appear in the zone file, that is, they may contain uppercase letters. Not
5855 all resolvers do this - some return always lower case. Exim was treating a
5856 change of case by a resolver as a change of domain, similar to a widening
5857 of a domain abbreviation. This triggered its re-routing code and so it was
5858 trying to route what was effectively the same domain again. This normally
5859 caused routing to fail (because the router wouldn't handle the domain
5860 twice). Now Exim checks for this case specially, and just changes the
5861 casing of the domain that it ultimately uses when it transmits the message
5864 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5867 50. If a filter generated a file delivery with a non-absolute name (possible if
5868 no home directory exists for the router), the forbid_file option was not
5871 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5872 the existing equality matching.
5874 52. Exim was using ints instead of ino_t variables in some places where it was
5875 dealing with inode numbers.
5877 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5878 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5879 called TMPDIR, and if it finds it is different, it changes its value.
5881 54. The smtp_printf() function is now made available to local_scan() so
5882 additional output lines can be written before returning. There is also an
5883 smtp_fflush() function to enable the detection of a dropped connection.
5884 The variables smtp_input and smtp_batched_input are exported to
5887 55. Changed the default runtime configuration: the message "Unknown user"
5888 has been removed from the ACL, and instead placed on the localuser router,
5889 using the cannot_route_message feature. This means that any verification
5890 failures that generate their own messages won't get overridden. Similarly,
5891 the "Unrouteable address" message that was in the ACL for unverifiable
5892 relay addresses has also been removed.
5894 56. Added hosts_avoid_esmtp to the smtp transport.
5896 57. The exicyclog script was not checking for the esoteric option
5897 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5898 will work only if exicyclog is run under the appropriate euid.
5900 58. Following a discussion on the list, the rules by which Exim recognises line
5901 endings on incoming messages have been changed. The -dropcr and drop_cr
5902 options are now no-ops, retained only for backwards compatibility. The
5903 following line terminators are recognized: LF CRLF CR. However, special
5904 processing applies to CR:
5906 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5907 nor a local message in the state where . is a terminator.
5909 (ii) If a bare CR is encountered in a header line, an extra space is added
5910 after the line terminator so as not to end the header. The reasoning
5911 behind this is that bare CRs in header lines are most likely either
5912 to be mistakes, or people trying to play silly games.
5914 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5915 was being incorrectly given as 18 bytes larger than it should have been.
5916 This is a VOB (very old bug).
5918 60. This may never have affected anything current, but just in case it has:
5919 When the local host is found other than at the start of a list of hosts,
5920 the local host, those with the same MX, and any that follow, are discarded.
5921 When the list in question was part of a longer list of hosts, the following
5922 hosts (not currently being processed) were also being discarded. This no
5923 longer happens. I'm not sure if this situation could ever has previously
5926 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5929 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5930 header. This is something that is recommended in a new Internet Draft, and
5931 is something that is documented as being done by Sendmail. There are two
5932 possible values. For messages generated by the autoreply transport, Exim
5935 Auto-Submitted: auto-replied
5937 whereas for all other generated messages (e.g. bounces) it adds
5939 Auto-Submitted: auto-generated
5941 63. The "personal" condition in filters now includes a test for the
5942 Auto-Submitted: header. If it contains the string "auto-" the message it
5943 not considered personal.
5945 64. Added rcpt_include_affixes as a generic transport option.
5947 65. Added queue_only_override (default true).
5949 66. Added the syslog_duplication option.
5951 67. If what should have been the first header line of a message consisted of
5952 a space followed by a colon, Exim was mis-interpreting it as a header line.
5953 It isn't of course - it is syntactically invalid and should therefore be
5954 treated as the start of the message body. The misbehaviour could have
5955 caused a number of strange effects, including loss of data in subsequent
5956 header lines, and spool format errors.
5958 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5959 client host had authenticated. This control can now be exercised by an ACL
5960 for more flexibility.
5962 69. By default, callouts do not happen when testing with -bh. There is now a
5963 variant, -bhc, which does actually run the callout code, including
5964 consulting and updating the callout cache.
5966 70. Added support for saslauthd authentication, courtesy of Alexander
5969 71. If statvfs() failed on the spool or log directories while checking their
5970 size for availability, Exim confusingly gave the error "space shortage".
5971 Furthermore, in debugging mode it crashed with a floating point exception.
5972 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5973 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5974 serious problem, Exim now writes to the main and panic logs when this
5975 happens, with details of the failure. It then refuses to accept the
5976 incoming message, giving the message "spool directory problem" or "log
5977 directory problem" with a 421 code for SMTP messages.
5979 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5980 0, 1, and 2 exist, because some OS complain for execs without them (see
5981 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5982 descriptors. However, the code omitted to check that the open succeeded,
5983 causing mysterious errors if for some reason the permissions on /dev/null
5984 got screwed. Now Exim writes a message to the main and panic logs, and
5985 bombs out if it can't open /dev/null.
5987 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5988 interact so that it is all more flexible. It is supposed to remain
5989 backwards compatible. Also added extra_local_interfaces.
5991 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5992 to bomb out with an assertion failure - to the client this appears as a
5993 connection drop. This problem occurs in the part of the code that was taken
5994 from the Samba project. Fortunately, the assertion is in a very simple
5995 function, so I have fixed this by reproducing the function inline in the
5996 one place where it is called, and arranging for authentication to fail
5997 instead of killing the process with assert().
5999 75. The SPA client code was not working when the server requested OEM rather
6000 than Unicode encoding.
6002 76. Added code to make require_files with a specific uid setting more usable in
6003 the case where statting the file as root fails - usually a non-root-mounted
6004 NFS file system. When this happens and the failure is EACCES, Exim now
6005 forks a subprocess and does the per-uid checking as the relevant uid.
6007 77. Added process_log_path.
6009 78. If log_file_path was not explicitly set, a setting of check_log_space or
6010 check_log_inodes was ignored.
6012 79. If a space check for the spool or log partitions fails, the incident is now
6013 logged. Of course, in the latter case the data may get lost...
6015 80. Added the %p formatting code to string_format() so that it can be used to
6016 print addresses in debug_print(). Adjusted all the address printing in the
6017 debugging in store.c to use %p rather than %d.
6019 81. There was a concern that a line of code in smtp_in.c could overflow a
6020 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6021 initially expressed, the concern was not well-founded, because trailing
6022 spaces are removed early. However, if the trailing spaces were followed by
6023 a NULL, they did not get removed, so the overflow was possible. Two fixes
6026 (a) I re-wrote the offending code in a cleaner fashion.
6027 (b) If an incoming SMTP command contains a NULL character, it is rejected
6030 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6031 runs initgroups(), so that if the Exim user is in any additional groups,
6032 they will be used during message reception.
6038 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob