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
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
54 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
55 is EXPERIMENTAL_EVENT, the main-configuration and transport options
56 both become "event_action", the variables become $event_name, $event_data
57 and $event_defer_errno. There is a new variable $verify_mode, usable in
58 routers, transports and related events. The tls:cert event is now also
59 raised for inbound connections, if the main configuration event_action
65 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
66 checkers that were complaining about end of non-void function with no
69 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
70 This was a regression intruduced in 4.83 by another bugfix.
72 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
74 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
75 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
81 TF/01 Correctly close the server side of TLS when forking for delivery.
83 When a message was received over SMTP with TLS, Exim failed to clear up
84 the incoming connection properly after forking off the child process to
85 deliver the message. In some situations the subsequent outgoing
86 delivery connection happened to have the same fd number as the incoming
87 connection previously had. Exim would try to use TLS and fail, logging
88 a "Bad file descriptor" error.
90 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
91 utilities have not been installed.
93 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
94 temporary space as the ACL may create new global variables.
96 TL/01 LDAP support uses per connection or global context settings, depending
97 upon the detected version of the libraries at build time.
99 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
100 to extract and use the src ip:port in logging and expansions as if it
101 were a direct connection from the outside internet. PPv2 support was
102 updated based on HAProxy spec change in May 2014.
104 JH/02 Add ${listextract {number}{list}{success}{fail}}.
106 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
107 Properly escape header and check for NULL return.
109 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
112 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
114 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
115 characters in header names, implemented as a verify condition.
116 Contributed by Michael Fischer v. Mollard.
118 TL/05 Rename SPF condition results err_perm and err_temp to standardized
119 results permerror and temperror. Previous values are deprecated but
120 still accepted. In a future release, err_perm and err_temp will be
121 completely removed, which will be a backward incompatibility if the
122 ACL tests for either of these two old results. Patch contributed by
123 user bes-internal on the mailing list.
125 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
127 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
128 selectors, in both main and reject logs.
130 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
133 JH/07 Add malware type "sock" for talking to simple daemon.
135 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
137 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
138 routers/transports under cutthrough routing.
140 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
141 numbers. Touch up "bool" conditional to keep the same definition.
143 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
145 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
147 JH/12 Expand items in router/transport headers_add or headers_remove lists
148 individually rather than the list as a whole. Bug 1452.
150 Required for reasonable handling of multiple headers_ options when
151 they may be empty; requires that headers_remove items with embedded
152 colons must have them doubled (or the list-separator changed).
154 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
155 view the policy declared in the DMARC record. Currently, $dmarc_status
156 is a combined value of both the record presence and the result of the
159 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
161 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
162 dnslookup router and the smtp transport (applying to the forward
165 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
166 of ldap servers used for a specific lookup. Patch provided by Heiko
169 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
170 New variable $lookup_dnssec_authenticated for observability.
172 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
173 Patch submitted by Lars Timman.
175 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
177 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
178 Requires trusted mode and valid format message id, aborts otherwise.
179 Patch contributed by Heiko Schlichting.
181 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
182 certextract with support for various fields. Bug 1358.
184 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
185 is requested by default, modifiable by smtp transport option
188 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
189 operate on certificate variables to give certificate fingerprints
190 Also new ${sha256:cert_variable}.
192 JH/23 The PRDR feature is moved from being Experimental into the mainline.
194 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
195 Christian Aistleitner.
197 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
199 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
200 file. Patch from Wolfgang Breyha.
202 JH/25 Expand the coverage of the delivery $host and $host_address to
203 client authenticators run in verify callout. Bug 1476.
205 JH/26 Port service names are now accepted for tls_on_connect_ports, to
206 align with daemon_smtp_ports. Bug 72.
208 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
209 support and error reporting did not work properly.
211 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
212 and is readable. Patch from Andrew Colin Kissa.
214 TL/14 Enhance documentation of ${run expansion and how it parses the
215 commandline after expansion, particularly in the case when an
216 unquoted variable expansion results in an empty value.
218 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
220 PP/02 Fix internal collision of T_APL on systems which support RFC3123
221 by renaming away from it. Addresses GH issue 15, reported by
224 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
226 TL/15 SECURITY: prevent double expansion in math comparison functions
227 (can expand unsanitized data). Not remotely exploitable.
234 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
236 PP/02 Make -n do something, by making it not do something.
237 When combined with -bP, the name of an option is not output.
239 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
242 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
243 $sender_host_name and config options to manage this, and basic check
246 PP/05 DSCP support for outbound connections and control modifier for inbound.
248 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
249 (Only plugin which currently uses this is kerberos4, which nobody should
250 be using, but we should make it available and other future plugins might
251 conceivably use it, even though it would break NAT; stuff *should* be
252 using channel bindings instead).
254 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
255 name; added for Sendmail compatibility; requires admin caller.
256 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
257 just ignore it); requires trusted caller.
258 Also parse but ignore: -Ac -Am -X<logfile>
261 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
263 TL/02 Add +smtp_confirmation as a default logging option.
265 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
266 Patch by Magnus Holmgren from 2007-02-20.
268 TL/04 Bugzilla 1281 - Spec typo.
269 Bugzilla 1283 - Spec typo.
270 Bugzilla 1290 - Spec grammar fixes.
272 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
274 TL/06 Add Experimental DMARC support using libopendmarc libraries.
276 TL/07 Fix an out of order global option causing a segfault. Reported to dev
277 mailing list by by Dmitry Isaikin.
279 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
281 JH/02 Support "G" suffix to numbers in ${if comparisons.
283 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
285 NM/01 Bugzilla 1197 - Spec typo
286 Bugzilla 1196 - Spec examples corrections
288 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
290 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
291 gnutls_enable_pkcs11, but renamed to more accurately indicate its
294 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
295 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
297 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
298 "acl {{name}{arg}...}", and optional args on acl condition
301 JH/05 Permit multiple router/transport headers_add/remove lines.
303 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
305 JH/07 Avoid using a waiting database for a single-message-only transport.
306 Performance patch from Paul Fisher. Bugzilla 1262.
308 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
311 JH/09 Add $headers_added variable, with content from use of ACL modifier
312 add_header (but not yet added to the message). Bugzilla 199.
314 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
315 Pulled from Bugzilla 817 by Wolfgang Breyha.
317 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
319 (nb: this is the same fix as in Exim 4.80.1)
321 JH/11 Add A= logging on delivery lines, and a client_set_id option on
324 JH/12 Add optional authenticated_sender logging to A= and a log_selector
327 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
329 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
330 advertise SMTP AUTH mechanism to us, instead of a generic
331 protocol violation error. Also, make Exim more robust to bad
332 data from the Dovecot auth socket.
334 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
336 When a queue runner is handling a message, Exim first routes the
337 recipient addresses, during which it prunes them based on the retry
338 hints database. After that it attempts to deliver the message to
339 any remaining recipients. It then updates the hints database using
342 So if a recipient address works intermittently, it can get repeatedly
343 deferred at routing time. The retry hints record remains fresh so the
344 address never reaches the final cutoff time.
346 This is a fairly common occurrence when a user is bumping up against
347 their storage quota. Exim had some logic in its local delivery code
348 to deal with this. However it did not apply to per-recipient defers
349 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
351 This change adds a proper retry rule check during routing so that the
352 final cutoff time is checked against the message's age. We only do
353 this check if there is an address retry record and there is not a
354 domain retry record; this implies that previous attempts to handle
355 the address had the retry_use_local_parts option turned on. We use
356 this as an approximation for the destination being like a local
357 delivery, as in LMTP.
359 I suspect this new check makes the old local delivery cutoff check
360 redundant, but I have not verified this so I left the code in place.
362 TF/02 Correct gecos expansion when From: is a prefix of the username.
364 Test 0254 submits a message to Exim with the header
368 When I ran the test suite under the user fanf2, Exim expanded
369 the header to contain my full name, whereas it should have added
370 a Resent-Sender: header. It erroneously treats any prefix of the
371 username as equal to the username.
373 This change corrects that bug.
375 GF/01 DCC debug and logging tidyup
376 Error conditions log to paniclog rather than rejectlog.
377 Debug lines prefixed by "DCC: " to remove any ambiguity.
379 TF/03 Avoid unnecessary rebuilds of lookup-related code.
381 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
382 Bug spotted by Jeremy Harris; was flawed since initial commit.
383 Would have resulted in OCSP responses post-SNI triggering an Exim
384 NULL dereference and crash.
386 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
388 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
389 Bug detection, analysis and fix by Samuel Thibault.
390 Bugzilla 1331, Debian bug #698092.
392 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
394 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
395 Server implementation by Todd Lyons, client by JH.
396 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
397 config variable "prdr_enable" controls whether the server
398 advertises the facility. If the client requests PRDR a new
399 acl_data_smtp_prdr ACL is called once for each recipient, after
400 the body content is received and before the acl_smtp_data ACL.
401 The client is controlled by bolth of: a hosts_try_prdr option
402 on the smtp transport, and the server advertisement.
403 Default client logging of deliveries and rejections involving
404 PRDR are flagged with the string "PRDR".
406 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
407 fclose(). Diagnosis by Todd Lyons.
409 PP/17 Update configure.default to handle IPv6 localhost better.
410 Patch by Alain Williams (plus minor tweaks).
413 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
414 This is now consistent with GnuTLS, and is now documented: the
415 previous undocumented portable approach to treating the option as
416 unset was to force an expansion failure. That still works, and
417 an empty string is now equivalent.
419 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
420 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
421 not performing validation itself.
423 PP/20 Added force_command boolean option to pipe transport.
424 Patch from Nick Koston, of cPanel Inc.
426 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
429 TF/04 Added udpsend ACL modifer and hexquote expansion operator
431 PP/21 Fix eximon continuous updating with timestamped log-files.
432 Broken in a format-string cleanup in 4.80, missed when I repaired the
433 other false fix of the same issue.
434 Report and fix from Heiko Schlichting.
437 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
438 Report from Prashanth Katuri.
440 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
441 It's SecureTransport, so affects any MacOS clients which use the
442 system-integrated TLS libraries, including email clients.
444 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
445 using a MIME ACL for non-SMTP local injection.
446 Report and assistance in diagnosis by Warren Baker.
448 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
450 JH/16 Fix comparisons for 64b. Bugzilla 1385.
452 TL/09 Add expansion variable $authenticated_fail_id to keep track of
453 last id that failed so it may be referenced in subsequent ACL's.
455 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
458 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
459 ldap library initialization, allowing self-signed CA's to be
460 used. Also properly sets require_cert option later in code by
461 using NULL (global ldap config) instead of ldap handle (per
462 session). Bug diagnosis and testing by alxgomz.
464 TL/12 Enhanced documentation in the ratelimit.pl script provided in
465 the src/util/ subdirectory.
467 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
468 renamed to Transport Post Delivery Action by Jeremy Harris, as
471 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
472 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
473 redis_servers = needs to be configured which will be used by the redis
474 lookup. Patch from Warren Baker, of The Packet Hub.
476 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
478 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
479 hostname or reverse DNS when processing a host list. Used suggestions
480 from multiple comments on this bug.
482 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
484 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
485 Missed a few lines, added it to make the runtest require no keyboard
488 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
489 contains upper case chars. Make router use caseful_local_part.
491 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
492 support when GnuTLS has been built with p11-kit.
498 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
500 This, or similar/improved, will also be change PP/11 of 4.82.
506 PP/01 Handle short writes when writing local log-files.
507 In practice, only affects FreeBSD (8 onwards).
508 Bugzilla 1053, with thanks to Dmitry Isaikin.
510 NM/01 Bugzilla 949 - Documentation tweak
512 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
515 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
517 PP/02 Implemented gsasl authenticator.
519 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
521 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
522 `pkg-config foo` for cflags/libs.
524 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
525 with rest of GSASL and with heimdal_gssapi.
527 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
528 `pkg-config foo` for cflags/libs for the TLS implementation.
530 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
531 properties get this fed in as external SSF. A number of robustness
532 and debugging improvements to the cyrus_sasl authenticator.
534 PP/08 cyrus_sasl server now expands the server_realm option.
536 PP/09 Bugzilla 1214 - Log authentication information in reject log.
537 Patch by Jeremy Harris.
539 PP/10 Added dbmjz lookup type.
541 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
543 PP/12 MAIL args handles TAB as well as SP, for better interop with
544 non-compliant senders.
545 Analysis and variant patch by Todd Lyons.
547 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
548 Bug report from Lars Müller <lars@samba.org> (via SUSE),
549 Patch from Dirk Mueller <dmueller@suse.com>
551 PP/13 tls_peerdn now print-escaped for spool files.
552 Observed some $tls_peerdn in wild which contained \n, which resulted
553 in spool file corruption.
555 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
556 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
557 or write after TLS renegotiation, which otherwise led to messages
560 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
561 as a tracking header (ie: a signed header comes before the signature).
562 Patch from Wolfgang Breyha.
564 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
565 comma-sep list; embedded commas doubled.
567 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
569 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
571 Report and patch from Dmitry Banschikov.
573 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
574 Removed SSL_clear() after SSL_new() which led to protocol negotiation
575 failures. We appear to now support TLS1.1+ with Exim.
577 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
578 lets Exim select keys and certificates based upon TLS SNI from client.
579 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
580 before an outbound SMTP session. New log_selector, +tls_sni.
582 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
583 NULL dereference. Report and patch from Alun Jones.
585 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
586 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
587 Not seeing resolver debug output on NetBSD, but suspect this is a
588 resolver implementation change.
590 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
591 Left warnings. Added "eximon gdb" invocation mode.
593 PP/21 Defaulting "accept_8bitmime" to true, not false.
595 PP/22 Added -bw for inetd wait mode support.
597 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
598 locate the relevant includes and libraries. Made this the default.
600 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
601 Bugzilla 1246, report and most of solution from Tomasz Kusy.
603 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
604 This may cause build issues on older platforms.
606 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
607 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
608 gnutls_require_mac & gnutls_require_protocols (no longer supported).
609 Added SNI support via GnuTLS too.
610 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
612 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
614 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
615 Applied second patch from Janne, implementing suggestion to default
616 multiple-strings-in-record handling to match SPF spec.
618 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
620 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
621 read-only, out of scope).
622 Patch from Wolfgang Breyha, report from Stuart Northfield.
624 PP/29 Fix three issues highlighted by clang analyser static analysis.
625 Only crash-plausible issue would require the Cambridge-specific
626 iplookup router and a misconfiguration.
627 Report from Marcin Mirosław.
629 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
631 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
632 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
633 As part of this, removing so much warning spew let me fix some minor
634 real issues in debug logging.
636 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
637 assignment on my part. Fixed.
639 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
640 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
641 Janne Snabb (who went above and beyond: thank you).
643 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
644 string otherwise requires a connection and a bunch more work and it's
645 relatively easy to get wrong. Should also expose TLS library linkage
648 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
649 64-bit ${eval} (JH/03).
651 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
652 GNU libc to support some of the 64-bit stuff, should not lead to
653 conflicts. Defined before os.h is pulled in, so if a given platform
654 needs to override this, it can.
656 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
657 protection layer was required, which is not implemented.
658 Bugzilla 1254, patch from Wolfgang Breyha.
660 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
661 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
662 tls_dhparam take prime identifiers. Also unbreak combination of
663 OpenSSL+DH_params+TLSSNI.
665 PP/39 Disable SSLv2 by default in OpenSSL support.
671 PP/01 Solaris build fix for Oracle's LDAP libraries.
672 Bugzilla 1109, patch from Stephen Usher.
674 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
676 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
679 TF/02 Fix a couple more cases where we did not log the error message
680 when unlink() failed. See also change 4.74-TF/03.
682 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
683 lock up or crash if it happened to be inside a call to libc when it
684 got a SIGUSR1 from exiwhat.
686 The SIGUSR1 handler appends the current process status to the process
687 log which is later printed by exiwhat. It used to use the general
688 purpose logging code to do this, but several functions it calls are
689 not safe for signals.
691 The new output code in the SIGUSR1 handler is specific to the process
692 log, and simple enough that it's easy to inspect for signal safety.
693 Removing some special cases also simplifies the general logging code.
694 Removing the spurious timestamps from the process log simplifies
697 TF/04 Improved ratelimit ACL condition.
699 The /noupdate option has been deprecated in favour of /readonly which
700 has clearer semantics. The /leaky, /strict, and /readonly update modes
701 are mutually exclusive. The update mode is no longer included in the
702 database key; it just determines when the database is updated. (This
703 means that when you upgrde Exim will forget old rate measurements.)
705 Exim now checks that the per_* options are used with an update mode that
706 makes sense for the current ACL. For example, when Exim is processing a
707 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
708 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
709 must specify per_mail/readonly. If you omit the update mode it defaults to
710 /leaky where that makes sense (as before) or /readonly where required.
712 The /noupdate option is now undocumented but still supported for
713 backwards compatibility. It is equivalent to /readonly except that in
714 ACLs where /readonly is required you may specify /leaky/noupdate or
715 /strict/noupdate which are treated the same as /readonly.
717 A useful new feature is the /count= option. This is a generalization
718 of the per_byte option, so that you can measure the throughput of other
719 aggregate values. For example, the per_byte option is now equivalent
720 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
722 The per_rcpt option has been generalized using the /count= mechanism
723 (though it's more complicated than the per_byte equivalence). When it is
724 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
725 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
726 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
727 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
728 Note that using per_rcpt with a non-readonly update mode in more than
729 one ACL will cause the recipients to be double-counted. (The per_mail
730 and per_byte options don't have this problem.)
732 The handling of very low rates has changed slightly. If the computed rate
733 is less than the event's count (usually one) then this event is the first
734 after a long gap. In this case the rate is set to the same as this event's
735 count, so that the first message of a spam run is counted properly.
737 The major new feature is a mechanism for counting the rate of unique
738 events. The new per_addr option counts the number of different
739 recipients that someone has sent messages to in the last time period. It
740 behaves like per_rcpt if all the recipient addresses are different, but
741 duplicate recipient addresses do not increase the measured rate. Like
742 the /count= option this is a general mechanism, so the per_addr option
743 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
744 example, measure the rate that a client uses different sender addresses
745 with the options per_mail/unique=$sender_address. There are further
746 details in the main documentation.
748 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
750 TF/06 Removed a few PCRE remnants.
752 TF/07 Automatically extract Exim's version number from tags in the git
753 repository when doing development or release builds.
755 PP/02 Raise smtp_cmd_buffer_size to 16kB.
756 Bugzilla 879. Patch from Paul Fisher.
758 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
759 Heavily based on revision 40f9a89a from Simon Arlott's tree.
762 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
764 PP/05 Variable $av_failed, true if the AV scanner deferred.
765 Bugzilla 1078. Patch from John Horne.
767 PP/06 Stop make process more reliably on build failure.
768 Bugzilla 1087. Patch from Heiko Schlittermann.
770 PP/07 Make maildir_use_size_file an _expandable_ boolean.
771 Bugzilla 1089. Patch from Heiko Schlittermann.
773 PP/08 Handle ${run} returning more data than OS pipe buffer size.
774 Bugzilla 1131. Patch from Holger Weiß.
776 PP/09 Handle IPv6 addresses with SPF.
777 Bugzilla 860. Patch from Wolfgang Breyha.
779 PP/10 GnuTLS: support TLS 1.2 & 1.1.
781 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
784 PP/11 match_* no longer expand right-hand-side by default.
785 New compile-time build option, EXPAND_LISTMATCH_RHS.
786 New expansion conditions, "inlist", "inlisti".
788 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
790 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
792 PP/14 fix log_write() format string regression from TF/03.
793 Bugzilla 1152. Patch from Dmitry Isaikin.
799 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
801 PP/02 Harmonised TLS library version reporting; only show if debugging.
802 Layout now matches that introduced for other libraries in 4.74 PP/03.
804 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
806 PP/04 New "dns_use_edns0" global option.
808 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
811 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
812 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
814 TK/01 Updated PolarSSL code to 0.14.2.
815 Bugzilla 1097. Patch from Andreas Metzler.
817 PP/07 Catch divide-by-zero in ${eval:...}.
820 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
823 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
824 format-string attack -- SECURITY: remote arbitrary code execution.
826 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
827 time unintentionally subject to list matching rules, letting the header
828 cause arbitrary Exim lookups (of items which can occur in lists, *not*
829 arbitrary string expansion). This allowed for information disclosure.
831 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
832 INT_MIN/-1 -- value coerced to INT_MAX.
838 NM/01 Workround for PCRE version dependency in version reporting
841 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
842 This fixes portability to compilers other than gcc, notably
843 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
845 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
846 makefiles for portability to HP-UX and POSIX correctness.
848 PP/01 Permit LOOKUP_foo enabling on the make command-line.
849 Also via indented variable definition in the Makefile.
850 (Debugging by Oliver Heesakkers).
852 PP/02 Restore caching of spamd results with expanded spamd_address.
853 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
855 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
856 Improves build reliability. Fix from: Frank Elsner
858 NM/02 Fix wide character breakage in the rfc2047 coding
859 Fixes bug 1064. Patch from Andrey N. Oktyabrski
861 NM/03 Allow underscore in dnslist lookups
862 Fixes bug 1026. Patch from Graeme Fowler
864 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
865 Code patches from Adam Ciarcinski of NetBSD.
867 NM/04 Fixed exiqgrep to cope with mailq missing size issue
870 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
871 is logged, to avoid truncation. Patch from John Horne.
873 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
874 Patch from Jakob Hirsch.
876 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
877 SQL string expansion failure details.
878 Patch from Andrey Oktyabrski.
880 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
881 Patch from Simon Arlott.
883 PP/09 New lookups functionality failed to compile on old gcc which rejects
884 extern declarations in function scope.
885 Patch from Oliver Fleischmann
887 PP/10 Use sig_atomic_t for flags set from signal handlers.
888 Check getgroups() return and improve debugging.
889 Fixed developed for diagnosis in bug 927 (which turned out to be
892 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
893 Patch from Mark Zealey.
895 PP/12 Bugzilla 1056: Improved spamd server selection.
896 Patch from Mark Zealey.
898 PP/13 Bugzilla 1086: Deal with maildir quota file races.
899 Based on patch from Heiko Schlittermann.
901 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
902 Patch from Uwe Doering, sign-off by Michael Haardt.
904 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
905 variable declaration deep within a block. Bug and patch from
908 PP/15 lookups-Makefile IRIX compatibilty coercion.
910 PP/16 Make DISABLE_DKIM build knob functional.
912 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
913 Patch by Simon Arlott
915 TF/03 Fix valgrind.h portability to C89 compilers that do not support
916 variable argument macros. Our copy now differs from upstream.
922 TF/01 Failure to get a lock on a hints database can have serious
923 consequences so log it to the panic log.
925 TF/02 Log LMTP confirmation messages in the same way as SMTP,
926 controlled using the smtp_confirmation log selector.
928 TF/03 Include the error message when we fail to unlink a spool file.
930 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
931 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
932 for maintaining out-of-tree patches for some time.
934 PP/01 Bugzilla 139: Documentation and portability issues.
935 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
936 Handle per-OS dynamic-module compilation flags.
938 PP/02 Let /dev/null have normal permissions.
939 The 4.73 fixes were a little too stringent and complained about the
940 permissions on /dev/null. Exempt it from some checks.
941 Reported by Andreas M. Kirchwitz.
943 PP/03 Report version information for many libraries, including
944 Exim version information for dynamically loaded libraries. Created
945 version.h, now support a version extension string for distributors
946 who patch heavily. Dynamic module ABI change.
948 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
949 privilege escalation vulnerability whereby the Exim run-time user
950 can cause root to append content of the attacker's choosing to
953 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
956 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
957 If dropping privileges for untrusted macros, we disabled normal logging
958 on the basis that it would fail; for the Exim run-time user, this is not
959 the case, and it resulted in successful deliveries going unlogged.
960 Fixed. Reported by Andreas Metzler.
966 PP/01 Date: & Message-Id: revert to normally being appended to a message,
967 only prepend for the Resent-* case. Fixes regression introduced in
968 Exim 4.70 by NM/22 for Bugzilla 607.
970 PP/02 Include check_rfc2047_length in configure.default because we're seeing
971 increasing numbers of administrators be bitten by this.
973 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
975 PP/03 Bugzilla 994: added openssl_options main configuration option.
977 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
979 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
981 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
983 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
984 without a peer certificate, leading to a segfault because of an
985 assumption that peers always have certificates. Be a little more
986 paranoid. Problem reported by Martin Tscholak.
988 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
989 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
990 NB: ClamAV planning to remove STREAM in "middle of 2010".
991 CL also introduces -bmalware, various -d+acl logging additions and
992 more caution in buffer sizes.
994 PP/09 Implemented reverse_ip expansion operator.
996 PP/10 Bugzilla 937: provide a "debug" ACL control.
998 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1000 PP/12 Bugzilla 973: Implement --version.
1002 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1004 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1006 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1008 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1009 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1010 ignore trailing whitespace.
1012 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1014 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1017 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1018 Notification from Dr Andrew Aitchison.
1020 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1021 ExtendedDetectionInfo response format.
1022 Notification from John Horne.
1024 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1027 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1028 XSL and documented dependency on system catalogs, with examples of how
1031 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1034 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1035 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1036 configuration file which is writeable by the Exim user or group.
1038 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1039 of configuration files to cover files specified with the -C option if
1040 they are going to be used with root privileges, not just the default
1043 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1044 option (effectively making it always true).
1046 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1047 files to be used while preserving root privileges.
1049 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1050 that rogue child processes cannot use them.
1052 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1053 run-time user, instead of root.
1055 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1056 Exim run-time user without dropping privileges.
1058 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1059 result string, instead of calling string_vformat() twice with the same
1062 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1063 for other users. Others should always drop root privileges if they use
1064 -C on the command line, even for a whitelisted configure file.
1066 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1068 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1074 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1075 $data_path, and $header_path variables; fixed documentation bugs and
1078 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1079 exipick to access non-standard spools, including the "frozen" queue
1082 NM/01 Bugzilla 965: Support mysql stored procedures.
1083 Patch from Alain Williams
1085 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1087 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1088 Patch from Andreas Metzler
1090 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1091 Patch from Kirill Miazine
1093 NM/05 Bugzilla 671: Added umask to procmail example.
1095 JJ/03 installed exipick 20100323.0, fixing doc bug
1097 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1098 directory. Notification and patch from Dan Rosenberg.
1100 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1102 TK/02 Improve log output when DKIM signing operation fails.
1104 MH/01 Treat the transport option dkim_domain as a colon separated
1105 list, not as a single string, and sign the message with each element,
1106 omitting multiple occurences of the same signer.
1108 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1109 Bugzilla 985, 986. Patch by Simon Arlott
1111 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1112 Patch by Simon Arlott
1114 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1115 MBX locking. Notification from Dan Rosenberg.
1121 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1123 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1125 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1127 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1129 NM/04 Bugzilla 915: Flag broken perl installation during build.
1135 TK/01 Added patch by Johannes Berg that expands the main option
1136 "spamd_address" if it starts with a dollar sign.
1138 TK/02 Write list of recipients to X-Envelope-Sender header when building
1139 the mbox-format spool file for content scanning (suggested by Jakob
1142 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1143 (http://www.dcc-servers.net/) support via dccifd. Activated by
1144 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1146 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1147 by Mark Daniel Reidel <mr@df.eu>.
1149 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1150 When building exim an external PCRE library is now needed -
1151 PCRE is a system library on the majority of modern systems.
1152 See entry on PCRE_LIBS in EDITME file.
1154 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1155 conversation. Added nologin parameter to request.
1156 Patch contributed by Kirill Miazine.
1158 TF/01 Do not log submission mode rewrites if they do not change the address.
1160 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1162 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1163 log files in place. Contributed by Roberto Lima.
1165 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1167 TF/03 Bugzilla 615: When checking the local_parts router precondition
1168 after a local_part_suffix or local_part_prefix option, Exim now
1169 does not use the address's named list lookup cache, since this
1170 contains cached lookups for the whole local part.
1172 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1173 Robert Millan. Documentation is in experimental-spec.txt.
1175 TF/04 Bugzilla 668: Fix parallel build (make -j).
1177 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1179 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1180 Patch provided by Jan Srzednicki.
1182 TF/05 Leading white space used to be stripped from $spam_report which
1183 wrecked the formatting. Now it is preserved.
1185 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1186 that they are available at delivery time.
1188 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1190 TF/08 TLS error reporting now respects the incoming_interface and
1191 incoming_port log selectors.
1193 TF/09 Produce a more useful error message if an SMTP transport's hosts
1194 setting expands to an empty string.
1196 NM/06 Bugzilla 744: EXPN did not work under TLS.
1197 Patch provided by Phil Pennock.
1199 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1200 Patch provided by Richard Godbee.
1202 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1203 acl_smtp_notquit, added index entry.
1205 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1206 Patch provided by Eugene Bujak.
1208 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1209 accept(). Patch provided by Maxim Dounin.
1211 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1212 Patch provided by Phil Pennock.
1214 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1216 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1217 Patch provided by Brad "anomie" Jorsch.
1219 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1220 Patch provided by Dean Brooks.
1222 TK/05 Add native DKIM support (does not depend on external libraries).
1224 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1225 Patch provided by Graeme Fowler.
1227 NM/16 Bugzilla 851: Documentation example syntax fix.
1229 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1231 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1234 NM/19 Bugzilla 745: TLS version reporting.
1235 Patch provided by Phil Pennock.
1237 NM/20 Bugzilla 167: bool: condition support.
1238 Patch provided by Phil Pennock.
1240 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1241 clients. Patch provided by Phil Pennock.
1243 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1244 Patch provided by Brad "anomie" Jorsch.
1246 NM/23 Bugzilla 687: Fix misparses in eximstats.
1247 Patch provided by Heiko Schlittermann.
1249 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1250 Patch provided by Heiko Schlittermann.
1252 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1253 plus update to original patch.
1255 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1257 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1258 Patch provided by David Brownlee.
1260 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1262 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1264 NM/30 Bugzilla 888: TLS documentation bugfixes.
1266 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1268 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1269 Unlike the original bugzilla I have changed all shell scripts in src tree.
1271 NM/33 Bugzilla 898: Transport filter timeout fix.
1272 Patch by Todd Rinaldo.
1274 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1275 Patch by Serge Demonchaux.
1277 NM/35 Bugzilla 39: Base64 decode bug fixes.
1278 Patch by Jakob Hirsch.
1280 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1282 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1284 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1286 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1292 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1293 ALT-N's libdkim that I have put here:
1294 http://duncanthrax.net/exim-experimental/
1296 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1297 were called 'true' and it seems that C99 defines that as a reserved
1298 keyword to be used with 'bool' variable types. That means you could
1299 not include C99-style headers which use bools without triggering
1300 build errors in sieve.c.
1302 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1303 as mailq or other aliases. Changed the --help handling significantly
1304 to do whats expected. exim_usage() emits usage/help information.
1306 SC/01 Added the -bylocaldomain option to eximstats.
1308 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1310 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1312 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1318 PH/01 Another patch from the Sieve maintainer.
1320 PH/02 When an IPv6 address is converted to a string for single-key lookup
1321 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1322 dots are used instead of colons so that keys in lsearch files need not
1323 contain colons. This was done some time before quoting was made available
1324 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1325 (notated using the quote facility) so as to distinguish them from IPv4
1326 keys. This meant that lookups for IP addresses in host lists did not work
1327 for iplsearch lookups.
1329 This has been fixed by arranging for IPv6 addresses to be expressed with
1330 colons if the lookup type is iplsearch. This is not incompatible, because
1331 previously such lookups could never work.
1333 The situation is now rather anomolous, since one *can* have colons in
1334 ordinary lsearch keys. However, making the change in all cases is
1335 incompatible and would probably break a number of configurations.
1337 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1340 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1341 conversion specification without a maximum field width, thereby enabling
1342 a rogue spamd server to cause a buffer overflow. While nobody in their
1343 right mind would setup Exim to query an untrusted spamd server, an
1344 attacker that gains access to a server running spamd could potentially
1345 exploit this vulnerability to run arbitrary code as the Exim user.
1347 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1348 $primary_hostname instead of what libspf2 thinks the hosts name is.
1350 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1351 a directory entry by the name of the lookup key. Previously, if a
1352 symlink pointed to a non-existing file or a file in a directory that
1353 Exim lacked permissions to read, a lookup for a key matching that
1354 symlink would fail. Now it is enough that a matching directory entry
1355 exists, symlink or not. (Bugzilla 503.)
1357 PH/03 The body_linecount and body_zerocount variables are now exported in the
1360 PH/04 Added the $dnslist_matched variable.
1362 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1363 This means they are set thereafter only if the connection becomes
1366 PH/06 Added the client_condition to authenticators so that some can be skipped
1367 by clients under certain conditions.
1369 PH/07 The error message for a badly-placed control=no_multiline_responses left
1370 "_responses" off the end of the name.
1372 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1374 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1375 (without spaces) instead of just copying the configuration text.
1377 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1379 PH/11 Added $max_received_linelength.
1381 PH/12 Added +ignore_defer and +include_defer to host lists.
1383 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1384 way in which PCRE > 7.0 is built.
1386 PH/14 Implemented queue_only_load_latch.
1388 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1389 MAIL command. The effect was to mangle the value on 64-bit systems.
1391 PH/16 Another patch from the Sieve maintainer.
1393 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1395 PH/18 If a system quota error occurred while trying to create the file for
1396 a maildir delivery, the message "Mailbox is full" was not appended to the
1397 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1398 only to a quota excession during the actual writing of the file.
1400 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1401 characters?) which causes problems in log lines. The DN values are now
1402 passed through string_printing() before being added to log lines.
1404 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1405 and InterBase are left for another time.)
1407 PH/21 Added message_body_newlines option.
1409 PH/22 Guard against possible overflow in moan_check_errorcopy().
1411 PH/23 POSIX allows open() to be a macro; guard against that.
1413 PH/24 If the recipient of an error message contained an @ in the local part
1414 (suitably quoted, of course), incorrect values were put in $domain and
1415 $local_part during the evaluation of errors_copy.
1421 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1422 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1425 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1426 issue a MAIL command.
1428 PH/02 In an ACL statement such as
1430 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1432 if a client was not listed at all, or was listed with a value other than
1433 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1434 the condition was not true (as it should be), so access was not denied.
1435 The bug was that the ! inversion was incorrectly passed on to the second
1436 item. This has been fixed.
1438 PH/03 Added additional dnslists conditions == and =& which are different from
1439 = and & when the dns lookup returns more than one IP address.
1441 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1442 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1444 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1445 FSYNC, which compiles an option called disable_fsync that allows for
1446 bypassing fsync(). The documentation is heavily laced with warnings.
1448 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1450 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1451 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1452 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1453 including adding "make clean"; (3) Added -fPIC when compiling the test
1454 dynamically loaded module, to get rid of a warning.
1456 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1457 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1458 = 0s. The bug is otherwise harmless.
1460 PH/07 There was a bug in the dovecot authenticator such that the value of
1461 $auth1 could be overwritten, and so not correctly preserved, after a
1462 successful authentication. This usually meant that the value preserved by
1463 the server_setid option was incorrect.
1465 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1467 PH/09 Installed PCRE release 7.0.
1469 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1470 run for batched SMTP input. It is now run at the start of every message
1471 in the batch. While fixing this I discovered that the process information
1472 (output by running exiwhat) was not always getting set for -bs and -bS
1473 input. This is fixed, and it now also says "batched" for BSMTP.
1475 PH/11 Added control=no_pipelining.
1477 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1478 patch, slightly modified), and move the expansion of helo_data till after
1479 the connection is made in the smtp transport (so it can use these
1482 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1484 PH/14 Added log_selector = +pid.
1486 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1488 PH/16 Add ${if forany and ${if forall.
1490 PH/17 Added dsn_from option to vary the From: line in DSNs.
1492 PH/18 Flush SMTP output before performing a callout, unless control =
1493 no_callout_flush is set.
1495 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1496 was true (the default) a successful delivery failed to delete the retry
1497 item, thus causing premature timeout of the address. The bug is now
1500 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1502 PH/21 Long custom messages for fakedefer and fakereject are now split up
1503 into multiline reponses in the same way that messages for "deny" and
1504 other ACL rejections are.
1506 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1507 with slight modification.
1509 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1510 draft, changing the syntax and factoring some duplicate code".
1512 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1513 for deliveries of the second and subsequent messages over the same SMTP
1516 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1517 ${reduce, with only minor "tidies".
1519 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1521 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1522 expansion side effects.
1524 PH/27 When a message times out after an over-quota error from an Exim-imposed
1525 quota, the bounce message says "mailbox is full". This message was not
1526 being given when it was a system quota that was exceeded. It now should
1529 MH/03 Made $recipients available in local_scan(). local_scan() already has
1530 better access to the recipient list through recipients_list[], but
1531 $recipients can be useful in postmaster-provided expansion strings.
1533 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1534 in the case of a MAIL command with additional options following the
1535 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1536 were accidentally chopped off.
1538 PH/29 SMTP synchronization checks are implemented when a command is read -
1539 there is a check that no more input is waiting when there shouldn't be
1540 any. However, for some commands, a delay in an ACL can mean that it is
1541 some time before the response is written. In this time, more input might
1542 arrive, invalidly. So now there are extra checks after an ACL has run for
1543 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1544 pipelining has not been advertised.
1546 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1548 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1549 returned in the SMTP response when smtp_return_error_details was false.
1550 This has been fixed.
1552 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1553 instead of the C I/O that was originally supplied, because problems were
1554 reported on Solaris.
1556 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1557 Exim which did not show up earlier: it was assuming that a call to
1558 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1559 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1560 was a macro that became an assignment, so it seemed to work. This has
1561 changed to a proper function call with a void return, hence the compile
1562 error. Exim's code has been fixed.
1564 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1567 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1569 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1571 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1572 because it turns out to be too big or there is a timeout) there is no
1573 "Completed" line in the log. When some messages of this type were
1574 selected by exigrep, they were listed as "not completed". Others were
1575 picked up by some special patterns. I have improved the selection
1576 criteria to be more general.
1578 PH/38 The host_find_failed option in the manualroute router can now be set
1579 to "ignore", to completely ignore a host whose IP address cannot be
1580 found. If all hosts are ignored, the behaviour is controlled by the new
1581 host_all_ignored option.
1583 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1584 homing or because of multiple MX records with /mx) generated more than
1585 one IP address, and the following item turned out to be the local host,
1586 all the secondary addresses of the first item were incorrectly removed
1587 from the list, along with the local host and any following hosts (which
1588 is what is supposed to happen).
1590 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1591 whoever called Exim into the -H file. In the case of the daemon it was
1592 behaving confusingly. When first started, it used values for whoever
1593 started the daemon, but after a SIGHUP it used the Exim user (because it
1594 calls itself on a restart). I have changed the code so that it now always
1597 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1598 message are rejected with the same error (e.g. no authentication or bad
1599 sender address), and a DATA command is nevertheless sent (as can happen
1600 with PIPELINING or a stupid MUA), the error message that was given to the
1601 RCPT commands is included in the rejection of the DATA command. This is
1602 intended to be helpful for MUAs that show only the final error to their
1605 PH/42 Another patch from the Sieve maintainer.
1607 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1608 Eximstats - Fixed some broken HTML links and added missing column headers
1610 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1611 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1613 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1615 PH/43 Yet another patch from the Sieve maintainer.
1617 PH/44 I found a way to check for a TCP/IP connection going away before sending
1618 the response to the final '.' that terminates a message, but only in the
1619 case where the client has not sent further data following the '.'
1620 (unfortunately, this is allowed). However, in many cases there won't be
1621 any further data because there won't be any more messages to send. A call
1622 to select() can be used: if it shows that the input is "ready", there is
1623 either input waiting, or the socket has been closed. An attempt to read
1624 the next input character can distinguish the two cases. Previously, Exim
1625 would have sent an OK response which the client would never have see.
1626 This could lead to message repetition. This fix should cure that, at
1627 least in a lot of common cases.
1629 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1630 advertised in response to EHLO.
1636 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1637 fixed by 4.65/MH/01 (is this a record?) are fixed:
1639 (i) An empty string was always treated as zero by the numeric comparison
1640 operators. This behaviour has been restored.
1642 (ii) It is documented that the numeric comparison operators always treat
1643 their arguments as decimal numbers. This was broken in that numbers
1644 starting with 0 were being interpreted as octal.
1646 While fixing these problems I realized that there was another issue that
1647 hadn't been noticed. Values of message_size_limit (both the global option
1648 and the transport option) were treated as octal if they started with 0.
1649 The documentation was vague. These values are now always treated as
1650 decimal, and I will make that clear in the documentation.
1656 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1657 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1660 MH/01 Don't check that the operands of numeric comparison operators are
1661 integers when their expansion is in "skipping" mode (fixes bug
1662 introduced by 4.64-PH/07).
1664 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1665 child addresses, Exim now panics and dies. Previously, because the count
1666 is held in a short int, deliveries were likely to be lost. As such a
1667 large number of recipients for a single message is ridiculous
1668 (performance will be very, very poor), I have chosen to impose a limit
1669 rather than extend the field.
1675 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1676 leftover -K file (the existence of which was triggered by #402).
1677 While we were at it, introduced process PID as part of the -K
1678 filename. This should rule out race conditions when creating
1681 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1682 processing considerably. Previous code took too long for large mails,
1683 triggering a timeout which in turn triggers #401.
1685 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1686 in the DK code in transports.c. sendfile() is not really portable,
1687 hence the _LINUX specificness.
1689 TF/01 In the add_headers option to the mail command in an Exim filter,
1690 there was a bug that Exim would claim a syntax error in any
1691 header after the first one which had an odd number of characters
1694 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1695 callout verification, Exim cached a "reject" for the entire domain. This
1696 is correct for most verifications, but it is not correct for a recipient
1697 verification with use_sender or use_postmaster set, because in that case
1698 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1699 case of MAIL FROM:<> rejection from other early rejections (e.g.
1700 rejection of HELO). When verifying a recipient using a non-null MAIL
1701 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1702 Whatever the result of the callout, the value of the domain cache is
1703 left unchanged (for any other kind of callout, getting as far as trying
1704 RCPT means that the domain itself is ok).
1706 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1709 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1710 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1711 session, the daemon ignores ECONNECT errors and logs others; it now
1712 ignores EPIPE as well.
1714 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1715 (quoted-printable decoding).
1717 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1718 later the small subsequent patch to fix an introduced bug.
1720 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1722 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1724 PH/08 An error is now given if message_size_limit is specified negative.
1726 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1727 to be given (somewhat) arbitrary names.
1729 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1732 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1733 miscellaneous code fixes
1735 PH/10 Added the log_reject_target ACL modifier to specify where to log
1738 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1739 hostname. This is wrong, because it relates to the incoming message (and
1740 probably the interface on which it is arriving) and not to the outgoing
1741 callout (which could be using a different interface). This has been
1742 changed to use the value of the helo_data option from the smtp transport
1743 instead - this is what is used when a message is actually being sent. If
1744 there is no remote transport (possible with a router that sets up host
1745 addresses), $smtp_active_hostname is used.
1747 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1748 tweaks were necessary in order to get it to work (see also 21 below):
1749 (a) The code assumed that strncpy() returns a negative number on buffer
1750 overflow, which isn't the case. Replaced with Exim's string_format()
1752 (b) There were several signed/unsigned issues. I just did the minimum
1753 hacking in of casts. There is scope for a larger refactoring.
1754 (c) The code used strcasecmp() which is not a standard C function.
1755 Replaced with Exim's strcmpic() function.
1756 (d) The code set only $1; it now sets $auth1 as well.
1757 (e) A simple test gave the error "authentication client didn't specify
1758 service in request". It would seem that Dovecot has changed its
1759 interface. Fortunately there's a specification; I followed it and
1760 changed what the client sends and it appears to be working now.
1762 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1765 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1766 address A is aliased to B and C, where B exists and C does not. Without
1767 -v the output is "A verified" because verification stops after a
1768 successful redirection if more than one address is generated. However,
1769 with -v the child addresses are also verified. Exim was outputting "A
1770 failed to verify" and then showing the successful verification for C,
1771 with its parentage. It now outputs "B failed to verify", showing B's
1772 parentage before showing the successful verification of C.
1774 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1775 look up a TXT record in a specific list after matching in a combined
1778 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1779 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1780 they consult the DNS. I had assumed they would set it the way they
1781 wanted; and indeed my experiments on Linux seem to show that in some
1782 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1783 To be on the safe side, however, I have now made the interface to
1784 host_find_byname() similar to host_find_bydns(), with an argument
1785 containing the DNS resolver options. The host_find_byname() function now
1786 sets these options at its start, just as host_find_bydns() does. The smtp
1787 transport options dns_qualify_single and dns_search_parents are passed to
1788 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1789 of host_find_byname() use the default settings of RES_DEFNAMES
1790 (qualify_single) but not RES_DNSRCH (search_parents).
1792 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1793 spool_read_header() do less string testing, by means of a preliminary
1794 switch on the second character of optional "-foo" lines. (This is
1795 overdue, caused by the large number of possibilities that now exist.
1796 Originally there were few.) While I was there, I also converted the
1797 str(n)cmp tests so they don't re-test the leading "-" and the first
1798 character, in the hope this might squeeze out yet more improvement.
1800 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1801 flag allowing group syntax was set by the header_syntax check but not
1802 turned off, possible causing trouble later; (2) The flag was not being
1803 set at all for the header_verify test, causing "group"-style headers to
1804 be rejected. I have now set it in this case, and also caused header_
1805 verify to ignore an empty address taken from a group. While doing this, I
1806 came across some other cases where the code for allowing group syntax
1807 while scanning a header line wasn't quite right (mostly, not resetting
1808 the flag correctly in the right place). These bugs could have caused
1809 trouble for malformed header lines. I hope it is now all correct.
1811 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1812 with the "reply" argument non-NULL. The code, however (which originally
1813 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1814 but it didn't always do it. This confused somebody who was copying the
1815 code for some other use. I have removed all the tests.
1817 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1818 feature that was used to support insecure browsers during the U.S. crypto
1819 embargo. It requires special client support, and Exim is probably the
1820 only MTA that supported it -- and would never use it because real RSA is
1821 always available. This code has been removed, because it had the bad
1822 effect of slowing Exim down by computing (never used) parameters for the
1823 RSA_EXPORT functionality.
1825 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1826 authenticator to fail if there's a tab character in the incoming data
1827 (there should never be unless someone is messing about, as it's supposed
1828 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1829 if the connection is using TLS or if the remote IP is the same as the
1830 local IP, and the "valid-client-cert option" if a client certificate has
1833 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1834 authenticators. This can be used for authorization after authentication
1835 succeeds. (In the case of plaintext, it servers for both authentication
1838 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1839 if any retry times were supplied.
1841 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1842 connection, where there is no client IP address to check. In this
1843 situation, the verify now always succeeds.
1845 PH/25 Applied John Jetmore's -Mset patch.
1847 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1849 PH/27 In a string expansion for a processed (not raw) header when multiple
1850 headers of the same name were present, leading whitespace was being
1851 removed from all of them, but trailing whitespace was being removed only
1852 from the last one. Now trailing whitespace is removed from each header
1853 before concatenation. Completely empty headers in a concatenation (as
1854 before) are ignored.
1856 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1857 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1859 PH/29 [Removed. This was a change that I later backed out, and forgot to
1860 correct the ChangeLog entry (that I had efficiently created) before
1861 committing the later change.]
1863 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1864 address errors (4xx response to RCPT) over the same connection as other
1865 messages routed to the same hosts. Such deliveries are always "forced",
1866 so retry times are not inspected. This resulted in far too many retries
1867 for the affected addresses. The effect occurred only when there were more
1868 hosts than the hosts_max_try setting in the smtp transport when it had
1869 the 4xx errors. Those hosts that it had tried were not added to the list
1870 of hosts for which the message was waiting, so if all were tried, there
1871 was no problem. Two fixes have been applied:
1873 (i) If there are any address or message errors in an SMTP delivery, none
1874 of the hosts (tried or untried) are now added to the list of hosts
1875 for which the message is waiting, so the message should not be a
1876 candidate for sending over the same connection that was used for a
1877 successful delivery of some other message. This seems entirely
1878 reasonable: after all the message is NOT "waiting for some host".
1879 This is so "obvious" that I'm not sure why it wasn't done
1880 previously. Hope I haven't missed anything, but it can't do any
1881 harm, as the worst effect is to miss an optimization.
1883 (ii) If, despite (i), such a delivery is accidentally attempted, the
1884 routing retry time is respected, so at least it doesn't keep
1885 hammering the server.
1887 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1888 in ${readsocket because some servers need this prod.
1890 PH/32 Added some extra debug output when updating a wait-xxx database.
1892 PH/33 The hint "could be header name not terminated by colon", which has been
1893 given for certain expansion errors for a long time, was not being given
1894 for the ${if def:h_colon_omitted{... case.
1896 PH/34 The spec says: "With one important exception, whenever a domain list is
1897 being scanned, $domain contains the subject domain." There was at least
1898 one case where this was not true.
1900 PH/35 The error "getsockname() failed: connection reset by peer" was being
1901 written to the panic log as well as the main log, but it isn't really
1902 panic-worthy as it just means the connection died rather early on. I have
1903 removed the panic log writing for the ECONNRESET error when getsockname()
1906 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1907 runs only) independently of the message's sender address. This meant
1908 that, if the 4xx error was in fact related to the sender, a different
1909 message to the same recipient with a different sender could confuse
1910 things. In particualar, this can happen when sending to a greylisting
1911 server, but other circumstances could also provoke similar problems.
1912 I have changed the default so that the retry time for these errors is now
1913 based a combination of the sender and recipient addresses. This change
1914 can be overridden by setting address_retry_include_sender=false in the
1917 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1918 remote server are returned as part of bounce messages. This was not
1919 happening for LMTP over a pipe (the lmtp transport), but now it is the
1920 same for both kinds of LMTP.
1922 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1923 in header lines that were in fact CNAMEs. This is no longer the case.
1925 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1926 and queue runs started by the daemon processed all messages. This has
1927 been fixed so that -R and -S can now usefully be given with -q<time>.
1929 PH/40 Import PCRE release 6.7 (fixes some bugs).
1931 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1933 PH/42 Give an error if -q is specified more than once.
1935 PH/43 Renamed the variables $interface_address and $interface_port as
1936 $received_ip_address and $received_port, to make it clear that these
1937 values apply to message reception, and not to the outgoing interface when
1938 a message is delivered. (The old names remain recognized, of course.)
1940 PH/44 There was no timeout on the connect() call when using a Unix domain
1941 socket in the ${readsocket expansion. There now is.
1943 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1944 be meaningful with "accept".
1946 SC/01 Eximstats V1.43
1947 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1949 SC/02 Eximstats V1.44
1950 Use a glob alias rather than an array ref in the generated
1951 parser. This improves both readability and performance.
1953 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1954 Collect SpamAssassin and rejection statistics.
1955 Don't display local sender or destination tables unless
1956 there is data to show.
1957 Added average volumes into the top table text output.
1959 SC/04 Eximstats V1.46
1960 Collect data on the number of addresses (recipients)
1961 as well as the number of messages.
1963 SC/05 Eximstats V1.47
1964 Added 'Message too big' to the list of mail rejection
1965 reasons (thanks to Marco Gaiarin).
1967 SC/06 Eximstats V1.48
1968 Mainlog lines which have GMT offsets and are too short to
1969 have a flag are now skipped.
1971 SC/07 Eximstats V1.49 (Alain Williams)
1972 Added the -emptyok flag.
1974 SC/08 Eximstats V1.50
1975 Fixes for obtaining the IP address from reject messages.
1977 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1978 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1979 whitesspace changes from 4.64-PH/27
1981 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1984 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1985 are found, allow negative numbers in numeric criteria)
1987 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1989 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1990 to match changes made in 4.64-PH/43
1992 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1994 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1995 "domain" is actually an IP address, and force a failure. This locks out
1996 those revolvers/nameservers that support "A-for-A" lookups, in
1997 contravention of the specifications.
1999 PH/48 When a host name was looked up from an IP address, and the subsequent
2000 forward lookup of the name timed out, the host name was left in
2001 $sender_host_name, contrary to the specification.
2003 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2004 restricted to single-key lookups, Exim was not diagnosing an error if
2005 * or *@ was used with a query-style lookup.
2007 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2009 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2010 long ago, but noone interested enough thought of it. Let's just say that
2011 the "1.1" means that there are some new functions that weren't there at
2012 some point in the past.
2014 PH/51 Error processing for expansion failure of helo_data from an smtp
2015 transport during callout processing was broken.
2017 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2018 tested/used via the -bh/-bhc/-bs options.
2020 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2021 bug, fixed in subsequent PCRE releases).
2023 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2024 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2026 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2032 SC/01 Use a glob alias rather than an array ref in eximstats generated
2033 parser. This improves both readability and performance.
2035 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2036 Don't display local sender or destination tables in eximstats unless
2037 there is data to show.
2038 Added average volumes into the eximstats top table text output.
2040 SC/03 Collect data on the number of addresses (recipients) as well
2041 as the number of messages in eximstats.
2043 TF/01 Correct an error in the documentation for the redirect router. Exim
2044 does (usually) call initgroups() when daemonizing.
2046 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2047 with consistent privilege compared to when running as a daemon.
2049 TF/03 Note in the spec that $authenticated_id is not set for local
2050 submissions from trusted users.
2052 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2053 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2055 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2056 by adding some example configuration directives to the default
2057 configuration file. A little bit of work is required to uncomment the
2058 directives and define how usernames and passwords are checked, but
2059 there is now a framework to start from.
2061 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2062 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2063 without this. I don't know how relevant this is to other LDAP libraries.
2065 PH/02 Add the verb name to the "unknown ACL verb" error.
2067 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2069 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2071 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2072 directory not expanded when it should be if an expanded home directory
2073 was set for the address (which is overridden by the transport).
2075 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2078 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2079 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2080 because it is too late at that time, and has no effect.
2082 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2083 security issue with \' (bugzilla #107). I could not use the
2084 PQescapeStringConn() function, because it needs a PGconn value as one of
2087 PH/08 When testing addresses using -bt, indicate those final addresses that
2088 are duplicates that would not cause an additional delivery. At least one
2089 person was confused, thinking that -bt output corresponded to deliveries.
2090 (Suppressing duplicates isn't a good idea as you lose the information
2091 about possibly different redirections that led to the duplicates.)
2093 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2094 systems where poll() doesn't work, in particular OS X.
2096 PH/10 Added more information to debugging output for retry time not reached.
2098 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2099 operations in malware.c.
2101 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2104 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2105 syslog_duplication set false, log lines that would normally be written
2106 both the the main log and to the reject log were not written to syslog at
2109 PH/14 In the default configuration, change the use of "message" in ACL warn
2110 statements to "add_header".
2112 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2113 not followed by a command (e.g. "seen endif").
2115 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2116 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2119 PH/17 Added extra conditions to the default value of delay_warning_condition
2123 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2124 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2125 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2128 The Auto-Submitted: and various List- headers are standardised, whereas I
2129 don't think Precedence: ever was.
2131 PH/18 Refactored debugging code in route_finduser() to show more information,
2132 in particular, the error code if getpwnam() issues one.
2134 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2135 This is apparently needed in addition to the PH/07 change above to avoid
2136 any possible encoding problems.
2138 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2139 but not after initializing Perl.
2141 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2142 output them only if debugging. By default they are written stderr,
2143 apparently, which is not desirable.
2145 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2148 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2151 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2153 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2154 authenticated or an ident call has been made. Suppress the default
2155 values for $authenticated_id and $authenticated_sender (but permit -oMai
2156 and -oMas) when testing with -bh.
2158 PH/24 Re-jigged the order of the tests in the default configuration so that the
2159 tests for valid domains and recipients precede the DNS black list and CSA
2160 tests, on the grounds that those ones are more expensive.
2162 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2163 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2164 This bug exists in every version of Exim that I still have, right back to
2167 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2168 However, an attempt to turn on case-sensitivity in a regex key by
2169 including (?-i) didn't work because the subject string was already
2170 lowercased, and the effects were non-intuitive. It turns out that a
2171 one-line patch can be used to allow (?-i) to work as expected.
2177 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2178 other effects) broke the use of negated acl sub-conditions.
2180 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2183 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2184 "Deny" causes Exim to reject the incoming connection with a 554 error.
2185 Unfortunately, if there is a major crisis, such as a disk failure,
2186 tcp-wrappers gives "deny", whereas what one would like would be some
2187 kind of temporary error. A kludge has been added to help with this.
2188 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2189 554 error is used if errno is still zero or contains ENOENT (which occurs
2190 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2193 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2195 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2196 errors. Otherwise a message that provokes a temporary error (when other
2197 messages do not) can cause a whole host to time out.
2199 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2200 addresses were routed directly to files or pipes from a redirect router.
2201 File deliveries just didn't batch; pipe deliveries might have suffered
2204 PH/06 A failure to get a lock for a hints database would erroneously always say
2205 "Failed to get write lock", even when it was really a read lock.
2207 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2208 of 0600. This has been changed to use the value of the lockfile_mode
2209 option (which defaults to 0600).
2211 PH/08 Applied small patch from the Sieve maintainer.
2213 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2214 folder from quota calculations, a direct delivery into this folder messed
2215 up the contents of the maildirsize file. This was because the regex was
2216 used only to exclude .Trash (or whatever) when the size of the mailbox
2217 was calculated. There was no check that a delivery was happening into an
2218 excluded directory. This bug has been fixed by ignoring all quota
2219 processing for deliveries into excluded directories.
2221 PH/10 Added the maildirfolder_create_regex option to appendfile.
2227 PH/01 The code for finding all the local interface addresses on a FreeBSD
2228 system running IPv6 was broken. This may well have applied to all BSD
2229 systems, as well as to others that have similar system calls. The broken
2230 code found IPv4 interfaces correctly, but gave incorrect values for the
2231 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2232 that it would not match correctly against @[] and not recognize the IPv6
2235 PH/02 The ipliteral router was not recognizing addresses of the form user@
2236 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2238 PH/03 Added disable_ipv6.
2240 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2241 decoded form, because it is most often used to construct To: headers
2242 lines in autoreplies, and the decoded form may well be syntactically
2243 invalid. However, $reply_address has leading white space removed, and all
2244 newlines turned into spaces so that the autoreply transport does not
2247 PH/05 If group was specified without a user on a router, and no group or user
2248 was specified on a transport, the group from the router was ignored.
2250 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2251 for visible compile-time settings that can be used to change these
2252 numbers, for those that want even more. Backwards compatibility with old
2253 spool files has been maintained. However, going back to a previous Exim
2254 release will lost any variables that are in spool files.
2256 PH/07 Two small changes when running in the test harness: increase delay when
2257 passing a TCP/IP connection to a new process, in case the original
2258 process has to generate a bounce, and remove special handling of
2259 127.0.0.2 (sic), which is no longer necessary.
2261 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2262 be the same on different OS.
2264 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2267 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2268 whether --show-vars was specified or not
2270 JJ/02 exipick: Added support for new ACL variable spool format introduced
2273 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2274 syntactically invalid From: or Reply-to: line, and a filter used this to
2275 generate an autoreply, and therefore failed to obtain an address for the
2276 autoreply, Exim could try to deliver to a non-existent relative file
2277 name, causing unrelated and misleading errors. What now happens is that
2278 it logs this as a hard delivery error, but does not attempt to create a
2281 PH/11 The exinext utility has a -C option for testing purposes, but although
2282 the given file was scanned by exinext itself; it wasn't being passed on
2283 when Exim was called.
2285 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2286 an end-of-file indication when reading a command response.
2288 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2289 compiled. In many other places in Exim, IPv6 addresses are always
2290 recognized, so I have changed this. It also means that IPv4 domain
2291 literals of the form [IPV4:n.n.n.n] are now always recognized.
2293 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2294 used if the router is not running as root, for example, when verifying at
2295 ACL time, or when using -bh. The debugging output from this situation was
2296 non-existent - all you got was a failure to exec. I have made two
2299 (a) Failures to set uid/gid, the current directory, or a process leader
2300 in a subprocess such as that created by queryprogram now generate
2301 suitable debugging ouput when -d is set.
2303 (b) The queryprogram router detects when it is not running as root,
2304 outputs suitable debugging information if -d is set, and then runs
2305 the subprocess without attempting to change uid/gid.
2307 PH/15 Minor change to Makefile for building test_host (undocumented testing
2310 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2311 additional section of a DNS packet that returns MX or SRV records.
2312 Instead, it always explicitly searches for A/AAAA records. This avoids
2313 major problems that occur when a DNS server includes only records of one
2314 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2315 fixed another bug: if SRV records were looked up and the corresponding
2316 address records were *not* found in the additional section, the port
2317 values from the SRV records were lost.
2319 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2320 using the correct key (the original address) when searching the retry
2321 rules in order to find which one to use for generating the retry hint.
2323 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2324 adding the default one. Similarly, if it contains a Reply-To: header, the
2325 errors_reply_to option, if set, is not used.
2327 PH/19 When calculating a retry time, Exim used to measure the "time since
2328 failure" by looking at the "first failed" field in the retry record. Now
2329 it does not use this if it is later than than the arrival time of the
2330 message. Instead it uses the arrival time. This makes for better
2331 behaviour in cases where some deliveries succeed, thus re-setting the
2332 "first failed" field. An example is a quota failure for a huge message
2333 when small messages continue to be delivered. Without this change, the
2334 "time since failure" will always be short, possible causing more frequent
2335 delivery attempts for the huge message than are intended.
2336 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2338 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2339 $1, $2, $3) because the numerical variables can be reset during some
2340 expansion items (e.g. "match"), thereby losing the authentication data.
2342 PH/21 Make -bV show the size of off_t variables so that the test suite can
2343 decide whether to run tests for quotas > 2G.
2345 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2346 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2347 filecount value is greater than 2G or if a quota value is greater than 2G
2348 on a system where the size of off_t is not greater than 4, a panic error
2351 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2352 never match. The debug and -bh output now contains an explicit error
2353 message indicating a malformed IPv4 address or mask.
2355 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2356 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2357 PH/23 above applies.
2359 PH/25 Do not write to syslog when running in the test harness. The only
2360 occasion when this arises is a failure to open the main or panic logs
2361 (for which there is an explicit test).
2363 PH/26 Added the /no_tell option to "control=freeze".
2365 PH/27 If a host name lookup failed very early in a connection, for example, if
2366 the IP address matched host_lookup and the reverse lookup yielded a name
2367 that did not have a forward lookup, an error message of the form "no IP
2368 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2369 could be logged. Now it outputs the IP address instead of "NULL".
2371 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2372 allows the sender and the authenticated sender to be set when
2373 submitting a message from within Exim. Since child_open_exim() is
2374 documented for local_scan(), the new function should be too.
2376 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2377 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2378 results in an empty string is now treated as unset.
2380 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2382 PH/31 Added sender_verify_fail logging option.
2384 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2385 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2386 tidied the source and removed it altogether.
2388 PH/33 When a queue run was abandoned because the load average was too high, a
2389 log line was always written; now it is written only if the queue_run log
2390 selector is set. In addition, the log line for abandonment now contains
2391 information about the queue run such as the pid. This is always present
2392 in "start" and "stop" lines but was omitted from the "abandon" line.
2394 PH/34 Omit spaces between a header name and the colon in the error message that
2395 is given when verify = headers_syntax fails (if there are lots of them,
2396 the message gets confusing).
2398 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2399 names, as there are now some PTR records that contain slashes. This check
2400 is only to protect against broken name servers that fall over on strange
2401 characters, so the fact that it applies to all lookups doesn't matter.
2403 PH/36 Now that the new test suite is complete, we can remove some of the
2404 special code in Exim that was needed for the old test suite. For example,
2405 sorting DNS records because real resolvers return them in an arbitrary
2406 order. The new test suite's fake resolver always returns records in the
2409 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2410 bounces) except when queue_only is set, to avoid logging races between
2411 the different processes.
2413 PH/38 Panic-die if .include specifies a non-absolute path.
2415 PH/39 A tweak to the "H" retry rule from its user.
2417 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2418 a label. They prevented compilation on older perls.
2420 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2421 a warning to be raised on newish perls.
2423 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2424 on queue. Changes to match documented behaviour of showing count of
2425 messages matching specified criteria.
2427 PH/40 Changed the default ident timeout from 30s to 5s.
2429 PH/41 Added support for the use of login_cap features, on those BSD systems
2430 that have them, for controlling the resources used by pipe deliveries.
2432 PH/42 The content-scanning code uses fopen() to create files in which to put
2433 message data. Previously it was not paying any attention to the mode of
2434 the files. Exim runs with umask(0) because the rest of the code creates
2435 files with open(), and sets the required mode explicitly. Thus, these
2436 files were ending up world-writeable. This was not a big issue, because,
2437 being within the spool directory, they were not world-accessible. I have
2438 created a function called modefopen, which takes an additional mode
2439 argument. It sets umask(777), creates the file, chmods it to the required
2440 mode, then resets the umask. All the relevant calls to fopen() in the
2441 content scanning code have been changed to use this function.
2443 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2444 to 24 hours. This avoids potential overflow problems when processing G
2445 and H retry rules. I suspect nobody ever tinkers with this value.
2447 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2449 PH/45 When the plaintext authenticator is running as a client, the server's
2450 challenges are checked to ensure they are valid base64 strings. By
2451 default, the authentication attempt is cancelled if an invalid string is
2452 received. Setting client_ignore_invalid_base64 true ignores these errors.
2453 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2454 they are received. Thus, the responses can be made to depend on the
2455 challenges. If an invalid string is ignored, an empty string is placed in
2458 PH/46 Messages that are created by the autoreply transport now contains a
2459 References: header, in accordance with RFCs 2822 and 3834.
2461 PH/47 Added authenticated_sender_force to the smtp transport.
2463 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2465 PH/49 Installed latest patch from the Sieve maintainer.
2467 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2468 was also set, the appendfile transport was unnecessarily scanning a
2469 directory of message files (e.g. for maildir delivery) to find the count
2470 of files (along with the size), even though it did not need this
2471 information. It now does the scan only if it needs to find either the
2472 size of the count of files.
2474 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2476 PH/52 Two bugs concerned with error handling when the smtp transport is
2479 (i) Exim was not creating retry information for temporary errors given
2480 for individual recipients after the DATA command when the smtp transport
2481 was used in LMTP mode. This meant that they could be retried too
2482 frequently, and not timed out correctly.
2484 (ii) Exim was setting the flag that allows error details to be returned
2485 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2486 recipients that were returned after the DATA command.
2488 PH/53 This is related to PH/52, but is more general: for any failing address,
2489 when detailed error information was permitted to be returned to the
2490 sender, but the error was temporary, then after the final timeout, only
2491 "retry timeout exceeded" was returned. Now it returns the full error as
2492 well as "retry timeout exceeded".
2494 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2495 do this, and (what is worse) MTAs that accept it.
2497 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2498 will now be deprecated.
2500 PH/56 New os.c-cygwin from the Cygwin maintainer.
2502 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2503 formats (previously only available in exim formats via -bpr, -bpru,
2504 and -bpra. Now also available in native and exiqgrep formats)
2506 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2507 with very large, slow to parse queues
2509 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2511 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2513 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2514 responses to authentication challenges, though it was showing the
2515 challenges; (ii) I've removed the CR characters from the debug output for
2518 PH/58 Allow for the insertion of a newline as well as a space when a string
2519 is turned into more than one encoded-word during RFC 2047 encoding. The
2520 Sieve code now uses this.
2522 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2523 data_4xx, lost_connection, tls_required.
2525 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2526 message was being sent as an SMTP response.
2528 PH/61 Add -l and -k options to exicyclog.
2530 PH/62 When verifying, if an address was redirected to one new address, so that
2531 verification continued, and the new address failed or deferred after
2532 having set something in $address_data, the value of $address_data was not
2533 passed back to the ACL. This was different to the case when no
2534 redirection occurred. The value is now passed back in both cases.
2536 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2537 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2538 use login_cap.h, so on its own it isn't the distinguishing feature. The
2539 new name refers directly to the setclassresources() function.
2541 PH/65 Added configuration files for NetBSD3.
2543 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2545 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2546 is preferred over IPv4.
2548 PH/68 The bounce_return_message and bounce_return_body options were not being
2549 honoured for bounces generated during the reception of non-SMTP messages.
2550 In particular, this applied to messages rejected by the ACL. This bug has
2551 been fixed. However, if bounce_return_message is true and bounce_return_
2552 body is false, the headers that are returned for a non-SMTP message
2553 include only those that have been read before the error was detected.
2554 (In the case of an ACL rejection, they have all been read.)
2556 PH/69 The HTML version of the specification is now built in a directory called
2557 spec_html instead of spec.html, because the latter looks like a path with
2558 a MIME-type, and this confuses some software.
2560 PH/70 Catch two compiler warnings in sieve.c.
2562 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2563 function verify_get_ident() calls ip_connect() to connect a socket, but
2564 if the "connect()" function timed out, ip_connect() used to close the
2565 socket. However, verify_get_ident() also closes the socket later, and in
2566 between Exim writes to the log, which may get opened at this point. When
2567 the socket was closed in ip_connect(), the log could get the same file
2568 descriptor number as the socket. This naturally causes chaos. The fix is
2569 not to close the socket in ip_connect(); the socket should be closed by
2570 the function that creates it. There was only one place in the code where
2571 this was missing, in the iplookup router, which I don't think anybody now
2572 uses, but I've fixed it anyway.
2574 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2575 well as to direct DNS lookups. Otherwise the handling of names in host
2576 lists is inconsistent and therefore confusing.
2582 PH/01 Two changes to the default runtime configuration:
2584 (1) Move the checks for relay_from_hosts and authenticated clients from
2585 after to before the (commented out) DNS black list checks.
2587 (2) Add control=submission to the relay_from_hosts and authenticated
2588 clients checks, on the grounds that messages accepted by these
2589 statements are most likely to be submissions.
2591 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2593 (1) Generate an error if the third argument for the ${prvs expansion is
2596 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2599 (3) Reset the variables that are obtained from the first argument of
2600 ${prvscheck and used in the second argument before leaving the code,
2601 because their memory is reclaimed, so using them afterwards may do
2604 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2605 one (it's much easier than Tom thought :-).
2607 (5) Because of (4), we can now allow for the use of $prvscheck_result
2608 inside the third argument.
2610 PH/03 For some reason, the default setting of PATH when running a command from
2611 a pipe transport was just "/usr/bin". I have changed it to
2614 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2615 anything to be listed in the output from -bV.
2617 PH/05 When a filter generated an autoreply, the entire To: header line was
2618 quoted in the delivery log line, like this:
2620 => >A.N.Other <ano@some.domain> <original@ddress> ...
2622 This has been changed so that it extracts the operative address. There
2623 may be more than one such address. If so, they are comma-separated, like
2626 => >ano@some.domain,ona@other.domain <original@ddress> ...
2628 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2629 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2630 not being looked up in the rDNS to get a host name, Exim was showing the
2631 IP address twice in Received: lines, even though the IP addresses were
2632 identical. For example:
2634 Received: from [1.2.3.4] (helo=[1.2.3.4])
2636 However, if the real host name was known, it was omitting the HELO data
2637 if it matched the actual IP address. This has been tidied up so that it
2638 doesn't show the same IP address twice.
2640 PH/07 When both +timestamp and +memory debugging was on, the value given by
2641 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2642 called by the debug printing, thereby overwriting the timestamp buffer.
2643 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2645 PH/08 When the original message was included in an autoreply transport, it
2646 always said "this is a copy of the message, including all the headers",
2647 even if body_only or headers_only was set. It now gives an appropriate
2650 PH/09 Applied a patch from the Sieve maintainer which:
2652 o fixes some comments
2653 o adds the (disabled) notify extension core
2654 o adds some debug output for the result of if/elsif tests
2655 o points to the current vacation draft in the documentation
2656 and documents the missing references header update
2660 o fixes a bug in processing the envelope test (when testing
2661 multiple envelope elements, the last element determinted the
2664 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2665 Electronic Mail") by including:
2667 Auto-submitted: auto-generated
2669 in the messages that it generates (bounce messages and others, such as
2670 warnings). In the case of bounce messages for non-SMTP mesages, there was
2671 also a typo: it was using "Auto_submitted" (underscore instead of
2672 hyphen). Since every message generated by Exim is necessarily in response
2673 to another message, thes have all been changed to:
2675 Auto-Submitted: auto-replied
2677 in accordance with these statements in the RFC:
2679 The auto-replied keyword:
2681 - SHOULD be used on messages sent in direct response to another
2682 message by an automatic process,
2684 - MUST NOT be used on manually-generated messages,
2686 - MAY be used on Delivery Status Notifications (DSNs) and Message
2687 Disposition Notifications (MDNs),
2689 - MUST NOT be used on messages generated by automatic or periodic
2690 processes, except for messages which are automatic responses to
2693 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2694 to the default Received: header definition.
2696 PH/12 Added log selector acl_warn_skipped (default on).
2698 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2699 variables because (a) they are in the wrong storage pool and (b) even if
2700 they were copied, it wouldn't work properly because of the caching.
2702 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2703 checking when decoding. Apparently there are clients that generate
2704 overlong encoded strings. Why am I not surprised?
2706 PH/15 If the first argument of "${if match_address" was not empty, but did not
2707 contain an "@" character, Exim crashed. Now it writes a panic log message
2708 and treats the condition as false.
2710 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2712 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2713 extension "envelope-auth". The code is finished and in agreement with
2714 other implementations, but there is no documentation so far and in fact,
2715 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2716 not changing the active code.
2718 Print executed "if" and "elsif" statements when debugging is used. This
2719 helps a great deal to understand what a filter does.
2721 Document more things not specified clearly in RFC3028. I had all this
2722 sorted out, when out of a sudden new issues came to my mind. Oops."
2724 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2727 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2728 canonical form (as documented). However, after a host name lookup from
2729 the IP address, check_host() was doing a simple string comparison with
2730 addresses acquired from the DNS when checking that the found name did
2731 have the original IP as one of its addresses. Since any found IPv6
2732 addresses are likely to be in abbreviated form, the comparison could
2733 fail. Luckily, there already exists a function for doing the comparison
2734 by converting both addresses to binary, so now that is used instead of
2735 the text comparison.
2737 PH/20 There was another similar case to PH/19, when a complete host name was
2738 given in a host list; looking up its IP address could give an abbreviated
2739 form, whereas the current host's name might or might not be abbreviated.
2740 The same fix has been applied.
2746 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2747 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2750 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2751 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2753 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2755 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2756 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2757 Also, for "Auto-Submitted", treat anything other than "no" as
2758 non-personal, in accordance with RFC 3834. (Previously it treated
2759 anything starting "auto-" as non-personal.)
2761 TF/01 The control=submission/name=... option had a problem with syntax
2762 errors if the name included a slash character. The /name= option
2763 now slurps the rest of the string, so it can include any characters
2764 but it must come last in the list of options (after /sender_retain
2767 PH/05 Some modifications to the interface to the fake nameserver for the new
2775 TK/01 Added the "success_on_redirect" address verification option. See
2776 NewStuff for rationale and an example.
2778 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2780 PH/02 Patch to exigrep to allow it to work on syslog lines.
2782 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2783 fread() to skip over the body file's header line, because in Cygwin the
2784 header line is locked and is inaccessible.
2786 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2787 co-exist for some time) to make it clear that it is the Exim ID that is
2788 referenced, not the Message-ID: header line.
2790 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2791 string_format() function, because snprintf() does not exist on all
2794 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2795 ${stat: expansion item.
2797 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2798 protocol synchronization error", to keep the pedants happy.
2800 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2801 well as for IRIX systems, when gcc is being used. See the host.c source
2804 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2806 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2809 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2810 transport and to the smtp transport in LMTP mode.
2812 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2814 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2815 run a filter in a subprocess. This could lead to confusion in subsequent
2816 lookups in the parent process. There should also be a search_tidyup() at
2817 the end of the subprocess.
2819 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2820 only if the host matched helo_try_verify_hosts, which caused the
2821 verification to occur when the EHLO/HELO command was issued. The ACL just
2822 tested the remembered result. Now, if a previous verification attempt has
2823 not happened, "verify = helo" does it there and then.
2825 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2827 TK/03 Fix log output including CR from clamd.
2829 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2830 exist provoked a memory error which could cause a segfault.
2832 PH/15 Installed PCRE 6.2
2834 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2836 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2837 of the problem. Specifically, suggested +O2 rather than +O1 for the
2840 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2842 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2843 record, the port was not correctly logged when the outgoing_port log
2844 selector was set (it logged the transort's default port).
2846 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2847 fallback_hosts, and "hosts" in the smtp transport.
2849 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2850 host errors such as "Connection refused".
2852 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2853 authentication with radiusclient 0.4.9:
2855 - Error returned from rc_read_config was caught wrongly
2856 - Username/password not passed on to radius server due to wrong length.
2858 The presumption is that some radiusclient API changes for 4.51/PH/17
2859 were not taken care of correctly. The code is still untested by me (my
2860 Linux distribution still has 0.3.2 of radiusclient), but it was
2861 contributed by a Radius user.
2863 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2864 expanding the "port" option of the smtp transport.
2866 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2867 while reading a MIME header. Thanks to Tom Hughes for a patch.
2869 PH/24 Include config.h inside local_scan.h so that configuration settings are
2872 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2873 that in an ACL for RCPT (for example), you can examine exactly what was
2876 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2877 commands, but it was not correctly comparing the address with the actual
2878 client host address. Thus, it would show the EHLO address in Received:
2879 header lines when this was not necessary.
2881 PH/27 Added the % operator to ${eval:}.
2883 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2884 it should be ignoring failures (because with -C, for example, it has lost
2885 privilege). It wasn't ignoring creation failures other than "already
2888 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2891 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2892 because an input line was too long, either on its own, or by virtue of
2893 too many continuations, the temporary file was not being removed, and the
2894 return code was incorrect.
2896 PH/31 Missing "BOOL" in function definition in filtertest.c.
2898 PH/32 Applied Sieve patches from the maintainer.
2900 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2902 PH/33 Added "verify = not_blind".
2904 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2905 Local/Makefile (with some defaults set). These are used in built scripts
2906 such as exicyclog, but they have never been used in the exim_install
2907 script (though there are many overriding facilities there). I have
2908 arranged that the exim_install script now takes note of these two
2911 PH/35 Installed configuration files for Dragonfly.
2913 PH/36 When a locally submitted message by a trusted user did not contain a
2914 From: header, and the sender address was obtained from -f or from an SMTP
2915 MAIL command, and the trusted user did not use -F to supply a sender
2916 name, $originator_name was incorrectly used when constructing a From:
2917 header. Furthermore, $originator_name was used for submission mode
2918 messages from external hosts without From: headers in a similar way,
2919 which is clearly wrong.
2921 PH/37 Added control=suppress_local_fixups.
2923 PH/38 When log_selector = +received_sender was set, and the addition of the
2924 sender made the log line's construction buffer exactly full, or one byte
2925 less than full, an overflow happened when the terminating "\n" was
2928 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2929 when the result of a list match is failure because a DNS lookup failed.
2931 PH/40 RM_COMMAND is now used in the building process.
2933 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2934 the "build-* directories that it finds.
2936 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2937 address in a domain literal was a prefix of an interface address.
2939 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2940 when verifying a sender address, unless rewrite_headers is false.
2942 PH/44 Wrote a long comment about why errors_to addresses are verified as
2943 recipients, not senders.
2945 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2946 the ratelimit ACL was added.
2948 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2950 PH/46 Added extra information about PostgreSQL errors to the error string.
2952 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2953 avoiding the need to install special zones in a real server. This is
2954 backwards compatible; if it can't find the fake resolver, it drops back.
2955 Thus, both old and new test suites can be run.
2957 TF/02 Added util/ratelimit.pl
2959 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2960 clock is set back in time.
2962 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2963 Candler <B.Candler@pobox.com>.
2965 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2966 OK for addresses that are the result of redirections.
2968 PH/48 A number of further additions for the benefit of the new test suite,
2969 including a fake gethostbyname() that interfaces to the fake DNS resolver
2972 TF/06 The fix for widen_domains has also been applied to qualify_single and
2973 search_parents which are the other dnslookup options that can cause
2976 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2979 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2981 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2982 service" if it encountered an MX record with an empty target hostname.
2983 The message is now "an MX or SRV record indicated no SMTP service".
2985 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2986 if the DNS of the sending site is misconfigured. This is quite a
2987 common situation. This change restores the behaviour of treating a
2988 helo verification defer as a failure.
2990 PH/51 If self=fail was set on a router, the bounce message did not include the
2991 actual error message.
2997 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2999 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3000 command itself ended in error, the underlying message about the transport
3001 filter timeout was being overwritten with the pipe command error. Now the
3002 underlying error message should be appended to the second error message.
3004 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3006 PH/02 Reduce the amount of output that "make" produces by default. Full output
3007 can still be requested.
3009 PH/03 The warning log line about a condition test deferring for a "warn" verb
3010 was being output only once per connection, rather than after each
3011 occurrence (because it was using the same function as for successful
3012 "warn" verbs). This seems wrong, so I have changed it.
3014 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3015 it should not have, which might have caused a crash in the right
3016 circumstances, but probably never did.
3018 PH/04 Installed a modified version of Tony Finch's patch to make submission
3019 mode fix the return path as well as the Sender: header line, and to
3020 add a /name= option so that you can make the user's friendly name appear
3023 TF/03 Added the control = fakedefer ACL modifier.
3025 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3026 Mark Lowes for thorough testing.
3028 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3030 TK/03 Merged latest SRS patch from Miles Wilton.
3032 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3033 with the definition in sysexits.h (which is #included earlier).
3034 Fortunately, Exim does not actually use EX_OK. The code used to try to
3035 preserve the sysexits.h value, by assumimg that macro definitions were
3036 scanned for macro replacements. I have been disabused of this notion,
3037 so now the code just undefines EX_OK before #including unistd.h.
3039 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3040 in the smtp transport. When a block could not be written in a single
3041 write() function, the timeout was being re-applied to each part-write.
3042 This seems wrong - if the receiver was accepting one byte at a time it
3043 would take for ever. The timeout is now adjusted when this happens. It
3044 doesn't have to be particularly precise.
3046 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3047 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3049 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3050 without a domain if the check to <postmaster@domain> fails.
3052 SC/01 Eximstats: added -xls and the ability to specify output files
3053 (patch written by Frank Heydlauf).
3055 SC/02 Eximstats: use FileHandles for outputing results.
3057 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3059 SC/04 Eximstats: fixed display of large numbers with -nvr option
3061 SC/05 Eximstats: fixed merging of reports with empty tables.
3063 SC/06 Eximstats: added the -include_original_destination flag
3065 SC/07 Eximstats: removed tabs and trailing whitespace.
3067 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3069 TK/06 MBOX spool code: Add real "From " MBOX separator line
3070 so the .eml file is really in mbox format (even though
3071 most programs do not really care). Patch from Alex Miller.
3073 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3074 The latter is generated from $received_to and is only set if the
3075 message has one envelope recipient. SA can use these headers,
3076 obviously out-of-the-box. Patch from Alex Miller.
3078 PH/08 The ${def test on a variable was returning false if the variable's
3079 value was "0", contrary to what the specification has always said!
3080 The result should be true unless the variable is empty.
3082 PH/09 The syntax error of a character other than { following "${if
3083 def:variable_name" (after optional whitespace) was not being diagnosed.
3084 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3085 accidental colon was present, for example, could give incorrect results.
3087 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3088 result is used (not including appendfile, where other changes are about
3091 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3092 This involved changing a lot of size variables from int to off_t. It
3093 should work with maildirs and everything.
3095 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3096 spamd dying while we are connected to it.
3098 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3101 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3102 function for BDB 4.3.
3104 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3106 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3107 the sources for PCRE have been re-arranged and more files are now
3110 PH/15 The code I had for printing potentially long long variables in PH/11
3111 above was not the best (it lost precision). The length of off_t variables
3112 is now inspected at build time, and an appropriate printing format (%ld
3113 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3114 to be "long long int" or "long int". This is needed for the internal
3115 formatting function string_vformat().
3117 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3118 the configuration file to be ":syslog", then the script "guesses" where
3119 the logs files are, rather than using the compiled in default. In our
3120 case the guess is not the same as the compiled default, so the script
3121 suddenly stopped working when I started to use syslog. The patch checks
3122 to see if log_file_path is "". If so, it attempts to read it from exim
3123 with no configuration file to get the compiled in version, before it
3124 falls back to the previous guessing code."
3126 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3127 implementing BATV in an Exim configuration. See NewStuff for the gory
3130 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3131 Makefile that are specific to HP-UX.
3133 PH/18 If the "use_postmaster" option was set for a recipient callout together
3134 with the "random" option, the postmaster address was used as the MAIL
3135 FROM address for the random test, but not for the subsequent recipient
3136 test. It is now used for both.
3138 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3139 patch removes a few documentation additions to RFC 3028, because the
3140 latest draft now contains them. It adds the new en;ascii-case comparator
3141 and a new error check for 8bit text in MIME parts. Comparator and
3142 require names are now matched exactly. I enabled the subaddress
3143 extension, but it is not well tested yet (read: it works for me)."
3145 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3146 rework some of the code of TK/09 above to avoid the hardwired use of
3147 "%lld" and "long long". Replaced the call to snprintf() with a call to
3150 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3151 records point to non-existent hosts", "retry timeout exceeded", and
3152 "retry time not reached for any host after a long failure period".
3154 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3155 experimental DomainKeys support:
3157 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3158 (2) On an error such as an illegally used "control", the wrong name for
3159 the control was given.
3161 These problems did NOT occur unless DomainKeys support was compiled.
3163 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3165 PH/24 Added ${if match_ip condition.
3167 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3168 databases so that it will be absolutely obvious if a crash occurs in the
3169 DB library. This is a regular occurrence (often caused by mis-matched
3172 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3173 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3174 picked up on a user's system that detects such things. There doesn't seem
3175 to be a gcc warning option for this - only an attribute that has to be
3176 put on the function's prototype. It seems that in Fedora Core 4 they have
3177 set this on a number of new functions. No doubt there will be more in due
3180 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3181 specify a transport. However, if an address that was verified by such a
3182 router was the subject of a callout, Exim crashed because it tried to
3183 read the rcpt_include_affixes from the non-existent transport. Now it
3184 just assumes that the setting of that option is false. This bug was
3185 introduced by 4.51/PH/31.
3187 PH/28 Changed -d+all to exclude +memory, because that information is very
3188 rarely of interest, but it makes the output a lot bigger. People tend to
3189 do -d+all out of habit.
3191 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3192 code in os-type was giving problems when libc.so lives in lib64, like on
3195 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3196 aren't the modern standard, and it seems that some systems' include files
3197 don't always have them. Exim was already checking for some of the newer
3198 ones like T_AAAA, and defining it itself. I've added checks for all the
3199 record types that Exim uses.
3201 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3202 not automatically generating a new one, as it is supposed to. This
3203 prevented TLS from working. If the file did exist, but contained invalid
3204 data, a new version was generated, as expected. It was only the case of a
3205 non-existent file that was broken.
3207 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3208 with a change in libdomainkeys > 0.64.
3210 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3211 from DNS. If the selector record carries the flag, it now has
3212 precedence over the domain-wide flag.
3214 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3216 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3217 the use of a port name that isn't defined in /etc/services) occurred, the
3218 message was deferred as in a normal delivery, and thus remained on the
3219 spool, instead of being failed because of the mua_wrapper setting. This
3220 is now fixed, and I tidied up some of the mua_wrapper messages at the
3223 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3224 the messages in a hash of arrays rather than using individual hashes.
3225 This is a bit cleaner and results in dramatic memory savings, albeit
3226 at a slight CPU cost.
3228 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3229 as requested by Marc Sherman.
3231 SC/10 Eximstats: added histograms for user specified patterns as requested
3234 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3236 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3237 fopen() in the content-scanning modules that did not already have it.
3243 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3244 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3246 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3248 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3250 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3251 to test Sieve filters that use "vacation".
3253 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3254 that changes the way the GnuTLS parameters are stored in the cache file.
3255 The new format can be generated externally. For backward compatibility,
3256 if the data in the cache doesn't make sense, Exim assumes it has read an
3257 old-format file, and it generates new data and writes a new file. This
3258 means that you can't go back to an older release without removing the
3261 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3262 work if there are any delivery delays because "one_time" forces the
3263 parent to be marked "delivered", so its unseen clone is never tried
3264 again. For this reason, Exim now forbids the simultaneous setting of
3267 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3268 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3269 entry if you want to know the details. The fix, however, neglected to
3270 consider the case when local delivery batching is involved. The test for
3271 "previously delivered" was not happening when checking to see if an
3272 address could be batched with a previous (undelivered) one; under
3273 certain circumstances this could lead to multiple deliveries to the same
3276 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3277 in its include files, and this causes problems building Exim.
3279 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3280 header_syntax) but Exim was just ignoring anything given after a slash.
3281 In particular, this caused confusion with an attempt to use "verify =
3282 reverse_host_lookup/defer_ok". An error is now given when options are
3283 supplied for verify items that do not have them. (Maybe reverse_host_
3284 lookup should have a defer_ok option, but that's a different point.)
3286 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3287 defined by RFC 821) to 2048, because there were problems with some AUTH
3288 commands, and RFC 1869 says the size should be increased for extended
3289 SMTP commands that take arguments.
3291 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3294 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3295 "unknown" error; now it says that the functionality isn't in the binary.
3297 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3298 an address' error message when a string expansion fails (syntax or
3299 whatever). Otherwise the password may appear in the log. Following change
3300 PH/42 below, there is no longer a chance of it appearing in a bounce
3303 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3305 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3306 of its addresses was ever tried. (Bugzilla bug #2.)
3308 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3309 the result incorrectly in the debug output. (It correctly added a newline
3310 to what was transported.)
3312 TF/01 Added $received_time.
3314 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3315 commented out examples of how to interface to a virus scanner and to
3316 SpamAssassin. Also added commented examples of av_scanner and
3317 spamd_address settings.
3319 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3320 and controls are allowed in which ACLs. There were a couple of minor
3321 errors. Some of the entries in the conditions table (which is a table of
3322 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3323 negation of where the condition IS allowed.
3325 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3327 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3328 header file does not have a version number, so I've had to invent a new
3329 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3330 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3331 radiusclient), but it was contributed by a Radius user.
3333 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3334 files or directories, for OpenSSL.
3336 PH/19 When an Exim process that is running as root has to create an Exim log
3337 file, it does so in a subprocess that runs as exim:exim so as to get the
3338 ownership right at creation (otherwise, other Exim processes might see
3339 the file with the wrong ownership). There was no test for failure of this
3340 fork() call, which would lead to the process getting stuck as it waited
3341 for a non-existent subprocess. Forks do occasionally fail when resources
3342 run out. I reviewed all the other calls to fork(); they all seem to check
3345 PH/20 When checking for unexpected SMTP input at connect time (before writing
3346 the banner), Exim was not dealing correctly with a non-positive return
3347 from the read() function. If the client had disconnected by this time,
3348 the result was a log entry for a synchronization error with an empty
3349 string after "input=" when read() returned zero. If read() returned -1
3350 (an event I could not check), uninitialized data bytes were printed.
3351 There were reports of junk text (parts of files, etc) appearing after
3354 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3356 PH/22 Added support for macro redefinition, and (re)definition in between
3357 driver and ACL definitions.
3359 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3360 forgetting to use the resulting value; it was using the unexpanded value.
3362 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3363 hadn't been configured. The fix is from Juergen Kreileder, who
3364 understands it better than I do:
3366 "Here's what I see happening with three configured cyrus_sasl
3367 authenticators configured (plain, login, cram-md5):
3369 On startup auth_cyrus_sasl_init() gets called for each of these.
3370 This means three calls to sasl_listmech() without a specified mech_list.
3371 => SASL tests which mechs of all available mechs actually work
3372 => three warnings about OTP not working
3373 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3375 With the patch, sasl_listmech() also gets called three times. But now
3376 SASL's mech_list option is set to the server_mech specified in the the
3377 authenticator. Or in other words, the answer from sasl_listmech()
3378 gets limited to just the mech you're testing for (which is different
3380 => the return list contains just 'plain' or 'login', 'cram-md5' or
3381 nothing depending on the value of ob->server_mech.
3383 I've just tested the patch: Authentication still works fine,
3384 unavailable mechs specified in the exim configuration are still
3385 caught, and the auth.log warnings about OTP are gone."
3387 PH/25 When debugging is enabled, the contents of the command line are added
3388 to the debugging output, even when log_selector=+arguments is not
3391 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3392 answer is "GNU", and only if the return is "GNU/something" is the answer
3395 PH/27 $acl_verify_message is now set immediately after the failure of a
3396 verification in an ACL, and so is available in subsequent modifiers. In
3397 particular, the message can be preserved by coding like this:
3399 warn !verify = sender
3400 set acl_m0 = $acl_verify_message
3402 Previously, $acl_verify_message was set only while expanding "message"
3403 and "log_message" when a very denied access.
3405 PH/28 Modified OS/os.c-Linux with
3407 -#ifndef OS_LOAD_AVERAGE
3408 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3410 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3411 nomenclature these days.)
3413 PH/29 Installed patch from the Sieve maintainer that adds the options
3414 sieve_useraddress and sieve_subaddress to the redirect router.
3416 PH/30 In these circumstances:
3417 . Two addresses routed to the same list of hosts;
3418 . First host does not offer TLS;
3419 . First host accepts first address;
3420 . First host gives temporary error to second address;
3421 . Second host offers TLS and a TLS session is established;
3422 . Second host accepts second address.
3423 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3424 and peerdn, if requested) that were in fact used only for the second
3427 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3428 attention to any local part prefix or suffix that was matched by the
3429 router that accepted the address. It now behaves in the same way as it
3430 does for delivery: the affixes are removed from the local part unless
3431 rcpt_include_affixes is set on the transport.
3433 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3434 timeout during the DATA phase of an incoming message.
3436 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3437 applied a patch sanctioned by the Sieve maintainer.
3439 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3440 the uid or gid is negative. A case of a negative gid caused this to be
3441 noticed. The fix allows for either to be negative.
3443 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3444 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3447 PH/36 The change PH/12 above was broken. Fixed it.
3449 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3450 the grounds that routing the same address twice would always produce the
3451 same answer. This might have been true once, but it is certainly no
3452 longer true now. Routing a child address may depend on the previous
3453 routing that produced that child. Some complicated redirection strategies
3454 went wrong when messages had multiple recipients, and made Exim's
3455 behaviour dependent on the order in which the addresses were given.
3457 I have moved the duplicate checking until after the routing is complete.
3458 Exim scans the addresses that are assigned to local and remote
3459 transports, and removes any duplicates. This means that more work will be
3460 done, as duplicates will always all be routed, but duplicates are
3461 presumably rare, so I don't expect this is of any significance.
3463 For deliveries to pipes, files, and autoreplies, the duplicate checking
3464 still happens during the routing process, since they are not going to be
3467 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3468 It corrects a timeout issue with spamd. This is Ian's comment: "The
3469 background is that sometimes spamd either never reads data from a
3470 connection it has accepted, or it never writes response data. The exiscan
3471 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3472 blindly assumes that writes won't block so it may never time out."
3474 PH/39 Allow G after quota size as well as K and M.
3476 PH/40 The value set for $authenticated_id in an authenticator may not contain
3477 binary zeroes or newlines because the value is written to log lines and
3478 to spool files. There was no check on this. Now the value is run through
3479 the string_printing() function so that such characters are converted to
3480 printable escape sequences.
3482 PH/41 $message_linecount is a new variable that contains the total number of
3483 lines in the message. Compare $body_linecount, which is the count for the
3486 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3487 bounce and delay warning messages, except in certain special cases, which
3490 (a) An SMTP error message from a remote host;
3491 (b) A message specified in a :fail: redirection;
3492 (c) A message specified in a "fail" command in a system filter;
3493 (d) A message specified in a FAIL return from the queryprogram router;
3494 (e) A message specified by the cannot_route_message router option.
3496 In these cases only, Exim does include the error details in bounce and
3497 warning messages. There are also a few cases where bland messages such
3498 as "unrouteable address" or "local delivery error" are given.
3500 PH/43 $value is now also set for the "else" part of a ${run expansion.
3502 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3503 being worked on, but at least Exim now implements the latest version to
3506 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3507 process to complete was treated as a delivery failure, a timeout while
3508 writing the message to the pipe was logged, but erroneously treated as a
3509 successful delivery. Such timeouts include transport filter timeouts. For
3510 consistency with the overall process timeout, these timeouts are now
3511 treated as errors, giving rise to delivery failures by default. However,
3512 there is now a new Boolean option for the pipe transport called
3513 timeout_defer, which, if set TRUE, converts the failures into defers for
3514 both kinds of timeout. A transport filter timeout is now identified in
3517 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3518 systems where "make" and "gmake" are different, calling "gmake" at top
3519 level broke things. I've arranged for the value of $(MAKE) to be passed
3520 from the Makefile to this script so that it can call the same version of
3524 A note about Exim versions 4.44 and 4.50
3525 ----------------------------------------
3527 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3528 changes of various kinds. As a consequence, a big documentation update was
3529 needed. This delayed the release for rather longer than seemed good, especially
3530 in the light of a couple of (minor) security issues. Therefore, the changes
3531 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3532 release. So 4.44 and 4.50 are in effect two different branches that both start
3535 I have left the 4.50 change log unchanged; it contains all the changes since
3536 4.43. The change log for 4.44 is below; many of its items are identical to
3537 those for 4.50. This seems to be the most sensible way to preserve the
3538 historical information.
3544 1. Minor wording change to the doc/README.SIEVE file.
3546 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3547 computation of the current number of files was incorrect.
3549 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3550 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3553 4. Give more explanation in the error message when the command for a transport
3554 filter fails to execute.
3556 5. There are several places where Exim runs a non-Exim command in a
3557 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3558 was being done only for the command run by the queryprogram router. It is
3559 now done for all such subprocesses. The other cases are: ${run, transport
3560 filters, and the commands run by the lmtp and pipe transports.
3562 6. Added CONFIGURE_GROUP build-time option.
3564 7. Some older OS have a limit of 256 on the maximum number of file
3565 descriptors. Exim was using setrlimit() to set 1000 as a large value
3566 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3567 systems. I've change it so that if it can't get 1000, it tries for 256.
3569 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3570 was an oversight, and furthermore, ever since the addition of extra
3571 controls (e.g. 4.43/32), the checks on when to allow different forms of
3572 "control" were broken. There should now be diagnostics for all cases when a
3573 control that does not make sense is encountered.
3575 9. Added the /retain_sender option to "control=submission".
3577 10. $recipients is now available in the predata ACL (oversight).
3579 11. Tidy the search cache before the fork to do a delivery from a message
3580 received from the command line. Otherwise the child will trigger a lookup
3581 failure and thereby defer the delivery if it tries to use (for example) a
3582 cached ldap connection that the parent has called unbind on.
3584 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3585 of $address_data from the recipient verification was clobbered by the
3586 sender verification.
3588 13. The value of address_data from a sender verification is now available in
3589 $sender_address_data in subsequent conditions in the ACL statement.
3591 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3593 15. Added a new option "connect=<time>" to callout options, to set a different
3596 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3597 was its contents. (It was OK if the option was not defined at all.)
3599 17. A "Completed" log line is now written for messages that are removed from
3600 the spool by the -Mrm option.
3602 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3603 information about exactly what failed.
3605 19. Added -dd to debug only the daemon process.
3607 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3608 handles timeouts, both on the server side and network timeouts. Renamed the
3609 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3611 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3612 It is now set to "smtps".
3614 22. $host_address is now set to the target address during the checking of
3615 ignore_target_hosts.
3617 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3618 being passed; this would have caused $sender_host_name to have been used if
3619 matching the list had actually called for a host name (not very likely,
3620 since this list is usually IP addresses). A host name is now passed as
3623 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3624 code that specifies a non-restarting handler (typically sigaction() in
3625 modern systems) in an attempt to fix a rare and obscure crash bug.
3627 25. Narrowed the window for a race in the daemon that could cause it to ignore
3628 SIGCHLD signals. This is not a major problem, because they are used only to
3629 wake it up if nothing else does.
3631 26. A malformed maildirsize file could cause Exim to calculate negative values
3632 for the mailbox size or file count. Odd effects could occur as a result.
3633 The maildirsize information is now recalculated if the size or filecount
3636 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3637 support for a long time. Removed HAVE_SYS_VFS_H.
3639 28. Installed the latest version of exipick from John Jetmore.
3641 29. In an address list, if the pattern was not a regular expression, an empty
3642 subject address (from a bounce message) matched only if the pattern was an
3643 empty string. Non-empty patterns were not even tested. This was the wrong
3644 because it is perfectly reasonable to use an empty address as part of a
3645 database query. An empty address is now tested by patterns that are
3646 lookups. However, all the other forms of pattern expect the subject to
3647 contain a local part and a domain, and therefore, for them, an empty
3648 address still always fails if the pattern is not itself empty.
3650 30. Exim went into a mad DNS loop when attempting to do a callout where the
3651 host was specified on an smtp transport, and looking it up yielded more
3652 than one IP address.
3654 31. Re-factored the code for checking spool and log partition space into a
3655 function that finds that data and another that does the check. The former
3656 is then used to implement four new variables: $spool_space, $log_space,
3657 $spool_inodes, and $log_inodes.
3659 32. The RFC2047 encoding function was originally intended for short strings
3660 such as real names; it was not keeping to the 75-character limit for
3661 encoded words that the RFC imposes. It now respects the limit, and
3662 generates multiple encoded words if necessary. To be on the safe side, I
3663 have increased the buffer size for the ${rfc2047: expansion operator from
3666 33. It is now permitted to omit both strings after an "if" condition; if the
3667 condition is true, the result is "true". As before, when the second string
3668 is omitted, a false condition yields an empty string. This makes it less
3669 cumbersome to write custom ACL and router conditions.
3671 34. Failure to deliver a bounce message always caused it to be frozen, even if
3672 there was an errors_to setting on the router. The errors_to setting is now
3675 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3676 canonical form (fully expanded) before being placed in
3677 $sender_host_address.
3679 36. The table in the code that translates DNS record types into text (T_A to
3680 "A" for instance) was missing entries for NS and CNAME. It is just possible
3681 that this could have caused confusion if both these types were looked up
3682 for the same domain, because the text type is used as part of Exim's
3683 per-process caching. But the chance of anyone hitting this buglet seems
3686 37. The dnsdb lookup has been extended in a number of ways.
3688 (1) There is a new type, "zns", which walks up the domain tree until it
3689 finds some nameserver records. It should be used with care.
3691 (2) There is a new type, "mxh", which is like "mx" except that it returns
3692 just the host names, not the priorities.
3694 (3) It is now possible to give a list of domains (or IP addresses) to be
3695 looked up. The behaviour when one of the lookups defers can be
3696 controlled by a keyword.
3698 (4) It is now possible to specify the separator character for use when
3699 multiple records are returned.
3701 38. The dnslists ACL condition has been extended: it is now possible to supply
3702 a list of IP addresses and/or domains to be looked up in a particular DNS
3705 39. Added log_selector=+queue_time_overall.
3707 40. When running the queue in the test harness, wait just a tad after forking a
3708 delivery process, to get repeatability of debugging output.
3710 41. Include certificate and key file names in error message when GnuTLS fails
3711 to set them up, because the GnuTLS error message doesn't include the name
3712 of the failing file when there is a problem reading it.
3714 42. Allow both -bf and -bF in the same test run.
3716 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3718 44. The "Exiscan patch" is now merged into the mainline Exim source.
3720 45. Sometimes the final signoff response after QUIT could fail to get
3721 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3722 < 0 before doing a fflush(). This bug looks as though it goes back to the
3723 introduction of TLS in release 3.20, but "sometimes" must have been rare
3724 because the tests only now provoked it.
3726 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3727 (this can affect the format of dates).
3729 47. exim_tidydb, when checking for the continued existence of a message for
3730 which it has found a message-specific retry record, was not finding
3731 messages that were in split spool directories. Consequently, it was
3732 deleting retry records that should have stayed in existence.
3734 48. Steve fixed some bugs in eximstats.
3736 49. The SPA authentication driver was not abandoning authentication and moving
3737 on to the next authenticator when an expansion was forced to fail,
3738 contradicting the general specification for all authenticators. Instead it
3739 was generating a temporary error. It now behaves as specified.
3741 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3742 (the order specifies the preference for clients). The order is now AES256,
3743 AES128, 3DES, ARCFOUR128.
3745 51. Small patch to Sieve code - explicitly set From: when generating an
3748 52. Exim crashed if a remote delivery caused a very long error message to be
3749 recorded - for instance if somebody sent an entire SpamAssassin report back
3750 as a large number of 550 error lines. This bug was coincidentally fixed by
3751 increasing the size of one of Exim's internal buffers (big_buffer) that
3752 happened as part of the Exiscan merge. However, to be on the safe side, I
3753 have made the code more robust (and fixed the comments that describe what
3756 53. Now that there can be additional text after "Completed" in log lines (if
3757 the queue_time_overall log selector is set), a one-byte patch to exigrep
3758 was needed to allow it to recognize "Completed" as not the last thing in
3761 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3762 patch that reportedly fixes this has been added. I am not expert enough to
3763 create a test for it. This is what the patch creator wrote:
3765 "I found a little strange behaviour of ldap code when working with
3766 Windows 2003 AD Domain, where users was placed in more than one
3767 Organization Units. When I tried to give exim partial DN, the exit code
3768 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3769 But simultaneously result of request was absolutely normal ldap result,
3770 so I produce this patch..."
3772 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3773 REFERENCE, so I have modified the code to exclude the patch when that macro
3776 55. Some experimental protocols are using DNS PTR records for new purposes. The
3777 keys for these records are domain names, not reversed IP addresses. The
3778 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3779 leaves it alone. Component reversal etc. now happens only for IP addresses.
3782 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3784 57. Double the size of the debug message buffer (to 2048) so that more of very
3785 long debug lines gets shown.
3787 58. The exicyclog utility now does better if the number of log files to keep
3788 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3790 59. Two changes related to the smtp_active_hostname option:
3792 (1) $smtp_active_hostname is now available as a variable.
3793 (2) The default for smtp_banner uses $smtp_active_hostname instead
3794 of $primary_hostname.
3796 60. The host_aton() function is supposed to be passed a string that is known
3797 to be a valid IP address. However, in the case of IPv6 addresses, it was
3798 not checking this. This is a hostage to fortune. Exim now panics and dies
3799 if the condition is not met. A case was found where this could be provoked
3800 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3801 components; fortuitously, this particular loophole had already been fixed
3802 by change 4.50/55 above.
3804 If there are any other similar loopholes, the new check in host_aton()
3805 itself should stop them being exploited. The report I received stated that
3806 data on the command line could provoke the exploit when Exim was running as
3807 exim, but did not say which command line option was involved. All I could
3808 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3809 running as the user.
3812 61. There was a buffer overflow vulnerability in the SPA authentication code
3813 (which came originally from the Samba project). I have added a test to the
3814 spa_base64_to_bits() function which I hope fixes it.
3817 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3818 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3820 63. The daemon start-up calls getloadavg() while still root for those OS that
3821 need the first call to be done as root, but it missed one case: when
3822 deliver_queue_load_max is set with deliver_drop_privilege. This is
3823 necessary for the benefit of the queue runner, because there is no re-exec
3824 when deliver_drop_privilege is set.
3826 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3827 This has been fixed.
3829 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3830 was in use, was not putting the data itself into the right store pool;
3831 consequently, it could be overwritten for a subsequent message in the same
3832 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3835 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3837 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3838 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3839 the function were treating the return as a boolean value, which happened to
3840 work because 0=false and not-0=true, but is not correct code.
3842 68. The host_aton() function was not handling scoped IPv6 addresses (those
3843 with, for example, "%eth0" on the end) correctly.
3845 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3846 negated items (that is, ~something) in unsigned ints. Some compilers
3847 apparently mutter when there is no cast.
3849 70. If an address verification called from an ACL failed, and did not produce a
3850 user-specific message (i.e. there was only a "system" message), nothing was
3851 put in $acl_verify_message. In this situation, it now puts the system
3854 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3855 session; change 4.31/43 added the unwanted input to the log line - except
3856 that it did not do this in the start of session case. It now does.
3858 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3859 This is wrong and can cause the other end to generate a synchronization
3860 error if it is another Exim or anything else that does the synchronization
3861 check. A QUIT command is no longer sent after a timeout.
3863 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3864 during host lookups.
3866 74. The defer_ok option of callout verification was not working if it was used
3867 when verifying addresses in header lines, that is, for this case:
3869 verify = header_sender/callout=defer_ok
3871 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3872 those file descriptors could be used for SMTP connections. If anything
3873 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3874 could be sent to the SMTP client, causing chaos. The daemon now opens
3875 stdin, stdout, and stderr to /dev/null when it puts itself into the
3878 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3879 log by default. The user can override this with suitable Perl magic.
3881 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3882 the log message when discard triggers, was not working for the DATA ACL or
3883 for the non-SMTP ACL.
3885 78. Error message wording change in sieve.c.
3887 79. If smtp_accept_max_per_host was set, the number of connections could be
3888 restricted to fewer than expected, because the daemon was trying to set up
3889 a new connection before checking whether the processes handling previous
3890 connections had finished. The check for completed processes is now done
3891 earlier. On busy systems, this bug wouldn't be noticed because something
3892 else would have woken the daemon, and it would have reaped the completed
3895 80. If a message was submitted locally by a user whose login name contained one
3896 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3897 It caused a spool format error. I have fixed the spool reading code. A
3898 related problem was that the "from" clause in the Received: line became
3899 illegal because of the space(s). It is now covered by ${quote_local_part.
3901 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3904 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3906 83. Patch from Sieve maintainer to fix unterminated string problem in
3907 "vacation" handling.
3909 84. Some minor changes to the Linux configuration files to help with other
3910 OS variants using glibc.
3912 85. One more patch for Sieve to update vacation handling to latest spec.
3915 ----------------------------------------------------
3916 See the note above about the 4.44 and 4.50 releases.
3917 ----------------------------------------------------
3923 1. Change 4.43/35 introduced a bug that caused file counts to be
3924 incorrectly computed when quota_filecount was set in an appendfile
3927 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3928 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3931 3. Give more explanation in the error message when the command for a transport
3932 filter fails to execute.
3934 4. There are several places where Exim runs a non-Exim command in a
3935 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3936 was being done only for the command run by the queryprogram router. It is
3937 now done for all such subprocesses. The other cases are: ${run, transport
3938 filters, and the commands run by the lmtp and pipe transports.
3940 5. Some older OS have a limit of 256 on the maximum number of file
3941 descriptors. Exim was using setrlimit() to set 1000 as a large value
3942 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3943 systems. I've change it so that if it can't get 1000, it tries for 256.
3945 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3946 was an oversight, and furthermore, ever since the addition of extra
3947 controls (e.g. 4.43/32), the checks on when to allow different forms of
3948 "control" were broken. There should now be diagnostics for all cases when a
3949 control that does not make sense is encountered.
3951 7. $recipients is now available in the predata ACL (oversight).
3953 8. Tidy the search cache before the fork to do a delivery from a message
3954 received from the command line. Otherwise the child will trigger a lookup
3955 failure and thereby defer the delivery if it tries to use (for example) a
3956 cached ldap connection that the parent has called unbind on.
3958 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3959 of $address_data from the recipient verification was clobbered by the
3960 sender verification.
3962 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3963 was its contents. (It was OK if the option was not defined at all.)
3965 11. A "Completed" log line is now written for messages that are removed from
3966 the spool by the -Mrm option.
3968 12. $host_address is now set to the target address during the checking of
3969 ignore_target_hosts.
3971 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3972 being passed; this would have caused $sender_host_name to have been used if
3973 matching the list had actually called for a host name (not very likely,
3974 since this list is usually IP addresses). A host name is now passed as
3977 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3978 code that specifies a non-restarting handler (typically sigaction() in
3979 modern systems) in an attempt to fix a rare and obscure crash bug.
3981 15. Narrowed the window for a race in the daemon that could cause it to ignore
3982 SIGCHLD signals. This is not a major problem, because they are used only to
3983 wake it up if nothing else does.
3985 16. A malformed maildirsize file could cause Exim to calculate negative values
3986 for the mailbox size or file count. Odd effects could occur as a result.
3987 The maildirsize information is now recalculated if the size or filecount
3990 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3991 support for a long time. Removed HAVE_SYS_VFS_H.
3993 18. Updated exipick to current release from John Jetmore.
3995 19. Allow an empty sender to be matched against a lookup in an address list.
3996 Previously the only cases considered were a regular expression, or an
3999 20. Exim went into a mad DNS lookup loop when doing a callout where the
4000 host was specified on the transport, if the DNS lookup yielded more than
4003 21. The RFC2047 encoding function was originally intended for short strings
4004 such as real names; it was not keeping to the 75-character limit for
4005 encoded words that the RFC imposes. It now respects the limit, and
4006 generates multiple encoded words if necessary. To be on the safe side, I
4007 have increased the buffer size for the ${rfc2047: expansion operator from
4010 22. Failure to deliver a bounce message always caused it to be frozen, even if
4011 there was an errors_to setting on the router. The errors_to setting is now
4014 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4015 canonical form (fully expanded) before being placed in
4016 $sender_host_address.
4018 24. Updated eximstats to version 1.33
4020 25. Include certificate and key file names in error message when GnuTLS fails
4021 to set them up, because the GnuTLS error message doesn't include the name
4022 of the failing file when there is a problem reading it.
4024 26. Expand error message when OpenSSL has problems setting up cert/key files.
4027 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4028 (this can affect the format of dates).
4030 28. exim_tidydb, when checking for the continued existence of a message for
4031 which it has found a message-specific retry record, was not finding
4032 messages that were in split spool directories. Consequently, it was
4033 deleting retry records that should have stayed in existence.
4035 29. eximstats updated to version 1.35
4036 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4037 1.35 - bugfix such that pie charts by volume are generated correctly
4039 30. The SPA authentication driver was not abandoning authentication and moving
4040 on to the next authenticator when an expansion was forced to fail,
4041 contradicting the general specification for all authenticators. Instead it
4042 was generating a temporary error. It now behaves as specified.
4044 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4045 (the order specifies the preference for clients). The order is now AES256,
4046 AES128, 3DES, ARCFOUR128.
4048 31. Small patch to Sieve code - explicitly set From: when generating an
4051 32. Exim crashed if a remote delivery caused a very long error message to be
4052 recorded - for instance if somebody sent an entire SpamAssassin report back
4053 as a large number of 550 error lines. This bug was coincidentally fixed by
4054 increasing the size of one of Exim's internal buffers (big_buffer) that
4055 happened as part of the Exiscan merge. However, to be on the safe side, I
4056 have made the code more robust (and fixed the comments that describe what
4059 33. Some experimental protocols are using DNS PTR records for new purposes. The
4060 keys for these records are domain names, not reversed IP addresses. The
4061 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4062 leaves it alone. Component reversal etc. now happens only for IP addresses.
4065 34. The host_aton() function is supposed to be passed a string that is known
4066 to be a valid IP address. However, in the case of IPv6 addresses, it was
4067 not checking this. This is a hostage to fortune. Exim now panics and dies
4068 if the condition is not met. A case was found where this could be provoked
4069 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4070 components; fortuitously, this particular loophole had already been fixed
4071 by change 4.50/55 or 4.44/33 above.
4073 If there are any other similar loopholes, the new check in host_aton()
4074 itself should stop them being exploited. The report I received stated that
4075 data on the command line could provoke the exploit when Exim was running as
4076 exim, but did not say which command line option was involved. All I could
4077 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4078 running as the user.
4081 35. There was a buffer overflow vulnerability in the SPA authentication code
4082 (which came originally from the Samba project). I have added a test to the
4083 spa_base64_to_bits() function which I hope fixes it.
4086 36. The daemon start-up calls getloadavg() while still root for those OS that
4087 need the first call to be done as root, but it missed one case: when
4088 deliver_queue_load_max is set with deliver_drop_privilege. This is
4089 necessary for the benefit of the queue runner, because there is no re-exec
4090 when deliver_drop_privilege is set.
4092 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4093 was in use, was not putting the data itself into the right store pool;
4094 consequently, it could be overwritten for a subsequent message in the same
4095 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4098 38. Sometimes the final signoff response after QUIT could fail to get
4099 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4100 < 0 before doing a fflush(). This bug looks as though it goes back to the
4101 introduction of TLS in release 3.20, but "sometimes" must have been rare
4102 because the tests only now provoked it.
4108 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4109 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4110 Now it returns an integer. A number of calls were still expecting a T/F
4111 return. Fortuitously, in all cases, the tests worked in OK situations,
4112 which is the norm. However, things would have gone wrong on any write
4113 failures on the smtp file descriptor. This function is used when sending
4114 messages over SMTP and also when doing verify callouts.
4116 2. When Exim is called to do synchronous delivery of a locally submitted
4117 message (the -odf or -odi options), it no longer closes stderr before doing
4120 3. Implemented the mua_wrapper option.
4122 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4124 5. Implemented the functions header_remove(), header_testname(),
4125 header_add_at_position(), and receive_remove_recipient(), and exported them
4128 6. If an ACL "warn" statement specified the addition of headers, Exim already
4129 inserted X-ACL-Warn: at the start if there was no header name. However, it
4130 was not making this test for the second and subsequent header lines if
4131 there were newlines in the string. This meant that an invalid header could
4132 be inserted if Exim was badly configured.
4134 7. Allow an ACL "warn" statement to add header lines at the start or after all
4135 the Received: headers, as well as at the end.
4137 8. Added the rcpt_4xx retry error code.
4139 9. Added postmaster_mailfrom=xxx to callout verification option.
4141 10. Added mailfrom=xxxx to the callout verification option, for verify=
4144 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4145 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4147 12. Inserted some casts to stop certain compilers warning when using pointer
4148 differences as field lengths or precisions in printf-type calls (mostly
4149 affecting debugging statements).
4151 13. Added optional readline() support for -be (dynamically loaded).
4153 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4154 same clock tick as a message's arrival, so that its received time was the
4155 same as the "first fail" time on the retry record, and that message
4156 remained on the queue past the ultimate address timeout, every queue runner
4157 would try a delivery (because it was past the ultimate address timeout) but
4158 after another failure, the ultimate address timeout, which should have then
4159 bounced the address, did not kick in. This was a "< instead of <=" error;
4160 in most cases the first failure would have been in the next clock tick
4161 after the received time, and all would be well.
4163 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4164 being recognized when the domain list was tested by the match_domain
4165 condition in an expansion string.
4167 16. Added the ${str2b64: operator.
4169 17. Exim was always calling setrlimit() to set a large limit for the number of
4170 processes, without checking whether the existing limit was already
4171 adequate. (It did check for the limit on file descriptors.) Furthermore,
4172 errors from getrlimit() and setrlimit() were being ignored. Now they are
4173 logged to the main and panic logs, but Exim does carry on, to try to do its
4174 job under whatever limits there are.
4176 18. Imported PCRE 5.0.
4178 19. Trivial typo in log message " temporarily refused connection" (the leading
4181 20. If the log selector return_path_on_delivery was set and an address was
4182 redirected to /dev/null, the delivery process crashed because it assumed
4183 that a return path would always be set for a "successful" delivery. In this
4184 case, the whole delivery is bypassed as an optimization, and therefore no
4187 21. Internal re-arrangement: the function for sending a challenge and reading
4188 a response while authentication was assuming a zero-terminated challenge
4189 string. It's now changed to take a pointer and a length, to allow for
4190 binary data in such strings.
4192 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4194 23. Exim was not respecting finduser_retries when seeking the login of the
4195 uid under which it was called; it was always trying 10 times. (The default
4196 setting of finduser_retries is zero.) Also, it was sleeping after the final
4197 failure, which is pointless.
4199 24. Implemented tls_on_connect_ports.
4201 25. Implemented acl_smtp_predata.
4203 26. If the domain in control=submission is set empty, Exim assumes that the
4204 authenticated id is a complete email address when it generates From: or
4205 Sender: header lines.
4207 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4208 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4209 chown and chgrp in /bin and hostname in /usr/bin.
4211 28. Exim was keeping the "process log" file open after each use, just as it
4212 does for the main log. This opens the possibility of it remaining open for
4213 long periods when the USR1 signal hits a daemon. Occasional processlog
4214 errors were reported, that could have been caused by this. Anyway, it seems
4215 much more sensible not to leave this file open at all, so that is what now
4218 29. The long-running daemon process does not normally write to the log once it
4219 has entered its main loop, and it closes the log before doing so. This is
4220 so that log files can straightforwardly be renamed and moved. However,
4221 there are a couple of unusual error situations where the daemon does write
4222 log entries, and I had neglected to close the log afterwards.
4224 30. The text of an SMTP error response that was received during a remote
4225 delivery was being truncated at 512 bytes. This is too short for some of
4226 the long messages that one sometimes sees. I've increased the limit to
4229 31. It is now possible to make retry rules that apply only when a message has a
4230 specific sender, in particular, an empty sender.
4232 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4233 it possible to be selective about when SMTP synchronization is enforced.
4235 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4237 32. Implemented hosts_connection_nolog.
4239 33. Added an ACL for QUIT.
4241 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4244 35. Added mailbox_size and mailbox_filecount to appendfile.
4246 36. Added control = no_multiline_responses to ACLs.
4248 37. There was a bug in the logic of the code that waits for the clock to tick
4249 in the case where the clock went backwards by a substantial amount such
4250 that the microsecond fraction of "now" was more than the microsecond
4251 fraction of "then" (but the whole seconds number was less).
4253 38. Added support for the libradius Radius client library this is found on
4254 FreeBSD (previously only the radiusclient library was supported).
4260 1. When certain lookups returned multiple values in the form name=value, the
4261 quoting of the values was not always being done properly. Specifically:
4262 (a) If the value started with a double quote, but contained no whitespace,
4264 (b) If the value contained whitespace other than a space character (i.e.
4265 tabs or newlines or carriage returns) it was not quoted.
4266 This fix has been applied to the mysql and pgsql lookups by writing a
4267 separate quoting function and calling it from the lookup code. The fix
4268 should probably also be applied to nisplus, ibase and oracle lookups, but
4269 since I cannot test any of those, I have not disturbed their existing code.
4271 2. A hit in the callout cache for a specific address caused a log line with no
4272 reason for rejecting RCPT. Now it says "Previous (cached) callout
4273 verification failure".
4275 3. There was an off-by-one bug in the queryprogram router. An over-long
4276 return line was truncated at 256 instead of 255 characters, thereby
4277 overflowing its buffer with the terminating zero. As well as fixing this, I
4278 have increased the buffer size to 1024 (and made a note to document this).
4280 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4281 when Exim is waiting for an SMTP response from a remote server, Exim
4282 restarts its select() call on the socket, thereby resetting its timeout.
4283 This is not a problem when such interrupts are rare. Somebody set up a cron
4284 job to run exiwhat every 2 minutes, which is less than the normal select()
4285 timeout (5 or 10 minutes). This meant that the select() timeout never
4286 kicked in because it was always reset. I have fixed this by comparing the
4287 time when an interrupt arrives with the time at the start of the first call
4288 to select(). If more time than the timeout has elapsed, the interrupt is
4289 treated as a timeout.
4291 5. Some internal re-factoring in preparation for the addition of Sieve
4292 extensions (by MH). In particular, the "personal" test is moved to a
4293 separate function, and given an option for scanning Cc: and Bcc: (which is
4294 not set for Exim filters).
4296 6. When Exim created an email address using the login of the caller as the
4297 local part (e.g. when creating a From: or Sender: header line), it was not
4298 quoting the local part when it contained special characters such as @.
4300 7. Installed new OpenBSD configuration files.
4302 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4303 try to make them clearer.
4305 9. Callout options, other than the timeout value, were being ignored when
4306 verifying sender addresses in header lines. For example, when using
4308 verify = header_sender/callout=no_cache
4310 the cache was (incorrectly) being used.
4312 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4313 only the Cygwin environment.
4315 11. When return_path_on_delivery was set as a log selector, if different remote
4316 addresses in the same message used different return paths and parallel
4317 remote delivery occurred, the wrong values would sometimes be logged.
4318 (Whenever a remote delivery process finished, the return path value from
4319 the most recently started remote delivery process was logged.)
4321 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4322 lines when AUTH and/or TLS are in use. This is the "received protocol"
4323 field. Exim used to use "asmtp" for authenticated SMTP, without any
4324 indication (in the protocol name) for TLS use. Now it follows the RFC and
4325 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4326 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4327 names appear in log lines as well as in Received: header lines.
4329 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4330 extensions, and comparison tests, and to fix some bugs.
4332 14. Changes to the "personal" filter test:
4334 (1) The test was buggy in that it was just doing the equivalent of
4335 "contains" tests on header lines. For example, if a user's address was
4336 anne@some.where, the "personal" test would incorrectly be true for
4338 To: susanne@some.where
4340 This test is now done by extracting each address from the header in turn,
4341 and checking the entire address. Other tests that are part of "personal"
4342 are now done using regular expressions (for example, to check local parts
4343 of addresses in From: header lines).
4345 (2) The list of non-personal local parts in From: addresses has been
4346 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4347 taken from the Sieve specification recommendations.
4349 (3) If the message contains any header line starting with "List-" it is
4350 treated as non-personal.
4352 (4) The test for "circular" in the Subject: header line has been removed
4353 because it now seems ill-conceived.
4355 15. Minor typos in src/EDITME comments corrected.
4357 16. Installed latest exipick from John Jetmore.
4359 17. If headers_add on a router specified a text string that was too long for
4360 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4361 of string_sprintf() is now avoided.
4363 18. $message_body_size was not set (it was always zero) when running the DATA
4364 ACL and the local_scan() function.
4366 19. For the "mail" command in an Exim filter, no default was being set for
4367 the once_repeat time, causing a random time value to be used if "once" was
4368 specified. (If the value happened to be <= 0, no repeat happened.) The
4369 default is now 0s, meaning "never repeat". The "vacation" command was OK
4370 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4371 (I found it when inspecting the code).
4373 20. There is now an overall timeout for performing a callout verification. It
4374 defaults to 4 times the callout timeout, which applies to individual SMTP
4375 commands during the callout. The overall timeout applies when there is more
4376 than one host that can be tried. The timeout is checked before trying the
4377 next host. This prevents very long delays if there are a large number of
4378 hosts and all are timing out (e.g. when the network connections are timing
4379 out). The value of the overall timeout can be changed by specifying an
4380 additional sub-option for "callout", called "maxwait". For example:
4382 verify = sender/callout=5s,maxwait=20s
4384 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4385 to the end before writing, but this should make it even safer).
4387 22. Exim was forgetting that it had advertised PIPELINING for the second and
4388 subsequent messages on an SMTP connection. It was also not resetting its
4389 memory on STARTTLS and an internal HELO.
4391 23. When Exim logs an SMTP synchronization error within a session, it now
4392 records whether PIPELINING has been advertised or not.
4394 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4395 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4396 rather than long int.
4398 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4400 26. Added the never_mail option to autoreply.
4406 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4407 crash if the getsockname() call failed; this can happen if a connection is
4408 closed very soon after it is established. The problem was simply in the
4409 order in which certain operations were done, causing Exim to try to write
4410 to the SMTP stream before it had set up the file descriptor. The bug has
4411 been fixed by making things happen in the correct order.
4417 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4418 before the connection was closed, thus losing the rejection response.
4420 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4421 some early Solaris releases, but causes trouble in current releases where
4422 socklen_t is defined.
4424 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4427 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4430 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4431 that releasing the top of it at the end releases what was used for sub-
4432 expansions (unless the block got too big). However, discard this block if
4433 the first thing is a variable or header, so that we can use its block when
4434 it is dynamic (useful for very large $message_headers, for example).
4436 6. Lookups now cache *every* query, not just the most recent. A new, separate
4437 store pool is used for this. It can be recovered when all lookup caches are
4438 flushed. Lookups now release memory at the end of their result strings.
4439 This has involved some general refactoring of the lookup sources.
4441 7. Some code has been added to the store_xxx() functions to reduce the amount
4442 of flapping under certain conditions.
4444 8. log_incoming_interface used to affect only the <= reception log lines. Now
4445 it causes the local interface and port to be added to several more SMTP log
4446 lines, for example "SMTP connection from", and rejection lines.
4448 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4450 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4452 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4453 could be overwritten at the end of the current message (or the start of a
4454 new message if it was set in a HELO ACL). The value is now preserved for
4455 the duration of the SMTP connection.
4457 12. If a transport had a headers_rewrite setting, and a matching header line
4458 contained an unqualified address, that address was qualified, even if it
4459 did not match any rewriting rules. The underlying bug was that the values
4460 of the flags that permit the existence of unqualified sender and recipient
4461 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4462 non-local messages, and by -bnq for local messages) were not being
4463 preserved with the message after it was received.
4465 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4466 "next input=" as part of the text comprising the host identity instead of
4467 the correct text. The code was using the same buffer for two different
4468 strings. However, depending on which order the printing function evaluated
4469 its arguments, the bug did not always show up. Under Linux, for example, my
4470 test suite worked just fine.
4472 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4473 doesn't work with some older versions of Perl. It has been changed to "my",
4474 which in any case is probably the better facility to use.
4476 15. A really picky compiler found some instances of statements for creating
4477 error messages that either had too many or two few arguments for the format
4480 16. The size of the buffer for calls to the DNS resolver has been increased
4481 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4482 for addresses that have a lot of PTR records. This alleviates a problem; it
4483 does not fully solve it.
4485 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4486 buffer now truncates the list and logs the incident, which is the same
4487 action as happens when Exim is looking up a host name and its aliases.
4488 Previously in this situation something unpredictable would happen;
4489 sometimes it was "internal error: store_reset failed".
4491 18. If a server dropped the connection unexpectedly when an Exim client was
4492 using GnuTLS and trying to read a response, the client delivery process
4493 crashed while trying to generate an error log message.
4495 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4496 string, for example:
4498 warn message = H1: something\nH2: something
4500 the text was added as a single header line from Exim's point of view
4501 though it ended up OK in the delivered message. However, searching for the
4502 second and subsequent header lines using $h_h2: did not work. This has been
4503 fixed. Similarly, if a system filter added multiple headers in this way,
4504 the routers could not see them.
4506 20. Expanded the error message when iplsearch is called with an invalid key to
4507 suggest using net-iplsearch in a host list.
4509 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4510 ACLs are no longer actually imposed (and a message to that effect is
4513 22. If a "gecos" field in a passwd entry contained escaped characters, in
4514 particular, if it contained a \" sequence, Exim got it wrong when building
4515 a From: or a Sender: header from that name. A second bug also caused
4516 incorrect handling when an unquoted " was present following a character
4517 that needed quoting.
4519 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4520 was not being matched caselessly.
4522 24. Arranged for all hyphens in the exim.8 source to be escaped with
4525 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4526 properly. Recipient callout cache records were still being keyed to include
4527 the sender, even when use_sender was set false. This led to far more
4528 callouts that were necessary. The sender is no longer included in the key
4529 when use_sender is false.
4531 26. Added "control = submission" modifier to ACLs.
4533 27. Added the ${base62d: operator to decode base 62 numbers.
4535 28. dnsdb lookups can now access SRV records.
4537 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4538 the configuration file.
4540 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4541 (-v) mode. This makes the output for a verbose queue run more intelligible.
4543 31. Added a use_postmaster feature to recipient callouts.
4545 32. Added the $body_zerocount variable, containing the number of binary zero
4546 bytes in the message body.
4548 33. The time of last modification of the "new" subdirectory is now used as the
4549 "mailbox time last read" when there is a quota error for a maildir
4552 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4554 35. Added +ignore_unknown as a special item in host lists.
4556 36. Code for decoding IPv6 addresses in host lists is now included, even if
4557 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4558 address was recognized as an IP address, but was then not correctly decoded
4559 into binary, causing unexpected and incorrect effects when compared with
4566 1. Very minor rewording of debugging text in manualroute to say "list of
4567 hosts" instead of "hostlist".
4569 2. If verify=header_syntax was set, and a header line with an unqualified
4570 address (no domain) and a large number of spaces between the end of the
4571 name and the colon was received, the reception process suffered a buffer
4572 overflow, and (when I tested it) crashed. This was caused by some obsolete
4573 code that should have been removed. The fix is to remove it!
4575 3. When running in the test harness, delay a bit after writing a bounce
4576 message to get a bit more predictability in the log output.
4578 4. Added a call to search_tidyup() just before forking a reception process. In
4579 theory, someone could use a lookup in the expansion of smtp_accept_max_
4580 per_host which, without the tidyup, could leave open a database connection.
4582 5. Added the variables $recipient_data and $sender_data which get set from a
4583 lookup success in an ACL "recipients" or "senders" condition, or a router
4584 "senders" option, similar to $domain_data and $local_part_data.
4586 6. Moved the writing of debug_print from before to after the "senders" test
4589 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4590 problems for message scanning, either using a data ACL, or using
4591 local_scan() because the Received: header was not generated till after they
4592 were called (in order to set the time as the time of reception completion).
4593 I have revised the way this works. The header is now generated after the
4594 body is received, but before the ACL or local_scan() are called. After they
4595 are run, the timestamp in the header is updated.
4601 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4602 before starting a queue runner without re-exec. This happened only when
4603 deliver_drop_privilege was set or when the Exim user was set to root. The
4604 effect of the bug was that timeouts during subsequent deliveries caused
4605 crashes instead of being properly handled. The handler is now left at its
4606 default (and expected) setting.
4608 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4609 message, again when deliver_drop_privilege is set or Exim is run as root.
4610 The bug described in (1) was not present in this case, but the tidying up
4611 of the other signals was missing. I have made the two cases consistent.
4613 3. The ignore_target_hosts setting on a manualroute router was being ignored
4614 for hosts that were looked up using the /MX notation.
4616 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4619 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4620 operated on the sender address. After changing the $sender_address to <>
4621 for the sender address verify, Exim was re-instated it as the original
4622 (before rewriting) address, but remembering that it had rewritten it, so it
4623 wasn't rewriting it again. This bug also had the effect of breaking the
4624 sender address verification caching when the sender address was rewritten.
4626 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4627 This has been changed so that if the ip literal address matches
4628 ignore_target_hosts, the router declines.
4630 7. Added expansion conditions match_domain, match_address, and match_local_
4631 part (NOT match_host).
4633 8. The placeholder for the Received: header didn't have a length field set.
4635 9. Added code to Exim itself and to exim_lock to test for a specific race
4636 condition that could lead to file corruption when using MBX delivery. The
4637 issue is with the lockfile that is created in /tmp. If this file is removed
4638 after a process has opened it but before that process has acquired a lock,
4639 there is the potential for a second process to recreate the file and also
4640 acquire a lock. This could lead to two Exim processes writing to the file
4641 at the same time. The added code performs the same test as UW imapd; it
4642 checks after acquiring the lock that its file descriptor still refers to
4643 the same named file.
4645 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4646 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4649 11. Added the smtp_active_hostname option. If used, this will typically be made
4650 to depend on the incoming interface address. Because $interface_address is
4651 not set up until the daemon has forked a reception process, error responses
4652 that can happen earlier (such as "too many connections") no longer contain
4655 12. If an expansion in a condition on a "warn" statement fails because a lookup
4656 defers, the "warn" statement is abandoned, and the next ACL statement is
4657 processed. Previously this caused the whole ACL to be aborted.
4659 13. Added the iplsearch lookup type.
4661 14. Added ident_timeout as a log selector.
4663 15. Added tls_certificate_verified as a log selector.
4665 16. Added a global option tls_require_ciphers (compare the smtp transport
4666 option of the same name). This controls incoming TLS connections.
4668 17. I finally figured out how to make tls_require_ciphers do a similar thing
4669 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4670 before starting the TLS session.
4672 18. Tabs are now shown as \t in -bP output.
4674 19. If the log selector return_path_on_delivery was set, Exim crashed when
4675 bouncing a message because it had too many Received: header lines.
4677 20. If two routers both had headers_remove settings, and the first one included
4678 a superfluous trailing colon, the final name in the first list and the
4679 first name in the second list were incorrectly joined into one item (with a
4680 colon in the middle).
4686 1. Added -C and -D options to the exinext utility, mainly to make it easier
4687 to include in the automated testing, but these could be helpful when
4688 multiple configurations are in use.
4690 2. The exinext utility was not formatting the output nicely when there was
4691 an alternate port involved in the retry record key, nor when there was a
4692 message id as well (for retries that were specific to a specific message
4693 and a specific host). It was also confused by IPv6 addresses, because of
4694 the additional colons they contain. I have fixed the IPv4 problem, and
4695 patched it up to do a reasonable job for IPv6.
4697 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4698 delivery, the log line now contains "pipelined" if PIPELINING was used.
4700 4. An SMTP transport process used to panic and die if the bind() call to set
4701 an explicit outgoing interface failed. This has been changed; it is now
4702 treated in the same way as a connect() failure.
4704 5. A reference to $sender_host_name in the part of a conditional expansion
4705 that was being skipped was still causing a DNS lookup. This no longer
4708 6. The def: expansion condition was not recognizing references to header lines
4709 that used bh_ and bheader_.
4711 7. Added the _cache feature to named lists.
4713 8. The code for checking quota_filecount in the appendfile transport was
4714 allowing one more file than it should have been.
4716 9. For compatibility with Sendmail, the command line option
4724 and sets the incoming protocol and host name (for trusted callers). The
4725 host name and its colon can be omitted when only the protocol is to be set.
4726 Note the Exim already has two private options, -pd and -ps, that refer to
4727 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4728 "s", but I don't think that's a major issue.
4730 10. A number of refactoring changes to the code, none of which should affect
4733 (a) The number of logging options was getting close to filling up the
4734 32-bit word that was used as a bit map. I have split them into two classes:
4735 those that are passed in the argument to log_write(), and those that are
4736 only ever tested independently outside of that function. These are now in
4737 separate 32-bit words, so there is plenty of room for expansion again.
4738 There is no change in the user interface or the logging behaviour.
4740 (b) When building, for example, log lines, the code previously used a
4741 macro that called string_cat() twice, in order to add two strings. This is
4742 not really sufficiently general. Furthermore, there was one instance where
4743 it was actually wrong because one of the argument was used twice, and in
4744 one call a function was used. (As it happened, calling the function twice
4745 did not affect the overall behaviour.) The macro has been replaced by a
4746 function that can join an arbitrary number of extra strings onto a growing
4749 (c) The code for expansion conditions now uses a table and a binary chop
4750 instead of a serial search (which was left over from when there were very
4751 few conditions). Also, it now recognizes conditions like "pam" even when
4752 the relevant support is not compiled in: a suitably worded error message is
4753 given if an attempt is made to use such a condition.
4755 11. Added ${time_interval:xxxxx}.
4757 12. A bug was causing one of the ddress fields not to be passed back correctly
4758 from remote delivery subprocesses. The field in question was not being
4759 subsequently used, so this caused to problems in practice.
4761 13. Added new log selectors queue_time and deliver_time.
4763 14. Might have fixed a bug in maildirsizefile handling that threw up
4764 "unexpected character" debug warnings, and recalculated the data
4765 unnecessarily. In any case, I expanded the warning message to give more
4768 15. Added the message "Restricted characters in address" to the statements in
4769 the default ACL that block characters like @ and % in local parts.
4771 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4772 Three changes have been made:
4774 (a) There was a serious bug; a negative response to MAIL caused the whole
4775 recipient domain to be cached as invalid, thereby blocking all messages
4776 to all local parts at the same domain, from all senders. This bug has
4777 been fixed. The domain is no longer cached after a negative response to
4778 MAIL if the sender used is not empty.
4780 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4783 (c) A new callout option, "use_sender" has been added for people who want
4784 the modified behaviour.
4790 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4793 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4794 indeed breaks things for older releases.
4796 3. Added additional logging to the case where there is a problem reading data
4797 from a filter that is running in a subprocess using a pipe, in order to
4798 try to track down a specific problem.
4800 4. Testing facility fudge: when running in the test harness and attempting
4801 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4802 getting "No route to host". Convert this to a timeout.
4804 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4807 6. Some OS don't have socklen_t but use size_t instead. This affects the
4808 fifth argument of getsockopt() amongst other things. This is now
4809 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4810 can be set for individual OS. I have set it for SunOS5, OSF1, and
4811 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4812 some earlier ones do not.
4814 7. Change 4.30/15 was not doing the test caselessly.
4816 8. The standard form for an IPv6 address literal was being rejected by address
4817 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4818 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4819 this, as well as the form without the "IPv6" on the front (but only when
4820 address literals are enabled, of course).
4822 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4824 10. Exim crashed if a message with an empty sender address specified by -f
4825 encountered a router with an errors_to setting. This could be provoked only
4826 by a command such as
4830 where an empty string was supplied; "<>" did not hit this bug.
4832 11. Installed PCRE release 4.5.
4834 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4835 remained set. It is now erased.
4837 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4838 times from message ids (which are base 36 rather than the normal 62).
4840 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4841 were being counted as actual protocol errors, and logged if the log
4842 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4843 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4844 and DATA following a set of rejected RCPTs do not count as protocol errors.
4845 In other words, Exim assumes they were pipelined, though this may not
4846 actually be the case. Of course, in all cases the client gets an
4847 appropriate error code.
4849 15. If a lookup fails in an ACL condition, a message about the failure may
4850 be available; it is used if testing the ACL cannot continue, because most
4851 such messages specify what the cause of the deferral is. However, some
4852 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4853 that caused an old message to be retained and used if a later statement
4854 caused a defer, replacing the real cause of the deferral.
4856 16. If an IP address had so many PTR records that the DNS lookup buffer
4857 was not large enough to hold them, Exim could crash while trying to process
4858 the truncated data. It now detects and logs this case.
4860 17. Further to 4.21/58, another change has been made: if (and only if) the
4861 first line of a message (the first header line) ends with CRLF, a bare LF
4862 in a subsequent header line has a space inserted after it, so as not to
4863 terminate the header.
4865 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4866 unnecessarily, used atoi() instead of strtol(), and didn't check the
4867 termination when getting file sizes from file names by regex.
4869 19. Completely re-implemented the support for maildirsize files, in the light
4870 of a number of problems with the previous contributed implementation
4871 (4.30/29). In particular:
4873 . If the quota is zero, the maildirsize file is maintained, but no quota is
4876 . If the maildir directory does not exist, it is created before any attempt
4877 to write a maildirsize file.
4879 . The quota value in the file is just a cache; if the quota is changed in
4880 the transport, the new value overrides.
4882 . A regular expression is available for excluding directories from the
4885 20. The autoreply transport checks the characters in options that define the
4886 message's headers; it allows continued headers, but it was checking with
4887 isspace() after an embedded newline instead of explicitly looking for a
4890 21. If all the "regular" hosts to which an address was routed had passed their
4891 expiry times, and had not reached their retry times, the address was
4892 bounced, even if fallback hosts were defined. Now Exim should go on to try
4895 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4896 equivalent code in the SMTP transport. Some hosts send humungous responses
4897 to HELO/EHLO, more than 1024 it seems.
4899 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4900 gives compiler warnings in some environments. I've now done it "properly",
4903 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4904 (because of problems with the built-in one) was declared to return uschar *
4905 instead of char *, causing compiler failure.
4907 25. Fixed a file descriptor leak when processing alias/forward files.
4909 26. Fixed a minor format string issue in dbfn.c.
4911 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4913 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4914 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4917 29. When a sender address is verified, it is cached, to save repeating the test
4918 when there is more than one recipient in a message. However, when the
4919 verification involves a callout, it is possible for different callout
4920 options to be set for different recipients. It is too complicated to keep
4921 track of this in the cache, so now Exim always runs a verification when a
4922 callout is required, relying on the callout cache for the optimization.
4923 The overhead is duplication of the address routing, but this should not be
4926 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4927 to be verified with callout=postmaster, and the main callout worked but the
4928 postmaster check failed, the verification correctly failed. However, if a
4929 subsequent RCPT command asked for sender verification *without* the
4930 postmaster check, incorrect caching caused this verification also to fail,
4933 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4934 it was not caching the DNS options (qualify_single, search_parents) that
4935 were used when the lookup failed. A subsequent lookup with different
4936 options therefore always gave the same answer, though there were cases
4937 where it should not have. (Example: a "domains = !$mx_any" option on a
4938 dnslookup router: the "domains" option is always processed without any
4939 widening, but the router might have qualify_single set.) Now Exim uses the
4940 cached value only when the same options are set.
4942 32. Added John Jetmore's "exipick" utility to the distribution.
4944 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4945 than a timeout (e.g. a certificate is required, and is not provided), an
4946 Exim server now closes the connection immediately. Previously it waited for
4947 the client to close - but if the client is SSL, it seems that they each
4948 wait for each other, leading to a delay before one of them times out.
4950 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4951 maintained 0.8.x compatibility because I don't think many are using it, and
4952 it is clearly obsolete.
4954 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4957 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4958 client certificate was expired. A simple patch fixes this, though I don't
4959 understand the full logic of why the verify callback is called multiple
4962 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4963 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4964 which causes problems with some clients (such as the Certicom SSL Plus
4965 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4966 disables the coutermeasure allowing Eudora to connect."
4968 38. Exim was not checking that a write() to a log file succeeded. This could
4969 lead to Bad Things if a log got too big, in particular if it hit a file
4970 size limit. Exim now panics and dies if it cannot write to a log file, just
4971 as it does if it cannot open a log file.
4973 39. Modified OS/Makefile-Linux so that it now contains
4975 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4977 The two -D definitions ensure that Exim is compiled with large file
4978 support, which makes it possible to handle log files that are bigger than
4981 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4982 instance) a domain was checked against a named list that involved a lookup,
4983 causing $domain_data to be set, then another domain was checked against the
4984 same list, then the first domain was re-checked, the value of $domain_data
4985 after the final check could be wrong. In particular, if the second check
4986 failed, it could be set empty. This bug probably also applied to
4989 41. The strip_trailing_dot option was not being applied to the address given
4990 with the -f command-line option.
4992 42. The code for reading a message's header from the spool was incrementing
4993 $received_count, but never initializing it. This meant that the value was
4994 incorrect (doubled) while delivering a message in the same process in which
4995 it was received. In the most common configuration of Exim, this never
4996 happens - a fresh exec is done - but it can happen when
4997 deliver_drop_privilege is set.
4999 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5000 it now includes up to 150 characters of the unexpected data in the log
5003 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5004 and building data strings. The size of both of these buffers was 10 000
5005 bytes - far larger than anybody would *ever* want, thought I. Needless to
5006 say, somebody hit the limit. I have increased the maximum line length to
5007 20 000 and the maximum data length of concatenated lines to 100 000. I have
5008 also fixed two bugs, because there was no checking on these buffers. Tsk,
5009 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5010 buffer is too small.
5012 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5013 lsearch lookups. Now it does.
5015 46. When parsing a route_list item in a manualroute router, a fixed-length
5016 buffer was used for the list of hosts. I made this 1024 bytes long,
5017 thinking that nobody would ever have a list of hosts that long. Wrong.
5018 Somebody had a whole pile of complicated expansion conditions, and the
5019 string was silently truncated, leading to an expansion error. It turns out
5020 that it is easier to change to an unlimited length (owing to other changes
5021 that have happened since this code was originally written) than to build
5022 structure for giving a limitation error. The length of the item that
5023 expands into the list of hosts is now unlimited.
5025 47. The lsearch lookup could not handle data where the length of text line was
5026 more than 4095 characters. Such lines were truncated, leading to shortened
5027 data being returned. It should now handle lines of any length.
5029 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5030 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5033 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5034 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5035 no longer happens in comment lines. A list of replacements is now placed
5036 at the head of all of the source files, except those whose only change is
5037 to replace PERL_COMMAND in the very first #! line.
5039 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5040 messages on the queue, with a bottom-up merge sort, using code contributed
5041 by Michael Haardt. This should make operations like -bp somewhat faster on
5042 large queues. It won't affect queue runners, except when queue_run_in_order
5045 51. Installed eximstats 1.31 in the distribution.
5047 52. Added support for SRV lookups to the dnslookup router.
5049 53. If an ACL referred to $message_body or $message_body_end, the value was not
5050 reset for any messages that followed in the same SMTP session.
5052 54. The store-handling optimization for building very long strings was not
5053 differentiating between the different store pools. I don't think this
5054 actually made any difference in practice, but I've tidied it.
5056 55. While running the routers to verify a sender address, $sender_address
5057 was still set to the sender address. This is wrong, because when routing to
5058 send a bounce to the sender, it would be empty. Therefore, I have changed
5059 it so that, while verifying a sender address, $sender_address is set to <>.
5060 (There is no change to what happens when verifying a recipient address.)
5062 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5063 target A or AAAA records (if not already returned) without resetting the
5064 qualify_single or search_parents options of the DNS resolver. These are
5065 inappropriate in this case because the targets of MX and SRV records must
5066 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5067 target that, when qualified, matched something in the local domain. These
5068 two options are now turned off when doing these lookups.
5070 57. It seems that at least some releases of Reiserfs (which does not have the
5071 concept of a fixed number of inodes) returns zero and not -1 for the
5072 number of available inodes. This interacted badly with check_spool_inodes,
5073 which assumed that -1 was the "no such thing" setting. What I have done is
5074 to check that the total number of inodes is greater than zero before doing
5075 the test of how many are available.
5077 58. When a "warn" ACL statement has a log_message modifier, the message is
5078 remembered, and not repeated. This is to avoid a lot of repetition when a
5079 message has many recipients that cause the same warning to be written.
5080 Howewer, Exim was preserving the list of already written lines for an
5081 entire SMTP session, which doesn't seem right. The memory is now reset if a
5082 new message is started.
5084 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5085 debugging output unless log_rewrite was also set.
5087 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5088 of (int)(handle) when we know that handle contains (void *)(-1).
5090 61. The Exim daemon panic-logs an error return when it closes the incoming
5091 connection. However "connection reset by peer" seems to be common, and
5092 isn't really an error worthy of noting specially, so that particular error
5095 62. When Exim is trying to find all the local interfaces, it used to panic and
5096 die if the ioctl to get the interface flags failed. However, it seems that
5097 on at least one OS (Solaris 9) it is possible to have an interface that is
5098 included in the list of interfaces, but for which you get a failure error
5099 for this call. This happens when the interface is not "plumbed" into a
5100 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5101 failure of the "get flags" call assumes that the interface is down.
5103 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5104 makes life easier for people who are doing arithmetic on fields extracted
5105 from dates, where you often get leading zeros that should not be
5106 interpreted as octal.
5108 64. Added qualify_domain to the redirect router, to override the global
5111 65. If a pathologically long header line contained very many addresses (the
5112 report of this problem mentioned 10 000) and each of them was rewritten,
5113 Exim could use up a very large amount of memory. (It kept on making new
5114 copies of the header line as it rewrote, and never released the old ones.)
5115 At the expense of a bit more processing, the header rewriting function has
5116 been changed so that it no longer eats memory in this way.
5118 66. The generation of the Received: header has been moved from the time that a
5119 message starts to be received, to the time that it finishes. The timestamp
5120 in the Received: header should now be very close to that of the <= log
5121 line. There are two side-effects of this change:
5123 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5124 logged header lines no longer include the local Received: line, because
5125 it has not yet been created. The same applies to a copy of the message
5126 that is returned to a non-SMTP sender when a message is rejected.
5128 (b) When a filter file is tested using -bf, no additional Received: header
5129 is added to the test message. After some thought, I decided that this
5132 This change does not affect the value of $received_for. It is still set
5133 after address rewriting, but before local_scan() is called.
5135 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5137 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5138 gave an unhelpful panic error message, and a defer error. I have managed to
5139 change this behaviour so that it now rejects any supplied certificate,
5140 which seems right, as the list of acceptable certificates is empty.
5142 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5143 gave an unhelpful defer error. I have not managed to make this reject any
5144 supplied certificates, but the error message it gives is "no certificate
5145 supplied", which is not helpful.
5147 70. exigrep's output now also includes lines that are not associated with any
5148 message, but which match the given pattern. Implemented by a patch from
5149 Martin Sluka, which also tidied up the Perl a bit.
5151 71. Recipient callout verification, like sender verification, was using <> in
5152 the MAIL FROM command. This isn't really the right thing, since the actual
5153 sender may affect whether the remote host accepts the recipient or not. I
5154 have changed it to use the actual sender in the callout; this means that
5155 the cache record is now keyed on a recipient/sender pair, not just the
5156 recipient address. There doesn't seem to be a real danger of callout loops,
5157 since a callout by the remote host to check the sender would use <>.
5158 [SEE ABOVE: changed after hitting problems.]
5160 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5161 temporary errors. However, in the case of such a code being given after
5162 the end of a data transmission (i.e. after ".") Exim was failing to write
5163 a retry record for the message. (Yes, there was some broken host that was
5164 actually sending 8xx at this point.)
5166 73. An unknown lookup type in a host list could cause Exim to panic-die when
5167 the list was checked. (An example that provoked this was putting <; in the
5168 middle of a list instead of at the start.) If this happened during a DATA
5169 ACL check, a -D file could be left lying around. This kind of configuration
5170 error no longer causes Exim to die; instead it causes a defer errror. The
5171 incident is still logged to the main and panic logs.
5173 74. Buglet left over from Exim 3 conversion. The message "too many messages
5174 in one connection" was written to the rejectlog but not the mainlog, except
5175 when address rewriting (yes!) was being logged.
5177 75. Added write_rejectlog option.
5179 76. When a system filter was run not as root (that is, when system_filter_user
5180 was set), the values of the $n variables were not being returned to the
5181 main process; thus, they were not subsequently available in the $sn
5184 77. Added +return_path_on_delivery log selector.
5186 78. A connection timeout was being treated differently from recipients deferred
5187 when testing hosts_max_try with a message that was older than the host's
5188 retry timeout. (The host should not be counted, thus allowing all hosts to
5189 be tried at least once before bouncing.) This may have been the cause of an
5190 occasionally reported bug whereby a message would remain on the queue
5191 longer than the retry timeout, but would be bounced if a delivery was
5192 forced. I say "may" because I never totally pinned down the problem;
5193 setting up timeout/retry tests is difficult. See also the next item.
5195 79. The ultimate address timeout was not being applied to errors that involved
5196 a combination of host plus message (for example, a timeout on a MAIL
5197 command). When an address resolved to a number of possible hosts, and they
5198 were not all tried for each delivery (e.g. because of hosts_max_try), a
5199 message could remain on the queue longer than the retry timeout.
5201 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5204 81. Fixed an obscure SMTP outgoing bug which required at least the following
5205 conditions: (a) there was another message waiting for the same server;
5206 (b) the server returned 5xx to all RCPT commands in the first message so
5207 that the message was not completed; (c) the server dropped the connection
5208 or gave a negative response to the RSET that Exim sends to abort the
5209 transaction. The observed case was a dropped connection after DATA that had
5210 been sent in pipelining mode. That is, the server had advertised PIPELINING
5211 but was not implementing it correctly. The effect of the bug was incorrect
5212 behaviour, such as trying another host, and this could lead to a crash.
5218 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5219 and daemon.c were passed as pointers to ints; they should have been
5220 pointers to socklen_t variables (which are typically unsigned ints).
5222 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5225 3. Fixed a really odd bug that affected only the testing scheme; patching a
5226 certain fixed string in the binary changed the value of another string that
5227 happened to be identical to the end of the original first string.
5229 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5230 name", it returns that address as the IP address. On some operating
5231 systems (e.g. Solaris), it also passes back the IP address string as the
5232 "host name". However, on others (e.g. Linux), it passes back an empty
5233 string. Exim wasn't checking for this, and was changing the host name to an
5234 empty string, assuming it had been canonicized.
5236 5. Although rare, it is permitted to have more than one PTR record for a given
5237 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5238 all the names associated with an address, because they do in Solaris.
5239 However, it seems that they do not in Linux for data that comes from the
5240 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5241 I found this out when I moved to a new Linux workstation and tried to run
5242 the Exim test suite.
5244 To get round this problem I have changed the code so that it now does its
5245 own call to the DNS to look up PTR records when searching for a host name.
5246 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5247 addresses that are only in /etc/hosts are still found.
5249 This behaviour is, however, controlled by an option called host_lookup_
5250 order, which defaults to "bydns:byaddr". If people want to use the other
5251 order, or indeed, just use one or the other means of lookup, they can
5252 specify it in this variable.
5254 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5255 some operating systems, this comes back from gethostbyaddr() as an empty
5256 string, and this is what Exim used to test for. However, it seems that in
5257 other systems, "." is yielded. Exim now tests for this case too.
5259 7. The values of check_spool_space and check_log_space are now held internally
5260 as a number of kilobytes instead of an absolute number of bytes. If a
5261 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5262 kilobyte. This means that much larger values can be stored.
5264 8. Exim monitor: an attempt to get the action menu when not actually pointing
5265 at a message produces an empty menu entitled "No message selected". This
5266 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5267 no entries in it ("Shell widget menu has zero width and/or height"). So I
5268 have added a single, blank menu entry in this case.
5270 9. Added ${quote_local_part.
5272 10. MIME decoding is now applied to the contents of Subject: header lines when
5275 11. Now that a reference to $sender_host_address automatically causes a reverse
5276 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5277 host lookup before query-style lookups in lists that might use this
5278 variable. This has therefore been abolished, and the "net-" prefix is no
5279 longer necessary for query-style lookups.
5281 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5282 have been a typo for LFLAGS, so it has been changed.
5284 13. The install script calls Exim with "-C /dev/null" in order to find the
5285 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5286 to be output. Howeve, since Exim outputs its version number before the
5287 error, it didn't break the script. It just looked ugly. I fixed this by
5288 always allowing "-C /dev/null" if the caller is root.
5290 14. Ignore overlarge ACL variable number when reading spool file - insurance
5291 against a later release with more variables having written the file.
5293 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5294 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5295 this, as well as the form without the "IPv6" on the front.
5297 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5298 OS/Makefile-Darwin file.
5300 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5301 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5303 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5304 variable. However, some OS use ints for inodes. I've added cast to long int
5305 to get rid of the compiler warning.
5307 19. I had forgotten to lock out "/../" in configuration file names when
5308 ALT_CONFIG_PREFIX was set.
5310 20. Routers used for verification do not need to specify transports. However,
5311 if such a router generated a host list, and callout was configured, Exim
5312 crashed, because it could not find a port number from the (non-existent)
5313 transport. It now assumes port 25 in this circumstance.
5315 21. Added the -t option to exigrep.
5317 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5318 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5319 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5320 them is set, LOOKUP_LSEARCH is forced.
5322 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5324 24. Added sender and host information to the "rejected by local_scan()" log
5325 line; previously there was no indication of these.
5327 25. Added .include_if_exists.
5329 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5330 Linux. It turns out that not all file systems support this. Apparently some
5331 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5332 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5333 not supported on the file descriptor, is now ignored when Exim is trying to
5334 sync a directory. This applies only to Linux.
5336 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5338 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5339 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5342 29. Added support for maildirsize files from supplied patch (modified a bit).
5344 30. The use of :fail: followed by an empty string could lead Exim to respond to
5345 sender verification failures with (e.g.):
5347 550 Verification failed for <xxx>
5348 550 Sender verify failed
5350 where the first response line was missing the '-' that indicates it is not
5351 the final line of the response.
5353 31. The loop for finding the name of the user that called Exim had a hardwired
5354 limit of 10; it now uses the value of finduser_retries, which is used for
5355 all other user lookups.
5357 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5360 33. Exim was neglecting to zero errno before one call of strtol() when
5361 expanding a string and expecting an integer value. On some systems this
5362 resulted in spurious "integer overflow" errors. Also, it was casting the
5363 result into an int without checking.
5365 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5366 did not work. The code looks as if it has *never* worked, though it appears
5367 to have been documented since at least releast 1.62. I have made it work.
5369 35. The "timeout_DNS" error in retry rules, also documented since at least
5370 1.62, also never worked. As it isn't clear exactly what this means, and
5371 clearly it isn't a major issue, I have abolished the feature by treating it
5372 as "timeout", and writing a warning to the main and panic logs.
5374 36. The display of retry rules for -brt wasn't always showing the error code
5377 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5378 timeout_connect_A, timeout_connect_MX.
5380 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5381 to the empty sender.
5383 39. The daemon was not analysing the content of -oX till after it had closed
5384 stderr and disconnected from the controlling terminal. This meant that any
5385 syntax errors were only noted on the panic log, and the return code from
5386 the command was 0. By re-arranging the code a little, I've made the
5387 decoding happen first, so such errors now appear on stderr, and the return
5388 code is 1. However, the actual setting up of the sockets still happens in
5389 the disconnected process, so errors there are still only recorded on the
5392 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5393 connections (as happens on some IP stacks) was logged at start up time as
5394 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5395 it from "IPv6 and IPv4", which means that two separate sockets are being
5398 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5399 says whether AF_INET or AF_INET6 was passed as an argument.
5401 42. Exiwhat output was messed up when time zones were included in log
5404 43. Exiwhat now gives more information about the daemon's listening ports,
5405 and whether -tls-on-connect was used.
5407 44. The "port" option of the smtp transport is now expanded.
5409 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5410 silently ignored. Now an error message is written to the main and panic
5413 46. There's a new ACL modifier called "logwrite" which writes to a log file
5414 as soon as it is encountered.
5416 47. Added $local_user_uid and $local_user_gid at routing time.
5418 48. Exim crashed when trying to verify a sender address that was being
5421 49. Exim was recognizing only a space character after ".include". It now also
5422 recognizes a tab character.
5424 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5425 extracting the relevant information from the specification. The man page no
5426 longer contains scrambled data for the -d option, and I've added a section
5427 at the front about calling Exim under different names.
5429 51. Added "extra_headers" argument to the "mail" command in filter files.
5431 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5434 53. Installed eximstats 1.29.
5436 54. Added transport_filter_timeout as a generic transport option.
5438 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5439 Cc: header lines. This was required by RFC 822, but it not required by RFC
5442 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5443 incoming messages that did not have them. Now it does so only if the
5444 message originates locally, that is, if there is no associated remote host
5445 address. When Resent- header lines are present, this applies to the Resent-
5446 lines rather than the non-Resent- lines.
5448 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5449 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5451 58. Messages for configuration errors now include the name of the main
5452 configuration file - useful now that there may be more than one file in a
5453 list (.included file names were always shown).
5455 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5456 for those rare installations that do not start the daemon as root or run it
5457 setuid root. I've cut out the call to initgroups() if the daemon is not
5460 60. The Exim user and group can now be bound into the binary as text strings
5461 that are looked up at the start of Exim's processing.
5463 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5465 62. Added $mailstore_basename variable.
5467 63. Installed patch to sieve.c from Michael Haardt.
5469 64. When Exim failed to open the panic log after failing to open the main log,
5470 the original message it was trying to log was written to stderr and debug
5471 output, but if they were not available (the usual case in production), it
5472 was lost. Now it is written to syslog before the two lines that record the
5473 failures to open the logs.
5475 65. Users' Exim filters run in subprocesses under the user's uid. It is
5476 possible for a "deliver" command or an alias in a "personal" command to
5477 provoke an address rewrite. If logging of address rewriting is configured,
5478 this fails because the process is not running as root or exim. There may be
5479 a better way of dealing with this, but for the moment (because 4.30 needs
5480 to be released), I have disabled address rewrite logging when running a
5481 filter in a non-root, non-exim process.
5487 1. The buildconfig auxiliary program wasn't quoting the value set for
5488 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5489 not defined. This bug was masked in 4.22 by the effect that was fixed in
5492 2. Some messages that were rejected after a message id was allocated were
5493 shown as "incomplete" by exigrep. It no longer does this for messages that
5494 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5496 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5497 have allow_domain_literals set, the ID did not get logged in the <= line.
5498 Domain literals are now always recognized in Message-ID: header lines.
5500 4. The first argument for a ${extract expansion item is the key name or field
5501 number. Leading and trailing spaces in this item were not being ignored,
5502 causing some misleading effects.
5504 5. When deliver_drop_privilege was set, single queue runner processes started
5505 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5506 same command in the process it spins off) were not dropping privilege.
5508 6. When the daemon running as "exim" started a queue runner, it always
5509 re-executed Exim in the spun-off process. This is a waste of effort when
5510 deliver_drop_privilege is set. The new process now just calls the
5511 queue-runner function directly.
5517 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5520 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5521 ignored. Though the use of -oP was forcing the writing of a pid file, it
5522 was always written to the default place.
5524 3. If the message "no IP address found for host xxxx" is generated during
5525 incoming verification, it is now followed by identification of the incoming
5526 connection (so you can more easily find what provoked it).
5528 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5530 5. Added some features to "harden" Exim a bit more against certain attacks:
5532 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5533 be put in Local/Makefile. This is like the never_users runtime option,
5534 but it cannot be overridden. The default setting is "root".
5536 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5537 prefix string with which any file named in a -C command line option
5540 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5541 is retained for -C and -D only if the caller of Exim is root. Without
5542 it, the exim user may also use -C and -D and retain privilege.
5544 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5545 command line option is disabled.
5547 6. Macro names set by the -D option must start with an upper case letter, just
5548 like macro names defined in the configuration file.
5550 7. Added "dereference=" facility to LDAP.
5552 8. Two instances of the typo "uknown" in the source files are fixed.
5554 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5555 the Configure-Makefile script screwed up while processing it.
5557 10. Incorporated PCRE 4.4.
5559 11. The SMTP synchronization check was not operating right at the start of an
5560 SMTP session. For example, it could not catch a HELO sent before the client
5561 waited for the greeting. There is now a check for outstanding input at the
5562 point when the greeting is written. Because of the duplex, asynchronous
5563 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5564 way, but not yet received, when the check is performed.
5566 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5567 on TCP/IP sockets, because this apparently causes some broken clients to
5570 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5571 unchanged) from the Cygwin maintainer.
5573 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5574 was supported instead of testing for mbx. Effectively a typo.
5576 15. The spa authenticator server code was not checking that the input it
5577 received was valid base64.
5579 16. The debug output line for the "set" modifier in ACLs was not showing the
5580 name of the variable that was being set.
5582 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5584 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5585 lookup on its own. Something else was needed to trigger the lookup. For
5586 example, a match in host_lookup or the need for a host name in a host list.
5587 Now, if $sender_host_name is referenced and the host name has not yet been
5588 looked up, a lookup is performed. If the lookup fails, the variable remains
5589 empty, and $host_lookup_failed is set to "1".
5591 19. Added "eqi" as a case-independent comparison operator.
5593 20. The saslauthd authentication condition could segfault if neither service
5594 nor realm was specified.
5596 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5597 error message that was logged was misleading, and incoming SMTP
5598 connections were dropped. The message is now more accurate, and temporary
5599 errors are given to SMTP connections.
5601 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5602 (except RSET) with a 421 error, until QUIT is received. However, it was
5603 failing to send a response to QUIT.
5605 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5606 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5607 ACL is now run after the helo_try_verify_hosts code.
5609 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5610 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5611 case-independent, so other case variants are also recognized). Apparently
5612 some systems use these upper case variants.
5614 25. If more than two messages were waiting for the same host, and a transport
5615 filter was specified for the transport, Exim sent two messages over the
5616 same TCP/IP connection, and then failed with "socket operation on non-
5617 socket" when it tried to send the third.
5619 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5621 27. The extern definition of crypt16() in expand.c was not being excluded when
5622 the OS had its own crypt16() function.
5624 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5625 as a preferred synonym for return_size_limit, both as an option and as an
5628 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5630 30. Changed the default configuration ACL to relax the local part checking rule
5631 for addresses that are not in any local domains. For these addresses,
5632 slashes and pipe symbols are allowed within local parts, but the sequence
5633 /../ is explicitly forbidden.
5635 31. SPA server authentication was not clearing the challenge buffer before
5638 32. log_message in a "warn" ACL statement was writing to the reject log as
5639 well as to the main log, which contradicts the documentation and doesn't
5640 seem right (because no rejection is happening). So I have stopped it.
5642 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5643 However, I am unable to do any testing of this.
5645 34. Fixed an infelicity in the appendfile transport. When checking directories
5646 for a mailbox, to see if any needed to be created, it was accidentally
5647 using path names with one or more superfluous leading slashes; tracing
5648 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5650 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5651 contents are added to the log line that is written for every discarded
5652 recipient. (Previously a log_message setting was ignored.)
5654 36. The ${quote: operator now quotes the string if it is empty.
5656 37. The install script runs exim in order to find its version number. If for
5657 some reason other than non-existence or emptiness, which it checks, it
5658 could not run './exim', it was installing it with an empty version number,
5659 i.e. as "exim-". This error state is now caught, and the installation is
5662 38. An argument was missing from the function that creates an error message
5663 when Exim fails to connect to the socket for saslauthd authentication.
5664 This could cause Exim to crash, or give a corrupted message.
5666 39. Added isip, isip4, and isip6 to ${if conditions.
5668 40. The ACL variables $acl_xx are now saved with the message, and can be
5669 accessed later in routers, transports, and filters.
5671 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5672 strings in the file are not string-expanded.
5674 42. If a MAIL command specified a SIZE value that was too large to fit into an
5675 int variable, the check against message_size_limit failed. Such values are
5676 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5677 day this will have to be increased, but I don't think I want to be around
5678 when emails are that large.
5685 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5686 iconv() is not standard in FreeBSD.
5688 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5689 IPv6 enabled. The observed symptom was a segmentation fault on return from
5690 the function os_common_find_running_interfaces() in src/os.c.
5692 3. In the check_special_case() function in daemon.c I had used "errno" as an
5693 argument name, which causes warnings on some systems. This was basically a
5694 typo, since it was named "eno" in the comments!
5696 4. The code that waits for the clock to tick (at a resolution of some fraction
5697 of a second) so as to ensure message-id uniqueness was always waiting for
5698 at least one whole tick, when it could have waited for less. [This is
5699 almost certainly not relevant at current processor speeds, where it is
5700 unlikely to ever wait at all. But we try to future-proof.]
5702 5. The function that sleeps for a time interval that includes fractions of a
5703 second contained a race. It did not block SIGALRM between setting the
5704 timer, and suspending (a couple of lines later). If the interval was short
5705 and the sigsuspend() was delayed until after it had expired, the suspension
5706 never ended. On busy systems this could lead to processes getting stuck for
5709 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5710 process, before it forks any delivery processes. The open lookup caching
5711 mechanism meant that the open file or database connection was passed into
5712 the delivery process. The problem was that delivery processes always tidy
5713 up cached lookup data. This could cause a problem for the next delivery
5714 process started by the queue runner, because the external queue runner
5715 process does not know about the closure. So the next delivery process
5716 still has data in the lookup cache. In the case of a file lookup, there was
5717 no problem because closing a file descriptor in a subprocess doesn't affect
5718 the parent. However, if the lookup was caching a connection to a database,
5719 the connection was closed, and the second delivery process was likely to
5720 see errors such as "PGSQL: query failed: server closed the connection
5721 unexpectedly". The problem has been fixed by closing all cached lookups
5722 in a queue runner before running a delivery process.
5724 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5725 seem to have the "const" qualifier which it has on other OS. I've
5728 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5729 *of the same type* (client or server) with the same public name that an
5730 error should be diagnosed.
5732 9. When Exim looked up a host name for an IP address, but failed to find the
5733 original IP address when looking up the host name (a safety check), it
5734 output the message "<ip address> does not match any IP for NULL", which was
5735 confusing, to say the least. The bug was that the host name should have
5736 appeared instead of "NULL".
5738 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5739 user that is built into the binary, and the -C or -D options is used, root
5740 privilege is dropped before the configuration file is read. In addition,
5741 logging is switched to stderr instead of the normal log files. If the
5742 configuration then re-defines the Exim user, the unprivileged environment
5743 is probably not what is expected, so Exim logs a panic warning message (but
5746 However, if deliver_drop_privilege is set, the unprivileged state may well
5747 be exactly what is intended, so the warning has been cut out in that case,
5748 and Exim is allowed to try to write to its normal log files.
5754 1. smtp_return_error_details was not giving details for temporary sender
5755 or receiver verification errors.
5757 2. Diagnose a configuration error if two authenticators have the same public
5760 3. Exim used not to create the message log file for a message until the first
5761 delivery attempt. This could be confusing when incoming messages were held
5762 for policy or load reasons. The message log file is now created at the time
5763 the message is received, and an initial "Received" line is written to it.
5765 4. The automatically generated man page for command line options had a minor
5766 bug that caused no ill effects; however, a more serious problem was that
5767 the procedure for building the man page automatically didn't always
5768 operate. Consequently, release 4.20 contains an out-of-date version. This
5769 shouldn't happen again.
5771 5. When building Exim with embedded Perl support, the script that builds the
5772 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5773 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5775 6. The freeze_tell option was not being used for messages that were frozen on
5776 arrival, either by an ACL or by local_scan().
5778 7. Added the smtp_incomplete_transaction log selector.
5780 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5781 was accepting AUTH without a new EHLO.
5783 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5784 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5787 10. Logging of TCP/IP connections (when configured) now happens in the main
5788 daemon process instead of the child process, so that the TCP/IP connection
5789 count is more accurate (but it can never be perfect).
5791 11. The use of "drop" in a nested ACL was not being handled correctly in the
5792 outer ACL. Now, if condition failure induced by the nested "drop" causes
5793 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5794 or "require"), the connection is dropped.
5796 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5797 that yield "discard" can now be used with an "accept" or a "discard" verb,
5798 but an error is generated for any others (because I can't see a useful way
5799 to define what should happen).
5801 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5802 are now processed in the same way as lines in the Exim configuration file.
5803 In particular, continuation lines are supported.
5805 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5807 15. Added -ti meaning -t -i.
5809 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5810 domains, and warn by logging if others are found.
5812 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5813 returned from GIFCONF when Exim is trying to find the list of interfaces on
5814 a host. The code in os.c has been modified to copy each ifreq to an aligned
5815 structure in all cases.
5817 Also, in some cases, the returned ifreq's were being copied to a 'struct
5818 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5819 means the last couple of bytes of an IPv6 address could be chopped if the
5820 ifreq contained only a normal sockaddr (14 bytes storage).
5822 18. Named domain lists were not supported in the hosts_treat_as_local option.
5823 An entry such as +xxxx was not recognized, and was treated as a literal
5826 19. Ensure that header lines added by a DATA ACL are included in the reject log
5827 if the ACL subsequently rejects the message.
5829 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5830 MD5 (which is deprecated).
5832 21. When testing a filter file using -bf, Exim was writing a message when it
5833 took the sender from a "From " line in the message, but it was not doing so
5834 when it took $return_path from a Return-Path: header line. It now does.
5836 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5837 with a valid header line field name (a series of printing characters
5838 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5840 23. Changed "disc" in the source to "disk" to conform to the documentation and
5841 the book and for uniformity.
5843 24. Ignore Sendmail's -Ooption=value command line item.
5845 25. When execve() failed while trying to run a command in a pipe transport,
5846 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5847 could be confused with a return value of 69 from the command itself. This
5848 has been changed to 127, the value the shell returns if it is asked to run
5849 a non-existent command. The wording for the related log line suggests a
5850 non-existent command as the problem.
5852 26. If received_header_text expands to an empty string, do not add a Received:
5853 header line to the message. (Well, it adds a token one on the spool, but
5854 marks it "old" so that it doesn't get used or transmitted.)
5856 27. Installed eximstats 1.28 (addition of -nt option).
5858 28. There was no check for failure on the call to getsockname() in the daemon
5859 code. This can fail if there is a shortage of resources on the system, with
5860 ENOMEM, for example. A temporary error is now given on failure.
5862 29. Contrary to the C standard, it seems that in some environments, the
5863 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5864 program. Exim now does this explicitly; it affects the formatting of
5865 timestamps using strftime().
5867 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5868 complaints. I've now initialized all the variables, to avoid this.
5870 32. Header lines added by a system filter were not being "seen" during
5871 transport-time rewrites.
5873 33. The info_callback() function passed to OpenSSL is set up with type void
5874 (*)(SSL *, int, int), as described somewhere. However, when calling the
5875 function (actually a macro) that sets it up, the type void(*)() is
5876 expected. I've put in a cast to prevent warnings from picky compilers.
5878 34. If a DNS black list lookup found a CNAME record, but there were no A
5879 records associated with the domain it pointed at, Exim crashed.
5881 35. If a DNS black list lookup returned more than one A record, Exim ignored
5882 all but the first. It now scans all returned addresses if a particular IP
5883 value is being sought. In this situation, the contents of the
5884 $dnslist_value variable are a list of all the addresses, separated by a
5887 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5888 to accept a host name and all its aliases if the forward lookup for any of
5889 them yielded the IP address of the incoming connection. Now it accepts only
5890 those names whose forward lookup yields the correct IP address. Any other
5891 names are discarded. This closes a loophole whereby a rogue DNS
5892 administrator could create reverse DNS records to break through a
5893 wildcarded host restriction in an ACL.
5895 37. If a user filter or a system filter that ran in a subprocess used any of
5896 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5897 the wrong values were passed to the pipe command ($thisaddress had the
5898 value of $0, $0 had the value of $1, etc). This bug was introduced by
5899 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5901 38. Improved the line breaking for long SMTP error messages from ACLs.
5902 Previously, if there was no break point between 40 and 75 characters, Exim
5903 left the rest of the message alone. Two changes have been made: (a) I've
5904 reduced the minimum length to 35 characters; (b) if it can't find a break
5905 point between 35 and 75 characters, it looks ahead and uses the first one
5906 that it finds. This may give the occasional overlong line, but at least the
5907 remaining text gets split now.
5909 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5910 file descriptors might be low, and that setting 1000 would always raise it.
5911 It turns out that in some environments, the limit is already over 1000 and
5912 that lowering it causes trouble. So now Exim takes care not to decrease it.
5914 40. When delivering a message, the value of $return_path is set to $sender_
5915 address at the start of routing (routers may change the value). By an
5916 oversight, this default was not being set up when an address was tested by
5917 -bt or -bv, which affected the outcome if any router or filter referred to
5920 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5921 log only when "message" or "log_message" are set. However, if one of the
5922 conditions was an address verification, or a call to a nested ACL, the
5923 messages generated by the underlying test were being passed through. This
5924 no longer happens. The underlying message is available in $acl_verify_
5925 message for both "message" and "log_message" expansions, so it can be
5926 passed through if needed.
5928 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5929 new expansion $bh_ to give the encoded byte string without charset
5930 translation. Translation happens only if iconv() is available; HAVE_ICONV
5931 indicates this at build time. HEADERS_CHARSET gives the charset to
5932 translate to; headers_charset can change it in the configuration, and
5933 "headers charset" can change it in an individual filter file.
5935 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5936 that creates RFC 2047 encoded "words" labels them as that charset instead
5937 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5938 expansion operator; (ii) when Exim creates a From: line for a local
5939 message; (iii) when a header line is rewritten to include a "phrase" part.
5941 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5942 buggy, causing it to skip the first lines of messages whose message ID
5943 ended in 'D'. This would not have bitten before Exim release 4.14, because
5944 message IDs were unlikely to end in 'D' before then. The effect was to have
5945 incorrect size information for certain domains.
5947 45. #include "config.h" was missing at the start of the crypt16.c module. This
5948 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5951 46. If there was a timeout during a "random" callout check, Exim treated it as
5952 a failure of the random address, and carried on sending RSET and the real
5953 address. If the delay was just some slowness somewhere, the response to the
5954 original RCPT would be taken as a response to RSET and so on, causing
5955 mayhem of various kinds.
5957 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5958 when I implemented it. It didn't allow for the fact that some option values
5959 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5960 the right test for positive values.
5962 48. Domain names in DNS records are case-independent. Exim always looks them up
5963 in lower case. Some resolvers return domain names in exactly the case they
5964 appear in the zone file, that is, they may contain uppercase letters. Not
5965 all resolvers do this - some return always lower case. Exim was treating a
5966 change of case by a resolver as a change of domain, similar to a widening
5967 of a domain abbreviation. This triggered its re-routing code and so it was
5968 trying to route what was effectively the same domain again. This normally
5969 caused routing to fail (because the router wouldn't handle the domain
5970 twice). Now Exim checks for this case specially, and just changes the
5971 casing of the domain that it ultimately uses when it transmits the message
5974 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5977 50. If a filter generated a file delivery with a non-absolute name (possible if
5978 no home directory exists for the router), the forbid_file option was not
5981 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5982 the existing equality matching.
5984 52. Exim was using ints instead of ino_t variables in some places where it was
5985 dealing with inode numbers.
5987 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5988 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5989 called TMPDIR, and if it finds it is different, it changes its value.
5991 54. The smtp_printf() function is now made available to local_scan() so
5992 additional output lines can be written before returning. There is also an
5993 smtp_fflush() function to enable the detection of a dropped connection.
5994 The variables smtp_input and smtp_batched_input are exported to
5997 55. Changed the default runtime configuration: the message "Unknown user"
5998 has been removed from the ACL, and instead placed on the localuser router,
5999 using the cannot_route_message feature. This means that any verification
6000 failures that generate their own messages won't get overridden. Similarly,
6001 the "Unrouteable address" message that was in the ACL for unverifiable
6002 relay addresses has also been removed.
6004 56. Added hosts_avoid_esmtp to the smtp transport.
6006 57. The exicyclog script was not checking for the esoteric option
6007 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6008 will work only if exicyclog is run under the appropriate euid.
6010 58. Following a discussion on the list, the rules by which Exim recognises line
6011 endings on incoming messages have been changed. The -dropcr and drop_cr
6012 options are now no-ops, retained only for backwards compatibility. The
6013 following line terminators are recognized: LF CRLF CR. However, special
6014 processing applies to CR:
6016 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6017 nor a local message in the state where . is a terminator.
6019 (ii) If a bare CR is encountered in a header line, an extra space is added
6020 after the line terminator so as not to end the header. The reasoning
6021 behind this is that bare CRs in header lines are most likely either
6022 to be mistakes, or people trying to play silly games.
6024 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6025 was being incorrectly given as 18 bytes larger than it should have been.
6026 This is a VOB (very old bug).
6028 60. This may never have affected anything current, but just in case it has:
6029 When the local host is found other than at the start of a list of hosts,
6030 the local host, those with the same MX, and any that follow, are discarded.
6031 When the list in question was part of a longer list of hosts, the following
6032 hosts (not currently being processed) were also being discarded. This no
6033 longer happens. I'm not sure if this situation could ever has previously
6036 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6039 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6040 header. This is something that is recommended in a new Internet Draft, and
6041 is something that is documented as being done by Sendmail. There are two
6042 possible values. For messages generated by the autoreply transport, Exim
6045 Auto-Submitted: auto-replied
6047 whereas for all other generated messages (e.g. bounces) it adds
6049 Auto-Submitted: auto-generated
6051 63. The "personal" condition in filters now includes a test for the
6052 Auto-Submitted: header. If it contains the string "auto-" the message it
6053 not considered personal.
6055 64. Added rcpt_include_affixes as a generic transport option.
6057 65. Added queue_only_override (default true).
6059 66. Added the syslog_duplication option.
6061 67. If what should have been the first header line of a message consisted of
6062 a space followed by a colon, Exim was mis-interpreting it as a header line.
6063 It isn't of course - it is syntactically invalid and should therefore be
6064 treated as the start of the message body. The misbehaviour could have
6065 caused a number of strange effects, including loss of data in subsequent
6066 header lines, and spool format errors.
6068 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6069 client host had authenticated. This control can now be exercised by an ACL
6070 for more flexibility.
6072 69. By default, callouts do not happen when testing with -bh. There is now a
6073 variant, -bhc, which does actually run the callout code, including
6074 consulting and updating the callout cache.
6076 70. Added support for saslauthd authentication, courtesy of Alexander
6079 71. If statvfs() failed on the spool or log directories while checking their
6080 size for availability, Exim confusingly gave the error "space shortage".
6081 Furthermore, in debugging mode it crashed with a floating point exception.
6082 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6083 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6084 serious problem, Exim now writes to the main and panic logs when this
6085 happens, with details of the failure. It then refuses to accept the
6086 incoming message, giving the message "spool directory problem" or "log
6087 directory problem" with a 421 code for SMTP messages.
6089 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6090 0, 1, and 2 exist, because some OS complain for execs without them (see
6091 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6092 descriptors. However, the code omitted to check that the open succeeded,
6093 causing mysterious errors if for some reason the permissions on /dev/null
6094 got screwed. Now Exim writes a message to the main and panic logs, and
6095 bombs out if it can't open /dev/null.
6097 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6098 interact so that it is all more flexible. It is supposed to remain
6099 backwards compatible. Also added extra_local_interfaces.
6101 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6102 to bomb out with an assertion failure - to the client this appears as a
6103 connection drop. This problem occurs in the part of the code that was taken
6104 from the Samba project. Fortunately, the assertion is in a very simple
6105 function, so I have fixed this by reproducing the function inline in the
6106 one place where it is called, and arranging for authentication to fail
6107 instead of killing the process with assert().
6109 75. The SPA client code was not working when the server requested OEM rather
6110 than Unicode encoding.
6112 76. Added code to make require_files with a specific uid setting more usable in
6113 the case where statting the file as root fails - usually a non-root-mounted
6114 NFS file system. When this happens and the failure is EACCES, Exim now
6115 forks a subprocess and does the per-uid checking as the relevant uid.
6117 77. Added process_log_path.
6119 78. If log_file_path was not explicitly set, a setting of check_log_space or
6120 check_log_inodes was ignored.
6122 79. If a space check for the spool or log partitions fails, the incident is now
6123 logged. Of course, in the latter case the data may get lost...
6125 80. Added the %p formatting code to string_format() so that it can be used to
6126 print addresses in debug_print(). Adjusted all the address printing in the
6127 debugging in store.c to use %p rather than %d.
6129 81. There was a concern that a line of code in smtp_in.c could overflow a
6130 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6131 initially expressed, the concern was not well-founded, because trailing
6132 spaces are removed early. However, if the trailing spaces were followed by
6133 a NULL, they did not get removed, so the overflow was possible. Two fixes
6136 (a) I re-wrote the offending code in a cleaner fashion.
6137 (b) If an incoming SMTP command contains a NULL character, it is rejected
6140 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6141 runs initgroups(), so that if the Exim user is in any additional groups,
6142 they will be used during message reception.
6148 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob