1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
54 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
55 checkers that were complaining about end of non-void function with no
58 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
59 This was a regression intruduced in 4.83 by another bugfix.
61 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
63 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
64 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
70 TF/01 Correctly close the server side of TLS when forking for delivery.
72 When a message was received over SMTP with TLS, Exim failed to clear up
73 the incoming connection properly after forking off the child process to
74 deliver the message. In some situations the subsequent outgoing
75 delivery connection happened to have the same fd number as the incoming
76 connection previously had. Exim would try to use TLS and fail, logging
77 a "Bad file descriptor" error.
79 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
80 utilities have not been installed.
82 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
83 temporary space as the ACL may create new global variables.
85 TL/01 LDAP support uses per connection or global context settings, depending
86 upon the detected version of the libraries at build time.
88 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
89 to extract and use the src ip:port in logging and expansions as if it
90 were a direct connection from the outside internet. PPv2 support was
91 updated based on HAProxy spec change in May 2014.
93 JH/02 Add ${listextract {number}{list}{success}{fail}}.
95 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
96 Properly escape header and check for NULL return.
98 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
101 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
103 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
104 characters in header names, implemented as a verify condition.
105 Contributed by Michael Fischer v. Mollard.
107 TL/05 Rename SPF condition results err_perm and err_temp to standardized
108 results permerror and temperror. Previous values are deprecated but
109 still accepted. In a future release, err_perm and err_temp will be
110 completely removed, which will be a backward incompatibility if the
111 ACL tests for either of these two old results. Patch contributed by
112 user bes-internal on the mailing list.
114 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
116 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
117 selectors, in both main and reject logs.
119 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
122 JH/07 Add malware type "sock" for talking to simple daemon.
124 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
126 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
127 routers/transports under cutthrough routing.
129 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
130 numbers. Touch up "bool" conditional to keep the same definition.
132 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
134 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
136 JH/12 Expand items in router/transport headers_add or headers_remove lists
137 individually rather than the list as a whole. Bug 1452.
139 Required for reasonable handling of multiple headers_ options when
140 they may be empty; requires that headers_remove items with embedded
141 colons must have them doubled (or the list-separator changed).
143 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
144 view the policy declared in the DMARC record. Currently, $dmarc_status
145 is a combined value of both the record presence and the result of the
148 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
150 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
151 dnslookup router and the smtp transport (applying to the forward
154 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
155 of ldap servers used for a specific lookup. Patch provided by Heiko
158 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
159 New variable $lookup_dnssec_authenticated for observability.
161 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
162 Patch submitted by Lars Timman.
164 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
166 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
167 Requires trusted mode and valid format message id, aborts otherwise.
168 Patch contributed by Heiko Schlichting.
170 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
171 certextract with support for various fields. Bug 1358.
173 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
174 is requested by default, modifiable by smtp transport option
177 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
178 operate on certificate variables to give certificate fingerprints
179 Also new ${sha256:cert_variable}.
181 JH/23 The PRDR feature is moved from being Experimental into the mainline.
183 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
184 Christian Aistleitner.
186 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
188 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
189 file. Patch from Wolfgang Breyha.
191 JH/25 Expand the coverage of the delivery $host and $host_address to
192 client authenticators run in verify callout. Bug 1476.
194 JH/26 Port service names are now accepted for tls_on_connect_ports, to
195 align with daemon_smtp_ports. Bug 72.
197 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
198 support and error reporting did not work properly.
200 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
201 and is readable. Patch from Andrew Colin Kissa.
203 TL/14 Enhance documentation of ${run expansion and how it parses the
204 commandline after expansion, particularly in the case when an
205 unquoted variable expansion results in an empty value.
207 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
209 PP/02 Fix internal collision of T_APL on systems which support RFC3123
210 by renaming away from it. Addresses GH issue 15, reported by
213 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
215 TL/15 SECURITY: prevent double expansion in math comparison functions
216 (can expand unsanitized data). Not remotely exploitable.
223 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
225 PP/02 Make -n do something, by making it not do something.
226 When combined with -bP, the name of an option is not output.
228 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
231 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
232 $sender_host_name and config options to manage this, and basic check
235 PP/05 DSCP support for outbound connections and control modifier for inbound.
237 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
238 (Only plugin which currently uses this is kerberos4, which nobody should
239 be using, but we should make it available and other future plugins might
240 conceivably use it, even though it would break NAT; stuff *should* be
241 using channel bindings instead).
243 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
244 name; added for Sendmail compatibility; requires admin caller.
245 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
246 just ignore it); requires trusted caller.
247 Also parse but ignore: -Ac -Am -X<logfile>
250 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
252 TL/02 Add +smtp_confirmation as a default logging option.
254 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
255 Patch by Magnus Holmgren from 2007-02-20.
257 TL/04 Bugzilla 1281 - Spec typo.
258 Bugzilla 1283 - Spec typo.
259 Bugzilla 1290 - Spec grammar fixes.
261 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
263 TL/06 Add Experimental DMARC support using libopendmarc libraries.
265 TL/07 Fix an out of order global option causing a segfault. Reported to dev
266 mailing list by by Dmitry Isaikin.
268 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
270 JH/02 Support "G" suffix to numbers in ${if comparisons.
272 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
274 NM/01 Bugzilla 1197 - Spec typo
275 Bugzilla 1196 - Spec examples corrections
277 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
279 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
280 gnutls_enable_pkcs11, but renamed to more accurately indicate its
283 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
284 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
286 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
287 "acl {{name}{arg}...}", and optional args on acl condition
290 JH/05 Permit multiple router/transport headers_add/remove lines.
292 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
294 JH/07 Avoid using a waiting database for a single-message-only transport.
295 Performance patch from Paul Fisher. Bugzilla 1262.
297 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
300 JH/09 Add $headers_added variable, with content from use of ACL modifier
301 add_header (but not yet added to the message). Bugzilla 199.
303 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
304 Pulled from Bugzilla 817 by Wolfgang Breyha.
306 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
308 (nb: this is the same fix as in Exim 4.80.1)
310 JH/11 Add A= logging on delivery lines, and a client_set_id option on
313 JH/12 Add optional authenticated_sender logging to A= and a log_selector
316 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
318 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
319 advertise SMTP AUTH mechanism to us, instead of a generic
320 protocol violation error. Also, make Exim more robust to bad
321 data from the Dovecot auth socket.
323 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
325 When a queue runner is handling a message, Exim first routes the
326 recipient addresses, during which it prunes them based on the retry
327 hints database. After that it attempts to deliver the message to
328 any remaining recipients. It then updates the hints database using
331 So if a recipient address works intermittently, it can get repeatedly
332 deferred at routing time. The retry hints record remains fresh so the
333 address never reaches the final cutoff time.
335 This is a fairly common occurrence when a user is bumping up against
336 their storage quota. Exim had some logic in its local delivery code
337 to deal with this. However it did not apply to per-recipient defers
338 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
340 This change adds a proper retry rule check during routing so that the
341 final cutoff time is checked against the message's age. We only do
342 this check if there is an address retry record and there is not a
343 domain retry record; this implies that previous attempts to handle
344 the address had the retry_use_local_parts option turned on. We use
345 this as an approximation for the destination being like a local
346 delivery, as in LMTP.
348 I suspect this new check makes the old local delivery cutoff check
349 redundant, but I have not verified this so I left the code in place.
351 TF/02 Correct gecos expansion when From: is a prefix of the username.
353 Test 0254 submits a message to Exim with the header
357 When I ran the test suite under the user fanf2, Exim expanded
358 the header to contain my full name, whereas it should have added
359 a Resent-Sender: header. It erroneously treats any prefix of the
360 username as equal to the username.
362 This change corrects that bug.
364 GF/01 DCC debug and logging tidyup
365 Error conditions log to paniclog rather than rejectlog.
366 Debug lines prefixed by "DCC: " to remove any ambiguity.
368 TF/03 Avoid unnecessary rebuilds of lookup-related code.
370 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
371 Bug spotted by Jeremy Harris; was flawed since initial commit.
372 Would have resulted in OCSP responses post-SNI triggering an Exim
373 NULL dereference and crash.
375 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
377 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
378 Bug detection, analysis and fix by Samuel Thibault.
379 Bugzilla 1331, Debian bug #698092.
381 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
383 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
384 Server implementation by Todd Lyons, client by JH.
385 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
386 config variable "prdr_enable" controls whether the server
387 advertises the facility. If the client requests PRDR a new
388 acl_data_smtp_prdr ACL is called once for each recipient, after
389 the body content is received and before the acl_smtp_data ACL.
390 The client is controlled by bolth of: a hosts_try_prdr option
391 on the smtp transport, and the server advertisement.
392 Default client logging of deliveries and rejections involving
393 PRDR are flagged with the string "PRDR".
395 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
396 fclose(). Diagnosis by Todd Lyons.
398 PP/17 Update configure.default to handle IPv6 localhost better.
399 Patch by Alain Williams (plus minor tweaks).
402 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
403 This is now consistent with GnuTLS, and is now documented: the
404 previous undocumented portable approach to treating the option as
405 unset was to force an expansion failure. That still works, and
406 an empty string is now equivalent.
408 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
409 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
410 not performing validation itself.
412 PP/20 Added force_command boolean option to pipe transport.
413 Patch from Nick Koston, of cPanel Inc.
415 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
418 TF/04 Added udpsend ACL modifer and hexquote expansion operator
420 PP/21 Fix eximon continuous updating with timestamped log-files.
421 Broken in a format-string cleanup in 4.80, missed when I repaired the
422 other false fix of the same issue.
423 Report and fix from Heiko Schlichting.
426 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
427 Report from Prashanth Katuri.
429 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
430 It's SecureTransport, so affects any MacOS clients which use the
431 system-integrated TLS libraries, including email clients.
433 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
434 using a MIME ACL for non-SMTP local injection.
435 Report and assistance in diagnosis by Warren Baker.
437 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
439 JH/16 Fix comparisons for 64b. Bugzilla 1385.
441 TL/09 Add expansion variable $authenticated_fail_id to keep track of
442 last id that failed so it may be referenced in subsequent ACL's.
444 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
447 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
448 ldap library initialization, allowing self-signed CA's to be
449 used. Also properly sets require_cert option later in code by
450 using NULL (global ldap config) instead of ldap handle (per
451 session). Bug diagnosis and testing by alxgomz.
453 TL/12 Enhanced documentation in the ratelimit.pl script provided in
454 the src/util/ subdirectory.
456 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
457 renamed to Transport Post Delivery Action by Jeremy Harris, as
460 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
461 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
462 redis_servers = needs to be configured which will be used by the redis
463 lookup. Patch from Warren Baker, of The Packet Hub.
465 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
467 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
468 hostname or reverse DNS when processing a host list. Used suggestions
469 from multiple comments on this bug.
471 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
473 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
474 Missed a few lines, added it to make the runtest require no keyboard
477 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
478 contains upper case chars. Make router use caseful_local_part.
480 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
481 support when GnuTLS has been built with p11-kit.
487 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
489 This, or similar/improved, will also be change PP/11 of 4.82.
495 PP/01 Handle short writes when writing local log-files.
496 In practice, only affects FreeBSD (8 onwards).
497 Bugzilla 1053, with thanks to Dmitry Isaikin.
499 NM/01 Bugzilla 949 - Documentation tweak
501 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
504 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
506 PP/02 Implemented gsasl authenticator.
508 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
510 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
511 `pkg-config foo` for cflags/libs.
513 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
514 with rest of GSASL and with heimdal_gssapi.
516 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
517 `pkg-config foo` for cflags/libs for the TLS implementation.
519 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
520 properties get this fed in as external SSF. A number of robustness
521 and debugging improvements to the cyrus_sasl authenticator.
523 PP/08 cyrus_sasl server now expands the server_realm option.
525 PP/09 Bugzilla 1214 - Log authentication information in reject log.
526 Patch by Jeremy Harris.
528 PP/10 Added dbmjz lookup type.
530 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
532 PP/12 MAIL args handles TAB as well as SP, for better interop with
533 non-compliant senders.
534 Analysis and variant patch by Todd Lyons.
536 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
537 Bug report from Lars Müller <lars@samba.org> (via SUSE),
538 Patch from Dirk Mueller <dmueller@suse.com>
540 PP/13 tls_peerdn now print-escaped for spool files.
541 Observed some $tls_peerdn in wild which contained \n, which resulted
542 in spool file corruption.
544 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
545 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
546 or write after TLS renegotiation, which otherwise led to messages
549 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
550 as a tracking header (ie: a signed header comes before the signature).
551 Patch from Wolfgang Breyha.
553 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
554 comma-sep list; embedded commas doubled.
556 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
558 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
560 Report and patch from Dmitry Banschikov.
562 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
563 Removed SSL_clear() after SSL_new() which led to protocol negotiation
564 failures. We appear to now support TLS1.1+ with Exim.
566 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
567 lets Exim select keys and certificates based upon TLS SNI from client.
568 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
569 before an outbound SMTP session. New log_selector, +tls_sni.
571 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
572 NULL dereference. Report and patch from Alun Jones.
574 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
575 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
576 Not seeing resolver debug output on NetBSD, but suspect this is a
577 resolver implementation change.
579 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
580 Left warnings. Added "eximon gdb" invocation mode.
582 PP/21 Defaulting "accept_8bitmime" to true, not false.
584 PP/22 Added -bw for inetd wait mode support.
586 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
587 locate the relevant includes and libraries. Made this the default.
589 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
590 Bugzilla 1246, report and most of solution from Tomasz Kusy.
592 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
593 This may cause build issues on older platforms.
595 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
596 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
597 gnutls_require_mac & gnutls_require_protocols (no longer supported).
598 Added SNI support via GnuTLS too.
599 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
601 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
603 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
604 Applied second patch from Janne, implementing suggestion to default
605 multiple-strings-in-record handling to match SPF spec.
607 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
609 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
610 read-only, out of scope).
611 Patch from Wolfgang Breyha, report from Stuart Northfield.
613 PP/29 Fix three issues highlighted by clang analyser static analysis.
614 Only crash-plausible issue would require the Cambridge-specific
615 iplookup router and a misconfiguration.
616 Report from Marcin Mirosław.
618 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
620 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
621 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
622 As part of this, removing so much warning spew let me fix some minor
623 real issues in debug logging.
625 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
626 assignment on my part. Fixed.
628 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
629 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
630 Janne Snabb (who went above and beyond: thank you).
632 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
633 string otherwise requires a connection and a bunch more work and it's
634 relatively easy to get wrong. Should also expose TLS library linkage
637 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
638 64-bit ${eval} (JH/03).
640 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
641 GNU libc to support some of the 64-bit stuff, should not lead to
642 conflicts. Defined before os.h is pulled in, so if a given platform
643 needs to override this, it can.
645 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
646 protection layer was required, which is not implemented.
647 Bugzilla 1254, patch from Wolfgang Breyha.
649 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
650 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
651 tls_dhparam take prime identifiers. Also unbreak combination of
652 OpenSSL+DH_params+TLSSNI.
654 PP/39 Disable SSLv2 by default in OpenSSL support.
660 PP/01 Solaris build fix for Oracle's LDAP libraries.
661 Bugzilla 1109, patch from Stephen Usher.
663 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
665 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
668 TF/02 Fix a couple more cases where we did not log the error message
669 when unlink() failed. See also change 4.74-TF/03.
671 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
672 lock up or crash if it happened to be inside a call to libc when it
673 got a SIGUSR1 from exiwhat.
675 The SIGUSR1 handler appends the current process status to the process
676 log which is later printed by exiwhat. It used to use the general
677 purpose logging code to do this, but several functions it calls are
678 not safe for signals.
680 The new output code in the SIGUSR1 handler is specific to the process
681 log, and simple enough that it's easy to inspect for signal safety.
682 Removing some special cases also simplifies the general logging code.
683 Removing the spurious timestamps from the process log simplifies
686 TF/04 Improved ratelimit ACL condition.
688 The /noupdate option has been deprecated in favour of /readonly which
689 has clearer semantics. The /leaky, /strict, and /readonly update modes
690 are mutually exclusive. The update mode is no longer included in the
691 database key; it just determines when the database is updated. (This
692 means that when you upgrde Exim will forget old rate measurements.)
694 Exim now checks that the per_* options are used with an update mode that
695 makes sense for the current ACL. For example, when Exim is processing a
696 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
697 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
698 must specify per_mail/readonly. If you omit the update mode it defaults to
699 /leaky where that makes sense (as before) or /readonly where required.
701 The /noupdate option is now undocumented but still supported for
702 backwards compatibility. It is equivalent to /readonly except that in
703 ACLs where /readonly is required you may specify /leaky/noupdate or
704 /strict/noupdate which are treated the same as /readonly.
706 A useful new feature is the /count= option. This is a generalization
707 of the per_byte option, so that you can measure the throughput of other
708 aggregate values. For example, the per_byte option is now equivalent
709 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
711 The per_rcpt option has been generalized using the /count= mechanism
712 (though it's more complicated than the per_byte equivalence). When it is
713 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
714 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
715 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
716 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
717 Note that using per_rcpt with a non-readonly update mode in more than
718 one ACL will cause the recipients to be double-counted. (The per_mail
719 and per_byte options don't have this problem.)
721 The handling of very low rates has changed slightly. If the computed rate
722 is less than the event's count (usually one) then this event is the first
723 after a long gap. In this case the rate is set to the same as this event's
724 count, so that the first message of a spam run is counted properly.
726 The major new feature is a mechanism for counting the rate of unique
727 events. The new per_addr option counts the number of different
728 recipients that someone has sent messages to in the last time period. It
729 behaves like per_rcpt if all the recipient addresses are different, but
730 duplicate recipient addresses do not increase the measured rate. Like
731 the /count= option this is a general mechanism, so the per_addr option
732 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
733 example, measure the rate that a client uses different sender addresses
734 with the options per_mail/unique=$sender_address. There are further
735 details in the main documentation.
737 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
739 TF/06 Removed a few PCRE remnants.
741 TF/07 Automatically extract Exim's version number from tags in the git
742 repository when doing development or release builds.
744 PP/02 Raise smtp_cmd_buffer_size to 16kB.
745 Bugzilla 879. Patch from Paul Fisher.
747 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
748 Heavily based on revision 40f9a89a from Simon Arlott's tree.
751 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
753 PP/05 Variable $av_failed, true if the AV scanner deferred.
754 Bugzilla 1078. Patch from John Horne.
756 PP/06 Stop make process more reliably on build failure.
757 Bugzilla 1087. Patch from Heiko Schlittermann.
759 PP/07 Make maildir_use_size_file an _expandable_ boolean.
760 Bugzilla 1089. Patch from Heiko Schlittermann.
762 PP/08 Handle ${run} returning more data than OS pipe buffer size.
763 Bugzilla 1131. Patch from Holger Weiß.
765 PP/09 Handle IPv6 addresses with SPF.
766 Bugzilla 860. Patch from Wolfgang Breyha.
768 PP/10 GnuTLS: support TLS 1.2 & 1.1.
770 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
773 PP/11 match_* no longer expand right-hand-side by default.
774 New compile-time build option, EXPAND_LISTMATCH_RHS.
775 New expansion conditions, "inlist", "inlisti".
777 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
779 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
781 PP/14 fix log_write() format string regression from TF/03.
782 Bugzilla 1152. Patch from Dmitry Isaikin.
788 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
790 PP/02 Harmonised TLS library version reporting; only show if debugging.
791 Layout now matches that introduced for other libraries in 4.74 PP/03.
793 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
795 PP/04 New "dns_use_edns0" global option.
797 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
800 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
801 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
803 TK/01 Updated PolarSSL code to 0.14.2.
804 Bugzilla 1097. Patch from Andreas Metzler.
806 PP/07 Catch divide-by-zero in ${eval:...}.
809 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
812 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
813 format-string attack -- SECURITY: remote arbitrary code execution.
815 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
816 time unintentionally subject to list matching rules, letting the header
817 cause arbitrary Exim lookups (of items which can occur in lists, *not*
818 arbitrary string expansion). This allowed for information disclosure.
820 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
821 INT_MIN/-1 -- value coerced to INT_MAX.
827 NM/01 Workround for PCRE version dependency in version reporting
830 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
831 This fixes portability to compilers other than gcc, notably
832 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
834 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
835 makefiles for portability to HP-UX and POSIX correctness.
837 PP/01 Permit LOOKUP_foo enabling on the make command-line.
838 Also via indented variable definition in the Makefile.
839 (Debugging by Oliver Heesakkers).
841 PP/02 Restore caching of spamd results with expanded spamd_address.
842 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
844 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
845 Improves build reliability. Fix from: Frank Elsner
847 NM/02 Fix wide character breakage in the rfc2047 coding
848 Fixes bug 1064. Patch from Andrey N. Oktyabrski
850 NM/03 Allow underscore in dnslist lookups
851 Fixes bug 1026. Patch from Graeme Fowler
853 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
854 Code patches from Adam Ciarcinski of NetBSD.
856 NM/04 Fixed exiqgrep to cope with mailq missing size issue
859 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
860 is logged, to avoid truncation. Patch from John Horne.
862 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
863 Patch from Jakob Hirsch.
865 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
866 SQL string expansion failure details.
867 Patch from Andrey Oktyabrski.
869 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
870 Patch from Simon Arlott.
872 PP/09 New lookups functionality failed to compile on old gcc which rejects
873 extern declarations in function scope.
874 Patch from Oliver Fleischmann
876 PP/10 Use sig_atomic_t for flags set from signal handlers.
877 Check getgroups() return and improve debugging.
878 Fixed developed for diagnosis in bug 927 (which turned out to be
881 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
882 Patch from Mark Zealey.
884 PP/12 Bugzilla 1056: Improved spamd server selection.
885 Patch from Mark Zealey.
887 PP/13 Bugzilla 1086: Deal with maildir quota file races.
888 Based on patch from Heiko Schlittermann.
890 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
891 Patch from Uwe Doering, sign-off by Michael Haardt.
893 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
894 variable declaration deep within a block. Bug and patch from
897 PP/15 lookups-Makefile IRIX compatibilty coercion.
899 PP/16 Make DISABLE_DKIM build knob functional.
901 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
902 Patch by Simon Arlott
904 TF/03 Fix valgrind.h portability to C89 compilers that do not support
905 variable argument macros. Our copy now differs from upstream.
911 TF/01 Failure to get a lock on a hints database can have serious
912 consequences so log it to the panic log.
914 TF/02 Log LMTP confirmation messages in the same way as SMTP,
915 controlled using the smtp_confirmation log selector.
917 TF/03 Include the error message when we fail to unlink a spool file.
919 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
920 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
921 for maintaining out-of-tree patches for some time.
923 PP/01 Bugzilla 139: Documentation and portability issues.
924 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
925 Handle per-OS dynamic-module compilation flags.
927 PP/02 Let /dev/null have normal permissions.
928 The 4.73 fixes were a little too stringent and complained about the
929 permissions on /dev/null. Exempt it from some checks.
930 Reported by Andreas M. Kirchwitz.
932 PP/03 Report version information for many libraries, including
933 Exim version information for dynamically loaded libraries. Created
934 version.h, now support a version extension string for distributors
935 who patch heavily. Dynamic module ABI change.
937 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
938 privilege escalation vulnerability whereby the Exim run-time user
939 can cause root to append content of the attacker's choosing to
942 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
945 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
946 If dropping privileges for untrusted macros, we disabled normal logging
947 on the basis that it would fail; for the Exim run-time user, this is not
948 the case, and it resulted in successful deliveries going unlogged.
949 Fixed. Reported by Andreas Metzler.
955 PP/01 Date: & Message-Id: revert to normally being appended to a message,
956 only prepend for the Resent-* case. Fixes regression introduced in
957 Exim 4.70 by NM/22 for Bugzilla 607.
959 PP/02 Include check_rfc2047_length in configure.default because we're seeing
960 increasing numbers of administrators be bitten by this.
962 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
964 PP/03 Bugzilla 994: added openssl_options main configuration option.
966 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
968 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
970 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
972 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
973 without a peer certificate, leading to a segfault because of an
974 assumption that peers always have certificates. Be a little more
975 paranoid. Problem reported by Martin Tscholak.
977 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
978 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
979 NB: ClamAV planning to remove STREAM in "middle of 2010".
980 CL also introduces -bmalware, various -d+acl logging additions and
981 more caution in buffer sizes.
983 PP/09 Implemented reverse_ip expansion operator.
985 PP/10 Bugzilla 937: provide a "debug" ACL control.
987 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
989 PP/12 Bugzilla 973: Implement --version.
991 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
993 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
995 PP/15 Bugzilla 816: support multiple condition rules on Routers.
997 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
998 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
999 ignore trailing whitespace.
1001 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1003 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1006 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1007 Notification from Dr Andrew Aitchison.
1009 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1010 ExtendedDetectionInfo response format.
1011 Notification from John Horne.
1013 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1016 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1017 XSL and documented dependency on system catalogs, with examples of how
1020 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1023 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1024 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1025 configuration file which is writeable by the Exim user or group.
1027 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1028 of configuration files to cover files specified with the -C option if
1029 they are going to be used with root privileges, not just the default
1032 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1033 option (effectively making it always true).
1035 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1036 files to be used while preserving root privileges.
1038 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1039 that rogue child processes cannot use them.
1041 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1042 run-time user, instead of root.
1044 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1045 Exim run-time user without dropping privileges.
1047 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1048 result string, instead of calling string_vformat() twice with the same
1051 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1052 for other users. Others should always drop root privileges if they use
1053 -C on the command line, even for a whitelisted configure file.
1055 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1057 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1063 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1064 $data_path, and $header_path variables; fixed documentation bugs and
1067 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1068 exipick to access non-standard spools, including the "frozen" queue
1071 NM/01 Bugzilla 965: Support mysql stored procedures.
1072 Patch from Alain Williams
1074 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1076 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1077 Patch from Andreas Metzler
1079 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1080 Patch from Kirill Miazine
1082 NM/05 Bugzilla 671: Added umask to procmail example.
1084 JJ/03 installed exipick 20100323.0, fixing doc bug
1086 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1087 directory. Notification and patch from Dan Rosenberg.
1089 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1091 TK/02 Improve log output when DKIM signing operation fails.
1093 MH/01 Treat the transport option dkim_domain as a colon separated
1094 list, not as a single string, and sign the message with each element,
1095 omitting multiple occurences of the same signer.
1097 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1098 Bugzilla 985, 986. Patch by Simon Arlott
1100 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1101 Patch by Simon Arlott
1103 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1104 MBX locking. Notification from Dan Rosenberg.
1110 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1112 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1114 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1116 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1118 NM/04 Bugzilla 915: Flag broken perl installation during build.
1124 TK/01 Added patch by Johannes Berg that expands the main option
1125 "spamd_address" if it starts with a dollar sign.
1127 TK/02 Write list of recipients to X-Envelope-Sender header when building
1128 the mbox-format spool file for content scanning (suggested by Jakob
1131 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1132 (http://www.dcc-servers.net/) support via dccifd. Activated by
1133 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1135 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1136 by Mark Daniel Reidel <mr@df.eu>.
1138 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1139 When building exim an external PCRE library is now needed -
1140 PCRE is a system library on the majority of modern systems.
1141 See entry on PCRE_LIBS in EDITME file.
1143 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1144 conversation. Added nologin parameter to request.
1145 Patch contributed by Kirill Miazine.
1147 TF/01 Do not log submission mode rewrites if they do not change the address.
1149 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1151 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1152 log files in place. Contributed by Roberto Lima.
1154 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1156 TF/03 Bugzilla 615: When checking the local_parts router precondition
1157 after a local_part_suffix or local_part_prefix option, Exim now
1158 does not use the address's named list lookup cache, since this
1159 contains cached lookups for the whole local part.
1161 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1162 Robert Millan. Documentation is in experimental-spec.txt.
1164 TF/04 Bugzilla 668: Fix parallel build (make -j).
1166 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1168 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1169 Patch provided by Jan Srzednicki.
1171 TF/05 Leading white space used to be stripped from $spam_report which
1172 wrecked the formatting. Now it is preserved.
1174 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1175 that they are available at delivery time.
1177 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1179 TF/08 TLS error reporting now respects the incoming_interface and
1180 incoming_port log selectors.
1182 TF/09 Produce a more useful error message if an SMTP transport's hosts
1183 setting expands to an empty string.
1185 NM/06 Bugzilla 744: EXPN did not work under TLS.
1186 Patch provided by Phil Pennock.
1188 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1189 Patch provided by Richard Godbee.
1191 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1192 acl_smtp_notquit, added index entry.
1194 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1195 Patch provided by Eugene Bujak.
1197 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1198 accept(). Patch provided by Maxim Dounin.
1200 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1201 Patch provided by Phil Pennock.
1203 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1205 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1206 Patch provided by Brad "anomie" Jorsch.
1208 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1209 Patch provided by Dean Brooks.
1211 TK/05 Add native DKIM support (does not depend on external libraries).
1213 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1214 Patch provided by Graeme Fowler.
1216 NM/16 Bugzilla 851: Documentation example syntax fix.
1218 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1220 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1223 NM/19 Bugzilla 745: TLS version reporting.
1224 Patch provided by Phil Pennock.
1226 NM/20 Bugzilla 167: bool: condition support.
1227 Patch provided by Phil Pennock.
1229 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1230 clients. Patch provided by Phil Pennock.
1232 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1233 Patch provided by Brad "anomie" Jorsch.
1235 NM/23 Bugzilla 687: Fix misparses in eximstats.
1236 Patch provided by Heiko Schlittermann.
1238 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1239 Patch provided by Heiko Schlittermann.
1241 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1242 plus update to original patch.
1244 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1246 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1247 Patch provided by David Brownlee.
1249 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1251 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1253 NM/30 Bugzilla 888: TLS documentation bugfixes.
1255 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1257 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1258 Unlike the original bugzilla I have changed all shell scripts in src tree.
1260 NM/33 Bugzilla 898: Transport filter timeout fix.
1261 Patch by Todd Rinaldo.
1263 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1264 Patch by Serge Demonchaux.
1266 NM/35 Bugzilla 39: Base64 decode bug fixes.
1267 Patch by Jakob Hirsch.
1269 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1271 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1273 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1275 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1281 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1282 ALT-N's libdkim that I have put here:
1283 http://duncanthrax.net/exim-experimental/
1285 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1286 were called 'true' and it seems that C99 defines that as a reserved
1287 keyword to be used with 'bool' variable types. That means you could
1288 not include C99-style headers which use bools without triggering
1289 build errors in sieve.c.
1291 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1292 as mailq or other aliases. Changed the --help handling significantly
1293 to do whats expected. exim_usage() emits usage/help information.
1295 SC/01 Added the -bylocaldomain option to eximstats.
1297 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1299 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1301 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1307 PH/01 Another patch from the Sieve maintainer.
1309 PH/02 When an IPv6 address is converted to a string for single-key lookup
1310 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1311 dots are used instead of colons so that keys in lsearch files need not
1312 contain colons. This was done some time before quoting was made available
1313 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1314 (notated using the quote facility) so as to distinguish them from IPv4
1315 keys. This meant that lookups for IP addresses in host lists did not work
1316 for iplsearch lookups.
1318 This has been fixed by arranging for IPv6 addresses to be expressed with
1319 colons if the lookup type is iplsearch. This is not incompatible, because
1320 previously such lookups could never work.
1322 The situation is now rather anomolous, since one *can* have colons in
1323 ordinary lsearch keys. However, making the change in all cases is
1324 incompatible and would probably break a number of configurations.
1326 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1329 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1330 conversion specification without a maximum field width, thereby enabling
1331 a rogue spamd server to cause a buffer overflow. While nobody in their
1332 right mind would setup Exim to query an untrusted spamd server, an
1333 attacker that gains access to a server running spamd could potentially
1334 exploit this vulnerability to run arbitrary code as the Exim user.
1336 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1337 $primary_hostname instead of what libspf2 thinks the hosts name is.
1339 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1340 a directory entry by the name of the lookup key. Previously, if a
1341 symlink pointed to a non-existing file or a file in a directory that
1342 Exim lacked permissions to read, a lookup for a key matching that
1343 symlink would fail. Now it is enough that a matching directory entry
1344 exists, symlink or not. (Bugzilla 503.)
1346 PH/03 The body_linecount and body_zerocount variables are now exported in the
1349 PH/04 Added the $dnslist_matched variable.
1351 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1352 This means they are set thereafter only if the connection becomes
1355 PH/06 Added the client_condition to authenticators so that some can be skipped
1356 by clients under certain conditions.
1358 PH/07 The error message for a badly-placed control=no_multiline_responses left
1359 "_responses" off the end of the name.
1361 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1363 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1364 (without spaces) instead of just copying the configuration text.
1366 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1368 PH/11 Added $max_received_linelength.
1370 PH/12 Added +ignore_defer and +include_defer to host lists.
1372 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1373 way in which PCRE > 7.0 is built.
1375 PH/14 Implemented queue_only_load_latch.
1377 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1378 MAIL command. The effect was to mangle the value on 64-bit systems.
1380 PH/16 Another patch from the Sieve maintainer.
1382 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1384 PH/18 If a system quota error occurred while trying to create the file for
1385 a maildir delivery, the message "Mailbox is full" was not appended to the
1386 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1387 only to a quota excession during the actual writing of the file.
1389 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1390 characters?) which causes problems in log lines. The DN values are now
1391 passed through string_printing() before being added to log lines.
1393 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1394 and InterBase are left for another time.)
1396 PH/21 Added message_body_newlines option.
1398 PH/22 Guard against possible overflow in moan_check_errorcopy().
1400 PH/23 POSIX allows open() to be a macro; guard against that.
1402 PH/24 If the recipient of an error message contained an @ in the local part
1403 (suitably quoted, of course), incorrect values were put in $domain and
1404 $local_part during the evaluation of errors_copy.
1410 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1411 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1414 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1415 issue a MAIL command.
1417 PH/02 In an ACL statement such as
1419 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1421 if a client was not listed at all, or was listed with a value other than
1422 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1423 the condition was not true (as it should be), so access was not denied.
1424 The bug was that the ! inversion was incorrectly passed on to the second
1425 item. This has been fixed.
1427 PH/03 Added additional dnslists conditions == and =& which are different from
1428 = and & when the dns lookup returns more than one IP address.
1430 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1431 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1433 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1434 FSYNC, which compiles an option called disable_fsync that allows for
1435 bypassing fsync(). The documentation is heavily laced with warnings.
1437 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1439 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1440 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1441 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1442 including adding "make clean"; (3) Added -fPIC when compiling the test
1443 dynamically loaded module, to get rid of a warning.
1445 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1446 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1447 = 0s. The bug is otherwise harmless.
1449 PH/07 There was a bug in the dovecot authenticator such that the value of
1450 $auth1 could be overwritten, and so not correctly preserved, after a
1451 successful authentication. This usually meant that the value preserved by
1452 the server_setid option was incorrect.
1454 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1456 PH/09 Installed PCRE release 7.0.
1458 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1459 run for batched SMTP input. It is now run at the start of every message
1460 in the batch. While fixing this I discovered that the process information
1461 (output by running exiwhat) was not always getting set for -bs and -bS
1462 input. This is fixed, and it now also says "batched" for BSMTP.
1464 PH/11 Added control=no_pipelining.
1466 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1467 patch, slightly modified), and move the expansion of helo_data till after
1468 the connection is made in the smtp transport (so it can use these
1471 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1473 PH/14 Added log_selector = +pid.
1475 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1477 PH/16 Add ${if forany and ${if forall.
1479 PH/17 Added dsn_from option to vary the From: line in DSNs.
1481 PH/18 Flush SMTP output before performing a callout, unless control =
1482 no_callout_flush is set.
1484 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1485 was true (the default) a successful delivery failed to delete the retry
1486 item, thus causing premature timeout of the address. The bug is now
1489 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1491 PH/21 Long custom messages for fakedefer and fakereject are now split up
1492 into multiline reponses in the same way that messages for "deny" and
1493 other ACL rejections are.
1495 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1496 with slight modification.
1498 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1499 draft, changing the syntax and factoring some duplicate code".
1501 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1502 for deliveries of the second and subsequent messages over the same SMTP
1505 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1506 ${reduce, with only minor "tidies".
1508 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1510 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1511 expansion side effects.
1513 PH/27 When a message times out after an over-quota error from an Exim-imposed
1514 quota, the bounce message says "mailbox is full". This message was not
1515 being given when it was a system quota that was exceeded. It now should
1518 MH/03 Made $recipients available in local_scan(). local_scan() already has
1519 better access to the recipient list through recipients_list[], but
1520 $recipients can be useful in postmaster-provided expansion strings.
1522 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1523 in the case of a MAIL command with additional options following the
1524 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1525 were accidentally chopped off.
1527 PH/29 SMTP synchronization checks are implemented when a command is read -
1528 there is a check that no more input is waiting when there shouldn't be
1529 any. However, for some commands, a delay in an ACL can mean that it is
1530 some time before the response is written. In this time, more input might
1531 arrive, invalidly. So now there are extra checks after an ACL has run for
1532 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1533 pipelining has not been advertised.
1535 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1537 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1538 returned in the SMTP response when smtp_return_error_details was false.
1539 This has been fixed.
1541 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1542 instead of the C I/O that was originally supplied, because problems were
1543 reported on Solaris.
1545 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1546 Exim which did not show up earlier: it was assuming that a call to
1547 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1548 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1549 was a macro that became an assignment, so it seemed to work. This has
1550 changed to a proper function call with a void return, hence the compile
1551 error. Exim's code has been fixed.
1553 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1556 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1558 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1560 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1561 because it turns out to be too big or there is a timeout) there is no
1562 "Completed" line in the log. When some messages of this type were
1563 selected by exigrep, they were listed as "not completed". Others were
1564 picked up by some special patterns. I have improved the selection
1565 criteria to be more general.
1567 PH/38 The host_find_failed option in the manualroute router can now be set
1568 to "ignore", to completely ignore a host whose IP address cannot be
1569 found. If all hosts are ignored, the behaviour is controlled by the new
1570 host_all_ignored option.
1572 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1573 homing or because of multiple MX records with /mx) generated more than
1574 one IP address, and the following item turned out to be the local host,
1575 all the secondary addresses of the first item were incorrectly removed
1576 from the list, along with the local host and any following hosts (which
1577 is what is supposed to happen).
1579 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1580 whoever called Exim into the -H file. In the case of the daemon it was
1581 behaving confusingly. When first started, it used values for whoever
1582 started the daemon, but after a SIGHUP it used the Exim user (because it
1583 calls itself on a restart). I have changed the code so that it now always
1586 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1587 message are rejected with the same error (e.g. no authentication or bad
1588 sender address), and a DATA command is nevertheless sent (as can happen
1589 with PIPELINING or a stupid MUA), the error message that was given to the
1590 RCPT commands is included in the rejection of the DATA command. This is
1591 intended to be helpful for MUAs that show only the final error to their
1594 PH/42 Another patch from the Sieve maintainer.
1596 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1597 Eximstats - Fixed some broken HTML links and added missing column headers
1599 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1600 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1602 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1604 PH/43 Yet another patch from the Sieve maintainer.
1606 PH/44 I found a way to check for a TCP/IP connection going away before sending
1607 the response to the final '.' that terminates a message, but only in the
1608 case where the client has not sent further data following the '.'
1609 (unfortunately, this is allowed). However, in many cases there won't be
1610 any further data because there won't be any more messages to send. A call
1611 to select() can be used: if it shows that the input is "ready", there is
1612 either input waiting, or the socket has been closed. An attempt to read
1613 the next input character can distinguish the two cases. Previously, Exim
1614 would have sent an OK response which the client would never have see.
1615 This could lead to message repetition. This fix should cure that, at
1616 least in a lot of common cases.
1618 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1619 advertised in response to EHLO.
1625 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1626 fixed by 4.65/MH/01 (is this a record?) are fixed:
1628 (i) An empty string was always treated as zero by the numeric comparison
1629 operators. This behaviour has been restored.
1631 (ii) It is documented that the numeric comparison operators always treat
1632 their arguments as decimal numbers. This was broken in that numbers
1633 starting with 0 were being interpreted as octal.
1635 While fixing these problems I realized that there was another issue that
1636 hadn't been noticed. Values of message_size_limit (both the global option
1637 and the transport option) were treated as octal if they started with 0.
1638 The documentation was vague. These values are now always treated as
1639 decimal, and I will make that clear in the documentation.
1645 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1646 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1649 MH/01 Don't check that the operands of numeric comparison operators are
1650 integers when their expansion is in "skipping" mode (fixes bug
1651 introduced by 4.64-PH/07).
1653 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1654 child addresses, Exim now panics and dies. Previously, because the count
1655 is held in a short int, deliveries were likely to be lost. As such a
1656 large number of recipients for a single message is ridiculous
1657 (performance will be very, very poor), I have chosen to impose a limit
1658 rather than extend the field.
1664 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1665 leftover -K file (the existence of which was triggered by #402).
1666 While we were at it, introduced process PID as part of the -K
1667 filename. This should rule out race conditions when creating
1670 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1671 processing considerably. Previous code took too long for large mails,
1672 triggering a timeout which in turn triggers #401.
1674 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1675 in the DK code in transports.c. sendfile() is not really portable,
1676 hence the _LINUX specificness.
1678 TF/01 In the add_headers option to the mail command in an Exim filter,
1679 there was a bug that Exim would claim a syntax error in any
1680 header after the first one which had an odd number of characters
1683 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1684 callout verification, Exim cached a "reject" for the entire domain. This
1685 is correct for most verifications, but it is not correct for a recipient
1686 verification with use_sender or use_postmaster set, because in that case
1687 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1688 case of MAIL FROM:<> rejection from other early rejections (e.g.
1689 rejection of HELO). When verifying a recipient using a non-null MAIL
1690 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1691 Whatever the result of the callout, the value of the domain cache is
1692 left unchanged (for any other kind of callout, getting as far as trying
1693 RCPT means that the domain itself is ok).
1695 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1698 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1699 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1700 session, the daemon ignores ECONNECT errors and logs others; it now
1701 ignores EPIPE as well.
1703 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1704 (quoted-printable decoding).
1706 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1707 later the small subsequent patch to fix an introduced bug.
1709 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1711 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1713 PH/08 An error is now given if message_size_limit is specified negative.
1715 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1716 to be given (somewhat) arbitrary names.
1718 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1721 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1722 miscellaneous code fixes
1724 PH/10 Added the log_reject_target ACL modifier to specify where to log
1727 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1728 hostname. This is wrong, because it relates to the incoming message (and
1729 probably the interface on which it is arriving) and not to the outgoing
1730 callout (which could be using a different interface). This has been
1731 changed to use the value of the helo_data option from the smtp transport
1732 instead - this is what is used when a message is actually being sent. If
1733 there is no remote transport (possible with a router that sets up host
1734 addresses), $smtp_active_hostname is used.
1736 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1737 tweaks were necessary in order to get it to work (see also 21 below):
1738 (a) The code assumed that strncpy() returns a negative number on buffer
1739 overflow, which isn't the case. Replaced with Exim's string_format()
1741 (b) There were several signed/unsigned issues. I just did the minimum
1742 hacking in of casts. There is scope for a larger refactoring.
1743 (c) The code used strcasecmp() which is not a standard C function.
1744 Replaced with Exim's strcmpic() function.
1745 (d) The code set only $1; it now sets $auth1 as well.
1746 (e) A simple test gave the error "authentication client didn't specify
1747 service in request". It would seem that Dovecot has changed its
1748 interface. Fortunately there's a specification; I followed it and
1749 changed what the client sends and it appears to be working now.
1751 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1754 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1755 address A is aliased to B and C, where B exists and C does not. Without
1756 -v the output is "A verified" because verification stops after a
1757 successful redirection if more than one address is generated. However,
1758 with -v the child addresses are also verified. Exim was outputting "A
1759 failed to verify" and then showing the successful verification for C,
1760 with its parentage. It now outputs "B failed to verify", showing B's
1761 parentage before showing the successful verification of C.
1763 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1764 look up a TXT record in a specific list after matching in a combined
1767 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1768 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1769 they consult the DNS. I had assumed they would set it the way they
1770 wanted; and indeed my experiments on Linux seem to show that in some
1771 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1772 To be on the safe side, however, I have now made the interface to
1773 host_find_byname() similar to host_find_bydns(), with an argument
1774 containing the DNS resolver options. The host_find_byname() function now
1775 sets these options at its start, just as host_find_bydns() does. The smtp
1776 transport options dns_qualify_single and dns_search_parents are passed to
1777 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1778 of host_find_byname() use the default settings of RES_DEFNAMES
1779 (qualify_single) but not RES_DNSRCH (search_parents).
1781 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1782 spool_read_header() do less string testing, by means of a preliminary
1783 switch on the second character of optional "-foo" lines. (This is
1784 overdue, caused by the large number of possibilities that now exist.
1785 Originally there were few.) While I was there, I also converted the
1786 str(n)cmp tests so they don't re-test the leading "-" and the first
1787 character, in the hope this might squeeze out yet more improvement.
1789 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1790 flag allowing group syntax was set by the header_syntax check but not
1791 turned off, possible causing trouble later; (2) The flag was not being
1792 set at all for the header_verify test, causing "group"-style headers to
1793 be rejected. I have now set it in this case, and also caused header_
1794 verify to ignore an empty address taken from a group. While doing this, I
1795 came across some other cases where the code for allowing group syntax
1796 while scanning a header line wasn't quite right (mostly, not resetting
1797 the flag correctly in the right place). These bugs could have caused
1798 trouble for malformed header lines. I hope it is now all correct.
1800 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1801 with the "reply" argument non-NULL. The code, however (which originally
1802 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1803 but it didn't always do it. This confused somebody who was copying the
1804 code for some other use. I have removed all the tests.
1806 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1807 feature that was used to support insecure browsers during the U.S. crypto
1808 embargo. It requires special client support, and Exim is probably the
1809 only MTA that supported it -- and would never use it because real RSA is
1810 always available. This code has been removed, because it had the bad
1811 effect of slowing Exim down by computing (never used) parameters for the
1812 RSA_EXPORT functionality.
1814 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1815 authenticator to fail if there's a tab character in the incoming data
1816 (there should never be unless someone is messing about, as it's supposed
1817 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1818 if the connection is using TLS or if the remote IP is the same as the
1819 local IP, and the "valid-client-cert option" if a client certificate has
1822 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1823 authenticators. This can be used for authorization after authentication
1824 succeeds. (In the case of plaintext, it servers for both authentication
1827 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1828 if any retry times were supplied.
1830 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1831 connection, where there is no client IP address to check. In this
1832 situation, the verify now always succeeds.
1834 PH/25 Applied John Jetmore's -Mset patch.
1836 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1838 PH/27 In a string expansion for a processed (not raw) header when multiple
1839 headers of the same name were present, leading whitespace was being
1840 removed from all of them, but trailing whitespace was being removed only
1841 from the last one. Now trailing whitespace is removed from each header
1842 before concatenation. Completely empty headers in a concatenation (as
1843 before) are ignored.
1845 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1846 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1848 PH/29 [Removed. This was a change that I later backed out, and forgot to
1849 correct the ChangeLog entry (that I had efficiently created) before
1850 committing the later change.]
1852 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1853 address errors (4xx response to RCPT) over the same connection as other
1854 messages routed to the same hosts. Such deliveries are always "forced",
1855 so retry times are not inspected. This resulted in far too many retries
1856 for the affected addresses. The effect occurred only when there were more
1857 hosts than the hosts_max_try setting in the smtp transport when it had
1858 the 4xx errors. Those hosts that it had tried were not added to the list
1859 of hosts for which the message was waiting, so if all were tried, there
1860 was no problem. Two fixes have been applied:
1862 (i) If there are any address or message errors in an SMTP delivery, none
1863 of the hosts (tried or untried) are now added to the list of hosts
1864 for which the message is waiting, so the message should not be a
1865 candidate for sending over the same connection that was used for a
1866 successful delivery of some other message. This seems entirely
1867 reasonable: after all the message is NOT "waiting for some host".
1868 This is so "obvious" that I'm not sure why it wasn't done
1869 previously. Hope I haven't missed anything, but it can't do any
1870 harm, as the worst effect is to miss an optimization.
1872 (ii) If, despite (i), such a delivery is accidentally attempted, the
1873 routing retry time is respected, so at least it doesn't keep
1874 hammering the server.
1876 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1877 in ${readsocket because some servers need this prod.
1879 PH/32 Added some extra debug output when updating a wait-xxx database.
1881 PH/33 The hint "could be header name not terminated by colon", which has been
1882 given for certain expansion errors for a long time, was not being given
1883 for the ${if def:h_colon_omitted{... case.
1885 PH/34 The spec says: "With one important exception, whenever a domain list is
1886 being scanned, $domain contains the subject domain." There was at least
1887 one case where this was not true.
1889 PH/35 The error "getsockname() failed: connection reset by peer" was being
1890 written to the panic log as well as the main log, but it isn't really
1891 panic-worthy as it just means the connection died rather early on. I have
1892 removed the panic log writing for the ECONNRESET error when getsockname()
1895 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1896 runs only) independently of the message's sender address. This meant
1897 that, if the 4xx error was in fact related to the sender, a different
1898 message to the same recipient with a different sender could confuse
1899 things. In particualar, this can happen when sending to a greylisting
1900 server, but other circumstances could also provoke similar problems.
1901 I have changed the default so that the retry time for these errors is now
1902 based a combination of the sender and recipient addresses. This change
1903 can be overridden by setting address_retry_include_sender=false in the
1906 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1907 remote server are returned as part of bounce messages. This was not
1908 happening for LMTP over a pipe (the lmtp transport), but now it is the
1909 same for both kinds of LMTP.
1911 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1912 in header lines that were in fact CNAMEs. This is no longer the case.
1914 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1915 and queue runs started by the daemon processed all messages. This has
1916 been fixed so that -R and -S can now usefully be given with -q<time>.
1918 PH/40 Import PCRE release 6.7 (fixes some bugs).
1920 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1922 PH/42 Give an error if -q is specified more than once.
1924 PH/43 Renamed the variables $interface_address and $interface_port as
1925 $received_ip_address and $received_port, to make it clear that these
1926 values apply to message reception, and not to the outgoing interface when
1927 a message is delivered. (The old names remain recognized, of course.)
1929 PH/44 There was no timeout on the connect() call when using a Unix domain
1930 socket in the ${readsocket expansion. There now is.
1932 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1933 be meaningful with "accept".
1935 SC/01 Eximstats V1.43
1936 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1938 SC/02 Eximstats V1.44
1939 Use a glob alias rather than an array ref in the generated
1940 parser. This improves both readability and performance.
1942 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1943 Collect SpamAssassin and rejection statistics.
1944 Don't display local sender or destination tables unless
1945 there is data to show.
1946 Added average volumes into the top table text output.
1948 SC/04 Eximstats V1.46
1949 Collect data on the number of addresses (recipients)
1950 as well as the number of messages.
1952 SC/05 Eximstats V1.47
1953 Added 'Message too big' to the list of mail rejection
1954 reasons (thanks to Marco Gaiarin).
1956 SC/06 Eximstats V1.48
1957 Mainlog lines which have GMT offsets and are too short to
1958 have a flag are now skipped.
1960 SC/07 Eximstats V1.49 (Alain Williams)
1961 Added the -emptyok flag.
1963 SC/08 Eximstats V1.50
1964 Fixes for obtaining the IP address from reject messages.
1966 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1967 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1968 whitesspace changes from 4.64-PH/27
1970 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1973 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1974 are found, allow negative numbers in numeric criteria)
1976 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1978 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1979 to match changes made in 4.64-PH/43
1981 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1983 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1984 "domain" is actually an IP address, and force a failure. This locks out
1985 those revolvers/nameservers that support "A-for-A" lookups, in
1986 contravention of the specifications.
1988 PH/48 When a host name was looked up from an IP address, and the subsequent
1989 forward lookup of the name timed out, the host name was left in
1990 $sender_host_name, contrary to the specification.
1992 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1993 restricted to single-key lookups, Exim was not diagnosing an error if
1994 * or *@ was used with a query-style lookup.
1996 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1998 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1999 long ago, but noone interested enough thought of it. Let's just say that
2000 the "1.1" means that there are some new functions that weren't there at
2001 some point in the past.
2003 PH/51 Error processing for expansion failure of helo_data from an smtp
2004 transport during callout processing was broken.
2006 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2007 tested/used via the -bh/-bhc/-bs options.
2009 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2010 bug, fixed in subsequent PCRE releases).
2012 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2013 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2015 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2021 SC/01 Use a glob alias rather than an array ref in eximstats generated
2022 parser. This improves both readability and performance.
2024 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2025 Don't display local sender or destination tables in eximstats unless
2026 there is data to show.
2027 Added average volumes into the eximstats top table text output.
2029 SC/03 Collect data on the number of addresses (recipients) as well
2030 as the number of messages in eximstats.
2032 TF/01 Correct an error in the documentation for the redirect router. Exim
2033 does (usually) call initgroups() when daemonizing.
2035 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2036 with consistent privilege compared to when running as a daemon.
2038 TF/03 Note in the spec that $authenticated_id is not set for local
2039 submissions from trusted users.
2041 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2042 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2044 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2045 by adding some example configuration directives to the default
2046 configuration file. A little bit of work is required to uncomment the
2047 directives and define how usernames and passwords are checked, but
2048 there is now a framework to start from.
2050 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2051 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2052 without this. I don't know how relevant this is to other LDAP libraries.
2054 PH/02 Add the verb name to the "unknown ACL verb" error.
2056 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2058 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2060 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2061 directory not expanded when it should be if an expanded home directory
2062 was set for the address (which is overridden by the transport).
2064 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2067 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2068 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2069 because it is too late at that time, and has no effect.
2071 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2072 security issue with \' (bugzilla #107). I could not use the
2073 PQescapeStringConn() function, because it needs a PGconn value as one of
2076 PH/08 When testing addresses using -bt, indicate those final addresses that
2077 are duplicates that would not cause an additional delivery. At least one
2078 person was confused, thinking that -bt output corresponded to deliveries.
2079 (Suppressing duplicates isn't a good idea as you lose the information
2080 about possibly different redirections that led to the duplicates.)
2082 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2083 systems where poll() doesn't work, in particular OS X.
2085 PH/10 Added more information to debugging output for retry time not reached.
2087 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2088 operations in malware.c.
2090 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2093 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2094 syslog_duplication set false, log lines that would normally be written
2095 both the the main log and to the reject log were not written to syslog at
2098 PH/14 In the default configuration, change the use of "message" in ACL warn
2099 statements to "add_header".
2101 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2102 not followed by a command (e.g. "seen endif").
2104 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2105 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2108 PH/17 Added extra conditions to the default value of delay_warning_condition
2112 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2113 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2114 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2117 The Auto-Submitted: and various List- headers are standardised, whereas I
2118 don't think Precedence: ever was.
2120 PH/18 Refactored debugging code in route_finduser() to show more information,
2121 in particular, the error code if getpwnam() issues one.
2123 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2124 This is apparently needed in addition to the PH/07 change above to avoid
2125 any possible encoding problems.
2127 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2128 but not after initializing Perl.
2130 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2131 output them only if debugging. By default they are written stderr,
2132 apparently, which is not desirable.
2134 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2137 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2140 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2142 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2143 authenticated or an ident call has been made. Suppress the default
2144 values for $authenticated_id and $authenticated_sender (but permit -oMai
2145 and -oMas) when testing with -bh.
2147 PH/24 Re-jigged the order of the tests in the default configuration so that the
2148 tests for valid domains and recipients precede the DNS black list and CSA
2149 tests, on the grounds that those ones are more expensive.
2151 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2152 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2153 This bug exists in every version of Exim that I still have, right back to
2156 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2157 However, an attempt to turn on case-sensitivity in a regex key by
2158 including (?-i) didn't work because the subject string was already
2159 lowercased, and the effects were non-intuitive. It turns out that a
2160 one-line patch can be used to allow (?-i) to work as expected.
2166 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2167 other effects) broke the use of negated acl sub-conditions.
2169 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2172 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2173 "Deny" causes Exim to reject the incoming connection with a 554 error.
2174 Unfortunately, if there is a major crisis, such as a disk failure,
2175 tcp-wrappers gives "deny", whereas what one would like would be some
2176 kind of temporary error. A kludge has been added to help with this.
2177 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2178 554 error is used if errno is still zero or contains ENOENT (which occurs
2179 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2182 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2184 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2185 errors. Otherwise a message that provokes a temporary error (when other
2186 messages do not) can cause a whole host to time out.
2188 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2189 addresses were routed directly to files or pipes from a redirect router.
2190 File deliveries just didn't batch; pipe deliveries might have suffered
2193 PH/06 A failure to get a lock for a hints database would erroneously always say
2194 "Failed to get write lock", even when it was really a read lock.
2196 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2197 of 0600. This has been changed to use the value of the lockfile_mode
2198 option (which defaults to 0600).
2200 PH/08 Applied small patch from the Sieve maintainer.
2202 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2203 folder from quota calculations, a direct delivery into this folder messed
2204 up the contents of the maildirsize file. This was because the regex was
2205 used only to exclude .Trash (or whatever) when the size of the mailbox
2206 was calculated. There was no check that a delivery was happening into an
2207 excluded directory. This bug has been fixed by ignoring all quota
2208 processing for deliveries into excluded directories.
2210 PH/10 Added the maildirfolder_create_regex option to appendfile.
2216 PH/01 The code for finding all the local interface addresses on a FreeBSD
2217 system running IPv6 was broken. This may well have applied to all BSD
2218 systems, as well as to others that have similar system calls. The broken
2219 code found IPv4 interfaces correctly, but gave incorrect values for the
2220 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2221 that it would not match correctly against @[] and not recognize the IPv6
2224 PH/02 The ipliteral router was not recognizing addresses of the form user@
2225 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2227 PH/03 Added disable_ipv6.
2229 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2230 decoded form, because it is most often used to construct To: headers
2231 lines in autoreplies, and the decoded form may well be syntactically
2232 invalid. However, $reply_address has leading white space removed, and all
2233 newlines turned into spaces so that the autoreply transport does not
2236 PH/05 If group was specified without a user on a router, and no group or user
2237 was specified on a transport, the group from the router was ignored.
2239 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2240 for visible compile-time settings that can be used to change these
2241 numbers, for those that want even more. Backwards compatibility with old
2242 spool files has been maintained. However, going back to a previous Exim
2243 release will lost any variables that are in spool files.
2245 PH/07 Two small changes when running in the test harness: increase delay when
2246 passing a TCP/IP connection to a new process, in case the original
2247 process has to generate a bounce, and remove special handling of
2248 127.0.0.2 (sic), which is no longer necessary.
2250 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2251 be the same on different OS.
2253 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2256 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2257 whether --show-vars was specified or not
2259 JJ/02 exipick: Added support for new ACL variable spool format introduced
2262 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2263 syntactically invalid From: or Reply-to: line, and a filter used this to
2264 generate an autoreply, and therefore failed to obtain an address for the
2265 autoreply, Exim could try to deliver to a non-existent relative file
2266 name, causing unrelated and misleading errors. What now happens is that
2267 it logs this as a hard delivery error, but does not attempt to create a
2270 PH/11 The exinext utility has a -C option for testing purposes, but although
2271 the given file was scanned by exinext itself; it wasn't being passed on
2272 when Exim was called.
2274 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2275 an end-of-file indication when reading a command response.
2277 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2278 compiled. In many other places in Exim, IPv6 addresses are always
2279 recognized, so I have changed this. It also means that IPv4 domain
2280 literals of the form [IPV4:n.n.n.n] are now always recognized.
2282 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2283 used if the router is not running as root, for example, when verifying at
2284 ACL time, or when using -bh. The debugging output from this situation was
2285 non-existent - all you got was a failure to exec. I have made two
2288 (a) Failures to set uid/gid, the current directory, or a process leader
2289 in a subprocess such as that created by queryprogram now generate
2290 suitable debugging ouput when -d is set.
2292 (b) The queryprogram router detects when it is not running as root,
2293 outputs suitable debugging information if -d is set, and then runs
2294 the subprocess without attempting to change uid/gid.
2296 PH/15 Minor change to Makefile for building test_host (undocumented testing
2299 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2300 additional section of a DNS packet that returns MX or SRV records.
2301 Instead, it always explicitly searches for A/AAAA records. This avoids
2302 major problems that occur when a DNS server includes only records of one
2303 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2304 fixed another bug: if SRV records were looked up and the corresponding
2305 address records were *not* found in the additional section, the port
2306 values from the SRV records were lost.
2308 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2309 using the correct key (the original address) when searching the retry
2310 rules in order to find which one to use for generating the retry hint.
2312 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2313 adding the default one. Similarly, if it contains a Reply-To: header, the
2314 errors_reply_to option, if set, is not used.
2316 PH/19 When calculating a retry time, Exim used to measure the "time since
2317 failure" by looking at the "first failed" field in the retry record. Now
2318 it does not use this if it is later than than the arrival time of the
2319 message. Instead it uses the arrival time. This makes for better
2320 behaviour in cases where some deliveries succeed, thus re-setting the
2321 "first failed" field. An example is a quota failure for a huge message
2322 when small messages continue to be delivered. Without this change, the
2323 "time since failure" will always be short, possible causing more frequent
2324 delivery attempts for the huge message than are intended.
2325 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2327 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2328 $1, $2, $3) because the numerical variables can be reset during some
2329 expansion items (e.g. "match"), thereby losing the authentication data.
2331 PH/21 Make -bV show the size of off_t variables so that the test suite can
2332 decide whether to run tests for quotas > 2G.
2334 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2335 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2336 filecount value is greater than 2G or if a quota value is greater than 2G
2337 on a system where the size of off_t is not greater than 4, a panic error
2340 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2341 never match. The debug and -bh output now contains an explicit error
2342 message indicating a malformed IPv4 address or mask.
2344 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2345 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2346 PH/23 above applies.
2348 PH/25 Do not write to syslog when running in the test harness. The only
2349 occasion when this arises is a failure to open the main or panic logs
2350 (for which there is an explicit test).
2352 PH/26 Added the /no_tell option to "control=freeze".
2354 PH/27 If a host name lookup failed very early in a connection, for example, if
2355 the IP address matched host_lookup and the reverse lookup yielded a name
2356 that did not have a forward lookup, an error message of the form "no IP
2357 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2358 could be logged. Now it outputs the IP address instead of "NULL".
2360 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2361 allows the sender and the authenticated sender to be set when
2362 submitting a message from within Exim. Since child_open_exim() is
2363 documented for local_scan(), the new function should be too.
2365 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2366 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2367 results in an empty string is now treated as unset.
2369 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2371 PH/31 Added sender_verify_fail logging option.
2373 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2374 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2375 tidied the source and removed it altogether.
2377 PH/33 When a queue run was abandoned because the load average was too high, a
2378 log line was always written; now it is written only if the queue_run log
2379 selector is set. In addition, the log line for abandonment now contains
2380 information about the queue run such as the pid. This is always present
2381 in "start" and "stop" lines but was omitted from the "abandon" line.
2383 PH/34 Omit spaces between a header name and the colon in the error message that
2384 is given when verify = headers_syntax fails (if there are lots of them,
2385 the message gets confusing).
2387 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2388 names, as there are now some PTR records that contain slashes. This check
2389 is only to protect against broken name servers that fall over on strange
2390 characters, so the fact that it applies to all lookups doesn't matter.
2392 PH/36 Now that the new test suite is complete, we can remove some of the
2393 special code in Exim that was needed for the old test suite. For example,
2394 sorting DNS records because real resolvers return them in an arbitrary
2395 order. The new test suite's fake resolver always returns records in the
2398 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2399 bounces) except when queue_only is set, to avoid logging races between
2400 the different processes.
2402 PH/38 Panic-die if .include specifies a non-absolute path.
2404 PH/39 A tweak to the "H" retry rule from its user.
2406 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2407 a label. They prevented compilation on older perls.
2409 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2410 a warning to be raised on newish perls.
2412 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2413 on queue. Changes to match documented behaviour of showing count of
2414 messages matching specified criteria.
2416 PH/40 Changed the default ident timeout from 30s to 5s.
2418 PH/41 Added support for the use of login_cap features, on those BSD systems
2419 that have them, for controlling the resources used by pipe deliveries.
2421 PH/42 The content-scanning code uses fopen() to create files in which to put
2422 message data. Previously it was not paying any attention to the mode of
2423 the files. Exim runs with umask(0) because the rest of the code creates
2424 files with open(), and sets the required mode explicitly. Thus, these
2425 files were ending up world-writeable. This was not a big issue, because,
2426 being within the spool directory, they were not world-accessible. I have
2427 created a function called modefopen, which takes an additional mode
2428 argument. It sets umask(777), creates the file, chmods it to the required
2429 mode, then resets the umask. All the relevant calls to fopen() in the
2430 content scanning code have been changed to use this function.
2432 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2433 to 24 hours. This avoids potential overflow problems when processing G
2434 and H retry rules. I suspect nobody ever tinkers with this value.
2436 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2438 PH/45 When the plaintext authenticator is running as a client, the server's
2439 challenges are checked to ensure they are valid base64 strings. By
2440 default, the authentication attempt is cancelled if an invalid string is
2441 received. Setting client_ignore_invalid_base64 true ignores these errors.
2442 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2443 they are received. Thus, the responses can be made to depend on the
2444 challenges. If an invalid string is ignored, an empty string is placed in
2447 PH/46 Messages that are created by the autoreply transport now contains a
2448 References: header, in accordance with RFCs 2822 and 3834.
2450 PH/47 Added authenticated_sender_force to the smtp transport.
2452 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2454 PH/49 Installed latest patch from the Sieve maintainer.
2456 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2457 was also set, the appendfile transport was unnecessarily scanning a
2458 directory of message files (e.g. for maildir delivery) to find the count
2459 of files (along with the size), even though it did not need this
2460 information. It now does the scan only if it needs to find either the
2461 size of the count of files.
2463 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2465 PH/52 Two bugs concerned with error handling when the smtp transport is
2468 (i) Exim was not creating retry information for temporary errors given
2469 for individual recipients after the DATA command when the smtp transport
2470 was used in LMTP mode. This meant that they could be retried too
2471 frequently, and not timed out correctly.
2473 (ii) Exim was setting the flag that allows error details to be returned
2474 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2475 recipients that were returned after the DATA command.
2477 PH/53 This is related to PH/52, but is more general: for any failing address,
2478 when detailed error information was permitted to be returned to the
2479 sender, but the error was temporary, then after the final timeout, only
2480 "retry timeout exceeded" was returned. Now it returns the full error as
2481 well as "retry timeout exceeded".
2483 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2484 do this, and (what is worse) MTAs that accept it.
2486 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2487 will now be deprecated.
2489 PH/56 New os.c-cygwin from the Cygwin maintainer.
2491 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2492 formats (previously only available in exim formats via -bpr, -bpru,
2493 and -bpra. Now also available in native and exiqgrep formats)
2495 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2496 with very large, slow to parse queues
2498 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2500 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2502 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2503 responses to authentication challenges, though it was showing the
2504 challenges; (ii) I've removed the CR characters from the debug output for
2507 PH/58 Allow for the insertion of a newline as well as a space when a string
2508 is turned into more than one encoded-word during RFC 2047 encoding. The
2509 Sieve code now uses this.
2511 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2512 data_4xx, lost_connection, tls_required.
2514 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2515 message was being sent as an SMTP response.
2517 PH/61 Add -l and -k options to exicyclog.
2519 PH/62 When verifying, if an address was redirected to one new address, so that
2520 verification continued, and the new address failed or deferred after
2521 having set something in $address_data, the value of $address_data was not
2522 passed back to the ACL. This was different to the case when no
2523 redirection occurred. The value is now passed back in both cases.
2525 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2526 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2527 use login_cap.h, so on its own it isn't the distinguishing feature. The
2528 new name refers directly to the setclassresources() function.
2530 PH/65 Added configuration files for NetBSD3.
2532 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2534 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2535 is preferred over IPv4.
2537 PH/68 The bounce_return_message and bounce_return_body options were not being
2538 honoured for bounces generated during the reception of non-SMTP messages.
2539 In particular, this applied to messages rejected by the ACL. This bug has
2540 been fixed. However, if bounce_return_message is true and bounce_return_
2541 body is false, the headers that are returned for a non-SMTP message
2542 include only those that have been read before the error was detected.
2543 (In the case of an ACL rejection, they have all been read.)
2545 PH/69 The HTML version of the specification is now built in a directory called
2546 spec_html instead of spec.html, because the latter looks like a path with
2547 a MIME-type, and this confuses some software.
2549 PH/70 Catch two compiler warnings in sieve.c.
2551 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2552 function verify_get_ident() calls ip_connect() to connect a socket, but
2553 if the "connect()" function timed out, ip_connect() used to close the
2554 socket. However, verify_get_ident() also closes the socket later, and in
2555 between Exim writes to the log, which may get opened at this point. When
2556 the socket was closed in ip_connect(), the log could get the same file
2557 descriptor number as the socket. This naturally causes chaos. The fix is
2558 not to close the socket in ip_connect(); the socket should be closed by
2559 the function that creates it. There was only one place in the code where
2560 this was missing, in the iplookup router, which I don't think anybody now
2561 uses, but I've fixed it anyway.
2563 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2564 well as to direct DNS lookups. Otherwise the handling of names in host
2565 lists is inconsistent and therefore confusing.
2571 PH/01 Two changes to the default runtime configuration:
2573 (1) Move the checks for relay_from_hosts and authenticated clients from
2574 after to before the (commented out) DNS black list checks.
2576 (2) Add control=submission to the relay_from_hosts and authenticated
2577 clients checks, on the grounds that messages accepted by these
2578 statements are most likely to be submissions.
2580 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2582 (1) Generate an error if the third argument for the ${prvs expansion is
2585 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2588 (3) Reset the variables that are obtained from the first argument of
2589 ${prvscheck and used in the second argument before leaving the code,
2590 because their memory is reclaimed, so using them afterwards may do
2593 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2594 one (it's much easier than Tom thought :-).
2596 (5) Because of (4), we can now allow for the use of $prvscheck_result
2597 inside the third argument.
2599 PH/03 For some reason, the default setting of PATH when running a command from
2600 a pipe transport was just "/usr/bin". I have changed it to
2603 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2604 anything to be listed in the output from -bV.
2606 PH/05 When a filter generated an autoreply, the entire To: header line was
2607 quoted in the delivery log line, like this:
2609 => >A.N.Other <ano@some.domain> <original@ddress> ...
2611 This has been changed so that it extracts the operative address. There
2612 may be more than one such address. If so, they are comma-separated, like
2615 => >ano@some.domain,ona@other.domain <original@ddress> ...
2617 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2618 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2619 not being looked up in the rDNS to get a host name, Exim was showing the
2620 IP address twice in Received: lines, even though the IP addresses were
2621 identical. For example:
2623 Received: from [1.2.3.4] (helo=[1.2.3.4])
2625 However, if the real host name was known, it was omitting the HELO data
2626 if it matched the actual IP address. This has been tidied up so that it
2627 doesn't show the same IP address twice.
2629 PH/07 When both +timestamp and +memory debugging was on, the value given by
2630 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2631 called by the debug printing, thereby overwriting the timestamp buffer.
2632 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2634 PH/08 When the original message was included in an autoreply transport, it
2635 always said "this is a copy of the message, including all the headers",
2636 even if body_only or headers_only was set. It now gives an appropriate
2639 PH/09 Applied a patch from the Sieve maintainer which:
2641 o fixes some comments
2642 o adds the (disabled) notify extension core
2643 o adds some debug output for the result of if/elsif tests
2644 o points to the current vacation draft in the documentation
2645 and documents the missing references header update
2649 o fixes a bug in processing the envelope test (when testing
2650 multiple envelope elements, the last element determinted the
2653 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2654 Electronic Mail") by including:
2656 Auto-submitted: auto-generated
2658 in the messages that it generates (bounce messages and others, such as
2659 warnings). In the case of bounce messages for non-SMTP mesages, there was
2660 also a typo: it was using "Auto_submitted" (underscore instead of
2661 hyphen). Since every message generated by Exim is necessarily in response
2662 to another message, thes have all been changed to:
2664 Auto-Submitted: auto-replied
2666 in accordance with these statements in the RFC:
2668 The auto-replied keyword:
2670 - SHOULD be used on messages sent in direct response to another
2671 message by an automatic process,
2673 - MUST NOT be used on manually-generated messages,
2675 - MAY be used on Delivery Status Notifications (DSNs) and Message
2676 Disposition Notifications (MDNs),
2678 - MUST NOT be used on messages generated by automatic or periodic
2679 processes, except for messages which are automatic responses to
2682 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2683 to the default Received: header definition.
2685 PH/12 Added log selector acl_warn_skipped (default on).
2687 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2688 variables because (a) they are in the wrong storage pool and (b) even if
2689 they were copied, it wouldn't work properly because of the caching.
2691 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2692 checking when decoding. Apparently there are clients that generate
2693 overlong encoded strings. Why am I not surprised?
2695 PH/15 If the first argument of "${if match_address" was not empty, but did not
2696 contain an "@" character, Exim crashed. Now it writes a panic log message
2697 and treats the condition as false.
2699 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2701 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2702 extension "envelope-auth". The code is finished and in agreement with
2703 other implementations, but there is no documentation so far and in fact,
2704 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2705 not changing the active code.
2707 Print executed "if" and "elsif" statements when debugging is used. This
2708 helps a great deal to understand what a filter does.
2710 Document more things not specified clearly in RFC3028. I had all this
2711 sorted out, when out of a sudden new issues came to my mind. Oops."
2713 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2716 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2717 canonical form (as documented). However, after a host name lookup from
2718 the IP address, check_host() was doing a simple string comparison with
2719 addresses acquired from the DNS when checking that the found name did
2720 have the original IP as one of its addresses. Since any found IPv6
2721 addresses are likely to be in abbreviated form, the comparison could
2722 fail. Luckily, there already exists a function for doing the comparison
2723 by converting both addresses to binary, so now that is used instead of
2724 the text comparison.
2726 PH/20 There was another similar case to PH/19, when a complete host name was
2727 given in a host list; looking up its IP address could give an abbreviated
2728 form, whereas the current host's name might or might not be abbreviated.
2729 The same fix has been applied.
2735 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2736 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2739 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2740 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2742 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2744 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2745 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2746 Also, for "Auto-Submitted", treat anything other than "no" as
2747 non-personal, in accordance with RFC 3834. (Previously it treated
2748 anything starting "auto-" as non-personal.)
2750 TF/01 The control=submission/name=... option had a problem with syntax
2751 errors if the name included a slash character. The /name= option
2752 now slurps the rest of the string, so it can include any characters
2753 but it must come last in the list of options (after /sender_retain
2756 PH/05 Some modifications to the interface to the fake nameserver for the new
2764 TK/01 Added the "success_on_redirect" address verification option. See
2765 NewStuff for rationale and an example.
2767 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2769 PH/02 Patch to exigrep to allow it to work on syslog lines.
2771 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2772 fread() to skip over the body file's header line, because in Cygwin the
2773 header line is locked and is inaccessible.
2775 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2776 co-exist for some time) to make it clear that it is the Exim ID that is
2777 referenced, not the Message-ID: header line.
2779 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2780 string_format() function, because snprintf() does not exist on all
2783 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2784 ${stat: expansion item.
2786 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2787 protocol synchronization error", to keep the pedants happy.
2789 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2790 well as for IRIX systems, when gcc is being used. See the host.c source
2793 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2795 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2798 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2799 transport and to the smtp transport in LMTP mode.
2801 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2803 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2804 run a filter in a subprocess. This could lead to confusion in subsequent
2805 lookups in the parent process. There should also be a search_tidyup() at
2806 the end of the subprocess.
2808 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2809 only if the host matched helo_try_verify_hosts, which caused the
2810 verification to occur when the EHLO/HELO command was issued. The ACL just
2811 tested the remembered result. Now, if a previous verification attempt has
2812 not happened, "verify = helo" does it there and then.
2814 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2816 TK/03 Fix log output including CR from clamd.
2818 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2819 exist provoked a memory error which could cause a segfault.
2821 PH/15 Installed PCRE 6.2
2823 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2825 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2826 of the problem. Specifically, suggested +O2 rather than +O1 for the
2829 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2831 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2832 record, the port was not correctly logged when the outgoing_port log
2833 selector was set (it logged the transort's default port).
2835 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2836 fallback_hosts, and "hosts" in the smtp transport.
2838 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2839 host errors such as "Connection refused".
2841 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2842 authentication with radiusclient 0.4.9:
2844 - Error returned from rc_read_config was caught wrongly
2845 - Username/password not passed on to radius server due to wrong length.
2847 The presumption is that some radiusclient API changes for 4.51/PH/17
2848 were not taken care of correctly. The code is still untested by me (my
2849 Linux distribution still has 0.3.2 of radiusclient), but it was
2850 contributed by a Radius user.
2852 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2853 expanding the "port" option of the smtp transport.
2855 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2856 while reading a MIME header. Thanks to Tom Hughes for a patch.
2858 PH/24 Include config.h inside local_scan.h so that configuration settings are
2861 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2862 that in an ACL for RCPT (for example), you can examine exactly what was
2865 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2866 commands, but it was not correctly comparing the address with the actual
2867 client host address. Thus, it would show the EHLO address in Received:
2868 header lines when this was not necessary.
2870 PH/27 Added the % operator to ${eval:}.
2872 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2873 it should be ignoring failures (because with -C, for example, it has lost
2874 privilege). It wasn't ignoring creation failures other than "already
2877 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2880 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2881 because an input line was too long, either on its own, or by virtue of
2882 too many continuations, the temporary file was not being removed, and the
2883 return code was incorrect.
2885 PH/31 Missing "BOOL" in function definition in filtertest.c.
2887 PH/32 Applied Sieve patches from the maintainer.
2889 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2891 PH/33 Added "verify = not_blind".
2893 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2894 Local/Makefile (with some defaults set). These are used in built scripts
2895 such as exicyclog, but they have never been used in the exim_install
2896 script (though there are many overriding facilities there). I have
2897 arranged that the exim_install script now takes note of these two
2900 PH/35 Installed configuration files for Dragonfly.
2902 PH/36 When a locally submitted message by a trusted user did not contain a
2903 From: header, and the sender address was obtained from -f or from an SMTP
2904 MAIL command, and the trusted user did not use -F to supply a sender
2905 name, $originator_name was incorrectly used when constructing a From:
2906 header. Furthermore, $originator_name was used for submission mode
2907 messages from external hosts without From: headers in a similar way,
2908 which is clearly wrong.
2910 PH/37 Added control=suppress_local_fixups.
2912 PH/38 When log_selector = +received_sender was set, and the addition of the
2913 sender made the log line's construction buffer exactly full, or one byte
2914 less than full, an overflow happened when the terminating "\n" was
2917 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2918 when the result of a list match is failure because a DNS lookup failed.
2920 PH/40 RM_COMMAND is now used in the building process.
2922 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2923 the "build-* directories that it finds.
2925 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2926 address in a domain literal was a prefix of an interface address.
2928 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2929 when verifying a sender address, unless rewrite_headers is false.
2931 PH/44 Wrote a long comment about why errors_to addresses are verified as
2932 recipients, not senders.
2934 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2935 the ratelimit ACL was added.
2937 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2939 PH/46 Added extra information about PostgreSQL errors to the error string.
2941 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2942 avoiding the need to install special zones in a real server. This is
2943 backwards compatible; if it can't find the fake resolver, it drops back.
2944 Thus, both old and new test suites can be run.
2946 TF/02 Added util/ratelimit.pl
2948 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2949 clock is set back in time.
2951 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2952 Candler <B.Candler@pobox.com>.
2954 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2955 OK for addresses that are the result of redirections.
2957 PH/48 A number of further additions for the benefit of the new test suite,
2958 including a fake gethostbyname() that interfaces to the fake DNS resolver
2961 TF/06 The fix for widen_domains has also been applied to qualify_single and
2962 search_parents which are the other dnslookup options that can cause
2965 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2968 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2970 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2971 service" if it encountered an MX record with an empty target hostname.
2972 The message is now "an MX or SRV record indicated no SMTP service".
2974 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2975 if the DNS of the sending site is misconfigured. This is quite a
2976 common situation. This change restores the behaviour of treating a
2977 helo verification defer as a failure.
2979 PH/51 If self=fail was set on a router, the bounce message did not include the
2980 actual error message.
2986 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2988 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2989 command itself ended in error, the underlying message about the transport
2990 filter timeout was being overwritten with the pipe command error. Now the
2991 underlying error message should be appended to the second error message.
2993 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2995 PH/02 Reduce the amount of output that "make" produces by default. Full output
2996 can still be requested.
2998 PH/03 The warning log line about a condition test deferring for a "warn" verb
2999 was being output only once per connection, rather than after each
3000 occurrence (because it was using the same function as for successful
3001 "warn" verbs). This seems wrong, so I have changed it.
3003 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3004 it should not have, which might have caused a crash in the right
3005 circumstances, but probably never did.
3007 PH/04 Installed a modified version of Tony Finch's patch to make submission
3008 mode fix the return path as well as the Sender: header line, and to
3009 add a /name= option so that you can make the user's friendly name appear
3012 TF/03 Added the control = fakedefer ACL modifier.
3014 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3015 Mark Lowes for thorough testing.
3017 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3019 TK/03 Merged latest SRS patch from Miles Wilton.
3021 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3022 with the definition in sysexits.h (which is #included earlier).
3023 Fortunately, Exim does not actually use EX_OK. The code used to try to
3024 preserve the sysexits.h value, by assumimg that macro definitions were
3025 scanned for macro replacements. I have been disabused of this notion,
3026 so now the code just undefines EX_OK before #including unistd.h.
3028 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3029 in the smtp transport. When a block could not be written in a single
3030 write() function, the timeout was being re-applied to each part-write.
3031 This seems wrong - if the receiver was accepting one byte at a time it
3032 would take for ever. The timeout is now adjusted when this happens. It
3033 doesn't have to be particularly precise.
3035 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3036 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3038 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3039 without a domain if the check to <postmaster@domain> fails.
3041 SC/01 Eximstats: added -xls and the ability to specify output files
3042 (patch written by Frank Heydlauf).
3044 SC/02 Eximstats: use FileHandles for outputing results.
3046 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3048 SC/04 Eximstats: fixed display of large numbers with -nvr option
3050 SC/05 Eximstats: fixed merging of reports with empty tables.
3052 SC/06 Eximstats: added the -include_original_destination flag
3054 SC/07 Eximstats: removed tabs and trailing whitespace.
3056 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3058 TK/06 MBOX spool code: Add real "From " MBOX separator line
3059 so the .eml file is really in mbox format (even though
3060 most programs do not really care). Patch from Alex Miller.
3062 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3063 The latter is generated from $received_to and is only set if the
3064 message has one envelope recipient. SA can use these headers,
3065 obviously out-of-the-box. Patch from Alex Miller.
3067 PH/08 The ${def test on a variable was returning false if the variable's
3068 value was "0", contrary to what the specification has always said!
3069 The result should be true unless the variable is empty.
3071 PH/09 The syntax error of a character other than { following "${if
3072 def:variable_name" (after optional whitespace) was not being diagnosed.
3073 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3074 accidental colon was present, for example, could give incorrect results.
3076 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3077 result is used (not including appendfile, where other changes are about
3080 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3081 This involved changing a lot of size variables from int to off_t. It
3082 should work with maildirs and everything.
3084 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3085 spamd dying while we are connected to it.
3087 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3090 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3091 function for BDB 4.3.
3093 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3095 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3096 the sources for PCRE have been re-arranged and more files are now
3099 PH/15 The code I had for printing potentially long long variables in PH/11
3100 above was not the best (it lost precision). The length of off_t variables
3101 is now inspected at build time, and an appropriate printing format (%ld
3102 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3103 to be "long long int" or "long int". This is needed for the internal
3104 formatting function string_vformat().
3106 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3107 the configuration file to be ":syslog", then the script "guesses" where
3108 the logs files are, rather than using the compiled in default. In our
3109 case the guess is not the same as the compiled default, so the script
3110 suddenly stopped working when I started to use syslog. The patch checks
3111 to see if log_file_path is "". If so, it attempts to read it from exim
3112 with no configuration file to get the compiled in version, before it
3113 falls back to the previous guessing code."
3115 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3116 implementing BATV in an Exim configuration. See NewStuff for the gory
3119 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3120 Makefile that are specific to HP-UX.
3122 PH/18 If the "use_postmaster" option was set for a recipient callout together
3123 with the "random" option, the postmaster address was used as the MAIL
3124 FROM address for the random test, but not for the subsequent recipient
3125 test. It is now used for both.
3127 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3128 patch removes a few documentation additions to RFC 3028, because the
3129 latest draft now contains them. It adds the new en;ascii-case comparator
3130 and a new error check for 8bit text in MIME parts. Comparator and
3131 require names are now matched exactly. I enabled the subaddress
3132 extension, but it is not well tested yet (read: it works for me)."
3134 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3135 rework some of the code of TK/09 above to avoid the hardwired use of
3136 "%lld" and "long long". Replaced the call to snprintf() with a call to
3139 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3140 records point to non-existent hosts", "retry timeout exceeded", and
3141 "retry time not reached for any host after a long failure period".
3143 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3144 experimental DomainKeys support:
3146 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3147 (2) On an error such as an illegally used "control", the wrong name for
3148 the control was given.
3150 These problems did NOT occur unless DomainKeys support was compiled.
3152 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3154 PH/24 Added ${if match_ip condition.
3156 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3157 databases so that it will be absolutely obvious if a crash occurs in the
3158 DB library. This is a regular occurrence (often caused by mis-matched
3161 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3162 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3163 picked up on a user's system that detects such things. There doesn't seem
3164 to be a gcc warning option for this - only an attribute that has to be
3165 put on the function's prototype. It seems that in Fedora Core 4 they have
3166 set this on a number of new functions. No doubt there will be more in due
3169 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3170 specify a transport. However, if an address that was verified by such a
3171 router was the subject of a callout, Exim crashed because it tried to
3172 read the rcpt_include_affixes from the non-existent transport. Now it
3173 just assumes that the setting of that option is false. This bug was
3174 introduced by 4.51/PH/31.
3176 PH/28 Changed -d+all to exclude +memory, because that information is very
3177 rarely of interest, but it makes the output a lot bigger. People tend to
3178 do -d+all out of habit.
3180 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3181 code in os-type was giving problems when libc.so lives in lib64, like on
3184 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3185 aren't the modern standard, and it seems that some systems' include files
3186 don't always have them. Exim was already checking for some of the newer
3187 ones like T_AAAA, and defining it itself. I've added checks for all the
3188 record types that Exim uses.
3190 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3191 not automatically generating a new one, as it is supposed to. This
3192 prevented TLS from working. If the file did exist, but contained invalid
3193 data, a new version was generated, as expected. It was only the case of a
3194 non-existent file that was broken.
3196 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3197 with a change in libdomainkeys > 0.64.
3199 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3200 from DNS. If the selector record carries the flag, it now has
3201 precedence over the domain-wide flag.
3203 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3205 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3206 the use of a port name that isn't defined in /etc/services) occurred, the
3207 message was deferred as in a normal delivery, and thus remained on the
3208 spool, instead of being failed because of the mua_wrapper setting. This
3209 is now fixed, and I tidied up some of the mua_wrapper messages at the
3212 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3213 the messages in a hash of arrays rather than using individual hashes.
3214 This is a bit cleaner and results in dramatic memory savings, albeit
3215 at a slight CPU cost.
3217 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3218 as requested by Marc Sherman.
3220 SC/10 Eximstats: added histograms for user specified patterns as requested
3223 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3225 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3226 fopen() in the content-scanning modules that did not already have it.
3232 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3233 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3235 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3237 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3239 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3240 to test Sieve filters that use "vacation".
3242 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3243 that changes the way the GnuTLS parameters are stored in the cache file.
3244 The new format can be generated externally. For backward compatibility,
3245 if the data in the cache doesn't make sense, Exim assumes it has read an
3246 old-format file, and it generates new data and writes a new file. This
3247 means that you can't go back to an older release without removing the
3250 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3251 work if there are any delivery delays because "one_time" forces the
3252 parent to be marked "delivered", so its unseen clone is never tried
3253 again. For this reason, Exim now forbids the simultaneous setting of
3256 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3257 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3258 entry if you want to know the details. The fix, however, neglected to
3259 consider the case when local delivery batching is involved. The test for
3260 "previously delivered" was not happening when checking to see if an
3261 address could be batched with a previous (undelivered) one; under
3262 certain circumstances this could lead to multiple deliveries to the same
3265 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3266 in its include files, and this causes problems building Exim.
3268 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3269 header_syntax) but Exim was just ignoring anything given after a slash.
3270 In particular, this caused confusion with an attempt to use "verify =
3271 reverse_host_lookup/defer_ok". An error is now given when options are
3272 supplied for verify items that do not have them. (Maybe reverse_host_
3273 lookup should have a defer_ok option, but that's a different point.)
3275 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3276 defined by RFC 821) to 2048, because there were problems with some AUTH
3277 commands, and RFC 1869 says the size should be increased for extended
3278 SMTP commands that take arguments.
3280 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3283 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3284 "unknown" error; now it says that the functionality isn't in the binary.
3286 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3287 an address' error message when a string expansion fails (syntax or
3288 whatever). Otherwise the password may appear in the log. Following change
3289 PH/42 below, there is no longer a chance of it appearing in a bounce
3292 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3294 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3295 of its addresses was ever tried. (Bugzilla bug #2.)
3297 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3298 the result incorrectly in the debug output. (It correctly added a newline
3299 to what was transported.)
3301 TF/01 Added $received_time.
3303 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3304 commented out examples of how to interface to a virus scanner and to
3305 SpamAssassin. Also added commented examples of av_scanner and
3306 spamd_address settings.
3308 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3309 and controls are allowed in which ACLs. There were a couple of minor
3310 errors. Some of the entries in the conditions table (which is a table of
3311 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3312 negation of where the condition IS allowed.
3314 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3316 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3317 header file does not have a version number, so I've had to invent a new
3318 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3319 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3320 radiusclient), but it was contributed by a Radius user.
3322 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3323 files or directories, for OpenSSL.
3325 PH/19 When an Exim process that is running as root has to create an Exim log
3326 file, it does so in a subprocess that runs as exim:exim so as to get the
3327 ownership right at creation (otherwise, other Exim processes might see
3328 the file with the wrong ownership). There was no test for failure of this
3329 fork() call, which would lead to the process getting stuck as it waited
3330 for a non-existent subprocess. Forks do occasionally fail when resources
3331 run out. I reviewed all the other calls to fork(); they all seem to check
3334 PH/20 When checking for unexpected SMTP input at connect time (before writing
3335 the banner), Exim was not dealing correctly with a non-positive return
3336 from the read() function. If the client had disconnected by this time,
3337 the result was a log entry for a synchronization error with an empty
3338 string after "input=" when read() returned zero. If read() returned -1
3339 (an event I could not check), uninitialized data bytes were printed.
3340 There were reports of junk text (parts of files, etc) appearing after
3343 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3345 PH/22 Added support for macro redefinition, and (re)definition in between
3346 driver and ACL definitions.
3348 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3349 forgetting to use the resulting value; it was using the unexpanded value.
3351 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3352 hadn't been configured. The fix is from Juergen Kreileder, who
3353 understands it better than I do:
3355 "Here's what I see happening with three configured cyrus_sasl
3356 authenticators configured (plain, login, cram-md5):
3358 On startup auth_cyrus_sasl_init() gets called for each of these.
3359 This means three calls to sasl_listmech() without a specified mech_list.
3360 => SASL tests which mechs of all available mechs actually work
3361 => three warnings about OTP not working
3362 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3364 With the patch, sasl_listmech() also gets called three times. But now
3365 SASL's mech_list option is set to the server_mech specified in the the
3366 authenticator. Or in other words, the answer from sasl_listmech()
3367 gets limited to just the mech you're testing for (which is different
3369 => the return list contains just 'plain' or 'login', 'cram-md5' or
3370 nothing depending on the value of ob->server_mech.
3372 I've just tested the patch: Authentication still works fine,
3373 unavailable mechs specified in the exim configuration are still
3374 caught, and the auth.log warnings about OTP are gone."
3376 PH/25 When debugging is enabled, the contents of the command line are added
3377 to the debugging output, even when log_selector=+arguments is not
3380 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3381 answer is "GNU", and only if the return is "GNU/something" is the answer
3384 PH/27 $acl_verify_message is now set immediately after the failure of a
3385 verification in an ACL, and so is available in subsequent modifiers. In
3386 particular, the message can be preserved by coding like this:
3388 warn !verify = sender
3389 set acl_m0 = $acl_verify_message
3391 Previously, $acl_verify_message was set only while expanding "message"
3392 and "log_message" when a very denied access.
3394 PH/28 Modified OS/os.c-Linux with
3396 -#ifndef OS_LOAD_AVERAGE
3397 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3399 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3400 nomenclature these days.)
3402 PH/29 Installed patch from the Sieve maintainer that adds the options
3403 sieve_useraddress and sieve_subaddress to the redirect router.
3405 PH/30 In these circumstances:
3406 . Two addresses routed to the same list of hosts;
3407 . First host does not offer TLS;
3408 . First host accepts first address;
3409 . First host gives temporary error to second address;
3410 . Second host offers TLS and a TLS session is established;
3411 . Second host accepts second address.
3412 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3413 and peerdn, if requested) that were in fact used only for the second
3416 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3417 attention to any local part prefix or suffix that was matched by the
3418 router that accepted the address. It now behaves in the same way as it
3419 does for delivery: the affixes are removed from the local part unless
3420 rcpt_include_affixes is set on the transport.
3422 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3423 timeout during the DATA phase of an incoming message.
3425 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3426 applied a patch sanctioned by the Sieve maintainer.
3428 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3429 the uid or gid is negative. A case of a negative gid caused this to be
3430 noticed. The fix allows for either to be negative.
3432 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3433 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3436 PH/36 The change PH/12 above was broken. Fixed it.
3438 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3439 the grounds that routing the same address twice would always produce the
3440 same answer. This might have been true once, but it is certainly no
3441 longer true now. Routing a child address may depend on the previous
3442 routing that produced that child. Some complicated redirection strategies
3443 went wrong when messages had multiple recipients, and made Exim's
3444 behaviour dependent on the order in which the addresses were given.
3446 I have moved the duplicate checking until after the routing is complete.
3447 Exim scans the addresses that are assigned to local and remote
3448 transports, and removes any duplicates. This means that more work will be
3449 done, as duplicates will always all be routed, but duplicates are
3450 presumably rare, so I don't expect this is of any significance.
3452 For deliveries to pipes, files, and autoreplies, the duplicate checking
3453 still happens during the routing process, since they are not going to be
3456 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3457 It corrects a timeout issue with spamd. This is Ian's comment: "The
3458 background is that sometimes spamd either never reads data from a
3459 connection it has accepted, or it never writes response data. The exiscan
3460 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3461 blindly assumes that writes won't block so it may never time out."
3463 PH/39 Allow G after quota size as well as K and M.
3465 PH/40 The value set for $authenticated_id in an authenticator may not contain
3466 binary zeroes or newlines because the value is written to log lines and
3467 to spool files. There was no check on this. Now the value is run through
3468 the string_printing() function so that such characters are converted to
3469 printable escape sequences.
3471 PH/41 $message_linecount is a new variable that contains the total number of
3472 lines in the message. Compare $body_linecount, which is the count for the
3475 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3476 bounce and delay warning messages, except in certain special cases, which
3479 (a) An SMTP error message from a remote host;
3480 (b) A message specified in a :fail: redirection;
3481 (c) A message specified in a "fail" command in a system filter;
3482 (d) A message specified in a FAIL return from the queryprogram router;
3483 (e) A message specified by the cannot_route_message router option.
3485 In these cases only, Exim does include the error details in bounce and
3486 warning messages. There are also a few cases where bland messages such
3487 as "unrouteable address" or "local delivery error" are given.
3489 PH/43 $value is now also set for the "else" part of a ${run expansion.
3491 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3492 being worked on, but at least Exim now implements the latest version to
3495 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3496 process to complete was treated as a delivery failure, a timeout while
3497 writing the message to the pipe was logged, but erroneously treated as a
3498 successful delivery. Such timeouts include transport filter timeouts. For
3499 consistency with the overall process timeout, these timeouts are now
3500 treated as errors, giving rise to delivery failures by default. However,
3501 there is now a new Boolean option for the pipe transport called
3502 timeout_defer, which, if set TRUE, converts the failures into defers for
3503 both kinds of timeout. A transport filter timeout is now identified in
3506 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3507 systems where "make" and "gmake" are different, calling "gmake" at top
3508 level broke things. I've arranged for the value of $(MAKE) to be passed
3509 from the Makefile to this script so that it can call the same version of
3513 A note about Exim versions 4.44 and 4.50
3514 ----------------------------------------
3516 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3517 changes of various kinds. As a consequence, a big documentation update was
3518 needed. This delayed the release for rather longer than seemed good, especially
3519 in the light of a couple of (minor) security issues. Therefore, the changes
3520 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3521 release. So 4.44 and 4.50 are in effect two different branches that both start
3524 I have left the 4.50 change log unchanged; it contains all the changes since
3525 4.43. The change log for 4.44 is below; many of its items are identical to
3526 those for 4.50. This seems to be the most sensible way to preserve the
3527 historical information.
3533 1. Minor wording change to the doc/README.SIEVE file.
3535 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3536 computation of the current number of files was incorrect.
3538 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3539 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3542 4. Give more explanation in the error message when the command for a transport
3543 filter fails to execute.
3545 5. There are several places where Exim runs a non-Exim command in a
3546 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3547 was being done only for the command run by the queryprogram router. It is
3548 now done for all such subprocesses. The other cases are: ${run, transport
3549 filters, and the commands run by the lmtp and pipe transports.
3551 6. Added CONFIGURE_GROUP build-time option.
3553 7. Some older OS have a limit of 256 on the maximum number of file
3554 descriptors. Exim was using setrlimit() to set 1000 as a large value
3555 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3556 systems. I've change it so that if it can't get 1000, it tries for 256.
3558 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3559 was an oversight, and furthermore, ever since the addition of extra
3560 controls (e.g. 4.43/32), the checks on when to allow different forms of
3561 "control" were broken. There should now be diagnostics for all cases when a
3562 control that does not make sense is encountered.
3564 9. Added the /retain_sender option to "control=submission".
3566 10. $recipients is now available in the predata ACL (oversight).
3568 11. Tidy the search cache before the fork to do a delivery from a message
3569 received from the command line. Otherwise the child will trigger a lookup
3570 failure and thereby defer the delivery if it tries to use (for example) a
3571 cached ldap connection that the parent has called unbind on.
3573 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3574 of $address_data from the recipient verification was clobbered by the
3575 sender verification.
3577 13. The value of address_data from a sender verification is now available in
3578 $sender_address_data in subsequent conditions in the ACL statement.
3580 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3582 15. Added a new option "connect=<time>" to callout options, to set a different
3585 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3586 was its contents. (It was OK if the option was not defined at all.)
3588 17. A "Completed" log line is now written for messages that are removed from
3589 the spool by the -Mrm option.
3591 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3592 information about exactly what failed.
3594 19. Added -dd to debug only the daemon process.
3596 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3597 handles timeouts, both on the server side and network timeouts. Renamed the
3598 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3600 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3601 It is now set to "smtps".
3603 22. $host_address is now set to the target address during the checking of
3604 ignore_target_hosts.
3606 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3607 being passed; this would have caused $sender_host_name to have been used if
3608 matching the list had actually called for a host name (not very likely,
3609 since this list is usually IP addresses). A host name is now passed as
3612 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3613 code that specifies a non-restarting handler (typically sigaction() in
3614 modern systems) in an attempt to fix a rare and obscure crash bug.
3616 25. Narrowed the window for a race in the daemon that could cause it to ignore
3617 SIGCHLD signals. This is not a major problem, because they are used only to
3618 wake it up if nothing else does.
3620 26. A malformed maildirsize file could cause Exim to calculate negative values
3621 for the mailbox size or file count. Odd effects could occur as a result.
3622 The maildirsize information is now recalculated if the size or filecount
3625 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3626 support for a long time. Removed HAVE_SYS_VFS_H.
3628 28. Installed the latest version of exipick from John Jetmore.
3630 29. In an address list, if the pattern was not a regular expression, an empty
3631 subject address (from a bounce message) matched only if the pattern was an
3632 empty string. Non-empty patterns were not even tested. This was the wrong
3633 because it is perfectly reasonable to use an empty address as part of a
3634 database query. An empty address is now tested by patterns that are
3635 lookups. However, all the other forms of pattern expect the subject to
3636 contain a local part and a domain, and therefore, for them, an empty
3637 address still always fails if the pattern is not itself empty.
3639 30. Exim went into a mad DNS loop when attempting to do a callout where the
3640 host was specified on an smtp transport, and looking it up yielded more
3641 than one IP address.
3643 31. Re-factored the code for checking spool and log partition space into a
3644 function that finds that data and another that does the check. The former
3645 is then used to implement four new variables: $spool_space, $log_space,
3646 $spool_inodes, and $log_inodes.
3648 32. The RFC2047 encoding function was originally intended for short strings
3649 such as real names; it was not keeping to the 75-character limit for
3650 encoded words that the RFC imposes. It now respects the limit, and
3651 generates multiple encoded words if necessary. To be on the safe side, I
3652 have increased the buffer size for the ${rfc2047: expansion operator from
3655 33. It is now permitted to omit both strings after an "if" condition; if the
3656 condition is true, the result is "true". As before, when the second string
3657 is omitted, a false condition yields an empty string. This makes it less
3658 cumbersome to write custom ACL and router conditions.
3660 34. Failure to deliver a bounce message always caused it to be frozen, even if
3661 there was an errors_to setting on the router. The errors_to setting is now
3664 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3665 canonical form (fully expanded) before being placed in
3666 $sender_host_address.
3668 36. The table in the code that translates DNS record types into text (T_A to
3669 "A" for instance) was missing entries for NS and CNAME. It is just possible
3670 that this could have caused confusion if both these types were looked up
3671 for the same domain, because the text type is used as part of Exim's
3672 per-process caching. But the chance of anyone hitting this buglet seems
3675 37. The dnsdb lookup has been extended in a number of ways.
3677 (1) There is a new type, "zns", which walks up the domain tree until it
3678 finds some nameserver records. It should be used with care.
3680 (2) There is a new type, "mxh", which is like "mx" except that it returns
3681 just the host names, not the priorities.
3683 (3) It is now possible to give a list of domains (or IP addresses) to be
3684 looked up. The behaviour when one of the lookups defers can be
3685 controlled by a keyword.
3687 (4) It is now possible to specify the separator character for use when
3688 multiple records are returned.
3690 38. The dnslists ACL condition has been extended: it is now possible to supply
3691 a list of IP addresses and/or domains to be looked up in a particular DNS
3694 39. Added log_selector=+queue_time_overall.
3696 40. When running the queue in the test harness, wait just a tad after forking a
3697 delivery process, to get repeatability of debugging output.
3699 41. Include certificate and key file names in error message when GnuTLS fails
3700 to set them up, because the GnuTLS error message doesn't include the name
3701 of the failing file when there is a problem reading it.
3703 42. Allow both -bf and -bF in the same test run.
3705 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3707 44. The "Exiscan patch" is now merged into the mainline Exim source.
3709 45. Sometimes the final signoff response after QUIT could fail to get
3710 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3711 < 0 before doing a fflush(). This bug looks as though it goes back to the
3712 introduction of TLS in release 3.20, but "sometimes" must have been rare
3713 because the tests only now provoked it.
3715 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3716 (this can affect the format of dates).
3718 47. exim_tidydb, when checking for the continued existence of a message for
3719 which it has found a message-specific retry record, was not finding
3720 messages that were in split spool directories. Consequently, it was
3721 deleting retry records that should have stayed in existence.
3723 48. Steve fixed some bugs in eximstats.
3725 49. The SPA authentication driver was not abandoning authentication and moving
3726 on to the next authenticator when an expansion was forced to fail,
3727 contradicting the general specification for all authenticators. Instead it
3728 was generating a temporary error. It now behaves as specified.
3730 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3731 (the order specifies the preference for clients). The order is now AES256,
3732 AES128, 3DES, ARCFOUR128.
3734 51. Small patch to Sieve code - explicitly set From: when generating an
3737 52. Exim crashed if a remote delivery caused a very long error message to be
3738 recorded - for instance if somebody sent an entire SpamAssassin report back
3739 as a large number of 550 error lines. This bug was coincidentally fixed by
3740 increasing the size of one of Exim's internal buffers (big_buffer) that
3741 happened as part of the Exiscan merge. However, to be on the safe side, I
3742 have made the code more robust (and fixed the comments that describe what
3745 53. Now that there can be additional text after "Completed" in log lines (if
3746 the queue_time_overall log selector is set), a one-byte patch to exigrep
3747 was needed to allow it to recognize "Completed" as not the last thing in
3750 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3751 patch that reportedly fixes this has been added. I am not expert enough to
3752 create a test for it. This is what the patch creator wrote:
3754 "I found a little strange behaviour of ldap code when working with
3755 Windows 2003 AD Domain, where users was placed in more than one
3756 Organization Units. When I tried to give exim partial DN, the exit code
3757 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3758 But simultaneously result of request was absolutely normal ldap result,
3759 so I produce this patch..."
3761 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3762 REFERENCE, so I have modified the code to exclude the patch when that macro
3765 55. Some experimental protocols are using DNS PTR records for new purposes. The
3766 keys for these records are domain names, not reversed IP addresses. The
3767 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3768 leaves it alone. Component reversal etc. now happens only for IP addresses.
3771 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3773 57. Double the size of the debug message buffer (to 2048) so that more of very
3774 long debug lines gets shown.
3776 58. The exicyclog utility now does better if the number of log files to keep
3777 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3779 59. Two changes related to the smtp_active_hostname option:
3781 (1) $smtp_active_hostname is now available as a variable.
3782 (2) The default for smtp_banner uses $smtp_active_hostname instead
3783 of $primary_hostname.
3785 60. The host_aton() function is supposed to be passed a string that is known
3786 to be a valid IP address. However, in the case of IPv6 addresses, it was
3787 not checking this. This is a hostage to fortune. Exim now panics and dies
3788 if the condition is not met. A case was found where this could be provoked
3789 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3790 components; fortuitously, this particular loophole had already been fixed
3791 by change 4.50/55 above.
3793 If there are any other similar loopholes, the new check in host_aton()
3794 itself should stop them being exploited. The report I received stated that
3795 data on the command line could provoke the exploit when Exim was running as
3796 exim, but did not say which command line option was involved. All I could
3797 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3798 running as the user.
3801 61. There was a buffer overflow vulnerability in the SPA authentication code
3802 (which came originally from the Samba project). I have added a test to the
3803 spa_base64_to_bits() function which I hope fixes it.
3806 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3807 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3809 63. The daemon start-up calls getloadavg() while still root for those OS that
3810 need the first call to be done as root, but it missed one case: when
3811 deliver_queue_load_max is set with deliver_drop_privilege. This is
3812 necessary for the benefit of the queue runner, because there is no re-exec
3813 when deliver_drop_privilege is set.
3815 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3816 This has been fixed.
3818 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3819 was in use, was not putting the data itself into the right store pool;
3820 consequently, it could be overwritten for a subsequent message in the same
3821 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3824 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3826 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3827 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3828 the function were treating the return as a boolean value, which happened to
3829 work because 0=false and not-0=true, but is not correct code.
3831 68. The host_aton() function was not handling scoped IPv6 addresses (those
3832 with, for example, "%eth0" on the end) correctly.
3834 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3835 negated items (that is, ~something) in unsigned ints. Some compilers
3836 apparently mutter when there is no cast.
3838 70. If an address verification called from an ACL failed, and did not produce a
3839 user-specific message (i.e. there was only a "system" message), nothing was
3840 put in $acl_verify_message. In this situation, it now puts the system
3843 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3844 session; change 4.31/43 added the unwanted input to the log line - except
3845 that it did not do this in the start of session case. It now does.
3847 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3848 This is wrong and can cause the other end to generate a synchronization
3849 error if it is another Exim or anything else that does the synchronization
3850 check. A QUIT command is no longer sent after a timeout.
3852 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3853 during host lookups.
3855 74. The defer_ok option of callout verification was not working if it was used
3856 when verifying addresses in header lines, that is, for this case:
3858 verify = header_sender/callout=defer_ok
3860 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3861 those file descriptors could be used for SMTP connections. If anything
3862 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3863 could be sent to the SMTP client, causing chaos. The daemon now opens
3864 stdin, stdout, and stderr to /dev/null when it puts itself into the
3867 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3868 log by default. The user can override this with suitable Perl magic.
3870 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3871 the log message when discard triggers, was not working for the DATA ACL or
3872 for the non-SMTP ACL.
3874 78. Error message wording change in sieve.c.
3876 79. If smtp_accept_max_per_host was set, the number of connections could be
3877 restricted to fewer than expected, because the daemon was trying to set up
3878 a new connection before checking whether the processes handling previous
3879 connections had finished. The check for completed processes is now done
3880 earlier. On busy systems, this bug wouldn't be noticed because something
3881 else would have woken the daemon, and it would have reaped the completed
3884 80. If a message was submitted locally by a user whose login name contained one
3885 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3886 It caused a spool format error. I have fixed the spool reading code. A
3887 related problem was that the "from" clause in the Received: line became
3888 illegal because of the space(s). It is now covered by ${quote_local_part.
3890 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3893 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3895 83. Patch from Sieve maintainer to fix unterminated string problem in
3896 "vacation" handling.
3898 84. Some minor changes to the Linux configuration files to help with other
3899 OS variants using glibc.
3901 85. One more patch for Sieve to update vacation handling to latest spec.
3904 ----------------------------------------------------
3905 See the note above about the 4.44 and 4.50 releases.
3906 ----------------------------------------------------
3912 1. Change 4.43/35 introduced a bug that caused file counts to be
3913 incorrectly computed when quota_filecount was set in an appendfile
3916 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3917 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3920 3. Give more explanation in the error message when the command for a transport
3921 filter fails to execute.
3923 4. There are several places where Exim runs a non-Exim command in a
3924 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3925 was being done only for the command run by the queryprogram router. It is
3926 now done for all such subprocesses. The other cases are: ${run, transport
3927 filters, and the commands run by the lmtp and pipe transports.
3929 5. Some older OS have a limit of 256 on the maximum number of file
3930 descriptors. Exim was using setrlimit() to set 1000 as a large value
3931 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3932 systems. I've change it so that if it can't get 1000, it tries for 256.
3934 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3935 was an oversight, and furthermore, ever since the addition of extra
3936 controls (e.g. 4.43/32), the checks on when to allow different forms of
3937 "control" were broken. There should now be diagnostics for all cases when a
3938 control that does not make sense is encountered.
3940 7. $recipients is now available in the predata ACL (oversight).
3942 8. Tidy the search cache before the fork to do a delivery from a message
3943 received from the command line. Otherwise the child will trigger a lookup
3944 failure and thereby defer the delivery if it tries to use (for example) a
3945 cached ldap connection that the parent has called unbind on.
3947 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3948 of $address_data from the recipient verification was clobbered by the
3949 sender verification.
3951 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3952 was its contents. (It was OK if the option was not defined at all.)
3954 11. A "Completed" log line is now written for messages that are removed from
3955 the spool by the -Mrm option.
3957 12. $host_address is now set to the target address during the checking of
3958 ignore_target_hosts.
3960 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3961 being passed; this would have caused $sender_host_name to have been used if
3962 matching the list had actually called for a host name (not very likely,
3963 since this list is usually IP addresses). A host name is now passed as
3966 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3967 code that specifies a non-restarting handler (typically sigaction() in
3968 modern systems) in an attempt to fix a rare and obscure crash bug.
3970 15. Narrowed the window for a race in the daemon that could cause it to ignore
3971 SIGCHLD signals. This is not a major problem, because they are used only to
3972 wake it up if nothing else does.
3974 16. A malformed maildirsize file could cause Exim to calculate negative values
3975 for the mailbox size or file count. Odd effects could occur as a result.
3976 The maildirsize information is now recalculated if the size or filecount
3979 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3980 support for a long time. Removed HAVE_SYS_VFS_H.
3982 18. Updated exipick to current release from John Jetmore.
3984 19. Allow an empty sender to be matched against a lookup in an address list.
3985 Previously the only cases considered were a regular expression, or an
3988 20. Exim went into a mad DNS lookup loop when doing a callout where the
3989 host was specified on the transport, if the DNS lookup yielded more than
3992 21. The RFC2047 encoding function was originally intended for short strings
3993 such as real names; it was not keeping to the 75-character limit for
3994 encoded words that the RFC imposes. It now respects the limit, and
3995 generates multiple encoded words if necessary. To be on the safe side, I
3996 have increased the buffer size for the ${rfc2047: expansion operator from
3999 22. Failure to deliver a bounce message always caused it to be frozen, even if
4000 there was an errors_to setting on the router. The errors_to setting is now
4003 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4004 canonical form (fully expanded) before being placed in
4005 $sender_host_address.
4007 24. Updated eximstats to version 1.33
4009 25. Include certificate and key file names in error message when GnuTLS fails
4010 to set them up, because the GnuTLS error message doesn't include the name
4011 of the failing file when there is a problem reading it.
4013 26. Expand error message when OpenSSL has problems setting up cert/key files.
4016 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4017 (this can affect the format of dates).
4019 28. exim_tidydb, when checking for the continued existence of a message for
4020 which it has found a message-specific retry record, was not finding
4021 messages that were in split spool directories. Consequently, it was
4022 deleting retry records that should have stayed in existence.
4024 29. eximstats updated to version 1.35
4025 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4026 1.35 - bugfix such that pie charts by volume are generated correctly
4028 30. The SPA authentication driver was not abandoning authentication and moving
4029 on to the next authenticator when an expansion was forced to fail,
4030 contradicting the general specification for all authenticators. Instead it
4031 was generating a temporary error. It now behaves as specified.
4033 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4034 (the order specifies the preference for clients). The order is now AES256,
4035 AES128, 3DES, ARCFOUR128.
4037 31. Small patch to Sieve code - explicitly set From: when generating an
4040 32. Exim crashed if a remote delivery caused a very long error message to be
4041 recorded - for instance if somebody sent an entire SpamAssassin report back
4042 as a large number of 550 error lines. This bug was coincidentally fixed by
4043 increasing the size of one of Exim's internal buffers (big_buffer) that
4044 happened as part of the Exiscan merge. However, to be on the safe side, I
4045 have made the code more robust (and fixed the comments that describe what
4048 33. Some experimental protocols are using DNS PTR records for new purposes. The
4049 keys for these records are domain names, not reversed IP addresses. The
4050 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4051 leaves it alone. Component reversal etc. now happens only for IP addresses.
4054 34. The host_aton() function is supposed to be passed a string that is known
4055 to be a valid IP address. However, in the case of IPv6 addresses, it was
4056 not checking this. This is a hostage to fortune. Exim now panics and dies
4057 if the condition is not met. A case was found where this could be provoked
4058 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4059 components; fortuitously, this particular loophole had already been fixed
4060 by change 4.50/55 or 4.44/33 above.
4062 If there are any other similar loopholes, the new check in host_aton()
4063 itself should stop them being exploited. The report I received stated that
4064 data on the command line could provoke the exploit when Exim was running as
4065 exim, but did not say which command line option was involved. All I could
4066 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4067 running as the user.
4070 35. There was a buffer overflow vulnerability in the SPA authentication code
4071 (which came originally from the Samba project). I have added a test to the
4072 spa_base64_to_bits() function which I hope fixes it.
4075 36. The daemon start-up calls getloadavg() while still root for those OS that
4076 need the first call to be done as root, but it missed one case: when
4077 deliver_queue_load_max is set with deliver_drop_privilege. This is
4078 necessary for the benefit of the queue runner, because there is no re-exec
4079 when deliver_drop_privilege is set.
4081 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4082 was in use, was not putting the data itself into the right store pool;
4083 consequently, it could be overwritten for a subsequent message in the same
4084 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4087 38. Sometimes the final signoff response after QUIT could fail to get
4088 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4089 < 0 before doing a fflush(). This bug looks as though it goes back to the
4090 introduction of TLS in release 3.20, but "sometimes" must have been rare
4091 because the tests only now provoked it.
4097 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4098 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4099 Now it returns an integer. A number of calls were still expecting a T/F
4100 return. Fortuitously, in all cases, the tests worked in OK situations,
4101 which is the norm. However, things would have gone wrong on any write
4102 failures on the smtp file descriptor. This function is used when sending
4103 messages over SMTP and also when doing verify callouts.
4105 2. When Exim is called to do synchronous delivery of a locally submitted
4106 message (the -odf or -odi options), it no longer closes stderr before doing
4109 3. Implemented the mua_wrapper option.
4111 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4113 5. Implemented the functions header_remove(), header_testname(),
4114 header_add_at_position(), and receive_remove_recipient(), and exported them
4117 6. If an ACL "warn" statement specified the addition of headers, Exim already
4118 inserted X-ACL-Warn: at the start if there was no header name. However, it
4119 was not making this test for the second and subsequent header lines if
4120 there were newlines in the string. This meant that an invalid header could
4121 be inserted if Exim was badly configured.
4123 7. Allow an ACL "warn" statement to add header lines at the start or after all
4124 the Received: headers, as well as at the end.
4126 8. Added the rcpt_4xx retry error code.
4128 9. Added postmaster_mailfrom=xxx to callout verification option.
4130 10. Added mailfrom=xxxx to the callout verification option, for verify=
4133 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4134 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4136 12. Inserted some casts to stop certain compilers warning when using pointer
4137 differences as field lengths or precisions in printf-type calls (mostly
4138 affecting debugging statements).
4140 13. Added optional readline() support for -be (dynamically loaded).
4142 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4143 same clock tick as a message's arrival, so that its received time was the
4144 same as the "first fail" time on the retry record, and that message
4145 remained on the queue past the ultimate address timeout, every queue runner
4146 would try a delivery (because it was past the ultimate address timeout) but
4147 after another failure, the ultimate address timeout, which should have then
4148 bounced the address, did not kick in. This was a "< instead of <=" error;
4149 in most cases the first failure would have been in the next clock tick
4150 after the received time, and all would be well.
4152 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4153 being recognized when the domain list was tested by the match_domain
4154 condition in an expansion string.
4156 16. Added the ${str2b64: operator.
4158 17. Exim was always calling setrlimit() to set a large limit for the number of
4159 processes, without checking whether the existing limit was already
4160 adequate. (It did check for the limit on file descriptors.) Furthermore,
4161 errors from getrlimit() and setrlimit() were being ignored. Now they are
4162 logged to the main and panic logs, but Exim does carry on, to try to do its
4163 job under whatever limits there are.
4165 18. Imported PCRE 5.0.
4167 19. Trivial typo in log message " temporarily refused connection" (the leading
4170 20. If the log selector return_path_on_delivery was set and an address was
4171 redirected to /dev/null, the delivery process crashed because it assumed
4172 that a return path would always be set for a "successful" delivery. In this
4173 case, the whole delivery is bypassed as an optimization, and therefore no
4176 21. Internal re-arrangement: the function for sending a challenge and reading
4177 a response while authentication was assuming a zero-terminated challenge
4178 string. It's now changed to take a pointer and a length, to allow for
4179 binary data in such strings.
4181 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4183 23. Exim was not respecting finduser_retries when seeking the login of the
4184 uid under which it was called; it was always trying 10 times. (The default
4185 setting of finduser_retries is zero.) Also, it was sleeping after the final
4186 failure, which is pointless.
4188 24. Implemented tls_on_connect_ports.
4190 25. Implemented acl_smtp_predata.
4192 26. If the domain in control=submission is set empty, Exim assumes that the
4193 authenticated id is a complete email address when it generates From: or
4194 Sender: header lines.
4196 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4197 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4198 chown and chgrp in /bin and hostname in /usr/bin.
4200 28. Exim was keeping the "process log" file open after each use, just as it
4201 does for the main log. This opens the possibility of it remaining open for
4202 long periods when the USR1 signal hits a daemon. Occasional processlog
4203 errors were reported, that could have been caused by this. Anyway, it seems
4204 much more sensible not to leave this file open at all, so that is what now
4207 29. The long-running daemon process does not normally write to the log once it
4208 has entered its main loop, and it closes the log before doing so. This is
4209 so that log files can straightforwardly be renamed and moved. However,
4210 there are a couple of unusual error situations where the daemon does write
4211 log entries, and I had neglected to close the log afterwards.
4213 30. The text of an SMTP error response that was received during a remote
4214 delivery was being truncated at 512 bytes. This is too short for some of
4215 the long messages that one sometimes sees. I've increased the limit to
4218 31. It is now possible to make retry rules that apply only when a message has a
4219 specific sender, in particular, an empty sender.
4221 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4222 it possible to be selective about when SMTP synchronization is enforced.
4224 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4226 32. Implemented hosts_connection_nolog.
4228 33. Added an ACL for QUIT.
4230 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4233 35. Added mailbox_size and mailbox_filecount to appendfile.
4235 36. Added control = no_multiline_responses to ACLs.
4237 37. There was a bug in the logic of the code that waits for the clock to tick
4238 in the case where the clock went backwards by a substantial amount such
4239 that the microsecond fraction of "now" was more than the microsecond
4240 fraction of "then" (but the whole seconds number was less).
4242 38. Added support for the libradius Radius client library this is found on
4243 FreeBSD (previously only the radiusclient library was supported).
4249 1. When certain lookups returned multiple values in the form name=value, the
4250 quoting of the values was not always being done properly. Specifically:
4251 (a) If the value started with a double quote, but contained no whitespace,
4253 (b) If the value contained whitespace other than a space character (i.e.
4254 tabs or newlines or carriage returns) it was not quoted.
4255 This fix has been applied to the mysql and pgsql lookups by writing a
4256 separate quoting function and calling it from the lookup code. The fix
4257 should probably also be applied to nisplus, ibase and oracle lookups, but
4258 since I cannot test any of those, I have not disturbed their existing code.
4260 2. A hit in the callout cache for a specific address caused a log line with no
4261 reason for rejecting RCPT. Now it says "Previous (cached) callout
4262 verification failure".
4264 3. There was an off-by-one bug in the queryprogram router. An over-long
4265 return line was truncated at 256 instead of 255 characters, thereby
4266 overflowing its buffer with the terminating zero. As well as fixing this, I
4267 have increased the buffer size to 1024 (and made a note to document this).
4269 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4270 when Exim is waiting for an SMTP response from a remote server, Exim
4271 restarts its select() call on the socket, thereby resetting its timeout.
4272 This is not a problem when such interrupts are rare. Somebody set up a cron
4273 job to run exiwhat every 2 minutes, which is less than the normal select()
4274 timeout (5 or 10 minutes). This meant that the select() timeout never
4275 kicked in because it was always reset. I have fixed this by comparing the
4276 time when an interrupt arrives with the time at the start of the first call
4277 to select(). If more time than the timeout has elapsed, the interrupt is
4278 treated as a timeout.
4280 5. Some internal re-factoring in preparation for the addition of Sieve
4281 extensions (by MH). In particular, the "personal" test is moved to a
4282 separate function, and given an option for scanning Cc: and Bcc: (which is
4283 not set for Exim filters).
4285 6. When Exim created an email address using the login of the caller as the
4286 local part (e.g. when creating a From: or Sender: header line), it was not
4287 quoting the local part when it contained special characters such as @.
4289 7. Installed new OpenBSD configuration files.
4291 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4292 try to make them clearer.
4294 9. Callout options, other than the timeout value, were being ignored when
4295 verifying sender addresses in header lines. For example, when using
4297 verify = header_sender/callout=no_cache
4299 the cache was (incorrectly) being used.
4301 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4302 only the Cygwin environment.
4304 11. When return_path_on_delivery was set as a log selector, if different remote
4305 addresses in the same message used different return paths and parallel
4306 remote delivery occurred, the wrong values would sometimes be logged.
4307 (Whenever a remote delivery process finished, the return path value from
4308 the most recently started remote delivery process was logged.)
4310 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4311 lines when AUTH and/or TLS are in use. This is the "received protocol"
4312 field. Exim used to use "asmtp" for authenticated SMTP, without any
4313 indication (in the protocol name) for TLS use. Now it follows the RFC and
4314 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4315 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4316 names appear in log lines as well as in Received: header lines.
4318 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4319 extensions, and comparison tests, and to fix some bugs.
4321 14. Changes to the "personal" filter test:
4323 (1) The test was buggy in that it was just doing the equivalent of
4324 "contains" tests on header lines. For example, if a user's address was
4325 anne@some.where, the "personal" test would incorrectly be true for
4327 To: susanne@some.where
4329 This test is now done by extracting each address from the header in turn,
4330 and checking the entire address. Other tests that are part of "personal"
4331 are now done using regular expressions (for example, to check local parts
4332 of addresses in From: header lines).
4334 (2) The list of non-personal local parts in From: addresses has been
4335 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4336 taken from the Sieve specification recommendations.
4338 (3) If the message contains any header line starting with "List-" it is
4339 treated as non-personal.
4341 (4) The test for "circular" in the Subject: header line has been removed
4342 because it now seems ill-conceived.
4344 15. Minor typos in src/EDITME comments corrected.
4346 16. Installed latest exipick from John Jetmore.
4348 17. If headers_add on a router specified a text string that was too long for
4349 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4350 of string_sprintf() is now avoided.
4352 18. $message_body_size was not set (it was always zero) when running the DATA
4353 ACL and the local_scan() function.
4355 19. For the "mail" command in an Exim filter, no default was being set for
4356 the once_repeat time, causing a random time value to be used if "once" was
4357 specified. (If the value happened to be <= 0, no repeat happened.) The
4358 default is now 0s, meaning "never repeat". The "vacation" command was OK
4359 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4360 (I found it when inspecting the code).
4362 20. There is now an overall timeout for performing a callout verification. It
4363 defaults to 4 times the callout timeout, which applies to individual SMTP
4364 commands during the callout. The overall timeout applies when there is more
4365 than one host that can be tried. The timeout is checked before trying the
4366 next host. This prevents very long delays if there are a large number of
4367 hosts and all are timing out (e.g. when the network connections are timing
4368 out). The value of the overall timeout can be changed by specifying an
4369 additional sub-option for "callout", called "maxwait". For example:
4371 verify = sender/callout=5s,maxwait=20s
4373 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4374 to the end before writing, but this should make it even safer).
4376 22. Exim was forgetting that it had advertised PIPELINING for the second and
4377 subsequent messages on an SMTP connection. It was also not resetting its
4378 memory on STARTTLS and an internal HELO.
4380 23. When Exim logs an SMTP synchronization error within a session, it now
4381 records whether PIPELINING has been advertised or not.
4383 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4384 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4385 rather than long int.
4387 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4389 26. Added the never_mail option to autoreply.
4395 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4396 crash if the getsockname() call failed; this can happen if a connection is
4397 closed very soon after it is established. The problem was simply in the
4398 order in which certain operations were done, causing Exim to try to write
4399 to the SMTP stream before it had set up the file descriptor. The bug has
4400 been fixed by making things happen in the correct order.
4406 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4407 before the connection was closed, thus losing the rejection response.
4409 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4410 some early Solaris releases, but causes trouble in current releases where
4411 socklen_t is defined.
4413 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4416 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4419 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4420 that releasing the top of it at the end releases what was used for sub-
4421 expansions (unless the block got too big). However, discard this block if
4422 the first thing is a variable or header, so that we can use its block when
4423 it is dynamic (useful for very large $message_headers, for example).
4425 6. Lookups now cache *every* query, not just the most recent. A new, separate
4426 store pool is used for this. It can be recovered when all lookup caches are
4427 flushed. Lookups now release memory at the end of their result strings.
4428 This has involved some general refactoring of the lookup sources.
4430 7. Some code has been added to the store_xxx() functions to reduce the amount
4431 of flapping under certain conditions.
4433 8. log_incoming_interface used to affect only the <= reception log lines. Now
4434 it causes the local interface and port to be added to several more SMTP log
4435 lines, for example "SMTP connection from", and rejection lines.
4437 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4439 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4441 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4442 could be overwritten at the end of the current message (or the start of a
4443 new message if it was set in a HELO ACL). The value is now preserved for
4444 the duration of the SMTP connection.
4446 12. If a transport had a headers_rewrite setting, and a matching header line
4447 contained an unqualified address, that address was qualified, even if it
4448 did not match any rewriting rules. The underlying bug was that the values
4449 of the flags that permit the existence of unqualified sender and recipient
4450 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4451 non-local messages, and by -bnq for local messages) were not being
4452 preserved with the message after it was received.
4454 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4455 "next input=" as part of the text comprising the host identity instead of
4456 the correct text. The code was using the same buffer for two different
4457 strings. However, depending on which order the printing function evaluated
4458 its arguments, the bug did not always show up. Under Linux, for example, my
4459 test suite worked just fine.
4461 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4462 doesn't work with some older versions of Perl. It has been changed to "my",
4463 which in any case is probably the better facility to use.
4465 15. A really picky compiler found some instances of statements for creating
4466 error messages that either had too many or two few arguments for the format
4469 16. The size of the buffer for calls to the DNS resolver has been increased
4470 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4471 for addresses that have a lot of PTR records. This alleviates a problem; it
4472 does not fully solve it.
4474 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4475 buffer now truncates the list and logs the incident, which is the same
4476 action as happens when Exim is looking up a host name and its aliases.
4477 Previously in this situation something unpredictable would happen;
4478 sometimes it was "internal error: store_reset failed".
4480 18. If a server dropped the connection unexpectedly when an Exim client was
4481 using GnuTLS and trying to read a response, the client delivery process
4482 crashed while trying to generate an error log message.
4484 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4485 string, for example:
4487 warn message = H1: something\nH2: something
4489 the text was added as a single header line from Exim's point of view
4490 though it ended up OK in the delivered message. However, searching for the
4491 second and subsequent header lines using $h_h2: did not work. This has been
4492 fixed. Similarly, if a system filter added multiple headers in this way,
4493 the routers could not see them.
4495 20. Expanded the error message when iplsearch is called with an invalid key to
4496 suggest using net-iplsearch in a host list.
4498 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4499 ACLs are no longer actually imposed (and a message to that effect is
4502 22. If a "gecos" field in a passwd entry contained escaped characters, in
4503 particular, if it contained a \" sequence, Exim got it wrong when building
4504 a From: or a Sender: header from that name. A second bug also caused
4505 incorrect handling when an unquoted " was present following a character
4506 that needed quoting.
4508 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4509 was not being matched caselessly.
4511 24. Arranged for all hyphens in the exim.8 source to be escaped with
4514 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4515 properly. Recipient callout cache records were still being keyed to include
4516 the sender, even when use_sender was set false. This led to far more
4517 callouts that were necessary. The sender is no longer included in the key
4518 when use_sender is false.
4520 26. Added "control = submission" modifier to ACLs.
4522 27. Added the ${base62d: operator to decode base 62 numbers.
4524 28. dnsdb lookups can now access SRV records.
4526 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4527 the configuration file.
4529 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4530 (-v) mode. This makes the output for a verbose queue run more intelligible.
4532 31. Added a use_postmaster feature to recipient callouts.
4534 32. Added the $body_zerocount variable, containing the number of binary zero
4535 bytes in the message body.
4537 33. The time of last modification of the "new" subdirectory is now used as the
4538 "mailbox time last read" when there is a quota error for a maildir
4541 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4543 35. Added +ignore_unknown as a special item in host lists.
4545 36. Code for decoding IPv6 addresses in host lists is now included, even if
4546 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4547 address was recognized as an IP address, but was then not correctly decoded
4548 into binary, causing unexpected and incorrect effects when compared with
4555 1. Very minor rewording of debugging text in manualroute to say "list of
4556 hosts" instead of "hostlist".
4558 2. If verify=header_syntax was set, and a header line with an unqualified
4559 address (no domain) and a large number of spaces between the end of the
4560 name and the colon was received, the reception process suffered a buffer
4561 overflow, and (when I tested it) crashed. This was caused by some obsolete
4562 code that should have been removed. The fix is to remove it!
4564 3. When running in the test harness, delay a bit after writing a bounce
4565 message to get a bit more predictability in the log output.
4567 4. Added a call to search_tidyup() just before forking a reception process. In
4568 theory, someone could use a lookup in the expansion of smtp_accept_max_
4569 per_host which, without the tidyup, could leave open a database connection.
4571 5. Added the variables $recipient_data and $sender_data which get set from a
4572 lookup success in an ACL "recipients" or "senders" condition, or a router
4573 "senders" option, similar to $domain_data and $local_part_data.
4575 6. Moved the writing of debug_print from before to after the "senders" test
4578 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4579 problems for message scanning, either using a data ACL, or using
4580 local_scan() because the Received: header was not generated till after they
4581 were called (in order to set the time as the time of reception completion).
4582 I have revised the way this works. The header is now generated after the
4583 body is received, but before the ACL or local_scan() are called. After they
4584 are run, the timestamp in the header is updated.
4590 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4591 before starting a queue runner without re-exec. This happened only when
4592 deliver_drop_privilege was set or when the Exim user was set to root. The
4593 effect of the bug was that timeouts during subsequent deliveries caused
4594 crashes instead of being properly handled. The handler is now left at its
4595 default (and expected) setting.
4597 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4598 message, again when deliver_drop_privilege is set or Exim is run as root.
4599 The bug described in (1) was not present in this case, but the tidying up
4600 of the other signals was missing. I have made the two cases consistent.
4602 3. The ignore_target_hosts setting on a manualroute router was being ignored
4603 for hosts that were looked up using the /MX notation.
4605 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4608 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4609 operated on the sender address. After changing the $sender_address to <>
4610 for the sender address verify, Exim was re-instated it as the original
4611 (before rewriting) address, but remembering that it had rewritten it, so it
4612 wasn't rewriting it again. This bug also had the effect of breaking the
4613 sender address verification caching when the sender address was rewritten.
4615 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4616 This has been changed so that if the ip literal address matches
4617 ignore_target_hosts, the router declines.
4619 7. Added expansion conditions match_domain, match_address, and match_local_
4620 part (NOT match_host).
4622 8. The placeholder for the Received: header didn't have a length field set.
4624 9. Added code to Exim itself and to exim_lock to test for a specific race
4625 condition that could lead to file corruption when using MBX delivery. The
4626 issue is with the lockfile that is created in /tmp. If this file is removed
4627 after a process has opened it but before that process has acquired a lock,
4628 there is the potential for a second process to recreate the file and also
4629 acquire a lock. This could lead to two Exim processes writing to the file
4630 at the same time. The added code performs the same test as UW imapd; it
4631 checks after acquiring the lock that its file descriptor still refers to
4632 the same named file.
4634 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4635 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4638 11. Added the smtp_active_hostname option. If used, this will typically be made
4639 to depend on the incoming interface address. Because $interface_address is
4640 not set up until the daemon has forked a reception process, error responses
4641 that can happen earlier (such as "too many connections") no longer contain
4644 12. If an expansion in a condition on a "warn" statement fails because a lookup
4645 defers, the "warn" statement is abandoned, and the next ACL statement is
4646 processed. Previously this caused the whole ACL to be aborted.
4648 13. Added the iplsearch lookup type.
4650 14. Added ident_timeout as a log selector.
4652 15. Added tls_certificate_verified as a log selector.
4654 16. Added a global option tls_require_ciphers (compare the smtp transport
4655 option of the same name). This controls incoming TLS connections.
4657 17. I finally figured out how to make tls_require_ciphers do a similar thing
4658 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4659 before starting the TLS session.
4661 18. Tabs are now shown as \t in -bP output.
4663 19. If the log selector return_path_on_delivery was set, Exim crashed when
4664 bouncing a message because it had too many Received: header lines.
4666 20. If two routers both had headers_remove settings, and the first one included
4667 a superfluous trailing colon, the final name in the first list and the
4668 first name in the second list were incorrectly joined into one item (with a
4669 colon in the middle).
4675 1. Added -C and -D options to the exinext utility, mainly to make it easier
4676 to include in the automated testing, but these could be helpful when
4677 multiple configurations are in use.
4679 2. The exinext utility was not formatting the output nicely when there was
4680 an alternate port involved in the retry record key, nor when there was a
4681 message id as well (for retries that were specific to a specific message
4682 and a specific host). It was also confused by IPv6 addresses, because of
4683 the additional colons they contain. I have fixed the IPv4 problem, and
4684 patched it up to do a reasonable job for IPv6.
4686 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4687 delivery, the log line now contains "pipelined" if PIPELINING was used.
4689 4. An SMTP transport process used to panic and die if the bind() call to set
4690 an explicit outgoing interface failed. This has been changed; it is now
4691 treated in the same way as a connect() failure.
4693 5. A reference to $sender_host_name in the part of a conditional expansion
4694 that was being skipped was still causing a DNS lookup. This no longer
4697 6. The def: expansion condition was not recognizing references to header lines
4698 that used bh_ and bheader_.
4700 7. Added the _cache feature to named lists.
4702 8. The code for checking quota_filecount in the appendfile transport was
4703 allowing one more file than it should have been.
4705 9. For compatibility with Sendmail, the command line option
4713 and sets the incoming protocol and host name (for trusted callers). The
4714 host name and its colon can be omitted when only the protocol is to be set.
4715 Note the Exim already has two private options, -pd and -ps, that refer to
4716 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4717 "s", but I don't think that's a major issue.
4719 10. A number of refactoring changes to the code, none of which should affect
4722 (a) The number of logging options was getting close to filling up the
4723 32-bit word that was used as a bit map. I have split them into two classes:
4724 those that are passed in the argument to log_write(), and those that are
4725 only ever tested independently outside of that function. These are now in
4726 separate 32-bit words, so there is plenty of room for expansion again.
4727 There is no change in the user interface or the logging behaviour.
4729 (b) When building, for example, log lines, the code previously used a
4730 macro that called string_cat() twice, in order to add two strings. This is
4731 not really sufficiently general. Furthermore, there was one instance where
4732 it was actually wrong because one of the argument was used twice, and in
4733 one call a function was used. (As it happened, calling the function twice
4734 did not affect the overall behaviour.) The macro has been replaced by a
4735 function that can join an arbitrary number of extra strings onto a growing
4738 (c) The code for expansion conditions now uses a table and a binary chop
4739 instead of a serial search (which was left over from when there were very
4740 few conditions). Also, it now recognizes conditions like "pam" even when
4741 the relevant support is not compiled in: a suitably worded error message is
4742 given if an attempt is made to use such a condition.
4744 11. Added ${time_interval:xxxxx}.
4746 12. A bug was causing one of the ddress fields not to be passed back correctly
4747 from remote delivery subprocesses. The field in question was not being
4748 subsequently used, so this caused to problems in practice.
4750 13. Added new log selectors queue_time and deliver_time.
4752 14. Might have fixed a bug in maildirsizefile handling that threw up
4753 "unexpected character" debug warnings, and recalculated the data
4754 unnecessarily. In any case, I expanded the warning message to give more
4757 15. Added the message "Restricted characters in address" to the statements in
4758 the default ACL that block characters like @ and % in local parts.
4760 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4761 Three changes have been made:
4763 (a) There was a serious bug; a negative response to MAIL caused the whole
4764 recipient domain to be cached as invalid, thereby blocking all messages
4765 to all local parts at the same domain, from all senders. This bug has
4766 been fixed. The domain is no longer cached after a negative response to
4767 MAIL if the sender used is not empty.
4769 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4772 (c) A new callout option, "use_sender" has been added for people who want
4773 the modified behaviour.
4779 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4782 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4783 indeed breaks things for older releases.
4785 3. Added additional logging to the case where there is a problem reading data
4786 from a filter that is running in a subprocess using a pipe, in order to
4787 try to track down a specific problem.
4789 4. Testing facility fudge: when running in the test harness and attempting
4790 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4791 getting "No route to host". Convert this to a timeout.
4793 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4796 6. Some OS don't have socklen_t but use size_t instead. This affects the
4797 fifth argument of getsockopt() amongst other things. This is now
4798 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4799 can be set for individual OS. I have set it for SunOS5, OSF1, and
4800 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4801 some earlier ones do not.
4803 7. Change 4.30/15 was not doing the test caselessly.
4805 8. The standard form for an IPv6 address literal was being rejected by address
4806 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4807 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4808 this, as well as the form without the "IPv6" on the front (but only when
4809 address literals are enabled, of course).
4811 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4813 10. Exim crashed if a message with an empty sender address specified by -f
4814 encountered a router with an errors_to setting. This could be provoked only
4815 by a command such as
4819 where an empty string was supplied; "<>" did not hit this bug.
4821 11. Installed PCRE release 4.5.
4823 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4824 remained set. It is now erased.
4826 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4827 times from message ids (which are base 36 rather than the normal 62).
4829 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4830 were being counted as actual protocol errors, and logged if the log
4831 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4832 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4833 and DATA following a set of rejected RCPTs do not count as protocol errors.
4834 In other words, Exim assumes they were pipelined, though this may not
4835 actually be the case. Of course, in all cases the client gets an
4836 appropriate error code.
4838 15. If a lookup fails in an ACL condition, a message about the failure may
4839 be available; it is used if testing the ACL cannot continue, because most
4840 such messages specify what the cause of the deferral is. However, some
4841 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4842 that caused an old message to be retained and used if a later statement
4843 caused a defer, replacing the real cause of the deferral.
4845 16. If an IP address had so many PTR records that the DNS lookup buffer
4846 was not large enough to hold them, Exim could crash while trying to process
4847 the truncated data. It now detects and logs this case.
4849 17. Further to 4.21/58, another change has been made: if (and only if) the
4850 first line of a message (the first header line) ends with CRLF, a bare LF
4851 in a subsequent header line has a space inserted after it, so as not to
4852 terminate the header.
4854 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4855 unnecessarily, used atoi() instead of strtol(), and didn't check the
4856 termination when getting file sizes from file names by regex.
4858 19. Completely re-implemented the support for maildirsize files, in the light
4859 of a number of problems with the previous contributed implementation
4860 (4.30/29). In particular:
4862 . If the quota is zero, the maildirsize file is maintained, but no quota is
4865 . If the maildir directory does not exist, it is created before any attempt
4866 to write a maildirsize file.
4868 . The quota value in the file is just a cache; if the quota is changed in
4869 the transport, the new value overrides.
4871 . A regular expression is available for excluding directories from the
4874 20. The autoreply transport checks the characters in options that define the
4875 message's headers; it allows continued headers, but it was checking with
4876 isspace() after an embedded newline instead of explicitly looking for a
4879 21. If all the "regular" hosts to which an address was routed had passed their
4880 expiry times, and had not reached their retry times, the address was
4881 bounced, even if fallback hosts were defined. Now Exim should go on to try
4884 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4885 equivalent code in the SMTP transport. Some hosts send humungous responses
4886 to HELO/EHLO, more than 1024 it seems.
4888 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4889 gives compiler warnings in some environments. I've now done it "properly",
4892 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4893 (because of problems with the built-in one) was declared to return uschar *
4894 instead of char *, causing compiler failure.
4896 25. Fixed a file descriptor leak when processing alias/forward files.
4898 26. Fixed a minor format string issue in dbfn.c.
4900 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4902 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4903 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4906 29. When a sender address is verified, it is cached, to save repeating the test
4907 when there is more than one recipient in a message. However, when the
4908 verification involves a callout, it is possible for different callout
4909 options to be set for different recipients. It is too complicated to keep
4910 track of this in the cache, so now Exim always runs a verification when a
4911 callout is required, relying on the callout cache for the optimization.
4912 The overhead is duplication of the address routing, but this should not be
4915 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4916 to be verified with callout=postmaster, and the main callout worked but the
4917 postmaster check failed, the verification correctly failed. However, if a
4918 subsequent RCPT command asked for sender verification *without* the
4919 postmaster check, incorrect caching caused this verification also to fail,
4922 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4923 it was not caching the DNS options (qualify_single, search_parents) that
4924 were used when the lookup failed. A subsequent lookup with different
4925 options therefore always gave the same answer, though there were cases
4926 where it should not have. (Example: a "domains = !$mx_any" option on a
4927 dnslookup router: the "domains" option is always processed without any
4928 widening, but the router might have qualify_single set.) Now Exim uses the
4929 cached value only when the same options are set.
4931 32. Added John Jetmore's "exipick" utility to the distribution.
4933 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4934 than a timeout (e.g. a certificate is required, and is not provided), an
4935 Exim server now closes the connection immediately. Previously it waited for
4936 the client to close - but if the client is SSL, it seems that they each
4937 wait for each other, leading to a delay before one of them times out.
4939 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4940 maintained 0.8.x compatibility because I don't think many are using it, and
4941 it is clearly obsolete.
4943 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4946 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4947 client certificate was expired. A simple patch fixes this, though I don't
4948 understand the full logic of why the verify callback is called multiple
4951 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4952 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4953 which causes problems with some clients (such as the Certicom SSL Plus
4954 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4955 disables the coutermeasure allowing Eudora to connect."
4957 38. Exim was not checking that a write() to a log file succeeded. This could
4958 lead to Bad Things if a log got too big, in particular if it hit a file
4959 size limit. Exim now panics and dies if it cannot write to a log file, just
4960 as it does if it cannot open a log file.
4962 39. Modified OS/Makefile-Linux so that it now contains
4964 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4966 The two -D definitions ensure that Exim is compiled with large file
4967 support, which makes it possible to handle log files that are bigger than
4970 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4971 instance) a domain was checked against a named list that involved a lookup,
4972 causing $domain_data to be set, then another domain was checked against the
4973 same list, then the first domain was re-checked, the value of $domain_data
4974 after the final check could be wrong. In particular, if the second check
4975 failed, it could be set empty. This bug probably also applied to
4978 41. The strip_trailing_dot option was not being applied to the address given
4979 with the -f command-line option.
4981 42. The code for reading a message's header from the spool was incrementing
4982 $received_count, but never initializing it. This meant that the value was
4983 incorrect (doubled) while delivering a message in the same process in which
4984 it was received. In the most common configuration of Exim, this never
4985 happens - a fresh exec is done - but it can happen when
4986 deliver_drop_privilege is set.
4988 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4989 it now includes up to 150 characters of the unexpected data in the log
4992 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4993 and building data strings. The size of both of these buffers was 10 000
4994 bytes - far larger than anybody would *ever* want, thought I. Needless to
4995 say, somebody hit the limit. I have increased the maximum line length to
4996 20 000 and the maximum data length of concatenated lines to 100 000. I have
4997 also fixed two bugs, because there was no checking on these buffers. Tsk,
4998 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4999 buffer is too small.
5001 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5002 lsearch lookups. Now it does.
5004 46. When parsing a route_list item in a manualroute router, a fixed-length
5005 buffer was used for the list of hosts. I made this 1024 bytes long,
5006 thinking that nobody would ever have a list of hosts that long. Wrong.
5007 Somebody had a whole pile of complicated expansion conditions, and the
5008 string was silently truncated, leading to an expansion error. It turns out
5009 that it is easier to change to an unlimited length (owing to other changes
5010 that have happened since this code was originally written) than to build
5011 structure for giving a limitation error. The length of the item that
5012 expands into the list of hosts is now unlimited.
5014 47. The lsearch lookup could not handle data where the length of text line was
5015 more than 4095 characters. Such lines were truncated, leading to shortened
5016 data being returned. It should now handle lines of any length.
5018 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5019 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5022 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5023 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5024 no longer happens in comment lines. A list of replacements is now placed
5025 at the head of all of the source files, except those whose only change is
5026 to replace PERL_COMMAND in the very first #! line.
5028 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5029 messages on the queue, with a bottom-up merge sort, using code contributed
5030 by Michael Haardt. This should make operations like -bp somewhat faster on
5031 large queues. It won't affect queue runners, except when queue_run_in_order
5034 51. Installed eximstats 1.31 in the distribution.
5036 52. Added support for SRV lookups to the dnslookup router.
5038 53. If an ACL referred to $message_body or $message_body_end, the value was not
5039 reset for any messages that followed in the same SMTP session.
5041 54. The store-handling optimization for building very long strings was not
5042 differentiating between the different store pools. I don't think this
5043 actually made any difference in practice, but I've tidied it.
5045 55. While running the routers to verify a sender address, $sender_address
5046 was still set to the sender address. This is wrong, because when routing to
5047 send a bounce to the sender, it would be empty. Therefore, I have changed
5048 it so that, while verifying a sender address, $sender_address is set to <>.
5049 (There is no change to what happens when verifying a recipient address.)
5051 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5052 target A or AAAA records (if not already returned) without resetting the
5053 qualify_single or search_parents options of the DNS resolver. These are
5054 inappropriate in this case because the targets of MX and SRV records must
5055 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5056 target that, when qualified, matched something in the local domain. These
5057 two options are now turned off when doing these lookups.
5059 57. It seems that at least some releases of Reiserfs (which does not have the
5060 concept of a fixed number of inodes) returns zero and not -1 for the
5061 number of available inodes. This interacted badly with check_spool_inodes,
5062 which assumed that -1 was the "no such thing" setting. What I have done is
5063 to check that the total number of inodes is greater than zero before doing
5064 the test of how many are available.
5066 58. When a "warn" ACL statement has a log_message modifier, the message is
5067 remembered, and not repeated. This is to avoid a lot of repetition when a
5068 message has many recipients that cause the same warning to be written.
5069 Howewer, Exim was preserving the list of already written lines for an
5070 entire SMTP session, which doesn't seem right. The memory is now reset if a
5071 new message is started.
5073 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5074 debugging output unless log_rewrite was also set.
5076 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5077 of (int)(handle) when we know that handle contains (void *)(-1).
5079 61. The Exim daemon panic-logs an error return when it closes the incoming
5080 connection. However "connection reset by peer" seems to be common, and
5081 isn't really an error worthy of noting specially, so that particular error
5084 62. When Exim is trying to find all the local interfaces, it used to panic and
5085 die if the ioctl to get the interface flags failed. However, it seems that
5086 on at least one OS (Solaris 9) it is possible to have an interface that is
5087 included in the list of interfaces, but for which you get a failure error
5088 for this call. This happens when the interface is not "plumbed" into a
5089 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5090 failure of the "get flags" call assumes that the interface is down.
5092 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5093 makes life easier for people who are doing arithmetic on fields extracted
5094 from dates, where you often get leading zeros that should not be
5095 interpreted as octal.
5097 64. Added qualify_domain to the redirect router, to override the global
5100 65. If a pathologically long header line contained very many addresses (the
5101 report of this problem mentioned 10 000) and each of them was rewritten,
5102 Exim could use up a very large amount of memory. (It kept on making new
5103 copies of the header line as it rewrote, and never released the old ones.)
5104 At the expense of a bit more processing, the header rewriting function has
5105 been changed so that it no longer eats memory in this way.
5107 66. The generation of the Received: header has been moved from the time that a
5108 message starts to be received, to the time that it finishes. The timestamp
5109 in the Received: header should now be very close to that of the <= log
5110 line. There are two side-effects of this change:
5112 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5113 logged header lines no longer include the local Received: line, because
5114 it has not yet been created. The same applies to a copy of the message
5115 that is returned to a non-SMTP sender when a message is rejected.
5117 (b) When a filter file is tested using -bf, no additional Received: header
5118 is added to the test message. After some thought, I decided that this
5121 This change does not affect the value of $received_for. It is still set
5122 after address rewriting, but before local_scan() is called.
5124 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5126 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5127 gave an unhelpful panic error message, and a defer error. I have managed to
5128 change this behaviour so that it now rejects any supplied certificate,
5129 which seems right, as the list of acceptable certificates is empty.
5131 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5132 gave an unhelpful defer error. I have not managed to make this reject any
5133 supplied certificates, but the error message it gives is "no certificate
5134 supplied", which is not helpful.
5136 70. exigrep's output now also includes lines that are not associated with any
5137 message, but which match the given pattern. Implemented by a patch from
5138 Martin Sluka, which also tidied up the Perl a bit.
5140 71. Recipient callout verification, like sender verification, was using <> in
5141 the MAIL FROM command. This isn't really the right thing, since the actual
5142 sender may affect whether the remote host accepts the recipient or not. I
5143 have changed it to use the actual sender in the callout; this means that
5144 the cache record is now keyed on a recipient/sender pair, not just the
5145 recipient address. There doesn't seem to be a real danger of callout loops,
5146 since a callout by the remote host to check the sender would use <>.
5147 [SEE ABOVE: changed after hitting problems.]
5149 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5150 temporary errors. However, in the case of such a code being given after
5151 the end of a data transmission (i.e. after ".") Exim was failing to write
5152 a retry record for the message. (Yes, there was some broken host that was
5153 actually sending 8xx at this point.)
5155 73. An unknown lookup type in a host list could cause Exim to panic-die when
5156 the list was checked. (An example that provoked this was putting <; in the
5157 middle of a list instead of at the start.) If this happened during a DATA
5158 ACL check, a -D file could be left lying around. This kind of configuration
5159 error no longer causes Exim to die; instead it causes a defer errror. The
5160 incident is still logged to the main and panic logs.
5162 74. Buglet left over from Exim 3 conversion. The message "too many messages
5163 in one connection" was written to the rejectlog but not the mainlog, except
5164 when address rewriting (yes!) was being logged.
5166 75. Added write_rejectlog option.
5168 76. When a system filter was run not as root (that is, when system_filter_user
5169 was set), the values of the $n variables were not being returned to the
5170 main process; thus, they were not subsequently available in the $sn
5173 77. Added +return_path_on_delivery log selector.
5175 78. A connection timeout was being treated differently from recipients deferred
5176 when testing hosts_max_try with a message that was older than the host's
5177 retry timeout. (The host should not be counted, thus allowing all hosts to
5178 be tried at least once before bouncing.) This may have been the cause of an
5179 occasionally reported bug whereby a message would remain on the queue
5180 longer than the retry timeout, but would be bounced if a delivery was
5181 forced. I say "may" because I never totally pinned down the problem;
5182 setting up timeout/retry tests is difficult. See also the next item.
5184 79. The ultimate address timeout was not being applied to errors that involved
5185 a combination of host plus message (for example, a timeout on a MAIL
5186 command). When an address resolved to a number of possible hosts, and they
5187 were not all tried for each delivery (e.g. because of hosts_max_try), a
5188 message could remain on the queue longer than the retry timeout.
5190 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5193 81. Fixed an obscure SMTP outgoing bug which required at least the following
5194 conditions: (a) there was another message waiting for the same server;
5195 (b) the server returned 5xx to all RCPT commands in the first message so
5196 that the message was not completed; (c) the server dropped the connection
5197 or gave a negative response to the RSET that Exim sends to abort the
5198 transaction. The observed case was a dropped connection after DATA that had
5199 been sent in pipelining mode. That is, the server had advertised PIPELINING
5200 but was not implementing it correctly. The effect of the bug was incorrect
5201 behaviour, such as trying another host, and this could lead to a crash.
5207 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5208 and daemon.c were passed as pointers to ints; they should have been
5209 pointers to socklen_t variables (which are typically unsigned ints).
5211 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5214 3. Fixed a really odd bug that affected only the testing scheme; patching a
5215 certain fixed string in the binary changed the value of another string that
5216 happened to be identical to the end of the original first string.
5218 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5219 name", it returns that address as the IP address. On some operating
5220 systems (e.g. Solaris), it also passes back the IP address string as the
5221 "host name". However, on others (e.g. Linux), it passes back an empty
5222 string. Exim wasn't checking for this, and was changing the host name to an
5223 empty string, assuming it had been canonicized.
5225 5. Although rare, it is permitted to have more than one PTR record for a given
5226 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5227 all the names associated with an address, because they do in Solaris.
5228 However, it seems that they do not in Linux for data that comes from the
5229 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5230 I found this out when I moved to a new Linux workstation and tried to run
5231 the Exim test suite.
5233 To get round this problem I have changed the code so that it now does its
5234 own call to the DNS to look up PTR records when searching for a host name.
5235 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5236 addresses that are only in /etc/hosts are still found.
5238 This behaviour is, however, controlled by an option called host_lookup_
5239 order, which defaults to "bydns:byaddr". If people want to use the other
5240 order, or indeed, just use one or the other means of lookup, they can
5241 specify it in this variable.
5243 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5244 some operating systems, this comes back from gethostbyaddr() as an empty
5245 string, and this is what Exim used to test for. However, it seems that in
5246 other systems, "." is yielded. Exim now tests for this case too.
5248 7. The values of check_spool_space and check_log_space are now held internally
5249 as a number of kilobytes instead of an absolute number of bytes. If a
5250 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5251 kilobyte. This means that much larger values can be stored.
5253 8. Exim monitor: an attempt to get the action menu when not actually pointing
5254 at a message produces an empty menu entitled "No message selected". This
5255 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5256 no entries in it ("Shell widget menu has zero width and/or height"). So I
5257 have added a single, blank menu entry in this case.
5259 9. Added ${quote_local_part.
5261 10. MIME decoding is now applied to the contents of Subject: header lines when
5264 11. Now that a reference to $sender_host_address automatically causes a reverse
5265 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5266 host lookup before query-style lookups in lists that might use this
5267 variable. This has therefore been abolished, and the "net-" prefix is no
5268 longer necessary for query-style lookups.
5270 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5271 have been a typo for LFLAGS, so it has been changed.
5273 13. The install script calls Exim with "-C /dev/null" in order to find the
5274 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5275 to be output. Howeve, since Exim outputs its version number before the
5276 error, it didn't break the script. It just looked ugly. I fixed this by
5277 always allowing "-C /dev/null" if the caller is root.
5279 14. Ignore overlarge ACL variable number when reading spool file - insurance
5280 against a later release with more variables having written the file.
5282 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5283 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5284 this, as well as the form without the "IPv6" on the front.
5286 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5287 OS/Makefile-Darwin file.
5289 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5290 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5292 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5293 variable. However, some OS use ints for inodes. I've added cast to long int
5294 to get rid of the compiler warning.
5296 19. I had forgotten to lock out "/../" in configuration file names when
5297 ALT_CONFIG_PREFIX was set.
5299 20. Routers used for verification do not need to specify transports. However,
5300 if such a router generated a host list, and callout was configured, Exim
5301 crashed, because it could not find a port number from the (non-existent)
5302 transport. It now assumes port 25 in this circumstance.
5304 21. Added the -t option to exigrep.
5306 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5307 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5308 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5309 them is set, LOOKUP_LSEARCH is forced.
5311 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5313 24. Added sender and host information to the "rejected by local_scan()" log
5314 line; previously there was no indication of these.
5316 25. Added .include_if_exists.
5318 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5319 Linux. It turns out that not all file systems support this. Apparently some
5320 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5321 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5322 not supported on the file descriptor, is now ignored when Exim is trying to
5323 sync a directory. This applies only to Linux.
5325 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5327 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5328 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5331 29. Added support for maildirsize files from supplied patch (modified a bit).
5333 30. The use of :fail: followed by an empty string could lead Exim to respond to
5334 sender verification failures with (e.g.):
5336 550 Verification failed for <xxx>
5337 550 Sender verify failed
5339 where the first response line was missing the '-' that indicates it is not
5340 the final line of the response.
5342 31. The loop for finding the name of the user that called Exim had a hardwired
5343 limit of 10; it now uses the value of finduser_retries, which is used for
5344 all other user lookups.
5346 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5349 33. Exim was neglecting to zero errno before one call of strtol() when
5350 expanding a string and expecting an integer value. On some systems this
5351 resulted in spurious "integer overflow" errors. Also, it was casting the
5352 result into an int without checking.
5354 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5355 did not work. The code looks as if it has *never* worked, though it appears
5356 to have been documented since at least releast 1.62. I have made it work.
5358 35. The "timeout_DNS" error in retry rules, also documented since at least
5359 1.62, also never worked. As it isn't clear exactly what this means, and
5360 clearly it isn't a major issue, I have abolished the feature by treating it
5361 as "timeout", and writing a warning to the main and panic logs.
5363 36. The display of retry rules for -brt wasn't always showing the error code
5366 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5367 timeout_connect_A, timeout_connect_MX.
5369 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5370 to the empty sender.
5372 39. The daemon was not analysing the content of -oX till after it had closed
5373 stderr and disconnected from the controlling terminal. This meant that any
5374 syntax errors were only noted on the panic log, and the return code from
5375 the command was 0. By re-arranging the code a little, I've made the
5376 decoding happen first, so such errors now appear on stderr, and the return
5377 code is 1. However, the actual setting up of the sockets still happens in
5378 the disconnected process, so errors there are still only recorded on the
5381 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5382 connections (as happens on some IP stacks) was logged at start up time as
5383 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5384 it from "IPv6 and IPv4", which means that two separate sockets are being
5387 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5388 says whether AF_INET or AF_INET6 was passed as an argument.
5390 42. Exiwhat output was messed up when time zones were included in log
5393 43. Exiwhat now gives more information about the daemon's listening ports,
5394 and whether -tls-on-connect was used.
5396 44. The "port" option of the smtp transport is now expanded.
5398 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5399 silently ignored. Now an error message is written to the main and panic
5402 46. There's a new ACL modifier called "logwrite" which writes to a log file
5403 as soon as it is encountered.
5405 47. Added $local_user_uid and $local_user_gid at routing time.
5407 48. Exim crashed when trying to verify a sender address that was being
5410 49. Exim was recognizing only a space character after ".include". It now also
5411 recognizes a tab character.
5413 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5414 extracting the relevant information from the specification. The man page no
5415 longer contains scrambled data for the -d option, and I've added a section
5416 at the front about calling Exim under different names.
5418 51. Added "extra_headers" argument to the "mail" command in filter files.
5420 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5423 53. Installed eximstats 1.29.
5425 54. Added transport_filter_timeout as a generic transport option.
5427 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5428 Cc: header lines. This was required by RFC 822, but it not required by RFC
5431 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5432 incoming messages that did not have them. Now it does so only if the
5433 message originates locally, that is, if there is no associated remote host
5434 address. When Resent- header lines are present, this applies to the Resent-
5435 lines rather than the non-Resent- lines.
5437 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5438 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5440 58. Messages for configuration errors now include the name of the main
5441 configuration file - useful now that there may be more than one file in a
5442 list (.included file names were always shown).
5444 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5445 for those rare installations that do not start the daemon as root or run it
5446 setuid root. I've cut out the call to initgroups() if the daemon is not
5449 60. The Exim user and group can now be bound into the binary as text strings
5450 that are looked up at the start of Exim's processing.
5452 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5454 62. Added $mailstore_basename variable.
5456 63. Installed patch to sieve.c from Michael Haardt.
5458 64. When Exim failed to open the panic log after failing to open the main log,
5459 the original message it was trying to log was written to stderr and debug
5460 output, but if they were not available (the usual case in production), it
5461 was lost. Now it is written to syslog before the two lines that record the
5462 failures to open the logs.
5464 65. Users' Exim filters run in subprocesses under the user's uid. It is
5465 possible for a "deliver" command or an alias in a "personal" command to
5466 provoke an address rewrite. If logging of address rewriting is configured,
5467 this fails because the process is not running as root or exim. There may be
5468 a better way of dealing with this, but for the moment (because 4.30 needs
5469 to be released), I have disabled address rewrite logging when running a
5470 filter in a non-root, non-exim process.
5476 1. The buildconfig auxiliary program wasn't quoting the value set for
5477 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5478 not defined. This bug was masked in 4.22 by the effect that was fixed in
5481 2. Some messages that were rejected after a message id was allocated were
5482 shown as "incomplete" by exigrep. It no longer does this for messages that
5483 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5485 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5486 have allow_domain_literals set, the ID did not get logged in the <= line.
5487 Domain literals are now always recognized in Message-ID: header lines.
5489 4. The first argument for a ${extract expansion item is the key name or field
5490 number. Leading and trailing spaces in this item were not being ignored,
5491 causing some misleading effects.
5493 5. When deliver_drop_privilege was set, single queue runner processes started
5494 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5495 same command in the process it spins off) were not dropping privilege.
5497 6. When the daemon running as "exim" started a queue runner, it always
5498 re-executed Exim in the spun-off process. This is a waste of effort when
5499 deliver_drop_privilege is set. The new process now just calls the
5500 queue-runner function directly.
5506 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5509 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5510 ignored. Though the use of -oP was forcing the writing of a pid file, it
5511 was always written to the default place.
5513 3. If the message "no IP address found for host xxxx" is generated during
5514 incoming verification, it is now followed by identification of the incoming
5515 connection (so you can more easily find what provoked it).
5517 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5519 5. Added some features to "harden" Exim a bit more against certain attacks:
5521 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5522 be put in Local/Makefile. This is like the never_users runtime option,
5523 but it cannot be overridden. The default setting is "root".
5525 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5526 prefix string with which any file named in a -C command line option
5529 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5530 is retained for -C and -D only if the caller of Exim is root. Without
5531 it, the exim user may also use -C and -D and retain privilege.
5533 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5534 command line option is disabled.
5536 6. Macro names set by the -D option must start with an upper case letter, just
5537 like macro names defined in the configuration file.
5539 7. Added "dereference=" facility to LDAP.
5541 8. Two instances of the typo "uknown" in the source files are fixed.
5543 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5544 the Configure-Makefile script screwed up while processing it.
5546 10. Incorporated PCRE 4.4.
5548 11. The SMTP synchronization check was not operating right at the start of an
5549 SMTP session. For example, it could not catch a HELO sent before the client
5550 waited for the greeting. There is now a check for outstanding input at the
5551 point when the greeting is written. Because of the duplex, asynchronous
5552 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5553 way, but not yet received, when the check is performed.
5555 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5556 on TCP/IP sockets, because this apparently causes some broken clients to
5559 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5560 unchanged) from the Cygwin maintainer.
5562 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5563 was supported instead of testing for mbx. Effectively a typo.
5565 15. The spa authenticator server code was not checking that the input it
5566 received was valid base64.
5568 16. The debug output line for the "set" modifier in ACLs was not showing the
5569 name of the variable that was being set.
5571 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5573 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5574 lookup on its own. Something else was needed to trigger the lookup. For
5575 example, a match in host_lookup or the need for a host name in a host list.
5576 Now, if $sender_host_name is referenced and the host name has not yet been
5577 looked up, a lookup is performed. If the lookup fails, the variable remains
5578 empty, and $host_lookup_failed is set to "1".
5580 19. Added "eqi" as a case-independent comparison operator.
5582 20. The saslauthd authentication condition could segfault if neither service
5583 nor realm was specified.
5585 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5586 error message that was logged was misleading, and incoming SMTP
5587 connections were dropped. The message is now more accurate, and temporary
5588 errors are given to SMTP connections.
5590 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5591 (except RSET) with a 421 error, until QUIT is received. However, it was
5592 failing to send a response to QUIT.
5594 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5595 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5596 ACL is now run after the helo_try_verify_hosts code.
5598 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5599 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5600 case-independent, so other case variants are also recognized). Apparently
5601 some systems use these upper case variants.
5603 25. If more than two messages were waiting for the same host, and a transport
5604 filter was specified for the transport, Exim sent two messages over the
5605 same TCP/IP connection, and then failed with "socket operation on non-
5606 socket" when it tried to send the third.
5608 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5610 27. The extern definition of crypt16() in expand.c was not being excluded when
5611 the OS had its own crypt16() function.
5613 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5614 as a preferred synonym for return_size_limit, both as an option and as an
5617 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5619 30. Changed the default configuration ACL to relax the local part checking rule
5620 for addresses that are not in any local domains. For these addresses,
5621 slashes and pipe symbols are allowed within local parts, but the sequence
5622 /../ is explicitly forbidden.
5624 31. SPA server authentication was not clearing the challenge buffer before
5627 32. log_message in a "warn" ACL statement was writing to the reject log as
5628 well as to the main log, which contradicts the documentation and doesn't
5629 seem right (because no rejection is happening). So I have stopped it.
5631 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5632 However, I am unable to do any testing of this.
5634 34. Fixed an infelicity in the appendfile transport. When checking directories
5635 for a mailbox, to see if any needed to be created, it was accidentally
5636 using path names with one or more superfluous leading slashes; tracing
5637 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5639 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5640 contents are added to the log line that is written for every discarded
5641 recipient. (Previously a log_message setting was ignored.)
5643 36. The ${quote: operator now quotes the string if it is empty.
5645 37. The install script runs exim in order to find its version number. If for
5646 some reason other than non-existence or emptiness, which it checks, it
5647 could not run './exim', it was installing it with an empty version number,
5648 i.e. as "exim-". This error state is now caught, and the installation is
5651 38. An argument was missing from the function that creates an error message
5652 when Exim fails to connect to the socket for saslauthd authentication.
5653 This could cause Exim to crash, or give a corrupted message.
5655 39. Added isip, isip4, and isip6 to ${if conditions.
5657 40. The ACL variables $acl_xx are now saved with the message, and can be
5658 accessed later in routers, transports, and filters.
5660 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5661 strings in the file are not string-expanded.
5663 42. If a MAIL command specified a SIZE value that was too large to fit into an
5664 int variable, the check against message_size_limit failed. Such values are
5665 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5666 day this will have to be increased, but I don't think I want to be around
5667 when emails are that large.
5674 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5675 iconv() is not standard in FreeBSD.
5677 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5678 IPv6 enabled. The observed symptom was a segmentation fault on return from
5679 the function os_common_find_running_interfaces() in src/os.c.
5681 3. In the check_special_case() function in daemon.c I had used "errno" as an
5682 argument name, which causes warnings on some systems. This was basically a
5683 typo, since it was named "eno" in the comments!
5685 4. The code that waits for the clock to tick (at a resolution of some fraction
5686 of a second) so as to ensure message-id uniqueness was always waiting for
5687 at least one whole tick, when it could have waited for less. [This is
5688 almost certainly not relevant at current processor speeds, where it is
5689 unlikely to ever wait at all. But we try to future-proof.]
5691 5. The function that sleeps for a time interval that includes fractions of a
5692 second contained a race. It did not block SIGALRM between setting the
5693 timer, and suspending (a couple of lines later). If the interval was short
5694 and the sigsuspend() was delayed until after it had expired, the suspension
5695 never ended. On busy systems this could lead to processes getting stuck for
5698 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5699 process, before it forks any delivery processes. The open lookup caching
5700 mechanism meant that the open file or database connection was passed into
5701 the delivery process. The problem was that delivery processes always tidy
5702 up cached lookup data. This could cause a problem for the next delivery
5703 process started by the queue runner, because the external queue runner
5704 process does not know about the closure. So the next delivery process
5705 still has data in the lookup cache. In the case of a file lookup, there was
5706 no problem because closing a file descriptor in a subprocess doesn't affect
5707 the parent. However, if the lookup was caching a connection to a database,
5708 the connection was closed, and the second delivery process was likely to
5709 see errors such as "PGSQL: query failed: server closed the connection
5710 unexpectedly". The problem has been fixed by closing all cached lookups
5711 in a queue runner before running a delivery process.
5713 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5714 seem to have the "const" qualifier which it has on other OS. I've
5717 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5718 *of the same type* (client or server) with the same public name that an
5719 error should be diagnosed.
5721 9. When Exim looked up a host name for an IP address, but failed to find the
5722 original IP address when looking up the host name (a safety check), it
5723 output the message "<ip address> does not match any IP for NULL", which was
5724 confusing, to say the least. The bug was that the host name should have
5725 appeared instead of "NULL".
5727 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5728 user that is built into the binary, and the -C or -D options is used, root
5729 privilege is dropped before the configuration file is read. In addition,
5730 logging is switched to stderr instead of the normal log files. If the
5731 configuration then re-defines the Exim user, the unprivileged environment
5732 is probably not what is expected, so Exim logs a panic warning message (but
5735 However, if deliver_drop_privilege is set, the unprivileged state may well
5736 be exactly what is intended, so the warning has been cut out in that case,
5737 and Exim is allowed to try to write to its normal log files.
5743 1. smtp_return_error_details was not giving details for temporary sender
5744 or receiver verification errors.
5746 2. Diagnose a configuration error if two authenticators have the same public
5749 3. Exim used not to create the message log file for a message until the first
5750 delivery attempt. This could be confusing when incoming messages were held
5751 for policy or load reasons. The message log file is now created at the time
5752 the message is received, and an initial "Received" line is written to it.
5754 4. The automatically generated man page for command line options had a minor
5755 bug that caused no ill effects; however, a more serious problem was that
5756 the procedure for building the man page automatically didn't always
5757 operate. Consequently, release 4.20 contains an out-of-date version. This
5758 shouldn't happen again.
5760 5. When building Exim with embedded Perl support, the script that builds the
5761 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5762 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5764 6. The freeze_tell option was not being used for messages that were frozen on
5765 arrival, either by an ACL or by local_scan().
5767 7. Added the smtp_incomplete_transaction log selector.
5769 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5770 was accepting AUTH without a new EHLO.
5772 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5773 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5776 10. Logging of TCP/IP connections (when configured) now happens in the main
5777 daemon process instead of the child process, so that the TCP/IP connection
5778 count is more accurate (but it can never be perfect).
5780 11. The use of "drop" in a nested ACL was not being handled correctly in the
5781 outer ACL. Now, if condition failure induced by the nested "drop" causes
5782 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5783 or "require"), the connection is dropped.
5785 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5786 that yield "discard" can now be used with an "accept" or a "discard" verb,
5787 but an error is generated for any others (because I can't see a useful way
5788 to define what should happen).
5790 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5791 are now processed in the same way as lines in the Exim configuration file.
5792 In particular, continuation lines are supported.
5794 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5796 15. Added -ti meaning -t -i.
5798 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5799 domains, and warn by logging if others are found.
5801 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5802 returned from GIFCONF when Exim is trying to find the list of interfaces on
5803 a host. The code in os.c has been modified to copy each ifreq to an aligned
5804 structure in all cases.
5806 Also, in some cases, the returned ifreq's were being copied to a 'struct
5807 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5808 means the last couple of bytes of an IPv6 address could be chopped if the
5809 ifreq contained only a normal sockaddr (14 bytes storage).
5811 18. Named domain lists were not supported in the hosts_treat_as_local option.
5812 An entry such as +xxxx was not recognized, and was treated as a literal
5815 19. Ensure that header lines added by a DATA ACL are included in the reject log
5816 if the ACL subsequently rejects the message.
5818 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5819 MD5 (which is deprecated).
5821 21. When testing a filter file using -bf, Exim was writing a message when it
5822 took the sender from a "From " line in the message, but it was not doing so
5823 when it took $return_path from a Return-Path: header line. It now does.
5825 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5826 with a valid header line field name (a series of printing characters
5827 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5829 23. Changed "disc" in the source to "disk" to conform to the documentation and
5830 the book and for uniformity.
5832 24. Ignore Sendmail's -Ooption=value command line item.
5834 25. When execve() failed while trying to run a command in a pipe transport,
5835 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5836 could be confused with a return value of 69 from the command itself. This
5837 has been changed to 127, the value the shell returns if it is asked to run
5838 a non-existent command. The wording for the related log line suggests a
5839 non-existent command as the problem.
5841 26. If received_header_text expands to an empty string, do not add a Received:
5842 header line to the message. (Well, it adds a token one on the spool, but
5843 marks it "old" so that it doesn't get used or transmitted.)
5845 27. Installed eximstats 1.28 (addition of -nt option).
5847 28. There was no check for failure on the call to getsockname() in the daemon
5848 code. This can fail if there is a shortage of resources on the system, with
5849 ENOMEM, for example. A temporary error is now given on failure.
5851 29. Contrary to the C standard, it seems that in some environments, the
5852 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5853 program. Exim now does this explicitly; it affects the formatting of
5854 timestamps using strftime().
5856 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5857 complaints. I've now initialized all the variables, to avoid this.
5859 32. Header lines added by a system filter were not being "seen" during
5860 transport-time rewrites.
5862 33. The info_callback() function passed to OpenSSL is set up with type void
5863 (*)(SSL *, int, int), as described somewhere. However, when calling the
5864 function (actually a macro) that sets it up, the type void(*)() is
5865 expected. I've put in a cast to prevent warnings from picky compilers.
5867 34. If a DNS black list lookup found a CNAME record, but there were no A
5868 records associated with the domain it pointed at, Exim crashed.
5870 35. If a DNS black list lookup returned more than one A record, Exim ignored
5871 all but the first. It now scans all returned addresses if a particular IP
5872 value is being sought. In this situation, the contents of the
5873 $dnslist_value variable are a list of all the addresses, separated by a
5876 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5877 to accept a host name and all its aliases if the forward lookup for any of
5878 them yielded the IP address of the incoming connection. Now it accepts only
5879 those names whose forward lookup yields the correct IP address. Any other
5880 names are discarded. This closes a loophole whereby a rogue DNS
5881 administrator could create reverse DNS records to break through a
5882 wildcarded host restriction in an ACL.
5884 37. If a user filter or a system filter that ran in a subprocess used any of
5885 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5886 the wrong values were passed to the pipe command ($thisaddress had the
5887 value of $0, $0 had the value of $1, etc). This bug was introduced by
5888 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5890 38. Improved the line breaking for long SMTP error messages from ACLs.
5891 Previously, if there was no break point between 40 and 75 characters, Exim
5892 left the rest of the message alone. Two changes have been made: (a) I've
5893 reduced the minimum length to 35 characters; (b) if it can't find a break
5894 point between 35 and 75 characters, it looks ahead and uses the first one
5895 that it finds. This may give the occasional overlong line, but at least the
5896 remaining text gets split now.
5898 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5899 file descriptors might be low, and that setting 1000 would always raise it.
5900 It turns out that in some environments, the limit is already over 1000 and
5901 that lowering it causes trouble. So now Exim takes care not to decrease it.
5903 40. When delivering a message, the value of $return_path is set to $sender_
5904 address at the start of routing (routers may change the value). By an
5905 oversight, this default was not being set up when an address was tested by
5906 -bt or -bv, which affected the outcome if any router or filter referred to
5909 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5910 log only when "message" or "log_message" are set. However, if one of the
5911 conditions was an address verification, or a call to a nested ACL, the
5912 messages generated by the underlying test were being passed through. This
5913 no longer happens. The underlying message is available in $acl_verify_
5914 message for both "message" and "log_message" expansions, so it can be
5915 passed through if needed.
5917 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5918 new expansion $bh_ to give the encoded byte string without charset
5919 translation. Translation happens only if iconv() is available; HAVE_ICONV
5920 indicates this at build time. HEADERS_CHARSET gives the charset to
5921 translate to; headers_charset can change it in the configuration, and
5922 "headers charset" can change it in an individual filter file.
5924 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5925 that creates RFC 2047 encoded "words" labels them as that charset instead
5926 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5927 expansion operator; (ii) when Exim creates a From: line for a local
5928 message; (iii) when a header line is rewritten to include a "phrase" part.
5930 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5931 buggy, causing it to skip the first lines of messages whose message ID
5932 ended in 'D'. This would not have bitten before Exim release 4.14, because
5933 message IDs were unlikely to end in 'D' before then. The effect was to have
5934 incorrect size information for certain domains.
5936 45. #include "config.h" was missing at the start of the crypt16.c module. This
5937 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5940 46. If there was a timeout during a "random" callout check, Exim treated it as
5941 a failure of the random address, and carried on sending RSET and the real
5942 address. If the delay was just some slowness somewhere, the response to the
5943 original RCPT would be taken as a response to RSET and so on, causing
5944 mayhem of various kinds.
5946 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5947 when I implemented it. It didn't allow for the fact that some option values
5948 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5949 the right test for positive values.
5951 48. Domain names in DNS records are case-independent. Exim always looks them up
5952 in lower case. Some resolvers return domain names in exactly the case they
5953 appear in the zone file, that is, they may contain uppercase letters. Not
5954 all resolvers do this - some return always lower case. Exim was treating a
5955 change of case by a resolver as a change of domain, similar to a widening
5956 of a domain abbreviation. This triggered its re-routing code and so it was
5957 trying to route what was effectively the same domain again. This normally
5958 caused routing to fail (because the router wouldn't handle the domain
5959 twice). Now Exim checks for this case specially, and just changes the
5960 casing of the domain that it ultimately uses when it transmits the message
5963 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5966 50. If a filter generated a file delivery with a non-absolute name (possible if
5967 no home directory exists for the router), the forbid_file option was not
5970 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5971 the existing equality matching.
5973 52. Exim was using ints instead of ino_t variables in some places where it was
5974 dealing with inode numbers.
5976 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5977 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5978 called TMPDIR, and if it finds it is different, it changes its value.
5980 54. The smtp_printf() function is now made available to local_scan() so
5981 additional output lines can be written before returning. There is also an
5982 smtp_fflush() function to enable the detection of a dropped connection.
5983 The variables smtp_input and smtp_batched_input are exported to
5986 55. Changed the default runtime configuration: the message "Unknown user"
5987 has been removed from the ACL, and instead placed on the localuser router,
5988 using the cannot_route_message feature. This means that any verification
5989 failures that generate their own messages won't get overridden. Similarly,
5990 the "Unrouteable address" message that was in the ACL for unverifiable
5991 relay addresses has also been removed.
5993 56. Added hosts_avoid_esmtp to the smtp transport.
5995 57. The exicyclog script was not checking for the esoteric option
5996 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5997 will work only if exicyclog is run under the appropriate euid.
5999 58. Following a discussion on the list, the rules by which Exim recognises line
6000 endings on incoming messages have been changed. The -dropcr and drop_cr
6001 options are now no-ops, retained only for backwards compatibility. The
6002 following line terminators are recognized: LF CRLF CR. However, special
6003 processing applies to CR:
6005 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6006 nor a local message in the state where . is a terminator.
6008 (ii) If a bare CR is encountered in a header line, an extra space is added
6009 after the line terminator so as not to end the header. The reasoning
6010 behind this is that bare CRs in header lines are most likely either
6011 to be mistakes, or people trying to play silly games.
6013 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6014 was being incorrectly given as 18 bytes larger than it should have been.
6015 This is a VOB (very old bug).
6017 60. This may never have affected anything current, but just in case it has:
6018 When the local host is found other than at the start of a list of hosts,
6019 the local host, those with the same MX, and any that follow, are discarded.
6020 When the list in question was part of a longer list of hosts, the following
6021 hosts (not currently being processed) were also being discarded. This no
6022 longer happens. I'm not sure if this situation could ever has previously
6025 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6028 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6029 header. This is something that is recommended in a new Internet Draft, and
6030 is something that is documented as being done by Sendmail. There are two
6031 possible values. For messages generated by the autoreply transport, Exim
6034 Auto-Submitted: auto-replied
6036 whereas for all other generated messages (e.g. bounces) it adds
6038 Auto-Submitted: auto-generated
6040 63. The "personal" condition in filters now includes a test for the
6041 Auto-Submitted: header. If it contains the string "auto-" the message it
6042 not considered personal.
6044 64. Added rcpt_include_affixes as a generic transport option.
6046 65. Added queue_only_override (default true).
6048 66. Added the syslog_duplication option.
6050 67. If what should have been the first header line of a message consisted of
6051 a space followed by a colon, Exim was mis-interpreting it as a header line.
6052 It isn't of course - it is syntactically invalid and should therefore be
6053 treated as the start of the message body. The misbehaviour could have
6054 caused a number of strange effects, including loss of data in subsequent
6055 header lines, and spool format errors.
6057 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6058 client host had authenticated. This control can now be exercised by an ACL
6059 for more flexibility.
6061 69. By default, callouts do not happen when testing with -bh. There is now a
6062 variant, -bhc, which does actually run the callout code, including
6063 consulting and updating the callout cache.
6065 70. Added support for saslauthd authentication, courtesy of Alexander
6068 71. If statvfs() failed on the spool or log directories while checking their
6069 size for availability, Exim confusingly gave the error "space shortage".
6070 Furthermore, in debugging mode it crashed with a floating point exception.
6071 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6072 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6073 serious problem, Exim now writes to the main and panic logs when this
6074 happens, with details of the failure. It then refuses to accept the
6075 incoming message, giving the message "spool directory problem" or "log
6076 directory problem" with a 421 code for SMTP messages.
6078 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6079 0, 1, and 2 exist, because some OS complain for execs without them (see
6080 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6081 descriptors. However, the code omitted to check that the open succeeded,
6082 causing mysterious errors if for some reason the permissions on /dev/null
6083 got screwed. Now Exim writes a message to the main and panic logs, and
6084 bombs out if it can't open /dev/null.
6086 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6087 interact so that it is all more flexible. It is supposed to remain
6088 backwards compatible. Also added extra_local_interfaces.
6090 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6091 to bomb out with an assertion failure - to the client this appears as a
6092 connection drop. This problem occurs in the part of the code that was taken
6093 from the Samba project. Fortunately, the assertion is in a very simple
6094 function, so I have fixed this by reproducing the function inline in the
6095 one place where it is called, and arranging for authentication to fail
6096 instead of killing the process with assert().
6098 75. The SPA client code was not working when the server requested OEM rather
6099 than Unicode encoding.
6101 76. Added code to make require_files with a specific uid setting more usable in
6102 the case where statting the file as root fails - usually a non-root-mounted
6103 NFS file system. When this happens and the failure is EACCES, Exim now
6104 forks a subprocess and does the per-uid checking as the relevant uid.
6106 77. Added process_log_path.
6108 78. If log_file_path was not explicitly set, a setting of check_log_space or
6109 check_log_inodes was ignored.
6111 79. If a space check for the spool or log partitions fails, the incident is now
6112 logged. Of course, in the latter case the data may get lost...
6114 80. Added the %p formatting code to string_format() so that it can be used to
6115 print addresses in debug_print(). Adjusted all the address printing in the
6116 debugging in store.c to use %p rather than %d.
6118 81. There was a concern that a line of code in smtp_in.c could overflow a
6119 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6120 initially expressed, the concern was not well-founded, because trailing
6121 spaces are removed early. However, if the trailing spaces were followed by
6122 a NULL, they did not get removed, so the overflow was possible. Two fixes
6125 (a) I re-wrote the offending code in a cleaner fashion.
6126 (b) If an incoming SMTP command contains a NULL character, it is rejected
6129 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6130 runs initgroups(), so that if the Exim user is in any additional groups,
6131 they will be used during message reception.
6137 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob