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. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1::string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
130 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
132 PP/02 Make -n do something, by making it not do something.
133 When combined with -bP, the name of an option is not output.
135 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
138 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
139 $sender_host_name and config options to manage this, and basic check
142 PP/05 DSCP support for outbound connections and control modifier for inbound.
144 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
145 (Only plugin which currently uses this is kerberos4, which nobody should
146 be using, but we should make it available and other future plugins might
147 conceivably use it, even though it would break NAT; stuff *should* be
148 using channel bindings instead).
150 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
151 name; added for Sendmail compatibility; requires admin caller.
152 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
153 just ignore it); requires trusted caller.
154 Also parse but ignore: -Ac -Am -X<logfile>
157 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
159 TL/02 Add +smtp_confirmation as a default logging option.
161 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
162 Patch by Magnus Holmgren from 2007-02-20.
164 TL/04 Bugzilla 1281 - Spec typo.
165 Bugzilla 1283 - Spec typo.
166 Bugzilla 1290 - Spec grammar fixes.
168 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
170 TL/06 Add Experimental DMARC support using libopendmarc libraries.
172 TL/07 Fix an out of order global option causing a segfault. Reported to dev
173 mailing list by by Dmitry Isaikin.
175 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
177 JH/02 Support "G" suffix to numbers in ${if comparisons.
179 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
181 NM/01 Bugzilla 1197 - Spec typo
182 Bugzilla 1196 - Spec examples corrections
184 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
186 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
187 gnutls_enable_pkcs11, but renamed to more accurately indicate its
190 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
191 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
193 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
194 "acl {{name}{arg}...}", and optional args on acl condition
197 JH/05 Permit multiple router/transport headers_add/remove lines.
199 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
201 JH/07 Avoid using a waiting database for a single-message-only transport.
202 Performance patch from Paul Fisher. Bugzilla 1262.
204 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
207 JH/09 Add $headers_added variable, with content from use of ACL modifier
208 add_header (but not yet added to the message). Bugzilla 199.
210 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
211 Pulled from Bugzilla 817 by Wolfgang Breyha.
213 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
215 (nb: this is the same fix as in Exim 4.80.1)
217 JH/11 Add A= logging on delivery lines, and a client_set_id option on
220 JH/12 Add optional authenticated_sender logging to A= and a log_selector
223 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
225 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
226 advertise SMTP AUTH mechanism to us, instead of a generic
227 protocol violation error. Also, make Exim more robust to bad
228 data from the Dovecot auth socket.
230 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
232 When a queue runner is handling a message, Exim first routes the
233 recipient addresses, during which it prunes them based on the retry
234 hints database. After that it attempts to deliver the message to
235 any remaining recipients. It then updates the hints database using
238 So if a recipient address works intermittently, it can get repeatedly
239 deferred at routing time. The retry hints record remains fresh so the
240 address never reaches the final cutoff time.
242 This is a fairly common occurrence when a user is bumping up against
243 their storage quota. Exim had some logic in its local delivery code
244 to deal with this. However it did not apply to per-recipient defers
245 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
247 This change adds a proper retry rule check during routing so that the
248 final cutoff time is checked against the message's age. We only do
249 this check if there is an address retry record and there is not a
250 domain retry record; this implies that previous attempts to handle
251 the address had the retry_use_local_parts option turned on. We use
252 this as an approximation for the destination being like a local
253 delivery, as in LMTP.
255 I suspect this new check makes the old local delivery cutoff check
256 redundant, but I have not verified this so I left the code in place.
258 TF/02 Correct gecos expansion when From: is a prefix of the username.
260 Test 0254 submits a message to Exim with the header
264 When I ran the test suite under the user fanf2, Exim expanded
265 the header to contain my full name, whereas it should have added
266 a Resent-Sender: header. It erroneously treats any prefix of the
267 username as equal to the username.
269 This change corrects that bug.
271 GF/01 DCC debug and logging tidyup
272 Error conditions log to paniclog rather than rejectlog.
273 Debug lines prefixed by "DCC: " to remove any ambiguity.
275 TF/03 Avoid unnecessary rebuilds of lookup-related code.
277 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
278 Bug spotted by Jeremy Harris; was flawed since initial commit.
279 Would have resulted in OCSP responses post-SNI triggering an Exim
280 NULL dereference and crash.
282 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
284 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
285 Bug detection, analysis and fix by Samuel Thibault.
286 Bugzilla 1331, Debian bug #698092.
288 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
290 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
291 Server implementation by Todd Lyons, client by JH.
292 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
293 config variable "prdr_enable" controls whether the server
294 advertises the facility. If the client requests PRDR a new
295 acl_data_smtp_prdr ACL is called once for each recipient, after
296 the body content is received and before the acl_smtp_data ACL.
297 The client is controlled by bolth of: a hosts_try_prdr option
298 on the smtp transport, and the server advertisement.
299 Default client logging of deliveries and rejections involving
300 PRDR are flagged with the string "PRDR".
302 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
303 fclose(). Diagnosis by Todd Lyons.
305 PP/17 Update configure.default to handle IPv6 localhost better.
306 Patch by Alain Williams (plus minor tweaks).
309 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
310 This is now consistent with GnuTLS, and is now documented: the
311 previous undocumented portable approach to treating the option as
312 unset was to force an expansion failure. That still works, and
313 an empty string is now equivalent.
315 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
316 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
317 not performing validation itself.
319 PP/20 Added force_command boolean option to pipe transport.
320 Patch from Nick Koston, of cPanel Inc.
322 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
325 TF/04 Added udpsend ACL modifer and hexquote expansion operator
327 PP/21 Fix eximon continuous updating with timestamped log-files.
328 Broken in a format-string cleanup in 4.80, missed when I repaired the
329 other false fix of the same issue.
330 Report and fix from Heiko Schlichting.
333 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
334 Report from Prashanth Katuri.
336 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
337 It's SecureTransport, so affects any MacOS clients which use the
338 system-integrated TLS libraries, including email clients.
340 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
341 using a MIME ACL for non-SMTP local injection.
342 Report and assistance in diagnosis by Warren Baker.
344 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
346 JH/16 Fix comparisons for 64b. Bugzilla 1385.
348 TL/09 Add expansion variable $authenticated_fail_id to keep track of
349 last id that failed so it may be referenced in subsequent ACL's.
351 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
354 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
355 ldap library initialization, allowing self-signed CA's to be
356 used. Also properly sets require_cert option later in code by
357 using NULL (global ldap config) instead of ldap handle (per
358 session). Bug diagnosis and testing by alxgomz.
360 TL/12 Enhanced documentation in the ratelimit.pl script provided in
361 the src/util/ subdirectory.
363 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
364 renamed to Transport Post Delivery Action by Jeremy Harris, as
367 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
368 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
369 redis_servers = needs to be configured which will be used by the redis
370 lookup. Patch from Warren Baker, of The Packet Hub.
372 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
374 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
375 hostname or reverse DNS when processing a host list. Used suggestions
376 from multiple comments on this bug.
378 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
380 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
381 Missed a few lines, added it to make the runtest require no keyboard
384 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
385 contains upper case chars. Make router use caseful_local_part.
387 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
388 support when GnuTLS has been built with p11-kit.
394 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
396 This, or similar/improved, will also be change PP/11 of 4.82.
402 PP/01 Handle short writes when writing local log-files.
403 In practice, only affects FreeBSD (8 onwards).
404 Bugzilla 1053, with thanks to Dmitry Isaikin.
406 NM/01 Bugzilla 949 - Documentation tweak
408 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
411 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
413 PP/02 Implemented gsasl authenticator.
415 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
417 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
418 `pkg-config foo` for cflags/libs.
420 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
421 with rest of GSASL and with heimdal_gssapi.
423 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
424 `pkg-config foo` for cflags/libs for the TLS implementation.
426 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
427 properties get this fed in as external SSF. A number of robustness
428 and debugging improvements to the cyrus_sasl authenticator.
430 PP/08 cyrus_sasl server now expands the server_realm option.
432 PP/09 Bugzilla 1214 - Log authentication information in reject log.
433 Patch by Jeremy Harris.
435 PP/10 Added dbmjz lookup type.
437 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
439 PP/12 MAIL args handles TAB as well as SP, for better interop with
440 non-compliant senders.
441 Analysis and variant patch by Todd Lyons.
443 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
444 Bug report from Lars Müller <lars@samba.org> (via SUSE),
445 Patch from Dirk Mueller <dmueller@suse.com>
447 PP/13 tls_peerdn now print-escaped for spool files.
448 Observed some $tls_peerdn in wild which contained \n, which resulted
449 in spool file corruption.
451 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
452 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
453 or write after TLS renegotiation, which otherwise led to messages
456 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
457 as a tracking header (ie: a signed header comes before the signature).
458 Patch from Wolfgang Breyha.
460 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
461 comma-sep list; embedded commas doubled.
463 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
465 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
467 Report and patch from Dmitry Banschikov.
469 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
470 Removed SSL_clear() after SSL_new() which led to protocol negotiation
471 failures. We appear to now support TLS1.1+ with Exim.
473 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
474 lets Exim select keys and certificates based upon TLS SNI from client.
475 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
476 before an outbound SMTP session. New log_selector, +tls_sni.
478 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
479 NULL dereference. Report and patch from Alun Jones.
481 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
482 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
483 Not seeing resolver debug output on NetBSD, but suspect this is a
484 resolver implementation change.
486 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
487 Left warnings. Added "eximon gdb" invocation mode.
489 PP/21 Defaulting "accept_8bitmime" to true, not false.
491 PP/22 Added -bw for inetd wait mode support.
493 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
494 locate the relevant includes and libraries. Made this the default.
496 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
497 Bugzilla 1246, report and most of solution from Tomasz Kusy.
499 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
500 This may cause build issues on older platforms.
502 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
503 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
504 gnutls_require_mac & gnutls_require_protocols (no longer supported).
505 Added SNI support via GnuTLS too.
506 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
508 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
510 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
511 Applied second patch from Janne, implementing suggestion to default
512 multiple-strings-in-record handling to match SPF spec.
514 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
516 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
517 read-only, out of scope).
518 Patch from Wolfgang Breyha, report from Stuart Northfield.
520 PP/29 Fix three issues highlighted by clang analyser static analysis.
521 Only crash-plausible issue would require the Cambridge-specific
522 iplookup router and a misconfiguration.
523 Report from Marcin Mirosław.
525 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
527 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
528 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
529 As part of this, removing so much warning spew let me fix some minor
530 real issues in debug logging.
532 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
533 assignment on my part. Fixed.
535 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
536 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
537 Janne Snabb (who went above and beyond: thank you).
539 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
540 string otherwise requires a connection and a bunch more work and it's
541 relatively easy to get wrong. Should also expose TLS library linkage
544 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
545 64-bit ${eval} (JH/03).
547 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
548 GNU libc to support some of the 64-bit stuff, should not lead to
549 conflicts. Defined before os.h is pulled in, so if a given platform
550 needs to override this, it can.
552 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
553 protection layer was required, which is not implemented.
554 Bugzilla 1254, patch from Wolfgang Breyha.
556 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
557 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
558 tls_dhparam take prime identifiers. Also unbreak combination of
559 OpenSSL+DH_params+TLSSNI.
561 PP/39 Disable SSLv2 by default in OpenSSL support.
567 PP/01 Solaris build fix for Oracle's LDAP libraries.
568 Bugzilla 1109, patch from Stephen Usher.
570 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
572 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
575 TF/02 Fix a couple more cases where we did not log the error message
576 when unlink() failed. See also change 4.74-TF/03.
578 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
579 lock up or crash if it happened to be inside a call to libc when it
580 got a SIGUSR1 from exiwhat.
582 The SIGUSR1 handler appends the current process status to the process
583 log which is later printed by exiwhat. It used to use the general
584 purpose logging code to do this, but several functions it calls are
585 not safe for signals.
587 The new output code in the SIGUSR1 handler is specific to the process
588 log, and simple enough that it's easy to inspect for signal safety.
589 Removing some special cases also simplifies the general logging code.
590 Removing the spurious timestamps from the process log simplifies
593 TF/04 Improved ratelimit ACL condition.
595 The /noupdate option has been deprecated in favour of /readonly which
596 has clearer semantics. The /leaky, /strict, and /readonly update modes
597 are mutually exclusive. The update mode is no longer included in the
598 database key; it just determines when the database is updated. (This
599 means that when you upgrde Exim will forget old rate measurements.)
601 Exim now checks that the per_* options are used with an update mode that
602 makes sense for the current ACL. For example, when Exim is processing a
603 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
604 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
605 must specify per_mail/readonly. If you omit the update mode it defaults to
606 /leaky where that makes sense (as before) or /readonly where required.
608 The /noupdate option is now undocumented but still supported for
609 backwards compatibility. It is equivalent to /readonly except that in
610 ACLs where /readonly is required you may specify /leaky/noupdate or
611 /strict/noupdate which are treated the same as /readonly.
613 A useful new feature is the /count= option. This is a generalization
614 of the per_byte option, so that you can measure the throughput of other
615 aggregate values. For example, the per_byte option is now equivalent
616 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
618 The per_rcpt option has been generalized using the /count= mechanism
619 (though it's more complicated than the per_byte equivalence). When it is
620 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
621 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
622 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
623 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
624 Note that using per_rcpt with a non-readonly update mode in more than
625 one ACL will cause the recipients to be double-counted. (The per_mail
626 and per_byte options don't have this problem.)
628 The handling of very low rates has changed slightly. If the computed rate
629 is less than the event's count (usually one) then this event is the first
630 after a long gap. In this case the rate is set to the same as this event's
631 count, so that the first message of a spam run is counted properly.
633 The major new feature is a mechanism for counting the rate of unique
634 events. The new per_addr option counts the number of different
635 recipients that someone has sent messages to in the last time period. It
636 behaves like per_rcpt if all the recipient addresses are different, but
637 duplicate recipient addresses do not increase the measured rate. Like
638 the /count= option this is a general mechanism, so the per_addr option
639 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
640 example, measure the rate that a client uses different sender addresses
641 with the options per_mail/unique=$sender_address. There are further
642 details in the main documentation.
644 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
646 TF/06 Removed a few PCRE remnants.
648 TF/07 Automatically extract Exim's version number from tags in the git
649 repository when doing development or release builds.
651 PP/02 Raise smtp_cmd_buffer_size to 16kB.
652 Bugzilla 879. Patch from Paul Fisher.
654 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
655 Heavily based on revision 40f9a89a from Simon Arlott's tree.
658 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
660 PP/05 Variable $av_failed, true if the AV scanner deferred.
661 Bugzilla 1078. Patch from John Horne.
663 PP/06 Stop make process more reliably on build failure.
664 Bugzilla 1087. Patch from Heiko Schlittermann.
666 PP/07 Make maildir_use_size_file an _expandable_ boolean.
667 Bugzilla 1089. Patch from Heiko Schlittermann.
669 PP/08 Handle ${run} returning more data than OS pipe buffer size.
670 Bugzilla 1131. Patch from Holger Weiß.
672 PP/09 Handle IPv6 addresses with SPF.
673 Bugzilla 860. Patch from Wolfgang Breyha.
675 PP/10 GnuTLS: support TLS 1.2 & 1.1.
677 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
680 PP/11 match_* no longer expand right-hand-side by default.
681 New compile-time build option, EXPAND_LISTMATCH_RHS.
682 New expansion conditions, "inlist", "inlisti".
684 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
686 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
688 PP/14 fix log_write() format string regression from TF/03.
689 Bugzilla 1152. Patch from Dmitry Isaikin.
695 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
697 PP/02 Harmonised TLS library version reporting; only show if debugging.
698 Layout now matches that introduced for other libraries in 4.74 PP/03.
700 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
702 PP/04 New "dns_use_edns0" global option.
704 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
707 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
708 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
710 TK/01 Updated PolarSSL code to 0.14.2.
711 Bugzilla 1097. Patch from Andreas Metzler.
713 PP/07 Catch divide-by-zero in ${eval:...}.
716 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
719 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
720 format-string attack -- SECURITY: remote arbitrary code execution.
722 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
723 time unintentionally subject to list matching rules, letting the header
724 cause arbitrary Exim lookups (of items which can occur in lists, *not*
725 arbitrary string expansion). This allowed for information disclosure.
727 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
728 INT_MIN/-1 -- value coerced to INT_MAX.
734 NM/01 Workround for PCRE version dependency in version reporting
737 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
738 This fixes portability to compilers other than gcc, notably
739 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
741 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
742 makefiles for portability to HP-UX and POSIX correctness.
744 PP/01 Permit LOOKUP_foo enabling on the make command-line.
745 Also via indented variable definition in the Makefile.
746 (Debugging by Oliver Heesakkers).
748 PP/02 Restore caching of spamd results with expanded spamd_address.
749 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
751 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
752 Improves build reliability. Fix from: Frank Elsner
754 NM/02 Fix wide character breakage in the rfc2047 coding
755 Fixes bug 1064. Patch from Andrey N. Oktyabrski
757 NM/03 Allow underscore in dnslist lookups
758 Fixes bug 1026. Patch from Graeme Fowler
760 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
761 Code patches from Adam Ciarcinski of NetBSD.
763 NM/04 Fixed exiqgrep to cope with mailq missing size issue
766 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
767 is logged, to avoid truncation. Patch from John Horne.
769 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
770 Patch from Jakob Hirsch.
772 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
773 SQL string expansion failure details.
774 Patch from Andrey Oktyabrski.
776 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
777 Patch from Simon Arlott.
779 PP/09 New lookups functionality failed to compile on old gcc which rejects
780 extern declarations in function scope.
781 Patch from Oliver Fleischmann
783 PP/10 Use sig_atomic_t for flags set from signal handlers.
784 Check getgroups() return and improve debugging.
785 Fixed developed for diagnosis in bug 927 (which turned out to be
788 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
789 Patch from Mark Zealey.
791 PP/12 Bugzilla 1056: Improved spamd server selection.
792 Patch from Mark Zealey.
794 PP/13 Bugzilla 1086: Deal with maildir quota file races.
795 Based on patch from Heiko Schlittermann.
797 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
798 Patch from Uwe Doering, sign-off by Michael Haardt.
800 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
801 variable declaration deep within a block. Bug and patch from
804 PP/15 lookups-Makefile IRIX compatibilty coercion.
806 PP/16 Make DISABLE_DKIM build knob functional.
808 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
809 Patch by Simon Arlott
811 TF/03 Fix valgrind.h portability to C89 compilers that do not support
812 variable argument macros. Our copy now differs from upstream.
818 TF/01 Failure to get a lock on a hints database can have serious
819 consequences so log it to the panic log.
821 TF/02 Log LMTP confirmation messages in the same way as SMTP,
822 controlled using the smtp_confirmation log selector.
824 TF/03 Include the error message when we fail to unlink a spool file.
826 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
827 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
828 for maintaining out-of-tree patches for some time.
830 PP/01 Bugzilla 139: Documentation and portability issues.
831 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
832 Handle per-OS dynamic-module compilation flags.
834 PP/02 Let /dev/null have normal permissions.
835 The 4.73 fixes were a little too stringent and complained about the
836 permissions on /dev/null. Exempt it from some checks.
837 Reported by Andreas M. Kirchwitz.
839 PP/03 Report version information for many libraries, including
840 Exim version information for dynamically loaded libraries. Created
841 version.h, now support a version extension string for distributors
842 who patch heavily. Dynamic module ABI change.
844 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
845 privilege escalation vulnerability whereby the Exim run-time user
846 can cause root to append content of the attacker's choosing to
849 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
852 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
853 If dropping privileges for untrusted macros, we disabled normal logging
854 on the basis that it would fail; for the Exim run-time user, this is not
855 the case, and it resulted in successful deliveries going unlogged.
856 Fixed. Reported by Andreas Metzler.
862 PP/01 Date: & Message-Id: revert to normally being appended to a message,
863 only prepend for the Resent-* case. Fixes regression introduced in
864 Exim 4.70 by NM/22 for Bugzilla 607.
866 PP/02 Include check_rfc2047_length in configure.default because we're seeing
867 increasing numbers of administrators be bitten by this.
869 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
871 PP/03 Bugzilla 994: added openssl_options main configuration option.
873 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
875 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
877 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
879 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
880 without a peer certificate, leading to a segfault because of an
881 assumption that peers always have certificates. Be a little more
882 paranoid. Problem reported by Martin Tscholak.
884 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
885 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
886 NB: ClamAV planning to remove STREAM in "middle of 2010".
887 CL also introduces -bmalware, various -d+acl logging additions and
888 more caution in buffer sizes.
890 PP/09 Implemented reverse_ip expansion operator.
892 PP/10 Bugzilla 937: provide a "debug" ACL control.
894 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
896 PP/12 Bugzilla 973: Implement --version.
898 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
900 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
902 PP/15 Bugzilla 816: support multiple condition rules on Routers.
904 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
905 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
906 ignore trailing whitespace.
908 JJ/02 prevent non-panic DKIM error from being sent to paniclog
910 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
913 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
914 Notification from Dr Andrew Aitchison.
916 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
917 ExtendedDetectionInfo response format.
918 Notification from John Horne.
920 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
923 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
924 XSL and documented dependency on system catalogs, with examples of how
927 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
930 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
931 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
932 configuration file which is writeable by the Exim user or group.
934 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
935 of configuration files to cover files specified with the -C option if
936 they are going to be used with root privileges, not just the default
939 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
940 option (effectively making it always true).
942 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
943 files to be used while preserving root privileges.
945 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
946 that rogue child processes cannot use them.
948 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
949 run-time user, instead of root.
951 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
952 Exim run-time user without dropping privileges.
954 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
955 result string, instead of calling string_vformat() twice with the same
958 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
959 for other users. Others should always drop root privileges if they use
960 -C on the command line, even for a whitelisted configure file.
962 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
964 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
970 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
971 $data_path, and $header_path variables; fixed documentation bugs and
974 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
975 exipick to access non-standard spools, including the "frozen" queue
978 NM/01 Bugzilla 965: Support mysql stored procedures.
979 Patch from Alain Williams
981 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
983 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
984 Patch from Andreas Metzler
986 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
987 Patch from Kirill Miazine
989 NM/05 Bugzilla 671: Added umask to procmail example.
991 JJ/03 installed exipick 20100323.0, fixing doc bug
993 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
994 directory. Notification and patch from Dan Rosenberg.
996 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
998 TK/02 Improve log output when DKIM signing operation fails.
1000 MH/01 Treat the transport option dkim_domain as a colon separated
1001 list, not as a single string, and sign the message with each element,
1002 omitting multiple occurences of the same signer.
1004 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1005 Bugzilla 985, 986. Patch by Simon Arlott
1007 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1008 Patch by Simon Arlott
1010 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1011 MBX locking. Notification from Dan Rosenberg.
1017 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1019 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1021 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1023 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1025 NM/04 Bugzilla 915: Flag broken perl installation during build.
1031 TK/01 Added patch by Johannes Berg that expands the main option
1032 "spamd_address" if it starts with a dollar sign.
1034 TK/02 Write list of recipients to X-Envelope-Sender header when building
1035 the mbox-format spool file for content scanning (suggested by Jakob
1038 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1039 (http://www.dcc-servers.net/) support via dccifd. Activated by
1040 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1042 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1043 by Mark Daniel Reidel <mr@df.eu>.
1045 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1046 When building exim an external PCRE library is now needed -
1047 PCRE is a system library on the majority of modern systems.
1048 See entry on PCRE_LIBS in EDITME file.
1050 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1051 conversation. Added nologin parameter to request.
1052 Patch contributed by Kirill Miazine.
1054 TF/01 Do not log submission mode rewrites if they do not change the address.
1056 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1058 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1059 log files in place. Contributed by Roberto Lima.
1061 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1063 TF/03 Bugzilla 615: When checking the local_parts router precondition
1064 after a local_part_suffix or local_part_prefix option, Exim now
1065 does not use the address's named list lookup cache, since this
1066 contains cached lookups for the whole local part.
1068 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1069 Robert Millan. Documentation is in experimental-spec.txt.
1071 TF/04 Bugzilla 668: Fix parallel build (make -j).
1073 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1075 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1076 Patch provided by Jan Srzednicki.
1078 TF/05 Leading white space used to be stripped from $spam_report which
1079 wrecked the formatting. Now it is preserved.
1081 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1082 that they are available at delivery time.
1084 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1086 TF/08 TLS error reporting now respects the incoming_interface and
1087 incoming_port log selectors.
1089 TF/09 Produce a more useful error message if an SMTP transport's hosts
1090 setting expands to an empty string.
1092 NM/06 Bugzilla 744: EXPN did not work under TLS.
1093 Patch provided by Phil Pennock.
1095 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1096 Patch provided by Richard Godbee.
1098 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1099 acl_smtp_notquit, added index entry.
1101 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1102 Patch provided by Eugene Bujak.
1104 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1105 accept(). Patch provided by Maxim Dounin.
1107 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1108 Patch provided by Phil Pennock.
1110 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1112 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1113 Patch provided by Brad "anomie" Jorsch.
1115 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1116 Patch provided by Dean Brooks.
1118 TK/05 Add native DKIM support (does not depend on external libraries).
1120 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1121 Patch provided by Graeme Fowler.
1123 NM/16 Bugzilla 851: Documentation example syntax fix.
1125 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1127 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1130 NM/19 Bugzilla 745: TLS version reporting.
1131 Patch provided by Phil Pennock.
1133 NM/20 Bugzilla 167: bool: condition support.
1134 Patch provided by Phil Pennock.
1136 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1137 clients. Patch provided by Phil Pennock.
1139 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1140 Patch provided by Brad "anomie" Jorsch.
1142 NM/23 Bugzilla 687: Fix misparses in eximstats.
1143 Patch provided by Heiko Schlittermann.
1145 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1146 Patch provided by Heiko Schlittermann.
1148 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1149 plus update to original patch.
1151 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1153 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1154 Patch provided by David Brownlee.
1156 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1158 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1160 NM/30 Bugzilla 888: TLS documentation bugfixes.
1162 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1164 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1165 Unlike the original bugzilla I have changed all shell scripts in src tree.
1167 NM/33 Bugzilla 898: Transport filter timeout fix.
1168 Patch by Todd Rinaldo.
1170 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1171 Patch by Serge Demonchaux.
1173 NM/35 Bugzilla 39: Base64 decode bug fixes.
1174 Patch by Jakob Hirsch.
1176 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1178 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1180 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1182 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1188 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1189 ALT-N's libdkim that I have put here:
1190 http://duncanthrax.net/exim-experimental/
1192 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1193 were called 'true' and it seems that C99 defines that as a reserved
1194 keyword to be used with 'bool' variable types. That means you could
1195 not include C99-style headers which use bools without triggering
1196 build errors in sieve.c.
1198 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1199 as mailq or other aliases. Changed the --help handling significantly
1200 to do whats expected. exim_usage() emits usage/help information.
1202 SC/01 Added the -bylocaldomain option to eximstats.
1204 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1206 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1208 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1214 PH/01 Another patch from the Sieve maintainer.
1216 PH/02 When an IPv6 address is converted to a string for single-key lookup
1217 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1218 dots are used instead of colons so that keys in lsearch files need not
1219 contain colons. This was done some time before quoting was made available
1220 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1221 (notated using the quote facility) so as to distinguish them from IPv4
1222 keys. This meant that lookups for IP addresses in host lists did not work
1223 for iplsearch lookups.
1225 This has been fixed by arranging for IPv6 addresses to be expressed with
1226 colons if the lookup type is iplsearch. This is not incompatible, because
1227 previously such lookups could never work.
1229 The situation is now rather anomolous, since one *can* have colons in
1230 ordinary lsearch keys. However, making the change in all cases is
1231 incompatible and would probably break a number of configurations.
1233 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1236 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1237 conversion specification without a maximum field width, thereby enabling
1238 a rogue spamd server to cause a buffer overflow. While nobody in their
1239 right mind would setup Exim to query an untrusted spamd server, an
1240 attacker that gains access to a server running spamd could potentially
1241 exploit this vulnerability to run arbitrary code as the Exim user.
1243 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1244 $primary_hostname instead of what libspf2 thinks the hosts name is.
1246 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1247 a directory entry by the name of the lookup key. Previously, if a
1248 symlink pointed to a non-existing file or a file in a directory that
1249 Exim lacked permissions to read, a lookup for a key matching that
1250 symlink would fail. Now it is enough that a matching directory entry
1251 exists, symlink or not. (Bugzilla 503.)
1253 PH/03 The body_linecount and body_zerocount variables are now exported in the
1256 PH/04 Added the $dnslist_matched variable.
1258 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1259 This means they are set thereafter only if the connection becomes
1262 PH/06 Added the client_condition to authenticators so that some can be skipped
1263 by clients under certain conditions.
1265 PH/07 The error message for a badly-placed control=no_multiline_responses left
1266 "_responses" off the end of the name.
1268 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1270 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1271 (without spaces) instead of just copying the configuration text.
1273 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1275 PH/11 Added $max_received_linelength.
1277 PH/12 Added +ignore_defer and +include_defer to host lists.
1279 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1280 way in which PCRE > 7.0 is built.
1282 PH/14 Implemented queue_only_load_latch.
1284 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1285 MAIL command. The effect was to mangle the value on 64-bit systems.
1287 PH/16 Another patch from the Sieve maintainer.
1289 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1291 PH/18 If a system quota error occurred while trying to create the file for
1292 a maildir delivery, the message "Mailbox is full" was not appended to the
1293 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1294 only to a quota excession during the actual writing of the file.
1296 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1297 characters?) which causes problems in log lines. The DN values are now
1298 passed through string_printing() before being added to log lines.
1300 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1301 and InterBase are left for another time.)
1303 PH/21 Added message_body_newlines option.
1305 PH/22 Guard against possible overflow in moan_check_errorcopy().
1307 PH/23 POSIX allows open() to be a macro; guard against that.
1309 PH/24 If the recipient of an error message contained an @ in the local part
1310 (suitably quoted, of course), incorrect values were put in $domain and
1311 $local_part during the evaluation of errors_copy.
1317 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1318 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1321 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1322 issue a MAIL command.
1324 PH/02 In an ACL statement such as
1326 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1328 if a client was not listed at all, or was listed with a value other than
1329 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1330 the condition was not true (as it should be), so access was not denied.
1331 The bug was that the ! inversion was incorrectly passed on to the second
1332 item. This has been fixed.
1334 PH/03 Added additional dnslists conditions == and =& which are different from
1335 = and & when the dns lookup returns more than one IP address.
1337 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1338 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1340 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1341 FSYNC, which compiles an option called disable_fsync that allows for
1342 bypassing fsync(). The documentation is heavily laced with warnings.
1344 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1346 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1347 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1348 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1349 including adding "make clean"; (3) Added -fPIC when compiling the test
1350 dynamically loaded module, to get rid of a warning.
1352 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1353 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1354 = 0s. The bug is otherwise harmless.
1356 PH/07 There was a bug in the dovecot authenticator such that the value of
1357 $auth1 could be overwritten, and so not correctly preserved, after a
1358 successful authentication. This usually meant that the value preserved by
1359 the server_setid option was incorrect.
1361 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1363 PH/09 Installed PCRE release 7.0.
1365 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1366 run for batched SMTP input. It is now run at the start of every message
1367 in the batch. While fixing this I discovered that the process information
1368 (output by running exiwhat) was not always getting set for -bs and -bS
1369 input. This is fixed, and it now also says "batched" for BSMTP.
1371 PH/11 Added control=no_pipelining.
1373 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1374 patch, slightly modified), and move the expansion of helo_data till after
1375 the connection is made in the smtp transport (so it can use these
1378 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1380 PH/14 Added log_selector = +pid.
1382 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1384 PH/16 Add ${if forany and ${if forall.
1386 PH/17 Added dsn_from option to vary the From: line in DSNs.
1388 PH/18 Flush SMTP output before performing a callout, unless control =
1389 no_callout_flush is set.
1391 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1392 was true (the default) a successful delivery failed to delete the retry
1393 item, thus causing premature timeout of the address. The bug is now
1396 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1398 PH/21 Long custom messages for fakedefer and fakereject are now split up
1399 into multiline reponses in the same way that messages for "deny" and
1400 other ACL rejections are.
1402 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1403 with slight modification.
1405 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1406 draft, changing the syntax and factoring some duplicate code".
1408 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1409 for deliveries of the second and subsequent messages over the same SMTP
1412 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1413 ${reduce, with only minor "tidies".
1415 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1417 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1418 expansion side effects.
1420 PH/27 When a message times out after an over-quota error from an Exim-imposed
1421 quota, the bounce message says "mailbox is full". This message was not
1422 being given when it was a system quota that was exceeded. It now should
1425 MH/03 Made $recipients available in local_scan(). local_scan() already has
1426 better access to the recipient list through recipients_list[], but
1427 $recipients can be useful in postmaster-provided expansion strings.
1429 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1430 in the case of a MAIL command with additional options following the
1431 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1432 were accidentally chopped off.
1434 PH/29 SMTP synchronization checks are implemented when a command is read -
1435 there is a check that no more input is waiting when there shouldn't be
1436 any. However, for some commands, a delay in an ACL can mean that it is
1437 some time before the response is written. In this time, more input might
1438 arrive, invalidly. So now there are extra checks after an ACL has run for
1439 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1440 pipelining has not been advertised.
1442 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1444 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1445 returned in the SMTP response when smtp_return_error_details was false.
1446 This has been fixed.
1448 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1449 instead of the C I/O that was originally supplied, because problems were
1450 reported on Solaris.
1452 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1453 Exim which did not show up earlier: it was assuming that a call to
1454 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1455 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1456 was a macro that became an assignment, so it seemed to work. This has
1457 changed to a proper function call with a void return, hence the compile
1458 error. Exim's code has been fixed.
1460 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1463 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1465 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1467 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1468 because it turns out to be too big or there is a timeout) there is no
1469 "Completed" line in the log. When some messages of this type were
1470 selected by exigrep, they were listed as "not completed". Others were
1471 picked up by some special patterns. I have improved the selection
1472 criteria to be more general.
1474 PH/38 The host_find_failed option in the manualroute router can now be set
1475 to "ignore", to completely ignore a host whose IP address cannot be
1476 found. If all hosts are ignored, the behaviour is controlled by the new
1477 host_all_ignored option.
1479 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1480 homing or because of multiple MX records with /mx) generated more than
1481 one IP address, and the following item turned out to be the local host,
1482 all the secondary addresses of the first item were incorrectly removed
1483 from the list, along with the local host and any following hosts (which
1484 is what is supposed to happen).
1486 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1487 whoever called Exim into the -H file. In the case of the daemon it was
1488 behaving confusingly. When first started, it used values for whoever
1489 started the daemon, but after a SIGHUP it used the Exim user (because it
1490 calls itself on a restart). I have changed the code so that it now always
1493 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1494 message are rejected with the same error (e.g. no authentication or bad
1495 sender address), and a DATA command is nevertheless sent (as can happen
1496 with PIPELINING or a stupid MUA), the error message that was given to the
1497 RCPT commands is included in the rejection of the DATA command. This is
1498 intended to be helpful for MUAs that show only the final error to their
1501 PH/42 Another patch from the Sieve maintainer.
1503 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1504 Eximstats - Fixed some broken HTML links and added missing column headers
1506 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1507 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1509 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1511 PH/43 Yet another patch from the Sieve maintainer.
1513 PH/44 I found a way to check for a TCP/IP connection going away before sending
1514 the response to the final '.' that terminates a message, but only in the
1515 case where the client has not sent further data following the '.'
1516 (unfortunately, this is allowed). However, in many cases there won't be
1517 any further data because there won't be any more messages to send. A call
1518 to select() can be used: if it shows that the input is "ready", there is
1519 either input waiting, or the socket has been closed. An attempt to read
1520 the next input character can distinguish the two cases. Previously, Exim
1521 would have sent an OK response which the client would never have see.
1522 This could lead to message repetition. This fix should cure that, at
1523 least in a lot of common cases.
1525 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1526 advertised in response to EHLO.
1532 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1533 fixed by 4.65/MH/01 (is this a record?) are fixed:
1535 (i) An empty string was always treated as zero by the numeric comparison
1536 operators. This behaviour has been restored.
1538 (ii) It is documented that the numeric comparison operators always treat
1539 their arguments as decimal numbers. This was broken in that numbers
1540 starting with 0 were being interpreted as octal.
1542 While fixing these problems I realized that there was another issue that
1543 hadn't been noticed. Values of message_size_limit (both the global option
1544 and the transport option) were treated as octal if they started with 0.
1545 The documentation was vague. These values are now always treated as
1546 decimal, and I will make that clear in the documentation.
1552 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1553 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1556 MH/01 Don't check that the operands of numeric comparison operators are
1557 integers when their expansion is in "skipping" mode (fixes bug
1558 introduced by 4.64-PH/07).
1560 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1561 child addresses, Exim now panics and dies. Previously, because the count
1562 is held in a short int, deliveries were likely to be lost. As such a
1563 large number of recipients for a single message is ridiculous
1564 (performance will be very, very poor), I have chosen to impose a limit
1565 rather than extend the field.
1571 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1572 leftover -K file (the existence of which was triggered by #402).
1573 While we were at it, introduced process PID as part of the -K
1574 filename. This should rule out race conditions when creating
1577 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1578 processing considerably. Previous code took too long for large mails,
1579 triggering a timeout which in turn triggers #401.
1581 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1582 in the DK code in transports.c. sendfile() is not really portable,
1583 hence the _LINUX specificness.
1585 TF/01 In the add_headers option to the mail command in an Exim filter,
1586 there was a bug that Exim would claim a syntax error in any
1587 header after the first one which had an odd number of characters
1590 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1591 callout verification, Exim cached a "reject" for the entire domain. This
1592 is correct for most verifications, but it is not correct for a recipient
1593 verification with use_sender or use_postmaster set, because in that case
1594 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1595 case of MAIL FROM:<> rejection from other early rejections (e.g.
1596 rejection of HELO). When verifying a recipient using a non-null MAIL
1597 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1598 Whatever the result of the callout, the value of the domain cache is
1599 left unchanged (for any other kind of callout, getting as far as trying
1600 RCPT means that the domain itself is ok).
1602 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1605 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1606 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1607 session, the daemon ignores ECONNECT errors and logs others; it now
1608 ignores EPIPE as well.
1610 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1611 (quoted-printable decoding).
1613 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1614 later the small subsequent patch to fix an introduced bug.
1616 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1618 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1620 PH/08 An error is now given if message_size_limit is specified negative.
1622 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1623 to be given (somewhat) arbitrary names.
1625 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1628 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1629 miscellaneous code fixes
1631 PH/10 Added the log_reject_target ACL modifier to specify where to log
1634 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1635 hostname. This is wrong, because it relates to the incoming message (and
1636 probably the interface on which it is arriving) and not to the outgoing
1637 callout (which could be using a different interface). This has been
1638 changed to use the value of the helo_data option from the smtp transport
1639 instead - this is what is used when a message is actually being sent. If
1640 there is no remote transport (possible with a router that sets up host
1641 addresses), $smtp_active_hostname is used.
1643 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1644 tweaks were necessary in order to get it to work (see also 21 below):
1645 (a) The code assumed that strncpy() returns a negative number on buffer
1646 overflow, which isn't the case. Replaced with Exim's string_format()
1648 (b) There were several signed/unsigned issues. I just did the minimum
1649 hacking in of casts. There is scope for a larger refactoring.
1650 (c) The code used strcasecmp() which is not a standard C function.
1651 Replaced with Exim's strcmpic() function.
1652 (d) The code set only $1; it now sets $auth1 as well.
1653 (e) A simple test gave the error "authentication client didn't specify
1654 service in request". It would seem that Dovecot has changed its
1655 interface. Fortunately there's a specification; I followed it and
1656 changed what the client sends and it appears to be working now.
1658 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1661 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1662 address A is aliased to B and C, where B exists and C does not. Without
1663 -v the output is "A verified" because verification stops after a
1664 successful redirection if more than one address is generated. However,
1665 with -v the child addresses are also verified. Exim was outputting "A
1666 failed to verify" and then showing the successful verification for C,
1667 with its parentage. It now outputs "B failed to verify", showing B's
1668 parentage before showing the successful verification of C.
1670 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1671 look up a TXT record in a specific list after matching in a combined
1674 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1675 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1676 they consult the DNS. I had assumed they would set it the way they
1677 wanted; and indeed my experiments on Linux seem to show that in some
1678 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1679 To be on the safe side, however, I have now made the interface to
1680 host_find_byname() similar to host_find_bydns(), with an argument
1681 containing the DNS resolver options. The host_find_byname() function now
1682 sets these options at its start, just as host_find_bydns() does. The smtp
1683 transport options dns_qualify_single and dns_search_parents are passed to
1684 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1685 of host_find_byname() use the default settings of RES_DEFNAMES
1686 (qualify_single) but not RES_DNSRCH (search_parents).
1688 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1689 spool_read_header() do less string testing, by means of a preliminary
1690 switch on the second character of optional "-foo" lines. (This is
1691 overdue, caused by the large number of possibilities that now exist.
1692 Originally there were few.) While I was there, I also converted the
1693 str(n)cmp tests so they don't re-test the leading "-" and the first
1694 character, in the hope this might squeeze out yet more improvement.
1696 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1697 flag allowing group syntax was set by the header_syntax check but not
1698 turned off, possible causing trouble later; (2) The flag was not being
1699 set at all for the header_verify test, causing "group"-style headers to
1700 be rejected. I have now set it in this case, and also caused header_
1701 verify to ignore an empty address taken from a group. While doing this, I
1702 came across some other cases where the code for allowing group syntax
1703 while scanning a header line wasn't quite right (mostly, not resetting
1704 the flag correctly in the right place). These bugs could have caused
1705 trouble for malformed header lines. I hope it is now all correct.
1707 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1708 with the "reply" argument non-NULL. The code, however (which originally
1709 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1710 but it didn't always do it. This confused somebody who was copying the
1711 code for some other use. I have removed all the tests.
1713 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1714 feature that was used to support insecure browsers during the U.S. crypto
1715 embargo. It requires special client support, and Exim is probably the
1716 only MTA that supported it -- and would never use it because real RSA is
1717 always available. This code has been removed, because it had the bad
1718 effect of slowing Exim down by computing (never used) parameters for the
1719 RSA_EXPORT functionality.
1721 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1722 authenticator to fail if there's a tab character in the incoming data
1723 (there should never be unless someone is messing about, as it's supposed
1724 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1725 if the connection is using TLS or if the remote IP is the same as the
1726 local IP, and the "valid-client-cert option" if a client certificate has
1729 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1730 authenticators. This can be used for authorization after authentication
1731 succeeds. (In the case of plaintext, it servers for both authentication
1734 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1735 if any retry times were supplied.
1737 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1738 connection, where there is no client IP address to check. In this
1739 situation, the verify now always succeeds.
1741 PH/25 Applied John Jetmore's -Mset patch.
1743 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1745 PH/27 In a string expansion for a processed (not raw) header when multiple
1746 headers of the same name were present, leading whitespace was being
1747 removed from all of them, but trailing whitespace was being removed only
1748 from the last one. Now trailing whitespace is removed from each header
1749 before concatenation. Completely empty headers in a concatenation (as
1750 before) are ignored.
1752 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1753 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1755 PH/29 [Removed. This was a change that I later backed out, and forgot to
1756 correct the ChangeLog entry (that I had efficiently created) before
1757 committing the later change.]
1759 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1760 address errors (4xx response to RCPT) over the same connection as other
1761 messages routed to the same hosts. Such deliveries are always "forced",
1762 so retry times are not inspected. This resulted in far too many retries
1763 for the affected addresses. The effect occurred only when there were more
1764 hosts than the hosts_max_try setting in the smtp transport when it had
1765 the 4xx errors. Those hosts that it had tried were not added to the list
1766 of hosts for which the message was waiting, so if all were tried, there
1767 was no problem. Two fixes have been applied:
1769 (i) If there are any address or message errors in an SMTP delivery, none
1770 of the hosts (tried or untried) are now added to the list of hosts
1771 for which the message is waiting, so the message should not be a
1772 candidate for sending over the same connection that was used for a
1773 successful delivery of some other message. This seems entirely
1774 reasonable: after all the message is NOT "waiting for some host".
1775 This is so "obvious" that I'm not sure why it wasn't done
1776 previously. Hope I haven't missed anything, but it can't do any
1777 harm, as the worst effect is to miss an optimization.
1779 (ii) If, despite (i), such a delivery is accidentally attempted, the
1780 routing retry time is respected, so at least it doesn't keep
1781 hammering the server.
1783 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1784 in ${readsocket because some servers need this prod.
1786 PH/32 Added some extra debug output when updating a wait-xxx database.
1788 PH/33 The hint "could be header name not terminated by colon", which has been
1789 given for certain expansion errors for a long time, was not being given
1790 for the ${if def:h_colon_omitted{... case.
1792 PH/34 The spec says: "With one important exception, whenever a domain list is
1793 being scanned, $domain contains the subject domain." There was at least
1794 one case where this was not true.
1796 PH/35 The error "getsockname() failed: connection reset by peer" was being
1797 written to the panic log as well as the main log, but it isn't really
1798 panic-worthy as it just means the connection died rather early on. I have
1799 removed the panic log writing for the ECONNRESET error when getsockname()
1802 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1803 runs only) independently of the message's sender address. This meant
1804 that, if the 4xx error was in fact related to the sender, a different
1805 message to the same recipient with a different sender could confuse
1806 things. In particualar, this can happen when sending to a greylisting
1807 server, but other circumstances could also provoke similar problems.
1808 I have changed the default so that the retry time for these errors is now
1809 based a combination of the sender and recipient addresses. This change
1810 can be overridden by setting address_retry_include_sender=false in the
1813 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1814 remote server are returned as part of bounce messages. This was not
1815 happening for LMTP over a pipe (the lmtp transport), but now it is the
1816 same for both kinds of LMTP.
1818 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1819 in header lines that were in fact CNAMEs. This is no longer the case.
1821 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1822 and queue runs started by the daemon processed all messages. This has
1823 been fixed so that -R and -S can now usefully be given with -q<time>.
1825 PH/40 Import PCRE release 6.7 (fixes some bugs).
1827 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1829 PH/42 Give an error if -q is specified more than once.
1831 PH/43 Renamed the variables $interface_address and $interface_port as
1832 $received_ip_address and $received_port, to make it clear that these
1833 values apply to message reception, and not to the outgoing interface when
1834 a message is delivered. (The old names remain recognized, of course.)
1836 PH/44 There was no timeout on the connect() call when using a Unix domain
1837 socket in the ${readsocket expansion. There now is.
1839 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1840 be meaningful with "accept".
1842 SC/01 Eximstats V1.43
1843 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1845 SC/02 Eximstats V1.44
1846 Use a glob alias rather than an array ref in the generated
1847 parser. This improves both readability and performance.
1849 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1850 Collect SpamAssassin and rejection statistics.
1851 Don't display local sender or destination tables unless
1852 there is data to show.
1853 Added average volumes into the top table text output.
1855 SC/04 Eximstats V1.46
1856 Collect data on the number of addresses (recipients)
1857 as well as the number of messages.
1859 SC/05 Eximstats V1.47
1860 Added 'Message too big' to the list of mail rejection
1861 reasons (thanks to Marco Gaiarin).
1863 SC/06 Eximstats V1.48
1864 Mainlog lines which have GMT offsets and are too short to
1865 have a flag are now skipped.
1867 SC/07 Eximstats V1.49 (Alain Williams)
1868 Added the -emptyok flag.
1870 SC/08 Eximstats V1.50
1871 Fixes for obtaining the IP address from reject messages.
1873 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1874 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1875 whitesspace changes from 4.64-PH/27
1877 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1880 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1881 are found, allow negative numbers in numeric criteria)
1883 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1885 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1886 to match changes made in 4.64-PH/43
1888 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1890 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1891 "domain" is actually an IP address, and force a failure. This locks out
1892 those revolvers/nameservers that support "A-for-A" lookups, in
1893 contravention of the specifications.
1895 PH/48 When a host name was looked up from an IP address, and the subsequent
1896 forward lookup of the name timed out, the host name was left in
1897 $sender_host_name, contrary to the specification.
1899 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1900 restricted to single-key lookups, Exim was not diagnosing an error if
1901 * or *@ was used with a query-style lookup.
1903 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1905 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1906 long ago, but noone interested enough thought of it. Let's just say that
1907 the "1.1" means that there are some new functions that weren't there at
1908 some point in the past.
1910 PH/51 Error processing for expansion failure of helo_data from an smtp
1911 transport during callout processing was broken.
1913 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1914 tested/used via the -bh/-bhc/-bs options.
1916 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1917 bug, fixed in subsequent PCRE releases).
1919 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1920 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1922 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1928 SC/01 Use a glob alias rather than an array ref in eximstats generated
1929 parser. This improves both readability and performance.
1931 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1932 Don't display local sender or destination tables in eximstats unless
1933 there is data to show.
1934 Added average volumes into the eximstats top table text output.
1936 SC/03 Collect data on the number of addresses (recipients) as well
1937 as the number of messages in eximstats.
1939 TF/01 Correct an error in the documentation for the redirect router. Exim
1940 does (usually) call initgroups() when daemonizing.
1942 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1943 with consistent privilege compared to when running as a daemon.
1945 TF/03 Note in the spec that $authenticated_id is not set for local
1946 submissions from trusted users.
1948 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1949 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1951 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1952 by adding some example configuration directives to the default
1953 configuration file. A little bit of work is required to uncomment the
1954 directives and define how usernames and passwords are checked, but
1955 there is now a framework to start from.
1957 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1958 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1959 without this. I don't know how relevant this is to other LDAP libraries.
1961 PH/02 Add the verb name to the "unknown ACL verb" error.
1963 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1965 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1967 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1968 directory not expanded when it should be if an expanded home directory
1969 was set for the address (which is overridden by the transport).
1971 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1974 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1975 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1976 because it is too late at that time, and has no effect.
1978 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1979 security issue with \' (bugzilla #107). I could not use the
1980 PQescapeStringConn() function, because it needs a PGconn value as one of
1983 PH/08 When testing addresses using -bt, indicate those final addresses that
1984 are duplicates that would not cause an additional delivery. At least one
1985 person was confused, thinking that -bt output corresponded to deliveries.
1986 (Suppressing duplicates isn't a good idea as you lose the information
1987 about possibly different redirections that led to the duplicates.)
1989 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1990 systems where poll() doesn't work, in particular OS X.
1992 PH/10 Added more information to debugging output for retry time not reached.
1994 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1995 operations in malware.c.
1997 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2000 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2001 syslog_duplication set false, log lines that would normally be written
2002 both the the main log and to the reject log were not written to syslog at
2005 PH/14 In the default configuration, change the use of "message" in ACL warn
2006 statements to "add_header".
2008 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2009 not followed by a command (e.g. "seen endif").
2011 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2012 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2015 PH/17 Added extra conditions to the default value of delay_warning_condition
2019 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2020 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2021 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2024 The Auto-Submitted: and various List- headers are standardised, whereas I
2025 don't think Precedence: ever was.
2027 PH/18 Refactored debugging code in route_finduser() to show more information,
2028 in particular, the error code if getpwnam() issues one.
2030 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2031 This is apparently needed in addition to the PH/07 change above to avoid
2032 any possible encoding problems.
2034 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2035 but not after initializing Perl.
2037 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2038 output them only if debugging. By default they are written stderr,
2039 apparently, which is not desirable.
2041 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2044 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2047 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2049 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2050 authenticated or an ident call has been made. Suppress the default
2051 values for $authenticated_id and $authenticated_sender (but permit -oMai
2052 and -oMas) when testing with -bh.
2054 PH/24 Re-jigged the order of the tests in the default configuration so that the
2055 tests for valid domains and recipients precede the DNS black list and CSA
2056 tests, on the grounds that those ones are more expensive.
2058 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2059 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2060 This bug exists in every version of Exim that I still have, right back to
2063 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2064 However, an attempt to turn on case-sensitivity in a regex key by
2065 including (?-i) didn't work because the subject string was already
2066 lowercased, and the effects were non-intuitive. It turns out that a
2067 one-line patch can be used to allow (?-i) to work as expected.
2073 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2074 other effects) broke the use of negated acl sub-conditions.
2076 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2079 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2080 "Deny" causes Exim to reject the incoming connection with a 554 error.
2081 Unfortunately, if there is a major crisis, such as a disk failure,
2082 tcp-wrappers gives "deny", whereas what one would like would be some
2083 kind of temporary error. A kludge has been added to help with this.
2084 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2085 554 error is used if errno is still zero or contains ENOENT (which occurs
2086 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2089 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2091 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2092 errors. Otherwise a message that provokes a temporary error (when other
2093 messages do not) can cause a whole host to time out.
2095 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2096 addresses were routed directly to files or pipes from a redirect router.
2097 File deliveries just didn't batch; pipe deliveries might have suffered
2100 PH/06 A failure to get a lock for a hints database would erroneously always say
2101 "Failed to get write lock", even when it was really a read lock.
2103 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2104 of 0600. This has been changed to use the value of the lockfile_mode
2105 option (which defaults to 0600).
2107 PH/08 Applied small patch from the Sieve maintainer.
2109 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2110 folder from quota calculations, a direct delivery into this folder messed
2111 up the contents of the maildirsize file. This was because the regex was
2112 used only to exclude .Trash (or whatever) when the size of the mailbox
2113 was calculated. There was no check that a delivery was happening into an
2114 excluded directory. This bug has been fixed by ignoring all quota
2115 processing for deliveries into excluded directories.
2117 PH/10 Added the maildirfolder_create_regex option to appendfile.
2123 PH/01 The code for finding all the local interface addresses on a FreeBSD
2124 system running IPv6 was broken. This may well have applied to all BSD
2125 systems, as well as to others that have similar system calls. The broken
2126 code found IPv4 interfaces correctly, but gave incorrect values for the
2127 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2128 that it would not match correctly against @[] and not recognize the IPv6
2131 PH/02 The ipliteral router was not recognizing addresses of the form user@
2132 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2134 PH/03 Added disable_ipv6.
2136 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2137 decoded form, because it is most often used to construct To: headers
2138 lines in autoreplies, and the decoded form may well be syntactically
2139 invalid. However, $reply_address has leading white space removed, and all
2140 newlines turned into spaces so that the autoreply transport does not
2143 PH/05 If group was specified without a user on a router, and no group or user
2144 was specified on a transport, the group from the router was ignored.
2146 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2147 for visible compile-time settings that can be used to change these
2148 numbers, for those that want even more. Backwards compatibility with old
2149 spool files has been maintained. However, going back to a previous Exim
2150 release will lost any variables that are in spool files.
2152 PH/07 Two small changes when running in the test harness: increase delay when
2153 passing a TCP/IP connection to a new process, in case the original
2154 process has to generate a bounce, and remove special handling of
2155 127.0.0.2 (sic), which is no longer necessary.
2157 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2158 be the same on different OS.
2160 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2163 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2164 whether --show-vars was specified or not
2166 JJ/02 exipick: Added support for new ACL variable spool format introduced
2169 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2170 syntactically invalid From: or Reply-to: line, and a filter used this to
2171 generate an autoreply, and therefore failed to obtain an address for the
2172 autoreply, Exim could try to deliver to a non-existent relative file
2173 name, causing unrelated and misleading errors. What now happens is that
2174 it logs this as a hard delivery error, but does not attempt to create a
2177 PH/11 The exinext utility has a -C option for testing purposes, but although
2178 the given file was scanned by exinext itself; it wasn't being passed on
2179 when Exim was called.
2181 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2182 an end-of-file indication when reading a command response.
2184 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2185 compiled. In many other places in Exim, IPv6 addresses are always
2186 recognized, so I have changed this. It also means that IPv4 domain
2187 literals of the form [IPV4:n.n.n.n] are now always recognized.
2189 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2190 used if the router is not running as root, for example, when verifying at
2191 ACL time, or when using -bh. The debugging output from this situation was
2192 non-existent - all you got was a failure to exec. I have made two
2195 (a) Failures to set uid/gid, the current directory, or a process leader
2196 in a subprocess such as that created by queryprogram now generate
2197 suitable debugging ouput when -d is set.
2199 (b) The queryprogram router detects when it is not running as root,
2200 outputs suitable debugging information if -d is set, and then runs
2201 the subprocess without attempting to change uid/gid.
2203 PH/15 Minor change to Makefile for building test_host (undocumented testing
2206 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2207 additional section of a DNS packet that returns MX or SRV records.
2208 Instead, it always explicitly searches for A/AAAA records. This avoids
2209 major problems that occur when a DNS server includes only records of one
2210 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2211 fixed another bug: if SRV records were looked up and the corresponding
2212 address records were *not* found in the additional section, the port
2213 values from the SRV records were lost.
2215 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2216 using the correct key (the original address) when searching the retry
2217 rules in order to find which one to use for generating the retry hint.
2219 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2220 adding the default one. Similarly, if it contains a Reply-To: header, the
2221 errors_reply_to option, if set, is not used.
2223 PH/19 When calculating a retry time, Exim used to measure the "time since
2224 failure" by looking at the "first failed" field in the retry record. Now
2225 it does not use this if it is later than than the arrival time of the
2226 message. Instead it uses the arrival time. This makes for better
2227 behaviour in cases where some deliveries succeed, thus re-setting the
2228 "first failed" field. An example is a quota failure for a huge message
2229 when small messages continue to be delivered. Without this change, the
2230 "time since failure" will always be short, possible causing more frequent
2231 delivery attempts for the huge message than are intended.
2232 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2234 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2235 $1, $2, $3) because the numerical variables can be reset during some
2236 expansion items (e.g. "match"), thereby losing the authentication data.
2238 PH/21 Make -bV show the size of off_t variables so that the test suite can
2239 decide whether to run tests for quotas > 2G.
2241 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2242 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2243 filecount value is greater than 2G or if a quota value is greater than 2G
2244 on a system where the size of off_t is not greater than 4, a panic error
2247 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2248 never match. The debug and -bh output now contains an explicit error
2249 message indicating a malformed IPv4 address or mask.
2251 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2252 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2253 PH/23 above applies.
2255 PH/25 Do not write to syslog when running in the test harness. The only
2256 occasion when this arises is a failure to open the main or panic logs
2257 (for which there is an explicit test).
2259 PH/26 Added the /no_tell option to "control=freeze".
2261 PH/27 If a host name lookup failed very early in a connection, for example, if
2262 the IP address matched host_lookup and the reverse lookup yielded a name
2263 that did not have a forward lookup, an error message of the form "no IP
2264 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2265 could be logged. Now it outputs the IP address instead of "NULL".
2267 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2268 allows the sender and the authenticated sender to be set when
2269 submitting a message from within Exim. Since child_open_exim() is
2270 documented for local_scan(), the new function should be too.
2272 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2273 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2274 results in an empty string is now treated as unset.
2276 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2278 PH/31 Added sender_verify_fail logging option.
2280 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2281 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2282 tidied the source and removed it altogether.
2284 PH/33 When a queue run was abandoned because the load average was too high, a
2285 log line was always written; now it is written only if the queue_run log
2286 selector is set. In addition, the log line for abandonment now contains
2287 information about the queue run such as the pid. This is always present
2288 in "start" and "stop" lines but was omitted from the "abandon" line.
2290 PH/34 Omit spaces between a header name and the colon in the error message that
2291 is given when verify = headers_syntax fails (if there are lots of them,
2292 the message gets confusing).
2294 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2295 names, as there are now some PTR records that contain slashes. This check
2296 is only to protect against broken name servers that fall over on strange
2297 characters, so the fact that it applies to all lookups doesn't matter.
2299 PH/36 Now that the new test suite is complete, we can remove some of the
2300 special code in Exim that was needed for the old test suite. For example,
2301 sorting DNS records because real resolvers return them in an arbitrary
2302 order. The new test suite's fake resolver always returns records in the
2305 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2306 bounces) except when queue_only is set, to avoid logging races between
2307 the different processes.
2309 PH/38 Panic-die if .include specifies a non-absolute path.
2311 PH/39 A tweak to the "H" retry rule from its user.
2313 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2314 a label. They prevented compilation on older perls.
2316 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2317 a warning to be raised on newish perls.
2319 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2320 on queue. Changes to match documented behaviour of showing count of
2321 messages matching specified criteria.
2323 PH/40 Changed the default ident timeout from 30s to 5s.
2325 PH/41 Added support for the use of login_cap features, on those BSD systems
2326 that have them, for controlling the resources used by pipe deliveries.
2328 PH/42 The content-scanning code uses fopen() to create files in which to put
2329 message data. Previously it was not paying any attention to the mode of
2330 the files. Exim runs with umask(0) because the rest of the code creates
2331 files with open(), and sets the required mode explicitly. Thus, these
2332 files were ending up world-writeable. This was not a big issue, because,
2333 being within the spool directory, they were not world-accessible. I have
2334 created a function called modefopen, which takes an additional mode
2335 argument. It sets umask(777), creates the file, chmods it to the required
2336 mode, then resets the umask. All the relevant calls to fopen() in the
2337 content scanning code have been changed to use this function.
2339 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2340 to 24 hours. This avoids potential overflow problems when processing G
2341 and H retry rules. I suspect nobody ever tinkers with this value.
2343 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2345 PH/45 When the plaintext authenticator is running as a client, the server's
2346 challenges are checked to ensure they are valid base64 strings. By
2347 default, the authentication attempt is cancelled if an invalid string is
2348 received. Setting client_ignore_invalid_base64 true ignores these errors.
2349 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2350 they are received. Thus, the responses can be made to depend on the
2351 challenges. If an invalid string is ignored, an empty string is placed in
2354 PH/46 Messages that are created by the autoreply transport now contains a
2355 References: header, in accordance with RFCs 2822 and 3834.
2357 PH/47 Added authenticated_sender_force to the smtp transport.
2359 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2361 PH/49 Installed latest patch from the Sieve maintainer.
2363 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2364 was also set, the appendfile transport was unnecessarily scanning a
2365 directory of message files (e.g. for maildir delivery) to find the count
2366 of files (along with the size), even though it did not need this
2367 information. It now does the scan only if it needs to find either the
2368 size of the count of files.
2370 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2372 PH/52 Two bugs concerned with error handling when the smtp transport is
2375 (i) Exim was not creating retry information for temporary errors given
2376 for individual recipients after the DATA command when the smtp transport
2377 was used in LMTP mode. This meant that they could be retried too
2378 frequently, and not timed out correctly.
2380 (ii) Exim was setting the flag that allows error details to be returned
2381 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2382 recipients that were returned after the DATA command.
2384 PH/53 This is related to PH/52, but is more general: for any failing address,
2385 when detailed error information was permitted to be returned to the
2386 sender, but the error was temporary, then after the final timeout, only
2387 "retry timeout exceeded" was returned. Now it returns the full error as
2388 well as "retry timeout exceeded".
2390 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2391 do this, and (what is worse) MTAs that accept it.
2393 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2394 will now be deprecated.
2396 PH/56 New os.c-cygwin from the Cygwin maintainer.
2398 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2399 formats (previously only available in exim formats via -bpr, -bpru,
2400 and -bpra. Now also available in native and exiqgrep formats)
2402 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2403 with very large, slow to parse queues
2405 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2407 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2409 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2410 responses to authentication challenges, though it was showing the
2411 challenges; (ii) I've removed the CR characters from the debug output for
2414 PH/58 Allow for the insertion of a newline as well as a space when a string
2415 is turned into more than one encoded-word during RFC 2047 encoding. The
2416 Sieve code now uses this.
2418 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2419 data_4xx, lost_connection, tls_required.
2421 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2422 message was being sent as an SMTP response.
2424 PH/61 Add -l and -k options to exicyclog.
2426 PH/62 When verifying, if an address was redirected to one new address, so that
2427 verification continued, and the new address failed or deferred after
2428 having set something in $address_data, the value of $address_data was not
2429 passed back to the ACL. This was different to the case when no
2430 redirection occurred. The value is now passed back in both cases.
2432 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2433 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2434 use login_cap.h, so on its own it isn't the distinguishing feature. The
2435 new name refers directly to the setclassresources() function.
2437 PH/65 Added configuration files for NetBSD3.
2439 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2441 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2442 is preferred over IPv4.
2444 PH/68 The bounce_return_message and bounce_return_body options were not being
2445 honoured for bounces generated during the reception of non-SMTP messages.
2446 In particular, this applied to messages rejected by the ACL. This bug has
2447 been fixed. However, if bounce_return_message is true and bounce_return_
2448 body is false, the headers that are returned for a non-SMTP message
2449 include only those that have been read before the error was detected.
2450 (In the case of an ACL rejection, they have all been read.)
2452 PH/69 The HTML version of the specification is now built in a directory called
2453 spec_html instead of spec.html, because the latter looks like a path with
2454 a MIME-type, and this confuses some software.
2456 PH/70 Catch two compiler warnings in sieve.c.
2458 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2459 function verify_get_ident() calls ip_connect() to connect a socket, but
2460 if the "connect()" function timed out, ip_connect() used to close the
2461 socket. However, verify_get_ident() also closes the socket later, and in
2462 between Exim writes to the log, which may get opened at this point. When
2463 the socket was closed in ip_connect(), the log could get the same file
2464 descriptor number as the socket. This naturally causes chaos. The fix is
2465 not to close the socket in ip_connect(); the socket should be closed by
2466 the function that creates it. There was only one place in the code where
2467 this was missing, in the iplookup router, which I don't think anybody now
2468 uses, but I've fixed it anyway.
2470 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2471 well as to direct DNS lookups. Otherwise the handling of names in host
2472 lists is inconsistent and therefore confusing.
2478 PH/01 Two changes to the default runtime configuration:
2480 (1) Move the checks for relay_from_hosts and authenticated clients from
2481 after to before the (commented out) DNS black list checks.
2483 (2) Add control=submission to the relay_from_hosts and authenticated
2484 clients checks, on the grounds that messages accepted by these
2485 statements are most likely to be submissions.
2487 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2489 (1) Generate an error if the third argument for the ${prvs expansion is
2492 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2495 (3) Reset the variables that are obtained from the first argument of
2496 ${prvscheck and used in the second argument before leaving the code,
2497 because their memory is reclaimed, so using them afterwards may do
2500 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2501 one (it's much easier than Tom thought :-).
2503 (5) Because of (4), we can now allow for the use of $prvscheck_result
2504 inside the third argument.
2506 PH/03 For some reason, the default setting of PATH when running a command from
2507 a pipe transport was just "/usr/bin". I have changed it to
2510 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2511 anything to be listed in the output from -bV.
2513 PH/05 When a filter generated an autoreply, the entire To: header line was
2514 quoted in the delivery log line, like this:
2516 => >A.N.Other <ano@some.domain> <original@ddress> ...
2518 This has been changed so that it extracts the operative address. There
2519 may be more than one such address. If so, they are comma-separated, like
2522 => >ano@some.domain,ona@other.domain <original@ddress> ...
2524 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2525 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2526 not being looked up in the rDNS to get a host name, Exim was showing the
2527 IP address twice in Received: lines, even though the IP addresses were
2528 identical. For example:
2530 Received: from [1.2.3.4] (helo=[1.2.3.4])
2532 However, if the real host name was known, it was omitting the HELO data
2533 if it matched the actual IP address. This has been tidied up so that it
2534 doesn't show the same IP address twice.
2536 PH/07 When both +timestamp and +memory debugging was on, the value given by
2537 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2538 called by the debug printing, thereby overwriting the timestamp buffer.
2539 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2541 PH/08 When the original message was included in an autoreply transport, it
2542 always said "this is a copy of the message, including all the headers",
2543 even if body_only or headers_only was set. It now gives an appropriate
2546 PH/09 Applied a patch from the Sieve maintainer which:
2548 o fixes some comments
2549 o adds the (disabled) notify extension core
2550 o adds some debug output for the result of if/elsif tests
2551 o points to the current vacation draft in the documentation
2552 and documents the missing references header update
2556 o fixes a bug in processing the envelope test (when testing
2557 multiple envelope elements, the last element determinted the
2560 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2561 Electronic Mail") by including:
2563 Auto-submitted: auto-generated
2565 in the messages that it generates (bounce messages and others, such as
2566 warnings). In the case of bounce messages for non-SMTP mesages, there was
2567 also a typo: it was using "Auto_submitted" (underscore instead of
2568 hyphen). Since every message generated by Exim is necessarily in response
2569 to another message, thes have all been changed to:
2571 Auto-Submitted: auto-replied
2573 in accordance with these statements in the RFC:
2575 The auto-replied keyword:
2577 - SHOULD be used on messages sent in direct response to another
2578 message by an automatic process,
2580 - MUST NOT be used on manually-generated messages,
2582 - MAY be used on Delivery Status Notifications (DSNs) and Message
2583 Disposition Notifications (MDNs),
2585 - MUST NOT be used on messages generated by automatic or periodic
2586 processes, except for messages which are automatic responses to
2589 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2590 to the default Received: header definition.
2592 PH/12 Added log selector acl_warn_skipped (default on).
2594 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2595 variables because (a) they are in the wrong storage pool and (b) even if
2596 they were copied, it wouldn't work properly because of the caching.
2598 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2599 checking when decoding. Apparently there are clients that generate
2600 overlong encoded strings. Why am I not surprised?
2602 PH/15 If the first argument of "${if match_address" was not empty, but did not
2603 contain an "@" character, Exim crashed. Now it writes a panic log message
2604 and treats the condition as false.
2606 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2608 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2609 extension "envelope-auth". The code is finished and in agreement with
2610 other implementations, but there is no documentation so far and in fact,
2611 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2612 not changing the active code.
2614 Print executed "if" and "elsif" statements when debugging is used. This
2615 helps a great deal to understand what a filter does.
2617 Document more things not specified clearly in RFC3028. I had all this
2618 sorted out, when out of a sudden new issues came to my mind. Oops."
2620 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2623 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2624 canonical form (as documented). However, after a host name lookup from
2625 the IP address, check_host() was doing a simple string comparison with
2626 addresses acquired from the DNS when checking that the found name did
2627 have the original IP as one of its addresses. Since any found IPv6
2628 addresses are likely to be in abbreviated form, the comparison could
2629 fail. Luckily, there already exists a function for doing the comparison
2630 by converting both addresses to binary, so now that is used instead of
2631 the text comparison.
2633 PH/20 There was another similar case to PH/19, when a complete host name was
2634 given in a host list; looking up its IP address could give an abbreviated
2635 form, whereas the current host's name might or might not be abbreviated.
2636 The same fix has been applied.
2642 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2643 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2646 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2647 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2649 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2651 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2652 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2653 Also, for "Auto-Submitted", treat anything other than "no" as
2654 non-personal, in accordance with RFC 3834. (Previously it treated
2655 anything starting "auto-" as non-personal.)
2657 TF/01 The control=submission/name=... option had a problem with syntax
2658 errors if the name included a slash character. The /name= option
2659 now slurps the rest of the string, so it can include any characters
2660 but it must come last in the list of options (after /sender_retain
2663 PH/05 Some modifications to the interface to the fake nameserver for the new
2671 TK/01 Added the "success_on_redirect" address verification option. See
2672 NewStuff for rationale and an example.
2674 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2676 PH/02 Patch to exigrep to allow it to work on syslog lines.
2678 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2679 fread() to skip over the body file's header line, because in Cygwin the
2680 header line is locked and is inaccessible.
2682 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2683 co-exist for some time) to make it clear that it is the Exim ID that is
2684 referenced, not the Message-ID: header line.
2686 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2687 string_format() function, because snprintf() does not exist on all
2690 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2691 ${stat: expansion item.
2693 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2694 protocol synchronization error", to keep the pedants happy.
2696 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2697 well as for IRIX systems, when gcc is being used. See the host.c source
2700 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2702 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2705 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2706 transport and to the smtp transport in LMTP mode.
2708 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2710 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2711 run a filter in a subprocess. This could lead to confusion in subsequent
2712 lookups in the parent process. There should also be a search_tidyup() at
2713 the end of the subprocess.
2715 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2716 only if the host matched helo_try_verify_hosts, which caused the
2717 verification to occur when the EHLO/HELO command was issued. The ACL just
2718 tested the remembered result. Now, if a previous verification attempt has
2719 not happened, "verify = helo" does it there and then.
2721 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2723 TK/03 Fix log output including CR from clamd.
2725 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2726 exist provoked a memory error which could cause a segfault.
2728 PH/15 Installed PCRE 6.2
2730 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2732 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2733 of the problem. Specifically, suggested +O2 rather than +O1 for the
2736 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2738 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2739 record, the port was not correctly logged when the outgoing_port log
2740 selector was set (it logged the transort's default port).
2742 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2743 fallback_hosts, and "hosts" in the smtp transport.
2745 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2746 host errors such as "Connection refused".
2748 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2749 authentication with radiusclient 0.4.9:
2751 - Error returned from rc_read_config was caught wrongly
2752 - Username/password not passed on to radius server due to wrong length.
2754 The presumption is that some radiusclient API changes for 4.51/PH/17
2755 were not taken care of correctly. The code is still untested by me (my
2756 Linux distribution still has 0.3.2 of radiusclient), but it was
2757 contributed by a Radius user.
2759 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2760 expanding the "port" option of the smtp transport.
2762 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2763 while reading a MIME header. Thanks to Tom Hughes for a patch.
2765 PH/24 Include config.h inside local_scan.h so that configuration settings are
2768 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2769 that in an ACL for RCPT (for example), you can examine exactly what was
2772 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2773 commands, but it was not correctly comparing the address with the actual
2774 client host address. Thus, it would show the EHLO address in Received:
2775 header lines when this was not necessary.
2777 PH/27 Added the % operator to ${eval:}.
2779 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2780 it should be ignoring failures (because with -C, for example, it has lost
2781 privilege). It wasn't ignoring creation failures other than "already
2784 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2787 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2788 because an input line was too long, either on its own, or by virtue of
2789 too many continuations, the temporary file was not being removed, and the
2790 return code was incorrect.
2792 PH/31 Missing "BOOL" in function definition in filtertest.c.
2794 PH/32 Applied Sieve patches from the maintainer.
2796 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2798 PH/33 Added "verify = not_blind".
2800 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2801 Local/Makefile (with some defaults set). These are used in built scripts
2802 such as exicyclog, but they have never been used in the exim_install
2803 script (though there are many overriding facilities there). I have
2804 arranged that the exim_install script now takes note of these two
2807 PH/35 Installed configuration files for Dragonfly.
2809 PH/36 When a locally submitted message by a trusted user did not contain a
2810 From: header, and the sender address was obtained from -f or from an SMTP
2811 MAIL command, and the trusted user did not use -F to supply a sender
2812 name, $originator_name was incorrectly used when constructing a From:
2813 header. Furthermore, $originator_name was used for submission mode
2814 messages from external hosts without From: headers in a similar way,
2815 which is clearly wrong.
2817 PH/37 Added control=suppress_local_fixups.
2819 PH/38 When log_selector = +received_sender was set, and the addition of the
2820 sender made the log line's construction buffer exactly full, or one byte
2821 less than full, an overflow happened when the terminating "\n" was
2824 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2825 when the result of a list match is failure because a DNS lookup failed.
2827 PH/40 RM_COMMAND is now used in the building process.
2829 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2830 the "build-* directories that it finds.
2832 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2833 address in a domain literal was a prefix of an interface address.
2835 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2836 when verifying a sender address, unless rewrite_headers is false.
2838 PH/44 Wrote a long comment about why errors_to addresses are verified as
2839 recipients, not senders.
2841 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2842 the ratelimit ACL was added.
2844 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2846 PH/46 Added extra information about PostgreSQL errors to the error string.
2848 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2849 avoiding the need to install special zones in a real server. This is
2850 backwards compatible; if it can't find the fake resolver, it drops back.
2851 Thus, both old and new test suites can be run.
2853 TF/02 Added util/ratelimit.pl
2855 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2856 clock is set back in time.
2858 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2859 Candler <B.Candler@pobox.com>.
2861 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2862 OK for addresses that are the result of redirections.
2864 PH/48 A number of further additions for the benefit of the new test suite,
2865 including a fake gethostbyname() that interfaces to the fake DNS resolver
2868 TF/06 The fix for widen_domains has also been applied to qualify_single and
2869 search_parents which are the other dnslookup options that can cause
2872 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2875 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2877 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2878 service" if it encountered an MX record with an empty target hostname.
2879 The message is now "an MX or SRV record indicated no SMTP service".
2881 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2882 if the DNS of the sending site is misconfigured. This is quite a
2883 common situation. This change restores the behaviour of treating a
2884 helo verification defer as a failure.
2886 PH/51 If self=fail was set on a router, the bounce message did not include the
2887 actual error message.
2893 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2895 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2896 command itself ended in error, the underlying message about the transport
2897 filter timeout was being overwritten with the pipe command error. Now the
2898 underlying error message should be appended to the second error message.
2900 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2902 PH/02 Reduce the amount of output that "make" produces by default. Full output
2903 can still be requested.
2905 PH/03 The warning log line about a condition test deferring for a "warn" verb
2906 was being output only once per connection, rather than after each
2907 occurrence (because it was using the same function as for successful
2908 "warn" verbs). This seems wrong, so I have changed it.
2910 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2911 it should not have, which might have caused a crash in the right
2912 circumstances, but probably never did.
2914 PH/04 Installed a modified version of Tony Finch's patch to make submission
2915 mode fix the return path as well as the Sender: header line, and to
2916 add a /name= option so that you can make the user's friendly name appear
2919 TF/03 Added the control = fakedefer ACL modifier.
2921 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2922 Mark Lowes for thorough testing.
2924 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2926 TK/03 Merged latest SRS patch from Miles Wilton.
2928 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2929 with the definition in sysexits.h (which is #included earlier).
2930 Fortunately, Exim does not actually use EX_OK. The code used to try to
2931 preserve the sysexits.h value, by assumimg that macro definitions were
2932 scanned for macro replacements. I have been disabused of this notion,
2933 so now the code just undefines EX_OK before #including unistd.h.
2935 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2936 in the smtp transport. When a block could not be written in a single
2937 write() function, the timeout was being re-applied to each part-write.
2938 This seems wrong - if the receiver was accepting one byte at a time it
2939 would take for ever. The timeout is now adjusted when this happens. It
2940 doesn't have to be particularly precise.
2942 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2943 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2945 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2946 without a domain if the check to <postmaster@domain> fails.
2948 SC/01 Eximstats: added -xls and the ability to specify output files
2949 (patch written by Frank Heydlauf).
2951 SC/02 Eximstats: use FileHandles for outputing results.
2953 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2955 SC/04 Eximstats: fixed display of large numbers with -nvr option
2957 SC/05 Eximstats: fixed merging of reports with empty tables.
2959 SC/06 Eximstats: added the -include_original_destination flag
2961 SC/07 Eximstats: removed tabs and trailing whitespace.
2963 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2965 TK/06 MBOX spool code: Add real "From " MBOX separator line
2966 so the .eml file is really in mbox format (even though
2967 most programs do not really care). Patch from Alex Miller.
2969 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2970 The latter is generated from $received_to and is only set if the
2971 message has one envelope recipient. SA can use these headers,
2972 obviously out-of-the-box. Patch from Alex Miller.
2974 PH/08 The ${def test on a variable was returning false if the variable's
2975 value was "0", contrary to what the specification has always said!
2976 The result should be true unless the variable is empty.
2978 PH/09 The syntax error of a character other than { following "${if
2979 def:variable_name" (after optional whitespace) was not being diagnosed.
2980 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2981 accidental colon was present, for example, could give incorrect results.
2983 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2984 result is used (not including appendfile, where other changes are about
2987 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2988 This involved changing a lot of size variables from int to off_t. It
2989 should work with maildirs and everything.
2991 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2992 spamd dying while we are connected to it.
2994 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2997 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2998 function for BDB 4.3.
3000 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3002 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3003 the sources for PCRE have been re-arranged and more files are now
3006 PH/15 The code I had for printing potentially long long variables in PH/11
3007 above was not the best (it lost precision). The length of off_t variables
3008 is now inspected at build time, and an appropriate printing format (%ld
3009 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3010 to be "long long int" or "long int". This is needed for the internal
3011 formatting function string_vformat().
3013 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3014 the configuration file to be ":syslog", then the script "guesses" where
3015 the logs files are, rather than using the compiled in default. In our
3016 case the guess is not the same as the compiled default, so the script
3017 suddenly stopped working when I started to use syslog. The patch checks
3018 to see if log_file_path is "". If so, it attempts to read it from exim
3019 with no configuration file to get the compiled in version, before it
3020 falls back to the previous guessing code."
3022 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3023 implementing BATV in an Exim configuration. See NewStuff for the gory
3026 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3027 Makefile that are specific to HP-UX.
3029 PH/18 If the "use_postmaster" option was set for a recipient callout together
3030 with the "random" option, the postmaster address was used as the MAIL
3031 FROM address for the random test, but not for the subsequent recipient
3032 test. It is now used for both.
3034 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3035 patch removes a few documentation additions to RFC 3028, because the
3036 latest draft now contains them. It adds the new en;ascii-case comparator
3037 and a new error check for 8bit text in MIME parts. Comparator and
3038 require names are now matched exactly. I enabled the subaddress
3039 extension, but it is not well tested yet (read: it works for me)."
3041 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3042 rework some of the code of TK/09 above to avoid the hardwired use of
3043 "%lld" and "long long". Replaced the call to snprintf() with a call to
3046 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3047 records point to non-existent hosts", "retry timeout exceeded", and
3048 "retry time not reached for any host after a long failure period".
3050 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3051 experimental DomainKeys support:
3053 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3054 (2) On an error such as an illegally used "control", the wrong name for
3055 the control was given.
3057 These problems did NOT occur unless DomainKeys support was compiled.
3059 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3061 PH/24 Added ${if match_ip condition.
3063 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3064 databases so that it will be absolutely obvious if a crash occurs in the
3065 DB library. This is a regular occurrence (often caused by mis-matched
3068 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3069 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3070 picked up on a user's system that detects such things. There doesn't seem
3071 to be a gcc warning option for this - only an attribute that has to be
3072 put on the function's prototype. It seems that in Fedora Core 4 they have
3073 set this on a number of new functions. No doubt there will be more in due
3076 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3077 specify a transport. However, if an address that was verified by such a
3078 router was the subject of a callout, Exim crashed because it tried to
3079 read the rcpt_include_affixes from the non-existent transport. Now it
3080 just assumes that the setting of that option is false. This bug was
3081 introduced by 4.51/PH/31.
3083 PH/28 Changed -d+all to exclude +memory, because that information is very
3084 rarely of interest, but it makes the output a lot bigger. People tend to
3085 do -d+all out of habit.
3087 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3088 code in os-type was giving problems when libc.so lives in lib64, like on
3091 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3092 aren't the modern standard, and it seems that some systems' include files
3093 don't always have them. Exim was already checking for some of the newer
3094 ones like T_AAAA, and defining it itself. I've added checks for all the
3095 record types that Exim uses.
3097 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3098 not automatically generating a new one, as it is supposed to. This
3099 prevented TLS from working. If the file did exist, but contained invalid
3100 data, a new version was generated, as expected. It was only the case of a
3101 non-existent file that was broken.
3103 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3104 with a change in libdomainkeys > 0.64.
3106 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3107 from DNS. If the selector record carries the flag, it now has
3108 precedence over the domain-wide flag.
3110 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3112 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3113 the use of a port name that isn't defined in /etc/services) occurred, the
3114 message was deferred as in a normal delivery, and thus remained on the
3115 spool, instead of being failed because of the mua_wrapper setting. This
3116 is now fixed, and I tidied up some of the mua_wrapper messages at the
3119 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3120 the messages in a hash of arrays rather than using individual hashes.
3121 This is a bit cleaner and results in dramatic memory savings, albeit
3122 at a slight CPU cost.
3124 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3125 as requested by Marc Sherman.
3127 SC/10 Eximstats: added histograms for user specified patterns as requested
3130 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3132 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3133 fopen() in the content-scanning modules that did not already have it.
3139 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3140 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3142 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3144 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3146 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3147 to test Sieve filters that use "vacation".
3149 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3150 that changes the way the GnuTLS parameters are stored in the cache file.
3151 The new format can be generated externally. For backward compatibility,
3152 if the data in the cache doesn't make sense, Exim assumes it has read an
3153 old-format file, and it generates new data and writes a new file. This
3154 means that you can't go back to an older release without removing the
3157 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3158 work if there are any delivery delays because "one_time" forces the
3159 parent to be marked "delivered", so its unseen clone is never tried
3160 again. For this reason, Exim now forbids the simultaneous setting of
3163 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3164 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3165 entry if you want to know the details. The fix, however, neglected to
3166 consider the case when local delivery batching is involved. The test for
3167 "previously delivered" was not happening when checking to see if an
3168 address could be batched with a previous (undelivered) one; under
3169 certain circumstances this could lead to multiple deliveries to the same
3172 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3173 in its include files, and this causes problems building Exim.
3175 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3176 header_syntax) but Exim was just ignoring anything given after a slash.
3177 In particular, this caused confusion with an attempt to use "verify =
3178 reverse_host_lookup/defer_ok". An error is now given when options are
3179 supplied for verify items that do not have them. (Maybe reverse_host_
3180 lookup should have a defer_ok option, but that's a different point.)
3182 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3183 defined by RFC 821) to 2048, because there were problems with some AUTH
3184 commands, and RFC 1869 says the size should be increased for extended
3185 SMTP commands that take arguments.
3187 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3190 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3191 "unknown" error; now it says that the functionality isn't in the binary.
3193 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3194 an address' error message when a string expansion fails (syntax or
3195 whatever). Otherwise the password may appear in the log. Following change
3196 PH/42 below, there is no longer a chance of it appearing in a bounce
3199 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3201 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3202 of its addresses was ever tried. (Bugzilla bug #2.)
3204 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3205 the result incorrectly in the debug output. (It correctly added a newline
3206 to what was transported.)
3208 TF/01 Added $received_time.
3210 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3211 commented out examples of how to interface to a virus scanner and to
3212 SpamAssassin. Also added commented examples of av_scanner and
3213 spamd_address settings.
3215 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3216 and controls are allowed in which ACLs. There were a couple of minor
3217 errors. Some of the entries in the conditions table (which is a table of
3218 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3219 negation of where the condition IS allowed.
3221 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3223 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3224 header file does not have a version number, so I've had to invent a new
3225 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3226 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3227 radiusclient), but it was contributed by a Radius user.
3229 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3230 files or directories, for OpenSSL.
3232 PH/19 When an Exim process that is running as root has to create an Exim log
3233 file, it does so in a subprocess that runs as exim:exim so as to get the
3234 ownership right at creation (otherwise, other Exim processes might see
3235 the file with the wrong ownership). There was no test for failure of this
3236 fork() call, which would lead to the process getting stuck as it waited
3237 for a non-existent subprocess. Forks do occasionally fail when resources
3238 run out. I reviewed all the other calls to fork(); they all seem to check
3241 PH/20 When checking for unexpected SMTP input at connect time (before writing
3242 the banner), Exim was not dealing correctly with a non-positive return
3243 from the read() function. If the client had disconnected by this time,
3244 the result was a log entry for a synchronization error with an empty
3245 string after "input=" when read() returned zero. If read() returned -1
3246 (an event I could not check), uninitialized data bytes were printed.
3247 There were reports of junk text (parts of files, etc) appearing after
3250 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3252 PH/22 Added support for macro redefinition, and (re)definition in between
3253 driver and ACL definitions.
3255 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3256 forgetting to use the resulting value; it was using the unexpanded value.
3258 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3259 hadn't been configured. The fix is from Juergen Kreileder, who
3260 understands it better than I do:
3262 "Here's what I see happening with three configured cyrus_sasl
3263 authenticators configured (plain, login, cram-md5):
3265 On startup auth_cyrus_sasl_init() gets called for each of these.
3266 This means three calls to sasl_listmech() without a specified mech_list.
3267 => SASL tests which mechs of all available mechs actually work
3268 => three warnings about OTP not working
3269 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3271 With the patch, sasl_listmech() also gets called three times. But now
3272 SASL's mech_list option is set to the server_mech specified in the the
3273 authenticator. Or in other words, the answer from sasl_listmech()
3274 gets limited to just the mech you're testing for (which is different
3276 => the return list contains just 'plain' or 'login', 'cram-md5' or
3277 nothing depending on the value of ob->server_mech.
3279 I've just tested the patch: Authentication still works fine,
3280 unavailable mechs specified in the exim configuration are still
3281 caught, and the auth.log warnings about OTP are gone."
3283 PH/25 When debugging is enabled, the contents of the command line are added
3284 to the debugging output, even when log_selector=+arguments is not
3287 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3288 answer is "GNU", and only if the return is "GNU/something" is the answer
3291 PH/27 $acl_verify_message is now set immediately after the failure of a
3292 verification in an ACL, and so is available in subsequent modifiers. In
3293 particular, the message can be preserved by coding like this:
3295 warn !verify = sender
3296 set acl_m0 = $acl_verify_message
3298 Previously, $acl_verify_message was set only while expanding "message"
3299 and "log_message" when a very denied access.
3301 PH/28 Modified OS/os.c-Linux with
3303 -#ifndef OS_LOAD_AVERAGE
3304 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3306 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3307 nomenclature these days.)
3309 PH/29 Installed patch from the Sieve maintainer that adds the options
3310 sieve_useraddress and sieve_subaddress to the redirect router.
3312 PH/30 In these circumstances:
3313 . Two addresses routed to the same list of hosts;
3314 . First host does not offer TLS;
3315 . First host accepts first address;
3316 . First host gives temporary error to second address;
3317 . Second host offers TLS and a TLS session is established;
3318 . Second host accepts second address.
3319 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3320 and peerdn, if requested) that were in fact used only for the second
3323 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3324 attention to any local part prefix or suffix that was matched by the
3325 router that accepted the address. It now behaves in the same way as it
3326 does for delivery: the affixes are removed from the local part unless
3327 rcpt_include_affixes is set on the transport.
3329 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3330 timeout during the DATA phase of an incoming message.
3332 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3333 applied a patch sanctioned by the Sieve maintainer.
3335 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3336 the uid or gid is negative. A case of a negative gid caused this to be
3337 noticed. The fix allows for either to be negative.
3339 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3340 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3343 PH/36 The change PH/12 above was broken. Fixed it.
3345 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3346 the grounds that routing the same address twice would always produce the
3347 same answer. This might have been true once, but it is certainly no
3348 longer true now. Routing a child address may depend on the previous
3349 routing that produced that child. Some complicated redirection strategies
3350 went wrong when messages had multiple recipients, and made Exim's
3351 behaviour dependent on the order in which the addresses were given.
3353 I have moved the duplicate checking until after the routing is complete.
3354 Exim scans the addresses that are assigned to local and remote
3355 transports, and removes any duplicates. This means that more work will be
3356 done, as duplicates will always all be routed, but duplicates are
3357 presumably rare, so I don't expect this is of any significance.
3359 For deliveries to pipes, files, and autoreplies, the duplicate checking
3360 still happens during the routing process, since they are not going to be
3363 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3364 It corrects a timeout issue with spamd. This is Ian's comment: "The
3365 background is that sometimes spamd either never reads data from a
3366 connection it has accepted, or it never writes response data. The exiscan
3367 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3368 blindly assumes that writes won't block so it may never time out."
3370 PH/39 Allow G after quota size as well as K and M.
3372 PH/40 The value set for $authenticated_id in an authenticator may not contain
3373 binary zeroes or newlines because the value is written to log lines and
3374 to spool files. There was no check on this. Now the value is run through
3375 the string_printing() function so that such characters are converted to
3376 printable escape sequences.
3378 PH/41 $message_linecount is a new variable that contains the total number of
3379 lines in the message. Compare $body_linecount, which is the count for the
3382 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3383 bounce and delay warning messages, except in certain special cases, which
3386 (a) An SMTP error message from a remote host;
3387 (b) A message specified in a :fail: redirection;
3388 (c) A message specified in a "fail" command in a system filter;
3389 (d) A message specified in a FAIL return from the queryprogram router;
3390 (e) A message specified by the cannot_route_message router option.
3392 In these cases only, Exim does include the error details in bounce and
3393 warning messages. There are also a few cases where bland messages such
3394 as "unrouteable address" or "local delivery error" are given.
3396 PH/43 $value is now also set for the "else" part of a ${run expansion.
3398 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3399 being worked on, but at least Exim now implements the latest version to
3402 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3403 process to complete was treated as a delivery failure, a timeout while
3404 writing the message to the pipe was logged, but erroneously treated as a
3405 successful delivery. Such timeouts include transport filter timeouts. For
3406 consistency with the overall process timeout, these timeouts are now
3407 treated as errors, giving rise to delivery failures by default. However,
3408 there is now a new Boolean option for the pipe transport called
3409 timeout_defer, which, if set TRUE, converts the failures into defers for
3410 both kinds of timeout. A transport filter timeout is now identified in
3413 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3414 systems where "make" and "gmake" are different, calling "gmake" at top
3415 level broke things. I've arranged for the value of $(MAKE) to be passed
3416 from the Makefile to this script so that it can call the same version of
3420 A note about Exim versions 4.44 and 4.50
3421 ----------------------------------------
3423 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3424 changes of various kinds. As a consequence, a big documentation update was
3425 needed. This delayed the release for rather longer than seemed good, especially
3426 in the light of a couple of (minor) security issues. Therefore, the changes
3427 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3428 release. So 4.44 and 4.50 are in effect two different branches that both start
3431 I have left the 4.50 change log unchanged; it contains all the changes since
3432 4.43. The change log for 4.44 is below; many of its items are identical to
3433 those for 4.50. This seems to be the most sensible way to preserve the
3434 historical information.
3440 1. Minor wording change to the doc/README.SIEVE file.
3442 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3443 computation of the current number of files was incorrect.
3445 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3446 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3449 4. Give more explanation in the error message when the command for a transport
3450 filter fails to execute.
3452 5. There are several places where Exim runs a non-Exim command in a
3453 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3454 was being done only for the command run by the queryprogram router. It is
3455 now done for all such subprocesses. The other cases are: ${run, transport
3456 filters, and the commands run by the lmtp and pipe transports.
3458 6. Added CONFIGURE_GROUP build-time option.
3460 7. Some older OS have a limit of 256 on the maximum number of file
3461 descriptors. Exim was using setrlimit() to set 1000 as a large value
3462 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3463 systems. I've change it so that if it can't get 1000, it tries for 256.
3465 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3466 was an oversight, and furthermore, ever since the addition of extra
3467 controls (e.g. 4.43/32), the checks on when to allow different forms of
3468 "control" were broken. There should now be diagnostics for all cases when a
3469 control that does not make sense is encountered.
3471 9. Added the /retain_sender option to "control=submission".
3473 10. $recipients is now available in the predata ACL (oversight).
3475 11. Tidy the search cache before the fork to do a delivery from a message
3476 received from the command line. Otherwise the child will trigger a lookup
3477 failure and thereby defer the delivery if it tries to use (for example) a
3478 cached ldap connection that the parent has called unbind on.
3480 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3481 of $address_data from the recipient verification was clobbered by the
3482 sender verification.
3484 13. The value of address_data from a sender verification is now available in
3485 $sender_address_data in subsequent conditions in the ACL statement.
3487 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3489 15. Added a new option "connect=<time>" to callout options, to set a different
3492 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3493 was its contents. (It was OK if the option was not defined at all.)
3495 17. A "Completed" log line is now written for messages that are removed from
3496 the spool by the -Mrm option.
3498 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3499 information about exactly what failed.
3501 19. Added -dd to debug only the daemon process.
3503 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3504 handles timeouts, both on the server side and network timeouts. Renamed the
3505 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3507 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3508 It is now set to "smtps".
3510 22. $host_address is now set to the target address during the checking of
3511 ignore_target_hosts.
3513 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3514 being passed; this would have caused $sender_host_name to have been used if
3515 matching the list had actually called for a host name (not very likely,
3516 since this list is usually IP addresses). A host name is now passed as
3519 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3520 code that specifies a non-restarting handler (typically sigaction() in
3521 modern systems) in an attempt to fix a rare and obscure crash bug.
3523 25. Narrowed the window for a race in the daemon that could cause it to ignore
3524 SIGCHLD signals. This is not a major problem, because they are used only to
3525 wake it up if nothing else does.
3527 26. A malformed maildirsize file could cause Exim to calculate negative values
3528 for the mailbox size or file count. Odd effects could occur as a result.
3529 The maildirsize information is now recalculated if the size or filecount
3532 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3533 support for a long time. Removed HAVE_SYS_VFS_H.
3535 28. Installed the latest version of exipick from John Jetmore.
3537 29. In an address list, if the pattern was not a regular expression, an empty
3538 subject address (from a bounce message) matched only if the pattern was an
3539 empty string. Non-empty patterns were not even tested. This was the wrong
3540 because it is perfectly reasonable to use an empty address as part of a
3541 database query. An empty address is now tested by patterns that are
3542 lookups. However, all the other forms of pattern expect the subject to
3543 contain a local part and a domain, and therefore, for them, an empty
3544 address still always fails if the pattern is not itself empty.
3546 30. Exim went into a mad DNS loop when attempting to do a callout where the
3547 host was specified on an smtp transport, and looking it up yielded more
3548 than one IP address.
3550 31. Re-factored the code for checking spool and log partition space into a
3551 function that finds that data and another that does the check. The former
3552 is then used to implement four new variables: $spool_space, $log_space,
3553 $spool_inodes, and $log_inodes.
3555 32. The RFC2047 encoding function was originally intended for short strings
3556 such as real names; it was not keeping to the 75-character limit for
3557 encoded words that the RFC imposes. It now respects the limit, and
3558 generates multiple encoded words if necessary. To be on the safe side, I
3559 have increased the buffer size for the ${rfc2047: expansion operator from
3562 33. It is now permitted to omit both strings after an "if" condition; if the
3563 condition is true, the result is "true". As before, when the second string
3564 is omitted, a false condition yields an empty string. This makes it less
3565 cumbersome to write custom ACL and router conditions.
3567 34. Failure to deliver a bounce message always caused it to be frozen, even if
3568 there was an errors_to setting on the router. The errors_to setting is now
3571 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3572 canonical form (fully expanded) before being placed in
3573 $sender_host_address.
3575 36. The table in the code that translates DNS record types into text (T_A to
3576 "A" for instance) was missing entries for NS and CNAME. It is just possible
3577 that this could have caused confusion if both these types were looked up
3578 for the same domain, because the text type is used as part of Exim's
3579 per-process caching. But the chance of anyone hitting this buglet seems
3582 37. The dnsdb lookup has been extended in a number of ways.
3584 (1) There is a new type, "zns", which walks up the domain tree until it
3585 finds some nameserver records. It should be used with care.
3587 (2) There is a new type, "mxh", which is like "mx" except that it returns
3588 just the host names, not the priorities.
3590 (3) It is now possible to give a list of domains (or IP addresses) to be
3591 looked up. The behaviour when one of the lookups defers can be
3592 controlled by a keyword.
3594 (4) It is now possible to specify the separator character for use when
3595 multiple records are returned.
3597 38. The dnslists ACL condition has been extended: it is now possible to supply
3598 a list of IP addresses and/or domains to be looked up in a particular DNS
3601 39. Added log_selector=+queue_time_overall.
3603 40. When running the queue in the test harness, wait just a tad after forking a
3604 delivery process, to get repeatability of debugging output.
3606 41. Include certificate and key file names in error message when GnuTLS fails
3607 to set them up, because the GnuTLS error message doesn't include the name
3608 of the failing file when there is a problem reading it.
3610 42. Allow both -bf and -bF in the same test run.
3612 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3614 44. The "Exiscan patch" is now merged into the mainline Exim source.
3616 45. Sometimes the final signoff response after QUIT could fail to get
3617 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3618 < 0 before doing a fflush(). This bug looks as though it goes back to the
3619 introduction of TLS in release 3.20, but "sometimes" must have been rare
3620 because the tests only now provoked it.
3622 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3623 (this can affect the format of dates).
3625 47. exim_tidydb, when checking for the continued existence of a message for
3626 which it has found a message-specific retry record, was not finding
3627 messages that were in split spool directories. Consequently, it was
3628 deleting retry records that should have stayed in existence.
3630 48. Steve fixed some bugs in eximstats.
3632 49. The SPA authentication driver was not abandoning authentication and moving
3633 on to the next authenticator when an expansion was forced to fail,
3634 contradicting the general specification for all authenticators. Instead it
3635 was generating a temporary error. It now behaves as specified.
3637 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3638 (the order specifies the preference for clients). The order is now AES256,
3639 AES128, 3DES, ARCFOUR128.
3641 51. Small patch to Sieve code - explicitly set From: when generating an
3644 52. Exim crashed if a remote delivery caused a very long error message to be
3645 recorded - for instance if somebody sent an entire SpamAssassin report back
3646 as a large number of 550 error lines. This bug was coincidentally fixed by
3647 increasing the size of one of Exim's internal buffers (big_buffer) that
3648 happened as part of the Exiscan merge. However, to be on the safe side, I
3649 have made the code more robust (and fixed the comments that describe what
3652 53. Now that there can be additional text after "Completed" in log lines (if
3653 the queue_time_overall log selector is set), a one-byte patch to exigrep
3654 was needed to allow it to recognize "Completed" as not the last thing in
3657 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3658 patch that reportedly fixes this has been added. I am not expert enough to
3659 create a test for it. This is what the patch creator wrote:
3661 "I found a little strange behaviour of ldap code when working with
3662 Windows 2003 AD Domain, where users was placed in more than one
3663 Organization Units. When I tried to give exim partial DN, the exit code
3664 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3665 But simultaneously result of request was absolutely normal ldap result,
3666 so I produce this patch..."
3668 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3669 REFERENCE, so I have modified the code to exclude the patch when that macro
3672 55. Some experimental protocols are using DNS PTR records for new purposes. The
3673 keys for these records are domain names, not reversed IP addresses. The
3674 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3675 leaves it alone. Component reversal etc. now happens only for IP addresses.
3678 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3680 57. Double the size of the debug message buffer (to 2048) so that more of very
3681 long debug lines gets shown.
3683 58. The exicyclog utility now does better if the number of log files to keep
3684 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3686 59. Two changes related to the smtp_active_hostname option:
3688 (1) $smtp_active_hostname is now available as a variable.
3689 (2) The default for smtp_banner uses $smtp_active_hostname instead
3690 of $primary_hostname.
3692 60. The host_aton() function is supposed to be passed a string that is known
3693 to be a valid IP address. However, in the case of IPv6 addresses, it was
3694 not checking this. This is a hostage to fortune. Exim now panics and dies
3695 if the condition is not met. A case was found where this could be provoked
3696 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3697 components; fortuitously, this particular loophole had already been fixed
3698 by change 4.50/55 above.
3700 If there are any other similar loopholes, the new check in host_aton()
3701 itself should stop them being exploited. The report I received stated that
3702 data on the command line could provoke the exploit when Exim was running as
3703 exim, but did not say which command line option was involved. All I could
3704 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3705 running as the user.
3708 61. There was a buffer overflow vulnerability in the SPA authentication code
3709 (which came originally from the Samba project). I have added a test to the
3710 spa_base64_to_bits() function which I hope fixes it.
3713 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3714 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3716 63. The daemon start-up calls getloadavg() while still root for those OS that
3717 need the first call to be done as root, but it missed one case: when
3718 deliver_queue_load_max is set with deliver_drop_privilege. This is
3719 necessary for the benefit of the queue runner, because there is no re-exec
3720 when deliver_drop_privilege is set.
3722 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3723 This has been fixed.
3725 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3726 was in use, was not putting the data itself into the right store pool;
3727 consequently, it could be overwritten for a subsequent message in the same
3728 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3731 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3733 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3734 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3735 the function were treating the return as a boolean value, which happened to
3736 work because 0=false and not-0=true, but is not correct code.
3738 68. The host_aton() function was not handling scoped IPv6 addresses (those
3739 with, for example, "%eth0" on the end) correctly.
3741 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3742 negated items (that is, ~something) in unsigned ints. Some compilers
3743 apparently mutter when there is no cast.
3745 70. If an address verification called from an ACL failed, and did not produce a
3746 user-specific message (i.e. there was only a "system" message), nothing was
3747 put in $acl_verify_message. In this situation, it now puts the system
3750 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3751 session; change 4.31/43 added the unwanted input to the log line - except
3752 that it did not do this in the start of session case. It now does.
3754 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3755 This is wrong and can cause the other end to generate a synchronization
3756 error if it is another Exim or anything else that does the synchronization
3757 check. A QUIT command is no longer sent after a timeout.
3759 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3760 during host lookups.
3762 74. The defer_ok option of callout verification was not working if it was used
3763 when verifying addresses in header lines, that is, for this case:
3765 verify = header_sender/callout=defer_ok
3767 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3768 those file descriptors could be used for SMTP connections. If anything
3769 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3770 could be sent to the SMTP client, causing chaos. The daemon now opens
3771 stdin, stdout, and stderr to /dev/null when it puts itself into the
3774 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3775 log by default. The user can override this with suitable Perl magic.
3777 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3778 the log message when discard triggers, was not working for the DATA ACL or
3779 for the non-SMTP ACL.
3781 78. Error message wording change in sieve.c.
3783 79. If smtp_accept_max_per_host was set, the number of connections could be
3784 restricted to fewer than expected, because the daemon was trying to set up
3785 a new connection before checking whether the processes handling previous
3786 connections had finished. The check for completed processes is now done
3787 earlier. On busy systems, this bug wouldn't be noticed because something
3788 else would have woken the daemon, and it would have reaped the completed
3791 80. If a message was submitted locally by a user whose login name contained one
3792 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3793 It caused a spool format error. I have fixed the spool reading code. A
3794 related problem was that the "from" clause in the Received: line became
3795 illegal because of the space(s). It is now covered by ${quote_local_part.
3797 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3800 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3802 83. Patch from Sieve maintainer to fix unterminated string problem in
3803 "vacation" handling.
3805 84. Some minor changes to the Linux configuration files to help with other
3806 OS variants using glibc.
3808 85. One more patch for Sieve to update vacation handling to latest spec.
3811 ----------------------------------------------------
3812 See the note above about the 4.44 and 4.50 releases.
3813 ----------------------------------------------------
3819 1. Change 4.43/35 introduced a bug that caused file counts to be
3820 incorrectly computed when quota_filecount was set in an appendfile
3823 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3824 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3827 3. Give more explanation in the error message when the command for a transport
3828 filter fails to execute.
3830 4. There are several places where Exim runs a non-Exim command in a
3831 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3832 was being done only for the command run by the queryprogram router. It is
3833 now done for all such subprocesses. The other cases are: ${run, transport
3834 filters, and the commands run by the lmtp and pipe transports.
3836 5. Some older OS have a limit of 256 on the maximum number of file
3837 descriptors. Exim was using setrlimit() to set 1000 as a large value
3838 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3839 systems. I've change it so that if it can't get 1000, it tries for 256.
3841 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3842 was an oversight, and furthermore, ever since the addition of extra
3843 controls (e.g. 4.43/32), the checks on when to allow different forms of
3844 "control" were broken. There should now be diagnostics for all cases when a
3845 control that does not make sense is encountered.
3847 7. $recipients is now available in the predata ACL (oversight).
3849 8. Tidy the search cache before the fork to do a delivery from a message
3850 received from the command line. Otherwise the child will trigger a lookup
3851 failure and thereby defer the delivery if it tries to use (for example) a
3852 cached ldap connection that the parent has called unbind on.
3854 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3855 of $address_data from the recipient verification was clobbered by the
3856 sender verification.
3858 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3859 was its contents. (It was OK if the option was not defined at all.)
3861 11. A "Completed" log line is now written for messages that are removed from
3862 the spool by the -Mrm option.
3864 12. $host_address is now set to the target address during the checking of
3865 ignore_target_hosts.
3867 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3868 being passed; this would have caused $sender_host_name to have been used if
3869 matching the list had actually called for a host name (not very likely,
3870 since this list is usually IP addresses). A host name is now passed as
3873 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3874 code that specifies a non-restarting handler (typically sigaction() in
3875 modern systems) in an attempt to fix a rare and obscure crash bug.
3877 15. Narrowed the window for a race in the daemon that could cause it to ignore
3878 SIGCHLD signals. This is not a major problem, because they are used only to
3879 wake it up if nothing else does.
3881 16. A malformed maildirsize file could cause Exim to calculate negative values
3882 for the mailbox size or file count. Odd effects could occur as a result.
3883 The maildirsize information is now recalculated if the size or filecount
3886 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3887 support for a long time. Removed HAVE_SYS_VFS_H.
3889 18. Updated exipick to current release from John Jetmore.
3891 19. Allow an empty sender to be matched against a lookup in an address list.
3892 Previously the only cases considered were a regular expression, or an
3895 20. Exim went into a mad DNS lookup loop when doing a callout where the
3896 host was specified on the transport, if the DNS lookup yielded more than
3899 21. The RFC2047 encoding function was originally intended for short strings
3900 such as real names; it was not keeping to the 75-character limit for
3901 encoded words that the RFC imposes. It now respects the limit, and
3902 generates multiple encoded words if necessary. To be on the safe side, I
3903 have increased the buffer size for the ${rfc2047: expansion operator from
3906 22. Failure to deliver a bounce message always caused it to be frozen, even if
3907 there was an errors_to setting on the router. The errors_to setting is now
3910 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3911 canonical form (fully expanded) before being placed in
3912 $sender_host_address.
3914 24. Updated eximstats to version 1.33
3916 25. Include certificate and key file names in error message when GnuTLS fails
3917 to set them up, because the GnuTLS error message doesn't include the name
3918 of the failing file when there is a problem reading it.
3920 26. Expand error message when OpenSSL has problems setting up cert/key files.
3923 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3924 (this can affect the format of dates).
3926 28. exim_tidydb, when checking for the continued existence of a message for
3927 which it has found a message-specific retry record, was not finding
3928 messages that were in split spool directories. Consequently, it was
3929 deleting retry records that should have stayed in existence.
3931 29. eximstats updated to version 1.35
3932 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3933 1.35 - bugfix such that pie charts by volume are generated correctly
3935 30. The SPA authentication driver was not abandoning authentication and moving
3936 on to the next authenticator when an expansion was forced to fail,
3937 contradicting the general specification for all authenticators. Instead it
3938 was generating a temporary error. It now behaves as specified.
3940 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3941 (the order specifies the preference for clients). The order is now AES256,
3942 AES128, 3DES, ARCFOUR128.
3944 31. Small patch to Sieve code - explicitly set From: when generating an
3947 32. Exim crashed if a remote delivery caused a very long error message to be
3948 recorded - for instance if somebody sent an entire SpamAssassin report back
3949 as a large number of 550 error lines. This bug was coincidentally fixed by
3950 increasing the size of one of Exim's internal buffers (big_buffer) that
3951 happened as part of the Exiscan merge. However, to be on the safe side, I
3952 have made the code more robust (and fixed the comments that describe what
3955 33. Some experimental protocols are using DNS PTR records for new purposes. The
3956 keys for these records are domain names, not reversed IP addresses. The
3957 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3958 leaves it alone. Component reversal etc. now happens only for IP addresses.
3961 34. The host_aton() function is supposed to be passed a string that is known
3962 to be a valid IP address. However, in the case of IPv6 addresses, it was
3963 not checking this. This is a hostage to fortune. Exim now panics and dies
3964 if the condition is not met. A case was found where this could be provoked
3965 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3966 components; fortuitously, this particular loophole had already been fixed
3967 by change 4.50/55 or 4.44/33 above.
3969 If there are any other similar loopholes, the new check in host_aton()
3970 itself should stop them being exploited. The report I received stated that
3971 data on the command line could provoke the exploit when Exim was running as
3972 exim, but did not say which command line option was involved. All I could
3973 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3974 running as the user.
3977 35. There was a buffer overflow vulnerability in the SPA authentication code
3978 (which came originally from the Samba project). I have added a test to the
3979 spa_base64_to_bits() function which I hope fixes it.
3982 36. The daemon start-up calls getloadavg() while still root for those OS that
3983 need the first call to be done as root, but it missed one case: when
3984 deliver_queue_load_max is set with deliver_drop_privilege. This is
3985 necessary for the benefit of the queue runner, because there is no re-exec
3986 when deliver_drop_privilege is set.
3988 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3989 was in use, was not putting the data itself into the right store pool;
3990 consequently, it could be overwritten for a subsequent message in the same
3991 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3994 38. Sometimes the final signoff response after QUIT could fail to get
3995 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3996 < 0 before doing a fflush(). This bug looks as though it goes back to the
3997 introduction of TLS in release 3.20, but "sometimes" must have been rare
3998 because the tests only now provoked it.
4004 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4005 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4006 Now it returns an integer. A number of calls were still expecting a T/F
4007 return. Fortuitously, in all cases, the tests worked in OK situations,
4008 which is the norm. However, things would have gone wrong on any write
4009 failures on the smtp file descriptor. This function is used when sending
4010 messages over SMTP and also when doing verify callouts.
4012 2. When Exim is called to do synchronous delivery of a locally submitted
4013 message (the -odf or -odi options), it no longer closes stderr before doing
4016 3. Implemented the mua_wrapper option.
4018 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4020 5. Implemented the functions header_remove(), header_testname(),
4021 header_add_at_position(), and receive_remove_recipient(), and exported them
4024 6. If an ACL "warn" statement specified the addition of headers, Exim already
4025 inserted X-ACL-Warn: at the start if there was no header name. However, it
4026 was not making this test for the second and subsequent header lines if
4027 there were newlines in the string. This meant that an invalid header could
4028 be inserted if Exim was badly configured.
4030 7. Allow an ACL "warn" statement to add header lines at the start or after all
4031 the Received: headers, as well as at the end.
4033 8. Added the rcpt_4xx retry error code.
4035 9. Added postmaster_mailfrom=xxx to callout verification option.
4037 10. Added mailfrom=xxxx to the callout verification option, for verify=
4040 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4041 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4043 12. Inserted some casts to stop certain compilers warning when using pointer
4044 differences as field lengths or precisions in printf-type calls (mostly
4045 affecting debugging statements).
4047 13. Added optional readline() support for -be (dynamically loaded).
4049 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4050 same clock tick as a message's arrival, so that its received time was the
4051 same as the "first fail" time on the retry record, and that message
4052 remained on the queue past the ultimate address timeout, every queue runner
4053 would try a delivery (because it was past the ultimate address timeout) but
4054 after another failure, the ultimate address timeout, which should have then
4055 bounced the address, did not kick in. This was a "< instead of <=" error;
4056 in most cases the first failure would have been in the next clock tick
4057 after the received time, and all would be well.
4059 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4060 being recognized when the domain list was tested by the match_domain
4061 condition in an expansion string.
4063 16. Added the ${str2b64: operator.
4065 17. Exim was always calling setrlimit() to set a large limit for the number of
4066 processes, without checking whether the existing limit was already
4067 adequate. (It did check for the limit on file descriptors.) Furthermore,
4068 errors from getrlimit() and setrlimit() were being ignored. Now they are
4069 logged to the main and panic logs, but Exim does carry on, to try to do its
4070 job under whatever limits there are.
4072 18. Imported PCRE 5.0.
4074 19. Trivial typo in log message " temporarily refused connection" (the leading
4077 20. If the log selector return_path_on_delivery was set and an address was
4078 redirected to /dev/null, the delivery process crashed because it assumed
4079 that a return path would always be set for a "successful" delivery. In this
4080 case, the whole delivery is bypassed as an optimization, and therefore no
4083 21. Internal re-arrangement: the function for sending a challenge and reading
4084 a response while authentication was assuming a zero-terminated challenge
4085 string. It's now changed to take a pointer and a length, to allow for
4086 binary data in such strings.
4088 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4090 23. Exim was not respecting finduser_retries when seeking the login of the
4091 uid under which it was called; it was always trying 10 times. (The default
4092 setting of finduser_retries is zero.) Also, it was sleeping after the final
4093 failure, which is pointless.
4095 24. Implemented tls_on_connect_ports.
4097 25. Implemented acl_smtp_predata.
4099 26. If the domain in control=submission is set empty, Exim assumes that the
4100 authenticated id is a complete email address when it generates From: or
4101 Sender: header lines.
4103 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4104 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4105 chown and chgrp in /bin and hostname in /usr/bin.
4107 28. Exim was keeping the "process log" file open after each use, just as it
4108 does for the main log. This opens the possibility of it remaining open for
4109 long periods when the USR1 signal hits a daemon. Occasional processlog
4110 errors were reported, that could have been caused by this. Anyway, it seems
4111 much more sensible not to leave this file open at all, so that is what now
4114 29. The long-running daemon process does not normally write to the log once it
4115 has entered its main loop, and it closes the log before doing so. This is
4116 so that log files can straightforwardly be renamed and moved. However,
4117 there are a couple of unusual error situations where the daemon does write
4118 log entries, and I had neglected to close the log afterwards.
4120 30. The text of an SMTP error response that was received during a remote
4121 delivery was being truncated at 512 bytes. This is too short for some of
4122 the long messages that one sometimes sees. I've increased the limit to
4125 31. It is now possible to make retry rules that apply only when a message has a
4126 specific sender, in particular, an empty sender.
4128 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4129 it possible to be selective about when SMTP synchronization is enforced.
4131 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4133 32. Implemented hosts_connection_nolog.
4135 33. Added an ACL for QUIT.
4137 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4140 35. Added mailbox_size and mailbox_filecount to appendfile.
4142 36. Added control = no_multiline_responses to ACLs.
4144 37. There was a bug in the logic of the code that waits for the clock to tick
4145 in the case where the clock went backwards by a substantial amount such
4146 that the microsecond fraction of "now" was more than the microsecond
4147 fraction of "then" (but the whole seconds number was less).
4149 38. Added support for the libradius Radius client library this is found on
4150 FreeBSD (previously only the radiusclient library was supported).
4156 1. When certain lookups returned multiple values in the form name=value, the
4157 quoting of the values was not always being done properly. Specifically:
4158 (a) If the value started with a double quote, but contained no whitespace,
4160 (b) If the value contained whitespace other than a space character (i.e.
4161 tabs or newlines or carriage returns) it was not quoted.
4162 This fix has been applied to the mysql and pgsql lookups by writing a
4163 separate quoting function and calling it from the lookup code. The fix
4164 should probably also be applied to nisplus, ibase and oracle lookups, but
4165 since I cannot test any of those, I have not disturbed their existing code.
4167 2. A hit in the callout cache for a specific address caused a log line with no
4168 reason for rejecting RCPT. Now it says "Previous (cached) callout
4169 verification failure".
4171 3. There was an off-by-one bug in the queryprogram router. An over-long
4172 return line was truncated at 256 instead of 255 characters, thereby
4173 overflowing its buffer with the terminating zero. As well as fixing this, I
4174 have increased the buffer size to 1024 (and made a note to document this).
4176 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4177 when Exim is waiting for an SMTP response from a remote server, Exim
4178 restarts its select() call on the socket, thereby resetting its timeout.
4179 This is not a problem when such interrupts are rare. Somebody set up a cron
4180 job to run exiwhat every 2 minutes, which is less than the normal select()
4181 timeout (5 or 10 minutes). This meant that the select() timeout never
4182 kicked in because it was always reset. I have fixed this by comparing the
4183 time when an interrupt arrives with the time at the start of the first call
4184 to select(). If more time than the timeout has elapsed, the interrupt is
4185 treated as a timeout.
4187 5. Some internal re-factoring in preparation for the addition of Sieve
4188 extensions (by MH). In particular, the "personal" test is moved to a
4189 separate function, and given an option for scanning Cc: and Bcc: (which is
4190 not set for Exim filters).
4192 6. When Exim created an email address using the login of the caller as the
4193 local part (e.g. when creating a From: or Sender: header line), it was not
4194 quoting the local part when it contained special characters such as @.
4196 7. Installed new OpenBSD configuration files.
4198 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4199 try to make them clearer.
4201 9. Callout options, other than the timeout value, were being ignored when
4202 verifying sender addresses in header lines. For example, when using
4204 verify = header_sender/callout=no_cache
4206 the cache was (incorrectly) being used.
4208 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4209 only the Cygwin environment.
4211 11. When return_path_on_delivery was set as a log selector, if different remote
4212 addresses in the same message used different return paths and parallel
4213 remote delivery occurred, the wrong values would sometimes be logged.
4214 (Whenever a remote delivery process finished, the return path value from
4215 the most recently started remote delivery process was logged.)
4217 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4218 lines when AUTH and/or TLS are in use. This is the "received protocol"
4219 field. Exim used to use "asmtp" for authenticated SMTP, without any
4220 indication (in the protocol name) for TLS use. Now it follows the RFC and
4221 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4222 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4223 names appear in log lines as well as in Received: header lines.
4225 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4226 extensions, and comparison tests, and to fix some bugs.
4228 14. Changes to the "personal" filter test:
4230 (1) The test was buggy in that it was just doing the equivalent of
4231 "contains" tests on header lines. For example, if a user's address was
4232 anne@some.where, the "personal" test would incorrectly be true for
4234 To: susanne@some.where
4236 This test is now done by extracting each address from the header in turn,
4237 and checking the entire address. Other tests that are part of "personal"
4238 are now done using regular expressions (for example, to check local parts
4239 of addresses in From: header lines).
4241 (2) The list of non-personal local parts in From: addresses has been
4242 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4243 taken from the Sieve specification recommendations.
4245 (3) If the message contains any header line starting with "List-" it is
4246 treated as non-personal.
4248 (4) The test for "circular" in the Subject: header line has been removed
4249 because it now seems ill-conceived.
4251 15. Minor typos in src/EDITME comments corrected.
4253 16. Installed latest exipick from John Jetmore.
4255 17. If headers_add on a router specified a text string that was too long for
4256 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4257 of string_sprintf() is now avoided.
4259 18. $message_body_size was not set (it was always zero) when running the DATA
4260 ACL and the local_scan() function.
4262 19. For the "mail" command in an Exim filter, no default was being set for
4263 the once_repeat time, causing a random time value to be used if "once" was
4264 specified. (If the value happened to be <= 0, no repeat happened.) The
4265 default is now 0s, meaning "never repeat". The "vacation" command was OK
4266 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4267 (I found it when inspecting the code).
4269 20. There is now an overall timeout for performing a callout verification. It
4270 defaults to 4 times the callout timeout, which applies to individual SMTP
4271 commands during the callout. The overall timeout applies when there is more
4272 than one host that can be tried. The timeout is checked before trying the
4273 next host. This prevents very long delays if there are a large number of
4274 hosts and all are timing out (e.g. when the network connections are timing
4275 out). The value of the overall timeout can be changed by specifying an
4276 additional sub-option for "callout", called "maxwait". For example:
4278 verify = sender/callout=5s,maxwait=20s
4280 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4281 to the end before writing, but this should make it even safer).
4283 22. Exim was forgetting that it had advertised PIPELINING for the second and
4284 subsequent messages on an SMTP connection. It was also not resetting its
4285 memory on STARTTLS and an internal HELO.
4287 23. When Exim logs an SMTP synchronization error within a session, it now
4288 records whether PIPELINING has been advertised or not.
4290 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4291 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4292 rather than long int.
4294 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4296 26. Added the never_mail option to autoreply.
4302 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4303 crash if the getsockname() call failed; this can happen if a connection is
4304 closed very soon after it is established. The problem was simply in the
4305 order in which certain operations were done, causing Exim to try to write
4306 to the SMTP stream before it had set up the file descriptor. The bug has
4307 been fixed by making things happen in the correct order.
4313 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4314 before the connection was closed, thus losing the rejection response.
4316 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4317 some early Solaris releases, but causes trouble in current releases where
4318 socklen_t is defined.
4320 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4323 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4326 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4327 that releasing the top of it at the end releases what was used for sub-
4328 expansions (unless the block got too big). However, discard this block if
4329 the first thing is a variable or header, so that we can use its block when
4330 it is dynamic (useful for very large $message_headers, for example).
4332 6. Lookups now cache *every* query, not just the most recent. A new, separate
4333 store pool is used for this. It can be recovered when all lookup caches are
4334 flushed. Lookups now release memory at the end of their result strings.
4335 This has involved some general refactoring of the lookup sources.
4337 7. Some code has been added to the store_xxx() functions to reduce the amount
4338 of flapping under certain conditions.
4340 8. log_incoming_interface used to affect only the <= reception log lines. Now
4341 it causes the local interface and port to be added to several more SMTP log
4342 lines, for example "SMTP connection from", and rejection lines.
4344 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4346 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4348 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4349 could be overwritten at the end of the current message (or the start of a
4350 new message if it was set in a HELO ACL). The value is now preserved for
4351 the duration of the SMTP connection.
4353 12. If a transport had a headers_rewrite setting, and a matching header line
4354 contained an unqualified address, that address was qualified, even if it
4355 did not match any rewriting rules. The underlying bug was that the values
4356 of the flags that permit the existence of unqualified sender and recipient
4357 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4358 non-local messages, and by -bnq for local messages) were not being
4359 preserved with the message after it was received.
4361 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4362 "next input=" as part of the text comprising the host identity instead of
4363 the correct text. The code was using the same buffer for two different
4364 strings. However, depending on which order the printing function evaluated
4365 its arguments, the bug did not always show up. Under Linux, for example, my
4366 test suite worked just fine.
4368 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4369 doesn't work with some older versions of Perl. It has been changed to "my",
4370 which in any case is probably the better facility to use.
4372 15. A really picky compiler found some instances of statements for creating
4373 error messages that either had too many or two few arguments for the format
4376 16. The size of the buffer for calls to the DNS resolver has been increased
4377 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4378 for addresses that have a lot of PTR records. This alleviates a problem; it
4379 does not fully solve it.
4381 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4382 buffer now truncates the list and logs the incident, which is the same
4383 action as happens when Exim is looking up a host name and its aliases.
4384 Previously in this situation something unpredictable would happen;
4385 sometimes it was "internal error: store_reset failed".
4387 18. If a server dropped the connection unexpectedly when an Exim client was
4388 using GnuTLS and trying to read a response, the client delivery process
4389 crashed while trying to generate an error log message.
4391 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4392 string, for example:
4394 warn message = H1: something\nH2: something
4396 the text was added as a single header line from Exim's point of view
4397 though it ended up OK in the delivered message. However, searching for the
4398 second and subsequent header lines using $h_h2: did not work. This has been
4399 fixed. Similarly, if a system filter added multiple headers in this way,
4400 the routers could not see them.
4402 20. Expanded the error message when iplsearch is called with an invalid key to
4403 suggest using net-iplsearch in a host list.
4405 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4406 ACLs are no longer actually imposed (and a message to that effect is
4409 22. If a "gecos" field in a passwd entry contained escaped characters, in
4410 particular, if it contained a \" sequence, Exim got it wrong when building
4411 a From: or a Sender: header from that name. A second bug also caused
4412 incorrect handling when an unquoted " was present following a character
4413 that needed quoting.
4415 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4416 was not being matched caselessly.
4418 24. Arranged for all hyphens in the exim.8 source to be escaped with
4421 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4422 properly. Recipient callout cache records were still being keyed to include
4423 the sender, even when use_sender was set false. This led to far more
4424 callouts that were necessary. The sender is no longer included in the key
4425 when use_sender is false.
4427 26. Added "control = submission" modifier to ACLs.
4429 27. Added the ${base62d: operator to decode base 62 numbers.
4431 28. dnsdb lookups can now access SRV records.
4433 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4434 the configuration file.
4436 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4437 (-v) mode. This makes the output for a verbose queue run more intelligible.
4439 31. Added a use_postmaster feature to recipient callouts.
4441 32. Added the $body_zerocount variable, containing the number of binary zero
4442 bytes in the message body.
4444 33. The time of last modification of the "new" subdirectory is now used as the
4445 "mailbox time last read" when there is a quota error for a maildir
4448 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4450 35. Added +ignore_unknown as a special item in host lists.
4452 36. Code for decoding IPv6 addresses in host lists is now included, even if
4453 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4454 address was recognized as an IP address, but was then not correctly decoded
4455 into binary, causing unexpected and incorrect effects when compared with
4462 1. Very minor rewording of debugging text in manualroute to say "list of
4463 hosts" instead of "hostlist".
4465 2. If verify=header_syntax was set, and a header line with an unqualified
4466 address (no domain) and a large number of spaces between the end of the
4467 name and the colon was received, the reception process suffered a buffer
4468 overflow, and (when I tested it) crashed. This was caused by some obsolete
4469 code that should have been removed. The fix is to remove it!
4471 3. When running in the test harness, delay a bit after writing a bounce
4472 message to get a bit more predictability in the log output.
4474 4. Added a call to search_tidyup() just before forking a reception process. In
4475 theory, someone could use a lookup in the expansion of smtp_accept_max_
4476 per_host which, without the tidyup, could leave open a database connection.
4478 5. Added the variables $recipient_data and $sender_data which get set from a
4479 lookup success in an ACL "recipients" or "senders" condition, or a router
4480 "senders" option, similar to $domain_data and $local_part_data.
4482 6. Moved the writing of debug_print from before to after the "senders" test
4485 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4486 problems for message scanning, either using a data ACL, or using
4487 local_scan() because the Received: header was not generated till after they
4488 were called (in order to set the time as the time of reception completion).
4489 I have revised the way this works. The header is now generated after the
4490 body is received, but before the ACL or local_scan() are called. After they
4491 are run, the timestamp in the header is updated.
4497 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4498 before starting a queue runner without re-exec. This happened only when
4499 deliver_drop_privilege was set or when the Exim user was set to root. The
4500 effect of the bug was that timeouts during subsequent deliveries caused
4501 crashes instead of being properly handled. The handler is now left at its
4502 default (and expected) setting.
4504 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4505 message, again when deliver_drop_privilege is set or Exim is run as root.
4506 The bug described in (1) was not present in this case, but the tidying up
4507 of the other signals was missing. I have made the two cases consistent.
4509 3. The ignore_target_hosts setting on a manualroute router was being ignored
4510 for hosts that were looked up using the /MX notation.
4512 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4515 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4516 operated on the sender address. After changing the $sender_address to <>
4517 for the sender address verify, Exim was re-instated it as the original
4518 (before rewriting) address, but remembering that it had rewritten it, so it
4519 wasn't rewriting it again. This bug also had the effect of breaking the
4520 sender address verification caching when the sender address was rewritten.
4522 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4523 This has been changed so that if the ip literal address matches
4524 ignore_target_hosts, the router declines.
4526 7. Added expansion conditions match_domain, match_address, and match_local_
4527 part (NOT match_host).
4529 8. The placeholder for the Received: header didn't have a length field set.
4531 9. Added code to Exim itself and to exim_lock to test for a specific race
4532 condition that could lead to file corruption when using MBX delivery. The
4533 issue is with the lockfile that is created in /tmp. If this file is removed
4534 after a process has opened it but before that process has acquired a lock,
4535 there is the potential for a second process to recreate the file and also
4536 acquire a lock. This could lead to two Exim processes writing to the file
4537 at the same time. The added code performs the same test as UW imapd; it
4538 checks after acquiring the lock that its file descriptor still refers to
4539 the same named file.
4541 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4542 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4545 11. Added the smtp_active_hostname option. If used, this will typically be made
4546 to depend on the incoming interface address. Because $interface_address is
4547 not set up until the daemon has forked a reception process, error responses
4548 that can happen earlier (such as "too many connections") no longer contain
4551 12. If an expansion in a condition on a "warn" statement fails because a lookup
4552 defers, the "warn" statement is abandoned, and the next ACL statement is
4553 processed. Previously this caused the whole ACL to be aborted.
4555 13. Added the iplsearch lookup type.
4557 14. Added ident_timeout as a log selector.
4559 15. Added tls_certificate_verified as a log selector.
4561 16. Added a global option tls_require_ciphers (compare the smtp transport
4562 option of the same name). This controls incoming TLS connections.
4564 17. I finally figured out how to make tls_require_ciphers do a similar thing
4565 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4566 before starting the TLS session.
4568 18. Tabs are now shown as \t in -bP output.
4570 19. If the log selector return_path_on_delivery was set, Exim crashed when
4571 bouncing a message because it had too many Received: header lines.
4573 20. If two routers both had headers_remove settings, and the first one included
4574 a superfluous trailing colon, the final name in the first list and the
4575 first name in the second list were incorrectly joined into one item (with a
4576 colon in the middle).
4582 1. Added -C and -D options to the exinext utility, mainly to make it easier
4583 to include in the automated testing, but these could be helpful when
4584 multiple configurations are in use.
4586 2. The exinext utility was not formatting the output nicely when there was
4587 an alternate port involved in the retry record key, nor when there was a
4588 message id as well (for retries that were specific to a specific message
4589 and a specific host). It was also confused by IPv6 addresses, because of
4590 the additional colons they contain. I have fixed the IPv4 problem, and
4591 patched it up to do a reasonable job for IPv6.
4593 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4594 delivery, the log line now contains "pipelined" if PIPELINING was used.
4596 4. An SMTP transport process used to panic and die if the bind() call to set
4597 an explicit outgoing interface failed. This has been changed; it is now
4598 treated in the same way as a connect() failure.
4600 5. A reference to $sender_host_name in the part of a conditional expansion
4601 that was being skipped was still causing a DNS lookup. This no longer
4604 6. The def: expansion condition was not recognizing references to header lines
4605 that used bh_ and bheader_.
4607 7. Added the _cache feature to named lists.
4609 8. The code for checking quota_filecount in the appendfile transport was
4610 allowing one more file than it should have been.
4612 9. For compatibility with Sendmail, the command line option
4620 and sets the incoming protocol and host name (for trusted callers). The
4621 host name and its colon can be omitted when only the protocol is to be set.
4622 Note the Exim already has two private options, -pd and -ps, that refer to
4623 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4624 "s", but I don't think that's a major issue.
4626 10. A number of refactoring changes to the code, none of which should affect
4629 (a) The number of logging options was getting close to filling up the
4630 32-bit word that was used as a bit map. I have split them into two classes:
4631 those that are passed in the argument to log_write(), and those that are
4632 only ever tested independently outside of that function. These are now in
4633 separate 32-bit words, so there is plenty of room for expansion again.
4634 There is no change in the user interface or the logging behaviour.
4636 (b) When building, for example, log lines, the code previously used a
4637 macro that called string_cat() twice, in order to add two strings. This is
4638 not really sufficiently general. Furthermore, there was one instance where
4639 it was actually wrong because one of the argument was used twice, and in
4640 one call a function was used. (As it happened, calling the function twice
4641 did not affect the overall behaviour.) The macro has been replaced by a
4642 function that can join an arbitrary number of extra strings onto a growing
4645 (c) The code for expansion conditions now uses a table and a binary chop
4646 instead of a serial search (which was left over from when there were very
4647 few conditions). Also, it now recognizes conditions like "pam" even when
4648 the relevant support is not compiled in: a suitably worded error message is
4649 given if an attempt is made to use such a condition.
4651 11. Added ${time_interval:xxxxx}.
4653 12. A bug was causing one of the ddress fields not to be passed back correctly
4654 from remote delivery subprocesses. The field in question was not being
4655 subsequently used, so this caused to problems in practice.
4657 13. Added new log selectors queue_time and deliver_time.
4659 14. Might have fixed a bug in maildirsizefile handling that threw up
4660 "unexpected character" debug warnings, and recalculated the data
4661 unnecessarily. In any case, I expanded the warning message to give more
4664 15. Added the message "Restricted characters in address" to the statements in
4665 the default ACL that block characters like @ and % in local parts.
4667 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4668 Three changes have been made:
4670 (a) There was a serious bug; a negative response to MAIL caused the whole
4671 recipient domain to be cached as invalid, thereby blocking all messages
4672 to all local parts at the same domain, from all senders. This bug has
4673 been fixed. The domain is no longer cached after a negative response to
4674 MAIL if the sender used is not empty.
4676 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4679 (c) A new callout option, "use_sender" has been added for people who want
4680 the modified behaviour.
4686 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4689 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4690 indeed breaks things for older releases.
4692 3. Added additional logging to the case where there is a problem reading data
4693 from a filter that is running in a subprocess using a pipe, in order to
4694 try to track down a specific problem.
4696 4. Testing facility fudge: when running in the test harness and attempting
4697 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4698 getting "No route to host". Convert this to a timeout.
4700 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4703 6. Some OS don't have socklen_t but use size_t instead. This affects the
4704 fifth argument of getsockopt() amongst other things. This is now
4705 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4706 can be set for individual OS. I have set it for SunOS5, OSF1, and
4707 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4708 some earlier ones do not.
4710 7. Change 4.30/15 was not doing the test caselessly.
4712 8. The standard form for an IPv6 address literal was being rejected by address
4713 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4714 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4715 this, as well as the form without the "IPv6" on the front (but only when
4716 address literals are enabled, of course).
4718 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4720 10. Exim crashed if a message with an empty sender address specified by -f
4721 encountered a router with an errors_to setting. This could be provoked only
4722 by a command such as
4726 where an empty string was supplied; "<>" did not hit this bug.
4728 11. Installed PCRE release 4.5.
4730 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4731 remained set. It is now erased.
4733 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4734 times from message ids (which are base 36 rather than the normal 62).
4736 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4737 were being counted as actual protocol errors, and logged if the log
4738 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4739 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4740 and DATA following a set of rejected RCPTs do not count as protocol errors.
4741 In other words, Exim assumes they were pipelined, though this may not
4742 actually be the case. Of course, in all cases the client gets an
4743 appropriate error code.
4745 15. If a lookup fails in an ACL condition, a message about the failure may
4746 be available; it is used if testing the ACL cannot continue, because most
4747 such messages specify what the cause of the deferral is. However, some
4748 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4749 that caused an old message to be retained and used if a later statement
4750 caused a defer, replacing the real cause of the deferral.
4752 16. If an IP address had so many PTR records that the DNS lookup buffer
4753 was not large enough to hold them, Exim could crash while trying to process
4754 the truncated data. It now detects and logs this case.
4756 17. Further to 4.21/58, another change has been made: if (and only if) the
4757 first line of a message (the first header line) ends with CRLF, a bare LF
4758 in a subsequent header line has a space inserted after it, so as not to
4759 terminate the header.
4761 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4762 unnecessarily, used atoi() instead of strtol(), and didn't check the
4763 termination when getting file sizes from file names by regex.
4765 19. Completely re-implemented the support for maildirsize files, in the light
4766 of a number of problems with the previous contributed implementation
4767 (4.30/29). In particular:
4769 . If the quota is zero, the maildirsize file is maintained, but no quota is
4772 . If the maildir directory does not exist, it is created before any attempt
4773 to write a maildirsize file.
4775 . The quota value in the file is just a cache; if the quota is changed in
4776 the transport, the new value overrides.
4778 . A regular expression is available for excluding directories from the
4781 20. The autoreply transport checks the characters in options that define the
4782 message's headers; it allows continued headers, but it was checking with
4783 isspace() after an embedded newline instead of explicitly looking for a
4786 21. If all the "regular" hosts to which an address was routed had passed their
4787 expiry times, and had not reached their retry times, the address was
4788 bounced, even if fallback hosts were defined. Now Exim should go on to try
4791 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4792 equivalent code in the SMTP transport. Some hosts send humungous responses
4793 to HELO/EHLO, more than 1024 it seems.
4795 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4796 gives compiler warnings in some environments. I've now done it "properly",
4799 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4800 (because of problems with the built-in one) was declared to return uschar *
4801 instead of char *, causing compiler failure.
4803 25. Fixed a file descriptor leak when processing alias/forward files.
4805 26. Fixed a minor format string issue in dbfn.c.
4807 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4809 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4810 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4813 29. When a sender address is verified, it is cached, to save repeating the test
4814 when there is more than one recipient in a message. However, when the
4815 verification involves a callout, it is possible for different callout
4816 options to be set for different recipients. It is too complicated to keep
4817 track of this in the cache, so now Exim always runs a verification when a
4818 callout is required, relying on the callout cache for the optimization.
4819 The overhead is duplication of the address routing, but this should not be
4822 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4823 to be verified with callout=postmaster, and the main callout worked but the
4824 postmaster check failed, the verification correctly failed. However, if a
4825 subsequent RCPT command asked for sender verification *without* the
4826 postmaster check, incorrect caching caused this verification also to fail,
4829 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4830 it was not caching the DNS options (qualify_single, search_parents) that
4831 were used when the lookup failed. A subsequent lookup with different
4832 options therefore always gave the same answer, though there were cases
4833 where it should not have. (Example: a "domains = !$mx_any" option on a
4834 dnslookup router: the "domains" option is always processed without any
4835 widening, but the router might have qualify_single set.) Now Exim uses the
4836 cached value only when the same options are set.
4838 32. Added John Jetmore's "exipick" utility to the distribution.
4840 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4841 than a timeout (e.g. a certificate is required, and is not provided), an
4842 Exim server now closes the connection immediately. Previously it waited for
4843 the client to close - but if the client is SSL, it seems that they each
4844 wait for each other, leading to a delay before one of them times out.
4846 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4847 maintained 0.8.x compatibility because I don't think many are using it, and
4848 it is clearly obsolete.
4850 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4853 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4854 client certificate was expired. A simple patch fixes this, though I don't
4855 understand the full logic of why the verify callback is called multiple
4858 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4859 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4860 which causes problems with some clients (such as the Certicom SSL Plus
4861 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4862 disables the coutermeasure allowing Eudora to connect."
4864 38. Exim was not checking that a write() to a log file succeeded. This could
4865 lead to Bad Things if a log got too big, in particular if it hit a file
4866 size limit. Exim now panics and dies if it cannot write to a log file, just
4867 as it does if it cannot open a log file.
4869 39. Modified OS/Makefile-Linux so that it now contains
4871 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4873 The two -D definitions ensure that Exim is compiled with large file
4874 support, which makes it possible to handle log files that are bigger than
4877 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4878 instance) a domain was checked against a named list that involved a lookup,
4879 causing $domain_data to be set, then another domain was checked against the
4880 same list, then the first domain was re-checked, the value of $domain_data
4881 after the final check could be wrong. In particular, if the second check
4882 failed, it could be set empty. This bug probably also applied to
4885 41. The strip_trailing_dot option was not being applied to the address given
4886 with the -f command-line option.
4888 42. The code for reading a message's header from the spool was incrementing
4889 $received_count, but never initializing it. This meant that the value was
4890 incorrect (doubled) while delivering a message in the same process in which
4891 it was received. In the most common configuration of Exim, this never
4892 happens - a fresh exec is done - but it can happen when
4893 deliver_drop_privilege is set.
4895 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4896 it now includes up to 150 characters of the unexpected data in the log
4899 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4900 and building data strings. The size of both of these buffers was 10 000
4901 bytes - far larger than anybody would *ever* want, thought I. Needless to
4902 say, somebody hit the limit. I have increased the maximum line length to
4903 20 000 and the maximum data length of concatenated lines to 100 000. I have
4904 also fixed two bugs, because there was no checking on these buffers. Tsk,
4905 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4906 buffer is too small.
4908 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4909 lsearch lookups. Now it does.
4911 46. When parsing a route_list item in a manualroute router, a fixed-length
4912 buffer was used for the list of hosts. I made this 1024 bytes long,
4913 thinking that nobody would ever have a list of hosts that long. Wrong.
4914 Somebody had a whole pile of complicated expansion conditions, and the
4915 string was silently truncated, leading to an expansion error. It turns out
4916 that it is easier to change to an unlimited length (owing to other changes
4917 that have happened since this code was originally written) than to build
4918 structure for giving a limitation error. The length of the item that
4919 expands into the list of hosts is now unlimited.
4921 47. The lsearch lookup could not handle data where the length of text line was
4922 more than 4095 characters. Such lines were truncated, leading to shortened
4923 data being returned. It should now handle lines of any length.
4925 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4926 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4929 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4930 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4931 no longer happens in comment lines. A list of replacements is now placed
4932 at the head of all of the source files, except those whose only change is
4933 to replace PERL_COMMAND in the very first #! line.
4935 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4936 messages on the queue, with a bottom-up merge sort, using code contributed
4937 by Michael Haardt. This should make operations like -bp somewhat faster on
4938 large queues. It won't affect queue runners, except when queue_run_in_order
4941 51. Installed eximstats 1.31 in the distribution.
4943 52. Added support for SRV lookups to the dnslookup router.
4945 53. If an ACL referred to $message_body or $message_body_end, the value was not
4946 reset for any messages that followed in the same SMTP session.
4948 54. The store-handling optimization for building very long strings was not
4949 differentiating between the different store pools. I don't think this
4950 actually made any difference in practice, but I've tidied it.
4952 55. While running the routers to verify a sender address, $sender_address
4953 was still set to the sender address. This is wrong, because when routing to
4954 send a bounce to the sender, it would be empty. Therefore, I have changed
4955 it so that, while verifying a sender address, $sender_address is set to <>.
4956 (There is no change to what happens when verifying a recipient address.)
4958 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4959 target A or AAAA records (if not already returned) without resetting the
4960 qualify_single or search_parents options of the DNS resolver. These are
4961 inappropriate in this case because the targets of MX and SRV records must
4962 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4963 target that, when qualified, matched something in the local domain. These
4964 two options are now turned off when doing these lookups.
4966 57. It seems that at least some releases of Reiserfs (which does not have the
4967 concept of a fixed number of inodes) returns zero and not -1 for the
4968 number of available inodes. This interacted badly with check_spool_inodes,
4969 which assumed that -1 was the "no such thing" setting. What I have done is
4970 to check that the total number of inodes is greater than zero before doing
4971 the test of how many are available.
4973 58. When a "warn" ACL statement has a log_message modifier, the message is
4974 remembered, and not repeated. This is to avoid a lot of repetition when a
4975 message has many recipients that cause the same warning to be written.
4976 Howewer, Exim was preserving the list of already written lines for an
4977 entire SMTP session, which doesn't seem right. The memory is now reset if a
4978 new message is started.
4980 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4981 debugging output unless log_rewrite was also set.
4983 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4984 of (int)(handle) when we know that handle contains (void *)(-1).
4986 61. The Exim daemon panic-logs an error return when it closes the incoming
4987 connection. However "connection reset by peer" seems to be common, and
4988 isn't really an error worthy of noting specially, so that particular error
4991 62. When Exim is trying to find all the local interfaces, it used to panic and
4992 die if the ioctl to get the interface flags failed. However, it seems that
4993 on at least one OS (Solaris 9) it is possible to have an interface that is
4994 included in the list of interfaces, but for which you get a failure error
4995 for this call. This happens when the interface is not "plumbed" into a
4996 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4997 failure of the "get flags" call assumes that the interface is down.
4999 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5000 makes life easier for people who are doing arithmetic on fields extracted
5001 from dates, where you often get leading zeros that should not be
5002 interpreted as octal.
5004 64. Added qualify_domain to the redirect router, to override the global
5007 65. If a pathologically long header line contained very many addresses (the
5008 report of this problem mentioned 10 000) and each of them was rewritten,
5009 Exim could use up a very large amount of memory. (It kept on making new
5010 copies of the header line as it rewrote, and never released the old ones.)
5011 At the expense of a bit more processing, the header rewriting function has
5012 been changed so that it no longer eats memory in this way.
5014 66. The generation of the Received: header has been moved from the time that a
5015 message starts to be received, to the time that it finishes. The timestamp
5016 in the Received: header should now be very close to that of the <= log
5017 line. There are two side-effects of this change:
5019 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5020 logged header lines no longer include the local Received: line, because
5021 it has not yet been created. The same applies to a copy of the message
5022 that is returned to a non-SMTP sender when a message is rejected.
5024 (b) When a filter file is tested using -bf, no additional Received: header
5025 is added to the test message. After some thought, I decided that this
5028 This change does not affect the value of $received_for. It is still set
5029 after address rewriting, but before local_scan() is called.
5031 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5033 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5034 gave an unhelpful panic error message, and a defer error. I have managed to
5035 change this behaviour so that it now rejects any supplied certificate,
5036 which seems right, as the list of acceptable certificates is empty.
5038 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5039 gave an unhelpful defer error. I have not managed to make this reject any
5040 supplied certificates, but the error message it gives is "no certificate
5041 supplied", which is not helpful.
5043 70. exigrep's output now also includes lines that are not associated with any
5044 message, but which match the given pattern. Implemented by a patch from
5045 Martin Sluka, which also tidied up the Perl a bit.
5047 71. Recipient callout verification, like sender verification, was using <> in
5048 the MAIL FROM command. This isn't really the right thing, since the actual
5049 sender may affect whether the remote host accepts the recipient or not. I
5050 have changed it to use the actual sender in the callout; this means that
5051 the cache record is now keyed on a recipient/sender pair, not just the
5052 recipient address. There doesn't seem to be a real danger of callout loops,
5053 since a callout by the remote host to check the sender would use <>.
5054 [SEE ABOVE: changed after hitting problems.]
5056 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5057 temporary errors. However, in the case of such a code being given after
5058 the end of a data transmission (i.e. after ".") Exim was failing to write
5059 a retry record for the message. (Yes, there was some broken host that was
5060 actually sending 8xx at this point.)
5062 73. An unknown lookup type in a host list could cause Exim to panic-die when
5063 the list was checked. (An example that provoked this was putting <; in the
5064 middle of a list instead of at the start.) If this happened during a DATA
5065 ACL check, a -D file could be left lying around. This kind of configuration
5066 error no longer causes Exim to die; instead it causes a defer errror. The
5067 incident is still logged to the main and panic logs.
5069 74. Buglet left over from Exim 3 conversion. The message "too many messages
5070 in one connection" was written to the rejectlog but not the mainlog, except
5071 when address rewriting (yes!) was being logged.
5073 75. Added write_rejectlog option.
5075 76. When a system filter was run not as root (that is, when system_filter_user
5076 was set), the values of the $n variables were not being returned to the
5077 main process; thus, they were not subsequently available in the $sn
5080 77. Added +return_path_on_delivery log selector.
5082 78. A connection timeout was being treated differently from recipients deferred
5083 when testing hosts_max_try with a message that was older than the host's
5084 retry timeout. (The host should not be counted, thus allowing all hosts to
5085 be tried at least once before bouncing.) This may have been the cause of an
5086 occasionally reported bug whereby a message would remain on the queue
5087 longer than the retry timeout, but would be bounced if a delivery was
5088 forced. I say "may" because I never totally pinned down the problem;
5089 setting up timeout/retry tests is difficult. See also the next item.
5091 79. The ultimate address timeout was not being applied to errors that involved
5092 a combination of host plus message (for example, a timeout on a MAIL
5093 command). When an address resolved to a number of possible hosts, and they
5094 were not all tried for each delivery (e.g. because of hosts_max_try), a
5095 message could remain on the queue longer than the retry timeout.
5097 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5100 81. Fixed an obscure SMTP outgoing bug which required at least the following
5101 conditions: (a) there was another message waiting for the same server;
5102 (b) the server returned 5xx to all RCPT commands in the first message so
5103 that the message was not completed; (c) the server dropped the connection
5104 or gave a negative response to the RSET that Exim sends to abort the
5105 transaction. The observed case was a dropped connection after DATA that had
5106 been sent in pipelining mode. That is, the server had advertised PIPELINING
5107 but was not implementing it correctly. The effect of the bug was incorrect
5108 behaviour, such as trying another host, and this could lead to a crash.
5114 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5115 and daemon.c were passed as pointers to ints; they should have been
5116 pointers to socklen_t variables (which are typically unsigned ints).
5118 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5121 3. Fixed a really odd bug that affected only the testing scheme; patching a
5122 certain fixed string in the binary changed the value of another string that
5123 happened to be identical to the end of the original first string.
5125 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5126 name", it returns that address as the IP address. On some operating
5127 systems (e.g. Solaris), it also passes back the IP address string as the
5128 "host name". However, on others (e.g. Linux), it passes back an empty
5129 string. Exim wasn't checking for this, and was changing the host name to an
5130 empty string, assuming it had been canonicized.
5132 5. Although rare, it is permitted to have more than one PTR record for a given
5133 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5134 all the names associated with an address, because they do in Solaris.
5135 However, it seems that they do not in Linux for data that comes from the
5136 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5137 I found this out when I moved to a new Linux workstation and tried to run
5138 the Exim test suite.
5140 To get round this problem I have changed the code so that it now does its
5141 own call to the DNS to look up PTR records when searching for a host name.
5142 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5143 addresses that are only in /etc/hosts are still found.
5145 This behaviour is, however, controlled by an option called host_lookup_
5146 order, which defaults to "bydns:byaddr". If people want to use the other
5147 order, or indeed, just use one or the other means of lookup, they can
5148 specify it in this variable.
5150 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5151 some operating systems, this comes back from gethostbyaddr() as an empty
5152 string, and this is what Exim used to test for. However, it seems that in
5153 other systems, "." is yielded. Exim now tests for this case too.
5155 7. The values of check_spool_space and check_log_space are now held internally
5156 as a number of kilobytes instead of an absolute number of bytes. If a
5157 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5158 kilobyte. This means that much larger values can be stored.
5160 8. Exim monitor: an attempt to get the action menu when not actually pointing
5161 at a message produces an empty menu entitled "No message selected". This
5162 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5163 no entries in it ("Shell widget menu has zero width and/or height"). So I
5164 have added a single, blank menu entry in this case.
5166 9. Added ${quote_local_part.
5168 10. MIME decoding is now applied to the contents of Subject: header lines when
5171 11. Now that a reference to $sender_host_address automatically causes a reverse
5172 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5173 host lookup before query-style lookups in lists that might use this
5174 variable. This has therefore been abolished, and the "net-" prefix is no
5175 longer necessary for query-style lookups.
5177 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5178 have been a typo for LFLAGS, so it has been changed.
5180 13. The install script calls Exim with "-C /dev/null" in order to find the
5181 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5182 to be output. Howeve, since Exim outputs its version number before the
5183 error, it didn't break the script. It just looked ugly. I fixed this by
5184 always allowing "-C /dev/null" if the caller is root.
5186 14. Ignore overlarge ACL variable number when reading spool file - insurance
5187 against a later release with more variables having written the file.
5189 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5190 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5191 this, as well as the form without the "IPv6" on the front.
5193 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5194 OS/Makefile-Darwin file.
5196 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5197 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5199 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5200 variable. However, some OS use ints for inodes. I've added cast to long int
5201 to get rid of the compiler warning.
5203 19. I had forgotten to lock out "/../" in configuration file names when
5204 ALT_CONFIG_PREFIX was set.
5206 20. Routers used for verification do not need to specify transports. However,
5207 if such a router generated a host list, and callout was configured, Exim
5208 crashed, because it could not find a port number from the (non-existent)
5209 transport. It now assumes port 25 in this circumstance.
5211 21. Added the -t option to exigrep.
5213 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5214 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5215 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5216 them is set, LOOKUP_LSEARCH is forced.
5218 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5220 24. Added sender and host information to the "rejected by local_scan()" log
5221 line; previously there was no indication of these.
5223 25. Added .include_if_exists.
5225 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5226 Linux. It turns out that not all file systems support this. Apparently some
5227 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5228 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5229 not supported on the file descriptor, is now ignored when Exim is trying to
5230 sync a directory. This applies only to Linux.
5232 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5234 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5235 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5238 29. Added support for maildirsize files from supplied patch (modified a bit).
5240 30. The use of :fail: followed by an empty string could lead Exim to respond to
5241 sender verification failures with (e.g.):
5243 550 Verification failed for <xxx>
5244 550 Sender verify failed
5246 where the first response line was missing the '-' that indicates it is not
5247 the final line of the response.
5249 31. The loop for finding the name of the user that called Exim had a hardwired
5250 limit of 10; it now uses the value of finduser_retries, which is used for
5251 all other user lookups.
5253 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5256 33. Exim was neglecting to zero errno before one call of strtol() when
5257 expanding a string and expecting an integer value. On some systems this
5258 resulted in spurious "integer overflow" errors. Also, it was casting the
5259 result into an int without checking.
5261 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5262 did not work. The code looks as if it has *never* worked, though it appears
5263 to have been documented since at least releast 1.62. I have made it work.
5265 35. The "timeout_DNS" error in retry rules, also documented since at least
5266 1.62, also never worked. As it isn't clear exactly what this means, and
5267 clearly it isn't a major issue, I have abolished the feature by treating it
5268 as "timeout", and writing a warning to the main and panic logs.
5270 36. The display of retry rules for -brt wasn't always showing the error code
5273 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5274 timeout_connect_A, timeout_connect_MX.
5276 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5277 to the empty sender.
5279 39. The daemon was not analysing the content of -oX till after it had closed
5280 stderr and disconnected from the controlling terminal. This meant that any
5281 syntax errors were only noted on the panic log, and the return code from
5282 the command was 0. By re-arranging the code a little, I've made the
5283 decoding happen first, so such errors now appear on stderr, and the return
5284 code is 1. However, the actual setting up of the sockets still happens in
5285 the disconnected process, so errors there are still only recorded on the
5288 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5289 connections (as happens on some IP stacks) was logged at start up time as
5290 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5291 it from "IPv6 and IPv4", which means that two separate sockets are being
5294 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5295 says whether AF_INET or AF_INET6 was passed as an argument.
5297 42. Exiwhat output was messed up when time zones were included in log
5300 43. Exiwhat now gives more information about the daemon's listening ports,
5301 and whether -tls-on-connect was used.
5303 44. The "port" option of the smtp transport is now expanded.
5305 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5306 silently ignored. Now an error message is written to the main and panic
5309 46. There's a new ACL modifier called "logwrite" which writes to a log file
5310 as soon as it is encountered.
5312 47. Added $local_user_uid and $local_user_gid at routing time.
5314 48. Exim crashed when trying to verify a sender address that was being
5317 49. Exim was recognizing only a space character after ".include". It now also
5318 recognizes a tab character.
5320 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5321 extracting the relevant information from the specification. The man page no
5322 longer contains scrambled data for the -d option, and I've added a section
5323 at the front about calling Exim under different names.
5325 51. Added "extra_headers" argument to the "mail" command in filter files.
5327 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5330 53. Installed eximstats 1.29.
5332 54. Added transport_filter_timeout as a generic transport option.
5334 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5335 Cc: header lines. This was required by RFC 822, but it not required by RFC
5338 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5339 incoming messages that did not have them. Now it does so only if the
5340 message originates locally, that is, if there is no associated remote host
5341 address. When Resent- header lines are present, this applies to the Resent-
5342 lines rather than the non-Resent- lines.
5344 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5345 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5347 58. Messages for configuration errors now include the name of the main
5348 configuration file - useful now that there may be more than one file in a
5349 list (.included file names were always shown).
5351 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5352 for those rare installations that do not start the daemon as root or run it
5353 setuid root. I've cut out the call to initgroups() if the daemon is not
5356 60. The Exim user and group can now be bound into the binary as text strings
5357 that are looked up at the start of Exim's processing.
5359 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5361 62. Added $mailstore_basename variable.
5363 63. Installed patch to sieve.c from Michael Haardt.
5365 64. When Exim failed to open the panic log after failing to open the main log,
5366 the original message it was trying to log was written to stderr and debug
5367 output, but if they were not available (the usual case in production), it
5368 was lost. Now it is written to syslog before the two lines that record the
5369 failures to open the logs.
5371 65. Users' Exim filters run in subprocesses under the user's uid. It is
5372 possible for a "deliver" command or an alias in a "personal" command to
5373 provoke an address rewrite. If logging of address rewriting is configured,
5374 this fails because the process is not running as root or exim. There may be
5375 a better way of dealing with this, but for the moment (because 4.30 needs
5376 to be released), I have disabled address rewrite logging when running a
5377 filter in a non-root, non-exim process.
5383 1. The buildconfig auxiliary program wasn't quoting the value set for
5384 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5385 not defined. This bug was masked in 4.22 by the effect that was fixed in
5388 2. Some messages that were rejected after a message id was allocated were
5389 shown as "incomplete" by exigrep. It no longer does this for messages that
5390 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5392 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5393 have allow_domain_literals set, the ID did not get logged in the <= line.
5394 Domain literals are now always recognized in Message-ID: header lines.
5396 4. The first argument for a ${extract expansion item is the key name or field
5397 number. Leading and trailing spaces in this item were not being ignored,
5398 causing some misleading effects.
5400 5. When deliver_drop_privilege was set, single queue runner processes started
5401 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5402 same command in the process it spins off) were not dropping privilege.
5404 6. When the daemon running as "exim" started a queue runner, it always
5405 re-executed Exim in the spun-off process. This is a waste of effort when
5406 deliver_drop_privilege is set. The new process now just calls the
5407 queue-runner function directly.
5413 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5416 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5417 ignored. Though the use of -oP was forcing the writing of a pid file, it
5418 was always written to the default place.
5420 3. If the message "no IP address found for host xxxx" is generated during
5421 incoming verification, it is now followed by identification of the incoming
5422 connection (so you can more easily find what provoked it).
5424 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5426 5. Added some features to "harden" Exim a bit more against certain attacks:
5428 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5429 be put in Local/Makefile. This is like the never_users runtime option,
5430 but it cannot be overridden. The default setting is "root".
5432 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5433 prefix string with which any file named in a -C command line option
5436 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5437 is retained for -C and -D only if the caller of Exim is root. Without
5438 it, the exim user may also use -C and -D and retain privilege.
5440 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5441 command line option is disabled.
5443 6. Macro names set by the -D option must start with an upper case letter, just
5444 like macro names defined in the configuration file.
5446 7. Added "dereference=" facility to LDAP.
5448 8. Two instances of the typo "uknown" in the source files are fixed.
5450 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5451 the Configure-Makefile script screwed up while processing it.
5453 10. Incorporated PCRE 4.4.
5455 11. The SMTP synchronization check was not operating right at the start of an
5456 SMTP session. For example, it could not catch a HELO sent before the client
5457 waited for the greeting. There is now a check for outstanding input at the
5458 point when the greeting is written. Because of the duplex, asynchronous
5459 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5460 way, but not yet received, when the check is performed.
5462 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5463 on TCP/IP sockets, because this apparently causes some broken clients to
5466 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5467 unchanged) from the Cygwin maintainer.
5469 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5470 was supported instead of testing for mbx. Effectively a typo.
5472 15. The spa authenticator server code was not checking that the input it
5473 received was valid base64.
5475 16. The debug output line for the "set" modifier in ACLs was not showing the
5476 name of the variable that was being set.
5478 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5480 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5481 lookup on its own. Something else was needed to trigger the lookup. For
5482 example, a match in host_lookup or the need for a host name in a host list.
5483 Now, if $sender_host_name is referenced and the host name has not yet been
5484 looked up, a lookup is performed. If the lookup fails, the variable remains
5485 empty, and $host_lookup_failed is set to "1".
5487 19. Added "eqi" as a case-independent comparison operator.
5489 20. The saslauthd authentication condition could segfault if neither service
5490 nor realm was specified.
5492 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5493 error message that was logged was misleading, and incoming SMTP
5494 connections were dropped. The message is now more accurate, and temporary
5495 errors are given to SMTP connections.
5497 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5498 (except RSET) with a 421 error, until QUIT is received. However, it was
5499 failing to send a response to QUIT.
5501 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5502 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5503 ACL is now run after the helo_try_verify_hosts code.
5505 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5506 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5507 case-independent, so other case variants are also recognized). Apparently
5508 some systems use these upper case variants.
5510 25. If more than two messages were waiting for the same host, and a transport
5511 filter was specified for the transport, Exim sent two messages over the
5512 same TCP/IP connection, and then failed with "socket operation on non-
5513 socket" when it tried to send the third.
5515 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5517 27. The extern definition of crypt16() in expand.c was not being excluded when
5518 the OS had its own crypt16() function.
5520 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5521 as a preferred synonym for return_size_limit, both as an option and as an
5524 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5526 30. Changed the default configuration ACL to relax the local part checking rule
5527 for addresses that are not in any local domains. For these addresses,
5528 slashes and pipe symbols are allowed within local parts, but the sequence
5529 /../ is explicitly forbidden.
5531 31. SPA server authentication was not clearing the challenge buffer before
5534 32. log_message in a "warn" ACL statement was writing to the reject log as
5535 well as to the main log, which contradicts the documentation and doesn't
5536 seem right (because no rejection is happening). So I have stopped it.
5538 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5539 However, I am unable to do any testing of this.
5541 34. Fixed an infelicity in the appendfile transport. When checking directories
5542 for a mailbox, to see if any needed to be created, it was accidentally
5543 using path names with one or more superfluous leading slashes; tracing
5544 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5546 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5547 contents are added to the log line that is written for every discarded
5548 recipient. (Previously a log_message setting was ignored.)
5550 36. The ${quote: operator now quotes the string if it is empty.
5552 37. The install script runs exim in order to find its version number. If for
5553 some reason other than non-existence or emptiness, which it checks, it
5554 could not run './exim', it was installing it with an empty version number,
5555 i.e. as "exim-". This error state is now caught, and the installation is
5558 38. An argument was missing from the function that creates an error message
5559 when Exim fails to connect to the socket for saslauthd authentication.
5560 This could cause Exim to crash, or give a corrupted message.
5562 39. Added isip, isip4, and isip6 to ${if conditions.
5564 40. The ACL variables $acl_xx are now saved with the message, and can be
5565 accessed later in routers, transports, and filters.
5567 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5568 strings in the file are not string-expanded.
5570 42. If a MAIL command specified a SIZE value that was too large to fit into an
5571 int variable, the check against message_size_limit failed. Such values are
5572 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5573 day this will have to be increased, but I don't think I want to be around
5574 when emails are that large.
5581 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5582 iconv() is not standard in FreeBSD.
5584 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5585 IPv6 enabled. The observed symptom was a segmentation fault on return from
5586 the function os_common_find_running_interfaces() in src/os.c.
5588 3. In the check_special_case() function in daemon.c I had used "errno" as an
5589 argument name, which causes warnings on some systems. This was basically a
5590 typo, since it was named "eno" in the comments!
5592 4. The code that waits for the clock to tick (at a resolution of some fraction
5593 of a second) so as to ensure message-id uniqueness was always waiting for
5594 at least one whole tick, when it could have waited for less. [This is
5595 almost certainly not relevant at current processor speeds, where it is
5596 unlikely to ever wait at all. But we try to future-proof.]
5598 5. The function that sleeps for a time interval that includes fractions of a
5599 second contained a race. It did not block SIGALRM between setting the
5600 timer, and suspending (a couple of lines later). If the interval was short
5601 and the sigsuspend() was delayed until after it had expired, the suspension
5602 never ended. On busy systems this could lead to processes getting stuck for
5605 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5606 process, before it forks any delivery processes. The open lookup caching
5607 mechanism meant that the open file or database connection was passed into
5608 the delivery process. The problem was that delivery processes always tidy
5609 up cached lookup data. This could cause a problem for the next delivery
5610 process started by the queue runner, because the external queue runner
5611 process does not know about the closure. So the next delivery process
5612 still has data in the lookup cache. In the case of a file lookup, there was
5613 no problem because closing a file descriptor in a subprocess doesn't affect
5614 the parent. However, if the lookup was caching a connection to a database,
5615 the connection was closed, and the second delivery process was likely to
5616 see errors such as "PGSQL: query failed: server closed the connection
5617 unexpectedly". The problem has been fixed by closing all cached lookups
5618 in a queue runner before running a delivery process.
5620 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5621 seem to have the "const" qualifier which it has on other OS. I've
5624 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5625 *of the same type* (client or server) with the same public name that an
5626 error should be diagnosed.
5628 9. When Exim looked up a host name for an IP address, but failed to find the
5629 original IP address when looking up the host name (a safety check), it
5630 output the message "<ip address> does not match any IP for NULL", which was
5631 confusing, to say the least. The bug was that the host name should have
5632 appeared instead of "NULL".
5634 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5635 user that is built into the binary, and the -C or -D options is used, root
5636 privilege is dropped before the configuration file is read. In addition,
5637 logging is switched to stderr instead of the normal log files. If the
5638 configuration then re-defines the Exim user, the unprivileged environment
5639 is probably not what is expected, so Exim logs a panic warning message (but
5642 However, if deliver_drop_privilege is set, the unprivileged state may well
5643 be exactly what is intended, so the warning has been cut out in that case,
5644 and Exim is allowed to try to write to its normal log files.
5650 1. smtp_return_error_details was not giving details for temporary sender
5651 or receiver verification errors.
5653 2. Diagnose a configuration error if two authenticators have the same public
5656 3. Exim used not to create the message log file for a message until the first
5657 delivery attempt. This could be confusing when incoming messages were held
5658 for policy or load reasons. The message log file is now created at the time
5659 the message is received, and an initial "Received" line is written to it.
5661 4. The automatically generated man page for command line options had a minor
5662 bug that caused no ill effects; however, a more serious problem was that
5663 the procedure for building the man page automatically didn't always
5664 operate. Consequently, release 4.20 contains an out-of-date version. This
5665 shouldn't happen again.
5667 5. When building Exim with embedded Perl support, the script that builds the
5668 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5669 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5671 6. The freeze_tell option was not being used for messages that were frozen on
5672 arrival, either by an ACL or by local_scan().
5674 7. Added the smtp_incomplete_transaction log selector.
5676 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5677 was accepting AUTH without a new EHLO.
5679 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5680 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5683 10. Logging of TCP/IP connections (when configured) now happens in the main
5684 daemon process instead of the child process, so that the TCP/IP connection
5685 count is more accurate (but it can never be perfect).
5687 11. The use of "drop" in a nested ACL was not being handled correctly in the
5688 outer ACL. Now, if condition failure induced by the nested "drop" causes
5689 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5690 or "require"), the connection is dropped.
5692 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5693 that yield "discard" can now be used with an "accept" or a "discard" verb,
5694 but an error is generated for any others (because I can't see a useful way
5695 to define what should happen).
5697 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5698 are now processed in the same way as lines in the Exim configuration file.
5699 In particular, continuation lines are supported.
5701 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5703 15. Added -ti meaning -t -i.
5705 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5706 domains, and warn by logging if others are found.
5708 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5709 returned from GIFCONF when Exim is trying to find the list of interfaces on
5710 a host. The code in os.c has been modified to copy each ifreq to an aligned
5711 structure in all cases.
5713 Also, in some cases, the returned ifreq's were being copied to a 'struct
5714 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5715 means the last couple of bytes of an IPv6 address could be chopped if the
5716 ifreq contained only a normal sockaddr (14 bytes storage).
5718 18. Named domain lists were not supported in the hosts_treat_as_local option.
5719 An entry such as +xxxx was not recognized, and was treated as a literal
5722 19. Ensure that header lines added by a DATA ACL are included in the reject log
5723 if the ACL subsequently rejects the message.
5725 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5726 MD5 (which is deprecated).
5728 21. When testing a filter file using -bf, Exim was writing a message when it
5729 took the sender from a "From " line in the message, but it was not doing so
5730 when it took $return_path from a Return-Path: header line. It now does.
5732 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5733 with a valid header line field name (a series of printing characters
5734 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5736 23. Changed "disc" in the source to "disk" to conform to the documentation and
5737 the book and for uniformity.
5739 24. Ignore Sendmail's -Ooption=value command line item.
5741 25. When execve() failed while trying to run a command in a pipe transport,
5742 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5743 could be confused with a return value of 69 from the command itself. This
5744 has been changed to 127, the value the shell returns if it is asked to run
5745 a non-existent command. The wording for the related log line suggests a
5746 non-existent command as the problem.
5748 26. If received_header_text expands to an empty string, do not add a Received:
5749 header line to the message. (Well, it adds a token one on the spool, but
5750 marks it "old" so that it doesn't get used or transmitted.)
5752 27. Installed eximstats 1.28 (addition of -nt option).
5754 28. There was no check for failure on the call to getsockname() in the daemon
5755 code. This can fail if there is a shortage of resources on the system, with
5756 ENOMEM, for example. A temporary error is now given on failure.
5758 29. Contrary to the C standard, it seems that in some environments, the
5759 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5760 program. Exim now does this explicitly; it affects the formatting of
5761 timestamps using strftime().
5763 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5764 complaints. I've now initialized all the variables, to avoid this.
5766 32. Header lines added by a system filter were not being "seen" during
5767 transport-time rewrites.
5769 33. The info_callback() function passed to OpenSSL is set up with type void
5770 (*)(SSL *, int, int), as described somewhere. However, when calling the
5771 function (actually a macro) that sets it up, the type void(*)() is
5772 expected. I've put in a cast to prevent warnings from picky compilers.
5774 34. If a DNS black list lookup found a CNAME record, but there were no A
5775 records associated with the domain it pointed at, Exim crashed.
5777 35. If a DNS black list lookup returned more than one A record, Exim ignored
5778 all but the first. It now scans all returned addresses if a particular IP
5779 value is being sought. In this situation, the contents of the
5780 $dnslist_value variable are a list of all the addresses, separated by a
5783 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5784 to accept a host name and all its aliases if the forward lookup for any of
5785 them yielded the IP address of the incoming connection. Now it accepts only
5786 those names whose forward lookup yields the correct IP address. Any other
5787 names are discarded. This closes a loophole whereby a rogue DNS
5788 administrator could create reverse DNS records to break through a
5789 wildcarded host restriction in an ACL.
5791 37. If a user filter or a system filter that ran in a subprocess used any of
5792 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5793 the wrong values were passed to the pipe command ($thisaddress had the
5794 value of $0, $0 had the value of $1, etc). This bug was introduced by
5795 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5797 38. Improved the line breaking for long SMTP error messages from ACLs.
5798 Previously, if there was no break point between 40 and 75 characters, Exim
5799 left the rest of the message alone. Two changes have been made: (a) I've
5800 reduced the minimum length to 35 characters; (b) if it can't find a break
5801 point between 35 and 75 characters, it looks ahead and uses the first one
5802 that it finds. This may give the occasional overlong line, but at least the
5803 remaining text gets split now.
5805 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5806 file descriptors might be low, and that setting 1000 would always raise it.
5807 It turns out that in some environments, the limit is already over 1000 and
5808 that lowering it causes trouble. So now Exim takes care not to decrease it.
5810 40. When delivering a message, the value of $return_path is set to $sender_
5811 address at the start of routing (routers may change the value). By an
5812 oversight, this default was not being set up when an address was tested by
5813 -bt or -bv, which affected the outcome if any router or filter referred to
5816 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5817 log only when "message" or "log_message" are set. However, if one of the
5818 conditions was an address verification, or a call to a nested ACL, the
5819 messages generated by the underlying test were being passed through. This
5820 no longer happens. The underlying message is available in $acl_verify_
5821 message for both "message" and "log_message" expansions, so it can be
5822 passed through if needed.
5824 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5825 new expansion $bh_ to give the encoded byte string without charset
5826 translation. Translation happens only if iconv() is available; HAVE_ICONV
5827 indicates this at build time. HEADERS_CHARSET gives the charset to
5828 translate to; headers_charset can change it in the configuration, and
5829 "headers charset" can change it in an individual filter file.
5831 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5832 that creates RFC 2047 encoded "words" labels them as that charset instead
5833 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5834 expansion operator; (ii) when Exim creates a From: line for a local
5835 message; (iii) when a header line is rewritten to include a "phrase" part.
5837 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5838 buggy, causing it to skip the first lines of messages whose message ID
5839 ended in 'D'. This would not have bitten before Exim release 4.14, because
5840 message IDs were unlikely to end in 'D' before then. The effect was to have
5841 incorrect size information for certain domains.
5843 45. #include "config.h" was missing at the start of the crypt16.c module. This
5844 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5847 46. If there was a timeout during a "random" callout check, Exim treated it as
5848 a failure of the random address, and carried on sending RSET and the real
5849 address. If the delay was just some slowness somewhere, the response to the
5850 original RCPT would be taken as a response to RSET and so on, causing
5851 mayhem of various kinds.
5853 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5854 when I implemented it. It didn't allow for the fact that some option values
5855 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5856 the right test for positive values.
5858 48. Domain names in DNS records are case-independent. Exim always looks them up
5859 in lower case. Some resolvers return domain names in exactly the case they
5860 appear in the zone file, that is, they may contain uppercase letters. Not
5861 all resolvers do this - some return always lower case. Exim was treating a
5862 change of case by a resolver as a change of domain, similar to a widening
5863 of a domain abbreviation. This triggered its re-routing code and so it was
5864 trying to route what was effectively the same domain again. This normally
5865 caused routing to fail (because the router wouldn't handle the domain
5866 twice). Now Exim checks for this case specially, and just changes the
5867 casing of the domain that it ultimately uses when it transmits the message
5870 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5873 50. If a filter generated a file delivery with a non-absolute name (possible if
5874 no home directory exists for the router), the forbid_file option was not
5877 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5878 the existing equality matching.
5880 52. Exim was using ints instead of ino_t variables in some places where it was
5881 dealing with inode numbers.
5883 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5884 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5885 called TMPDIR, and if it finds it is different, it changes its value.
5887 54. The smtp_printf() function is now made available to local_scan() so
5888 additional output lines can be written before returning. There is also an
5889 smtp_fflush() function to enable the detection of a dropped connection.
5890 The variables smtp_input and smtp_batched_input are exported to
5893 55. Changed the default runtime configuration: the message "Unknown user"
5894 has been removed from the ACL, and instead placed on the localuser router,
5895 using the cannot_route_message feature. This means that any verification
5896 failures that generate their own messages won't get overridden. Similarly,
5897 the "Unrouteable address" message that was in the ACL for unverifiable
5898 relay addresses has also been removed.
5900 56. Added hosts_avoid_esmtp to the smtp transport.
5902 57. The exicyclog script was not checking for the esoteric option
5903 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5904 will work only if exicyclog is run under the appropriate euid.
5906 58. Following a discussion on the list, the rules by which Exim recognises line
5907 endings on incoming messages have been changed. The -dropcr and drop_cr
5908 options are now no-ops, retained only for backwards compatibility. The
5909 following line terminators are recognized: LF CRLF CR. However, special
5910 processing applies to CR:
5912 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5913 nor a local message in the state where . is a terminator.
5915 (ii) If a bare CR is encountered in a header line, an extra space is added
5916 after the line terminator so as not to end the header. The reasoning
5917 behind this is that bare CRs in header lines are most likely either
5918 to be mistakes, or people trying to play silly games.
5920 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5921 was being incorrectly given as 18 bytes larger than it should have been.
5922 This is a VOB (very old bug).
5924 60. This may never have affected anything current, but just in case it has:
5925 When the local host is found other than at the start of a list of hosts,
5926 the local host, those with the same MX, and any that follow, are discarded.
5927 When the list in question was part of a longer list of hosts, the following
5928 hosts (not currently being processed) were also being discarded. This no
5929 longer happens. I'm not sure if this situation could ever has previously
5932 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5935 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5936 header. This is something that is recommended in a new Internet Draft, and
5937 is something that is documented as being done by Sendmail. There are two
5938 possible values. For messages generated by the autoreply transport, Exim
5941 Auto-Submitted: auto-replied
5943 whereas for all other generated messages (e.g. bounces) it adds
5945 Auto-Submitted: auto-generated
5947 63. The "personal" condition in filters now includes a test for the
5948 Auto-Submitted: header. If it contains the string "auto-" the message it
5949 not considered personal.
5951 64. Added rcpt_include_affixes as a generic transport option.
5953 65. Added queue_only_override (default true).
5955 66. Added the syslog_duplication option.
5957 67. If what should have been the first header line of a message consisted of
5958 a space followed by a colon, Exim was mis-interpreting it as a header line.
5959 It isn't of course - it is syntactically invalid and should therefore be
5960 treated as the start of the message body. The misbehaviour could have
5961 caused a number of strange effects, including loss of data in subsequent
5962 header lines, and spool format errors.
5964 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5965 client host had authenticated. This control can now be exercised by an ACL
5966 for more flexibility.
5968 69. By default, callouts do not happen when testing with -bh. There is now a
5969 variant, -bhc, which does actually run the callout code, including
5970 consulting and updating the callout cache.
5972 70. Added support for saslauthd authentication, courtesy of Alexander
5975 71. If statvfs() failed on the spool or log directories while checking their
5976 size for availability, Exim confusingly gave the error "space shortage".
5977 Furthermore, in debugging mode it crashed with a floating point exception.
5978 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5979 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5980 serious problem, Exim now writes to the main and panic logs when this
5981 happens, with details of the failure. It then refuses to accept the
5982 incoming message, giving the message "spool directory problem" or "log
5983 directory problem" with a 421 code for SMTP messages.
5985 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5986 0, 1, and 2 exist, because some OS complain for execs without them (see
5987 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5988 descriptors. However, the code omitted to check that the open succeeded,
5989 causing mysterious errors if for some reason the permissions on /dev/null
5990 got screwed. Now Exim writes a message to the main and panic logs, and
5991 bombs out if it can't open /dev/null.
5993 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5994 interact so that it is all more flexible. It is supposed to remain
5995 backwards compatible. Also added extra_local_interfaces.
5997 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5998 to bomb out with an assertion failure - to the client this appears as a
5999 connection drop. This problem occurs in the part of the code that was taken
6000 from the Samba project. Fortunately, the assertion is in a very simple
6001 function, so I have fixed this by reproducing the function inline in the
6002 one place where it is called, and arranging for authentication to fail
6003 instead of killing the process with assert().
6005 75. The SPA client code was not working when the server requested OEM rather
6006 than Unicode encoding.
6008 76. Added code to make require_files with a specific uid setting more usable in
6009 the case where statting the file as root fails - usually a non-root-mounted
6010 NFS file system. When this happens and the failure is EACCES, Exim now
6011 forks a subprocess and does the per-uid checking as the relevant uid.
6013 77. Added process_log_path.
6015 78. If log_file_path was not explicitly set, a setting of check_log_space or
6016 check_log_inodes was ignored.
6018 79. If a space check for the spool or log partitions fails, the incident is now
6019 logged. Of course, in the latter case the data may get lost...
6021 80. Added the %p formatting code to string_format() so that it can be used to
6022 print addresses in debug_print(). Adjusted all the address printing in the
6023 debugging in store.c to use %p rather than %d.
6025 81. There was a concern that a line of code in smtp_in.c could overflow a
6026 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6027 initially expressed, the concern was not well-founded, because trailing
6028 spaces are removed early. However, if the trailing spaces were followed by
6029 a NULL, they did not get removed, so the overflow was possible. Two fixes
6032 (a) I re-wrote the offending code in a cleaner fashion.
6033 (b) If an incoming SMTP command contains a NULL character, it is rejected
6036 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6037 runs initgroups(), so that if the Exim user is in any additional groups,
6038 they will be used during message reception.
6044 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob