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.
62 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
63 checkers that were complaining about end of non-void function with no
66 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
67 This was a regression intruduced in 4.83 by another bugfix.
69 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
71 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
72 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
78 TF/01 Correctly close the server side of TLS when forking for delivery.
80 When a message was received over SMTP with TLS, Exim failed to clear up
81 the incoming connection properly after forking off the child process to
82 deliver the message. In some situations the subsequent outgoing
83 delivery connection happened to have the same fd number as the incoming
84 connection previously had. Exim would try to use TLS and fail, logging
85 a "Bad file descriptor" error.
87 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
88 utilities have not been installed.
90 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
91 temporary space as the ACL may create new global variables.
93 TL/01 LDAP support uses per connection or global context settings, depending
94 upon the detected version of the libraries at build time.
96 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
97 to extract and use the src ip:port in logging and expansions as if it
98 were a direct connection from the outside internet. PPv2 support was
99 updated based on HAProxy spec change in May 2014.
101 JH/02 Add ${listextract {number}{list}{success}{fail}}.
103 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
104 Properly escape header and check for NULL return.
106 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
109 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
111 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
112 characters in header names, implemented as a verify condition.
113 Contributed by Michael Fischer v. Mollard.
115 TL/05 Rename SPF condition results err_perm and err_temp to standardized
116 results permerror and temperror. Previous values are deprecated but
117 still accepted. In a future release, err_perm and err_temp will be
118 completely removed, which will be a backward incompatibility if the
119 ACL tests for either of these two old results. Patch contributed by
120 user bes-internal on the mailing list.
122 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
124 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
125 selectors, in both main and reject logs.
127 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
130 JH/07 Add malware type "sock" for talking to simple daemon.
132 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
134 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
135 routers/transports under cutthrough routing.
137 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
138 numbers. Touch up "bool" conditional to keep the same definition.
140 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
142 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
144 JH/12 Expand items in router/transport headers_add or headers_remove lists
145 individually rather than the list as a whole. Bug 1452.
147 Required for reasonable handling of multiple headers_ options when
148 they may be empty; requires that headers_remove items with embedded
149 colons must have them doubled (or the list-separator changed).
151 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
152 view the policy declared in the DMARC record. Currently, $dmarc_status
153 is a combined value of both the record presence and the result of the
156 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
158 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
159 dnslookup router and the smtp transport (applying to the forward
162 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
163 of ldap servers used for a specific lookup. Patch provided by Heiko
166 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
167 New variable $lookup_dnssec_authenticated for observability.
169 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
170 Patch submitted by Lars Timman.
172 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
174 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
175 Requires trusted mode and valid format message id, aborts otherwise.
176 Patch contributed by Heiko Schlichting.
178 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
179 certextract with support for various fields. Bug 1358.
181 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
182 is requested by default, modifiable by smtp transport option
185 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
186 operate on certificate variables to give certificate fingerprints
187 Also new ${sha256:cert_variable}.
189 JH/23 The PRDR feature is moved from being Experimental into the mainline.
191 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
192 Christian Aistleitner.
194 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
196 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
197 file. Patch from Wolfgang Breyha.
199 JH/25 Expand the coverage of the delivery $host and $host_address to
200 client authenticators run in verify callout. Bug 1476.
202 JH/26 Port service names are now accepted for tls_on_connect_ports, to
203 align with daemon_smtp_ports. Bug 72.
205 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
206 support and error reporting did not work properly.
208 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
209 and is readable. Patch from Andrew Colin Kissa.
211 TL/14 Enhance documentation of ${run expansion and how it parses the
212 commandline after expansion, particularly in the case when an
213 unquoted variable expansion results in an empty value.
215 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
217 PP/02 Fix internal collision of T_APL on systems which support RFC3123
218 by renaming away from it. Addresses GH issue 15, reported by
221 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
223 TL/15 SECURITY: prevent double expansion in math comparison functions
224 (can expand unsanitized data). Not remotely exploitable.
231 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
233 PP/02 Make -n do something, by making it not do something.
234 When combined with -bP, the name of an option is not output.
236 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
239 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
240 $sender_host_name and config options to manage this, and basic check
243 PP/05 DSCP support for outbound connections and control modifier for inbound.
245 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
246 (Only plugin which currently uses this is kerberos4, which nobody should
247 be using, but we should make it available and other future plugins might
248 conceivably use it, even though it would break NAT; stuff *should* be
249 using channel bindings instead).
251 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
252 name; added for Sendmail compatibility; requires admin caller.
253 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
254 just ignore it); requires trusted caller.
255 Also parse but ignore: -Ac -Am -X<logfile>
258 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
260 TL/02 Add +smtp_confirmation as a default logging option.
262 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
263 Patch by Magnus Holmgren from 2007-02-20.
265 TL/04 Bugzilla 1281 - Spec typo.
266 Bugzilla 1283 - Spec typo.
267 Bugzilla 1290 - Spec grammar fixes.
269 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
271 TL/06 Add Experimental DMARC support using libopendmarc libraries.
273 TL/07 Fix an out of order global option causing a segfault. Reported to dev
274 mailing list by by Dmitry Isaikin.
276 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
278 JH/02 Support "G" suffix to numbers in ${if comparisons.
280 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
282 NM/01 Bugzilla 1197 - Spec typo
283 Bugzilla 1196 - Spec examples corrections
285 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
287 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
288 gnutls_enable_pkcs11, but renamed to more accurately indicate its
291 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
292 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
294 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
295 "acl {{name}{arg}...}", and optional args on acl condition
298 JH/05 Permit multiple router/transport headers_add/remove lines.
300 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
302 JH/07 Avoid using a waiting database for a single-message-only transport.
303 Performance patch from Paul Fisher. Bugzilla 1262.
305 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
308 JH/09 Add $headers_added variable, with content from use of ACL modifier
309 add_header (but not yet added to the message). Bugzilla 199.
311 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
312 Pulled from Bugzilla 817 by Wolfgang Breyha.
314 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
316 (nb: this is the same fix as in Exim 4.80.1)
318 JH/11 Add A= logging on delivery lines, and a client_set_id option on
321 JH/12 Add optional authenticated_sender logging to A= and a log_selector
324 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
326 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
327 advertise SMTP AUTH mechanism to us, instead of a generic
328 protocol violation error. Also, make Exim more robust to bad
329 data from the Dovecot auth socket.
331 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
333 When a queue runner is handling a message, Exim first routes the
334 recipient addresses, during which it prunes them based on the retry
335 hints database. After that it attempts to deliver the message to
336 any remaining recipients. It then updates the hints database using
339 So if a recipient address works intermittently, it can get repeatedly
340 deferred at routing time. The retry hints record remains fresh so the
341 address never reaches the final cutoff time.
343 This is a fairly common occurrence when a user is bumping up against
344 their storage quota. Exim had some logic in its local delivery code
345 to deal with this. However it did not apply to per-recipient defers
346 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
348 This change adds a proper retry rule check during routing so that the
349 final cutoff time is checked against the message's age. We only do
350 this check if there is an address retry record and there is not a
351 domain retry record; this implies that previous attempts to handle
352 the address had the retry_use_local_parts option turned on. We use
353 this as an approximation for the destination being like a local
354 delivery, as in LMTP.
356 I suspect this new check makes the old local delivery cutoff check
357 redundant, but I have not verified this so I left the code in place.
359 TF/02 Correct gecos expansion when From: is a prefix of the username.
361 Test 0254 submits a message to Exim with the header
365 When I ran the test suite under the user fanf2, Exim expanded
366 the header to contain my full name, whereas it should have added
367 a Resent-Sender: header. It erroneously treats any prefix of the
368 username as equal to the username.
370 This change corrects that bug.
372 GF/01 DCC debug and logging tidyup
373 Error conditions log to paniclog rather than rejectlog.
374 Debug lines prefixed by "DCC: " to remove any ambiguity.
376 TF/03 Avoid unnecessary rebuilds of lookup-related code.
378 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
379 Bug spotted by Jeremy Harris; was flawed since initial commit.
380 Would have resulted in OCSP responses post-SNI triggering an Exim
381 NULL dereference and crash.
383 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
385 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
386 Bug detection, analysis and fix by Samuel Thibault.
387 Bugzilla 1331, Debian bug #698092.
389 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
391 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
392 Server implementation by Todd Lyons, client by JH.
393 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
394 config variable "prdr_enable" controls whether the server
395 advertises the facility. If the client requests PRDR a new
396 acl_data_smtp_prdr ACL is called once for each recipient, after
397 the body content is received and before the acl_smtp_data ACL.
398 The client is controlled by bolth of: a hosts_try_prdr option
399 on the smtp transport, and the server advertisement.
400 Default client logging of deliveries and rejections involving
401 PRDR are flagged with the string "PRDR".
403 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
404 fclose(). Diagnosis by Todd Lyons.
406 PP/17 Update configure.default to handle IPv6 localhost better.
407 Patch by Alain Williams (plus minor tweaks).
410 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
411 This is now consistent with GnuTLS, and is now documented: the
412 previous undocumented portable approach to treating the option as
413 unset was to force an expansion failure. That still works, and
414 an empty string is now equivalent.
416 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
417 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
418 not performing validation itself.
420 PP/20 Added force_command boolean option to pipe transport.
421 Patch from Nick Koston, of cPanel Inc.
423 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
426 TF/04 Added udpsend ACL modifer and hexquote expansion operator
428 PP/21 Fix eximon continuous updating with timestamped log-files.
429 Broken in a format-string cleanup in 4.80, missed when I repaired the
430 other false fix of the same issue.
431 Report and fix from Heiko Schlichting.
434 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
435 Report from Prashanth Katuri.
437 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
438 It's SecureTransport, so affects any MacOS clients which use the
439 system-integrated TLS libraries, including email clients.
441 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
442 using a MIME ACL for non-SMTP local injection.
443 Report and assistance in diagnosis by Warren Baker.
445 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
447 JH/16 Fix comparisons for 64b. Bugzilla 1385.
449 TL/09 Add expansion variable $authenticated_fail_id to keep track of
450 last id that failed so it may be referenced in subsequent ACL's.
452 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
455 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
456 ldap library initialization, allowing self-signed CA's to be
457 used. Also properly sets require_cert option later in code by
458 using NULL (global ldap config) instead of ldap handle (per
459 session). Bug diagnosis and testing by alxgomz.
461 TL/12 Enhanced documentation in the ratelimit.pl script provided in
462 the src/util/ subdirectory.
464 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
465 renamed to Transport Post Delivery Action by Jeremy Harris, as
468 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
469 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
470 redis_servers = needs to be configured which will be used by the redis
471 lookup. Patch from Warren Baker, of The Packet Hub.
473 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
475 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
476 hostname or reverse DNS when processing a host list. Used suggestions
477 from multiple comments on this bug.
479 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
481 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
482 Missed a few lines, added it to make the runtest require no keyboard
485 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
486 contains upper case chars. Make router use caseful_local_part.
488 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
489 support when GnuTLS has been built with p11-kit.
495 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
497 This, or similar/improved, will also be change PP/11 of 4.82.
503 PP/01 Handle short writes when writing local log-files.
504 In practice, only affects FreeBSD (8 onwards).
505 Bugzilla 1053, with thanks to Dmitry Isaikin.
507 NM/01 Bugzilla 949 - Documentation tweak
509 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
512 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
514 PP/02 Implemented gsasl authenticator.
516 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
518 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
519 `pkg-config foo` for cflags/libs.
521 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
522 with rest of GSASL and with heimdal_gssapi.
524 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
525 `pkg-config foo` for cflags/libs for the TLS implementation.
527 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
528 properties get this fed in as external SSF. A number of robustness
529 and debugging improvements to the cyrus_sasl authenticator.
531 PP/08 cyrus_sasl server now expands the server_realm option.
533 PP/09 Bugzilla 1214 - Log authentication information in reject log.
534 Patch by Jeremy Harris.
536 PP/10 Added dbmjz lookup type.
538 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
540 PP/12 MAIL args handles TAB as well as SP, for better interop with
541 non-compliant senders.
542 Analysis and variant patch by Todd Lyons.
544 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
545 Bug report from Lars Müller <lars@samba.org> (via SUSE),
546 Patch from Dirk Mueller <dmueller@suse.com>
548 PP/13 tls_peerdn now print-escaped for spool files.
549 Observed some $tls_peerdn in wild which contained \n, which resulted
550 in spool file corruption.
552 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
553 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
554 or write after TLS renegotiation, which otherwise led to messages
557 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
558 as a tracking header (ie: a signed header comes before the signature).
559 Patch from Wolfgang Breyha.
561 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
562 comma-sep list; embedded commas doubled.
564 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
566 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
568 Report and patch from Dmitry Banschikov.
570 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
571 Removed SSL_clear() after SSL_new() which led to protocol negotiation
572 failures. We appear to now support TLS1.1+ with Exim.
574 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
575 lets Exim select keys and certificates based upon TLS SNI from client.
576 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
577 before an outbound SMTP session. New log_selector, +tls_sni.
579 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
580 NULL dereference. Report and patch from Alun Jones.
582 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
583 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
584 Not seeing resolver debug output on NetBSD, but suspect this is a
585 resolver implementation change.
587 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
588 Left warnings. Added "eximon gdb" invocation mode.
590 PP/21 Defaulting "accept_8bitmime" to true, not false.
592 PP/22 Added -bw for inetd wait mode support.
594 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
595 locate the relevant includes and libraries. Made this the default.
597 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
598 Bugzilla 1246, report and most of solution from Tomasz Kusy.
600 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
601 This may cause build issues on older platforms.
603 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
604 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
605 gnutls_require_mac & gnutls_require_protocols (no longer supported).
606 Added SNI support via GnuTLS too.
607 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
609 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
611 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
612 Applied second patch from Janne, implementing suggestion to default
613 multiple-strings-in-record handling to match SPF spec.
615 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
617 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
618 read-only, out of scope).
619 Patch from Wolfgang Breyha, report from Stuart Northfield.
621 PP/29 Fix three issues highlighted by clang analyser static analysis.
622 Only crash-plausible issue would require the Cambridge-specific
623 iplookup router and a misconfiguration.
624 Report from Marcin Mirosław.
626 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
628 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
629 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
630 As part of this, removing so much warning spew let me fix some minor
631 real issues in debug logging.
633 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
634 assignment on my part. Fixed.
636 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
637 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
638 Janne Snabb (who went above and beyond: thank you).
640 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
641 string otherwise requires a connection and a bunch more work and it's
642 relatively easy to get wrong. Should also expose TLS library linkage
645 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
646 64-bit ${eval} (JH/03).
648 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
649 GNU libc to support some of the 64-bit stuff, should not lead to
650 conflicts. Defined before os.h is pulled in, so if a given platform
651 needs to override this, it can.
653 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
654 protection layer was required, which is not implemented.
655 Bugzilla 1254, patch from Wolfgang Breyha.
657 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
658 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
659 tls_dhparam take prime identifiers. Also unbreak combination of
660 OpenSSL+DH_params+TLSSNI.
662 PP/39 Disable SSLv2 by default in OpenSSL support.
668 PP/01 Solaris build fix for Oracle's LDAP libraries.
669 Bugzilla 1109, patch from Stephen Usher.
671 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
673 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
676 TF/02 Fix a couple more cases where we did not log the error message
677 when unlink() failed. See also change 4.74-TF/03.
679 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
680 lock up or crash if it happened to be inside a call to libc when it
681 got a SIGUSR1 from exiwhat.
683 The SIGUSR1 handler appends the current process status to the process
684 log which is later printed by exiwhat. It used to use the general
685 purpose logging code to do this, but several functions it calls are
686 not safe for signals.
688 The new output code in the SIGUSR1 handler is specific to the process
689 log, and simple enough that it's easy to inspect for signal safety.
690 Removing some special cases also simplifies the general logging code.
691 Removing the spurious timestamps from the process log simplifies
694 TF/04 Improved ratelimit ACL condition.
696 The /noupdate option has been deprecated in favour of /readonly which
697 has clearer semantics. The /leaky, /strict, and /readonly update modes
698 are mutually exclusive. The update mode is no longer included in the
699 database key; it just determines when the database is updated. (This
700 means that when you upgrde Exim will forget old rate measurements.)
702 Exim now checks that the per_* options are used with an update mode that
703 makes sense for the current ACL. For example, when Exim is processing a
704 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
705 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
706 must specify per_mail/readonly. If you omit the update mode it defaults to
707 /leaky where that makes sense (as before) or /readonly where required.
709 The /noupdate option is now undocumented but still supported for
710 backwards compatibility. It is equivalent to /readonly except that in
711 ACLs where /readonly is required you may specify /leaky/noupdate or
712 /strict/noupdate which are treated the same as /readonly.
714 A useful new feature is the /count= option. This is a generalization
715 of the per_byte option, so that you can measure the throughput of other
716 aggregate values. For example, the per_byte option is now equivalent
717 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
719 The per_rcpt option has been generalized using the /count= mechanism
720 (though it's more complicated than the per_byte equivalence). When it is
721 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
722 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
723 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
724 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
725 Note that using per_rcpt with a non-readonly update mode in more than
726 one ACL will cause the recipients to be double-counted. (The per_mail
727 and per_byte options don't have this problem.)
729 The handling of very low rates has changed slightly. If the computed rate
730 is less than the event's count (usually one) then this event is the first
731 after a long gap. In this case the rate is set to the same as this event's
732 count, so that the first message of a spam run is counted properly.
734 The major new feature is a mechanism for counting the rate of unique
735 events. The new per_addr option counts the number of different
736 recipients that someone has sent messages to in the last time period. It
737 behaves like per_rcpt if all the recipient addresses are different, but
738 duplicate recipient addresses do not increase the measured rate. Like
739 the /count= option this is a general mechanism, so the per_addr option
740 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
741 example, measure the rate that a client uses different sender addresses
742 with the options per_mail/unique=$sender_address. There are further
743 details in the main documentation.
745 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
747 TF/06 Removed a few PCRE remnants.
749 TF/07 Automatically extract Exim's version number from tags in the git
750 repository when doing development or release builds.
752 PP/02 Raise smtp_cmd_buffer_size to 16kB.
753 Bugzilla 879. Patch from Paul Fisher.
755 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
756 Heavily based on revision 40f9a89a from Simon Arlott's tree.
759 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
761 PP/05 Variable $av_failed, true if the AV scanner deferred.
762 Bugzilla 1078. Patch from John Horne.
764 PP/06 Stop make process more reliably on build failure.
765 Bugzilla 1087. Patch from Heiko Schlittermann.
767 PP/07 Make maildir_use_size_file an _expandable_ boolean.
768 Bugzilla 1089. Patch from Heiko Schlittermann.
770 PP/08 Handle ${run} returning more data than OS pipe buffer size.
771 Bugzilla 1131. Patch from Holger Weiß.
773 PP/09 Handle IPv6 addresses with SPF.
774 Bugzilla 860. Patch from Wolfgang Breyha.
776 PP/10 GnuTLS: support TLS 1.2 & 1.1.
778 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
781 PP/11 match_* no longer expand right-hand-side by default.
782 New compile-time build option, EXPAND_LISTMATCH_RHS.
783 New expansion conditions, "inlist", "inlisti".
785 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
787 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
789 PP/14 fix log_write() format string regression from TF/03.
790 Bugzilla 1152. Patch from Dmitry Isaikin.
796 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
798 PP/02 Harmonised TLS library version reporting; only show if debugging.
799 Layout now matches that introduced for other libraries in 4.74 PP/03.
801 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
803 PP/04 New "dns_use_edns0" global option.
805 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
808 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
809 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
811 TK/01 Updated PolarSSL code to 0.14.2.
812 Bugzilla 1097. Patch from Andreas Metzler.
814 PP/07 Catch divide-by-zero in ${eval:...}.
817 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
820 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
821 format-string attack -- SECURITY: remote arbitrary code execution.
823 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
824 time unintentionally subject to list matching rules, letting the header
825 cause arbitrary Exim lookups (of items which can occur in lists, *not*
826 arbitrary string expansion). This allowed for information disclosure.
828 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
829 INT_MIN/-1 -- value coerced to INT_MAX.
835 NM/01 Workround for PCRE version dependency in version reporting
838 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
839 This fixes portability to compilers other than gcc, notably
840 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
842 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
843 makefiles for portability to HP-UX and POSIX correctness.
845 PP/01 Permit LOOKUP_foo enabling on the make command-line.
846 Also via indented variable definition in the Makefile.
847 (Debugging by Oliver Heesakkers).
849 PP/02 Restore caching of spamd results with expanded spamd_address.
850 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
852 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
853 Improves build reliability. Fix from: Frank Elsner
855 NM/02 Fix wide character breakage in the rfc2047 coding
856 Fixes bug 1064. Patch from Andrey N. Oktyabrski
858 NM/03 Allow underscore in dnslist lookups
859 Fixes bug 1026. Patch from Graeme Fowler
861 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
862 Code patches from Adam Ciarcinski of NetBSD.
864 NM/04 Fixed exiqgrep to cope with mailq missing size issue
867 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
868 is logged, to avoid truncation. Patch from John Horne.
870 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
871 Patch from Jakob Hirsch.
873 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
874 SQL string expansion failure details.
875 Patch from Andrey Oktyabrski.
877 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
878 Patch from Simon Arlott.
880 PP/09 New lookups functionality failed to compile on old gcc which rejects
881 extern declarations in function scope.
882 Patch from Oliver Fleischmann
884 PP/10 Use sig_atomic_t for flags set from signal handlers.
885 Check getgroups() return and improve debugging.
886 Fixed developed for diagnosis in bug 927 (which turned out to be
889 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
890 Patch from Mark Zealey.
892 PP/12 Bugzilla 1056: Improved spamd server selection.
893 Patch from Mark Zealey.
895 PP/13 Bugzilla 1086: Deal with maildir quota file races.
896 Based on patch from Heiko Schlittermann.
898 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
899 Patch from Uwe Doering, sign-off by Michael Haardt.
901 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
902 variable declaration deep within a block. Bug and patch from
905 PP/15 lookups-Makefile IRIX compatibilty coercion.
907 PP/16 Make DISABLE_DKIM build knob functional.
909 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
910 Patch by Simon Arlott
912 TF/03 Fix valgrind.h portability to C89 compilers that do not support
913 variable argument macros. Our copy now differs from upstream.
919 TF/01 Failure to get a lock on a hints database can have serious
920 consequences so log it to the panic log.
922 TF/02 Log LMTP confirmation messages in the same way as SMTP,
923 controlled using the smtp_confirmation log selector.
925 TF/03 Include the error message when we fail to unlink a spool file.
927 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
928 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
929 for maintaining out-of-tree patches for some time.
931 PP/01 Bugzilla 139: Documentation and portability issues.
932 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
933 Handle per-OS dynamic-module compilation flags.
935 PP/02 Let /dev/null have normal permissions.
936 The 4.73 fixes were a little too stringent and complained about the
937 permissions on /dev/null. Exempt it from some checks.
938 Reported by Andreas M. Kirchwitz.
940 PP/03 Report version information for many libraries, including
941 Exim version information for dynamically loaded libraries. Created
942 version.h, now support a version extension string for distributors
943 who patch heavily. Dynamic module ABI change.
945 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
946 privilege escalation vulnerability whereby the Exim run-time user
947 can cause root to append content of the attacker's choosing to
950 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
953 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
954 If dropping privileges for untrusted macros, we disabled normal logging
955 on the basis that it would fail; for the Exim run-time user, this is not
956 the case, and it resulted in successful deliveries going unlogged.
957 Fixed. Reported by Andreas Metzler.
963 PP/01 Date: & Message-Id: revert to normally being appended to a message,
964 only prepend for the Resent-* case. Fixes regression introduced in
965 Exim 4.70 by NM/22 for Bugzilla 607.
967 PP/02 Include check_rfc2047_length in configure.default because we're seeing
968 increasing numbers of administrators be bitten by this.
970 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
972 PP/03 Bugzilla 994: added openssl_options main configuration option.
974 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
976 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
978 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
980 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
981 without a peer certificate, leading to a segfault because of an
982 assumption that peers always have certificates. Be a little more
983 paranoid. Problem reported by Martin Tscholak.
985 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
986 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
987 NB: ClamAV planning to remove STREAM in "middle of 2010".
988 CL also introduces -bmalware, various -d+acl logging additions and
989 more caution in buffer sizes.
991 PP/09 Implemented reverse_ip expansion operator.
993 PP/10 Bugzilla 937: provide a "debug" ACL control.
995 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
997 PP/12 Bugzilla 973: Implement --version.
999 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1001 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1003 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1005 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1006 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1007 ignore trailing whitespace.
1009 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1011 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1014 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1015 Notification from Dr Andrew Aitchison.
1017 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1018 ExtendedDetectionInfo response format.
1019 Notification from John Horne.
1021 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1024 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1025 XSL and documented dependency on system catalogs, with examples of how
1028 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1031 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1032 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1033 configuration file which is writeable by the Exim user or group.
1035 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1036 of configuration files to cover files specified with the -C option if
1037 they are going to be used with root privileges, not just the default
1040 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1041 option (effectively making it always true).
1043 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1044 files to be used while preserving root privileges.
1046 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1047 that rogue child processes cannot use them.
1049 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1050 run-time user, instead of root.
1052 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1053 Exim run-time user without dropping privileges.
1055 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1056 result string, instead of calling string_vformat() twice with the same
1059 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1060 for other users. Others should always drop root privileges if they use
1061 -C on the command line, even for a whitelisted configure file.
1063 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1065 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1071 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1072 $data_path, and $header_path variables; fixed documentation bugs and
1075 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1076 exipick to access non-standard spools, including the "frozen" queue
1079 NM/01 Bugzilla 965: Support mysql stored procedures.
1080 Patch from Alain Williams
1082 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1084 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1085 Patch from Andreas Metzler
1087 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1088 Patch from Kirill Miazine
1090 NM/05 Bugzilla 671: Added umask to procmail example.
1092 JJ/03 installed exipick 20100323.0, fixing doc bug
1094 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1095 directory. Notification and patch from Dan Rosenberg.
1097 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1099 TK/02 Improve log output when DKIM signing operation fails.
1101 MH/01 Treat the transport option dkim_domain as a colon separated
1102 list, not as a single string, and sign the message with each element,
1103 omitting multiple occurences of the same signer.
1105 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1106 Bugzilla 985, 986. Patch by Simon Arlott
1108 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1109 Patch by Simon Arlott
1111 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1112 MBX locking. Notification from Dan Rosenberg.
1118 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1120 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1122 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1124 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1126 NM/04 Bugzilla 915: Flag broken perl installation during build.
1132 TK/01 Added patch by Johannes Berg that expands the main option
1133 "spamd_address" if it starts with a dollar sign.
1135 TK/02 Write list of recipients to X-Envelope-Sender header when building
1136 the mbox-format spool file for content scanning (suggested by Jakob
1139 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1140 (http://www.dcc-servers.net/) support via dccifd. Activated by
1141 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1143 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1144 by Mark Daniel Reidel <mr@df.eu>.
1146 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1147 When building exim an external PCRE library is now needed -
1148 PCRE is a system library on the majority of modern systems.
1149 See entry on PCRE_LIBS in EDITME file.
1151 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1152 conversation. Added nologin parameter to request.
1153 Patch contributed by Kirill Miazine.
1155 TF/01 Do not log submission mode rewrites if they do not change the address.
1157 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1159 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1160 log files in place. Contributed by Roberto Lima.
1162 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1164 TF/03 Bugzilla 615: When checking the local_parts router precondition
1165 after a local_part_suffix or local_part_prefix option, Exim now
1166 does not use the address's named list lookup cache, since this
1167 contains cached lookups for the whole local part.
1169 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1170 Robert Millan. Documentation is in experimental-spec.txt.
1172 TF/04 Bugzilla 668: Fix parallel build (make -j).
1174 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1176 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1177 Patch provided by Jan Srzednicki.
1179 TF/05 Leading white space used to be stripped from $spam_report which
1180 wrecked the formatting. Now it is preserved.
1182 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1183 that they are available at delivery time.
1185 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1187 TF/08 TLS error reporting now respects the incoming_interface and
1188 incoming_port log selectors.
1190 TF/09 Produce a more useful error message if an SMTP transport's hosts
1191 setting expands to an empty string.
1193 NM/06 Bugzilla 744: EXPN did not work under TLS.
1194 Patch provided by Phil Pennock.
1196 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1197 Patch provided by Richard Godbee.
1199 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1200 acl_smtp_notquit, added index entry.
1202 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1203 Patch provided by Eugene Bujak.
1205 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1206 accept(). Patch provided by Maxim Dounin.
1208 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1209 Patch provided by Phil Pennock.
1211 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1213 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1214 Patch provided by Brad "anomie" Jorsch.
1216 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1217 Patch provided by Dean Brooks.
1219 TK/05 Add native DKIM support (does not depend on external libraries).
1221 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1222 Patch provided by Graeme Fowler.
1224 NM/16 Bugzilla 851: Documentation example syntax fix.
1226 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1228 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1231 NM/19 Bugzilla 745: TLS version reporting.
1232 Patch provided by Phil Pennock.
1234 NM/20 Bugzilla 167: bool: condition support.
1235 Patch provided by Phil Pennock.
1237 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1238 clients. Patch provided by Phil Pennock.
1240 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1241 Patch provided by Brad "anomie" Jorsch.
1243 NM/23 Bugzilla 687: Fix misparses in eximstats.
1244 Patch provided by Heiko Schlittermann.
1246 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1247 Patch provided by Heiko Schlittermann.
1249 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1250 plus update to original patch.
1252 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1254 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1255 Patch provided by David Brownlee.
1257 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1259 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1261 NM/30 Bugzilla 888: TLS documentation bugfixes.
1263 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1265 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1266 Unlike the original bugzilla I have changed all shell scripts in src tree.
1268 NM/33 Bugzilla 898: Transport filter timeout fix.
1269 Patch by Todd Rinaldo.
1271 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1272 Patch by Serge Demonchaux.
1274 NM/35 Bugzilla 39: Base64 decode bug fixes.
1275 Patch by Jakob Hirsch.
1277 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1279 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1281 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1283 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1289 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1290 ALT-N's libdkim that I have put here:
1291 http://duncanthrax.net/exim-experimental/
1293 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1294 were called 'true' and it seems that C99 defines that as a reserved
1295 keyword to be used with 'bool' variable types. That means you could
1296 not include C99-style headers which use bools without triggering
1297 build errors in sieve.c.
1299 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1300 as mailq or other aliases. Changed the --help handling significantly
1301 to do whats expected. exim_usage() emits usage/help information.
1303 SC/01 Added the -bylocaldomain option to eximstats.
1305 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1307 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1309 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1315 PH/01 Another patch from the Sieve maintainer.
1317 PH/02 When an IPv6 address is converted to a string for single-key lookup
1318 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1319 dots are used instead of colons so that keys in lsearch files need not
1320 contain colons. This was done some time before quoting was made available
1321 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1322 (notated using the quote facility) so as to distinguish them from IPv4
1323 keys. This meant that lookups for IP addresses in host lists did not work
1324 for iplsearch lookups.
1326 This has been fixed by arranging for IPv6 addresses to be expressed with
1327 colons if the lookup type is iplsearch. This is not incompatible, because
1328 previously such lookups could never work.
1330 The situation is now rather anomolous, since one *can* have colons in
1331 ordinary lsearch keys. However, making the change in all cases is
1332 incompatible and would probably break a number of configurations.
1334 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1337 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1338 conversion specification without a maximum field width, thereby enabling
1339 a rogue spamd server to cause a buffer overflow. While nobody in their
1340 right mind would setup Exim to query an untrusted spamd server, an
1341 attacker that gains access to a server running spamd could potentially
1342 exploit this vulnerability to run arbitrary code as the Exim user.
1344 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1345 $primary_hostname instead of what libspf2 thinks the hosts name is.
1347 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1348 a directory entry by the name of the lookup key. Previously, if a
1349 symlink pointed to a non-existing file or a file in a directory that
1350 Exim lacked permissions to read, a lookup for a key matching that
1351 symlink would fail. Now it is enough that a matching directory entry
1352 exists, symlink or not. (Bugzilla 503.)
1354 PH/03 The body_linecount and body_zerocount variables are now exported in the
1357 PH/04 Added the $dnslist_matched variable.
1359 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1360 This means they are set thereafter only if the connection becomes
1363 PH/06 Added the client_condition to authenticators so that some can be skipped
1364 by clients under certain conditions.
1366 PH/07 The error message for a badly-placed control=no_multiline_responses left
1367 "_responses" off the end of the name.
1369 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1371 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1372 (without spaces) instead of just copying the configuration text.
1374 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1376 PH/11 Added $max_received_linelength.
1378 PH/12 Added +ignore_defer and +include_defer to host lists.
1380 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1381 way in which PCRE > 7.0 is built.
1383 PH/14 Implemented queue_only_load_latch.
1385 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1386 MAIL command. The effect was to mangle the value on 64-bit systems.
1388 PH/16 Another patch from the Sieve maintainer.
1390 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1392 PH/18 If a system quota error occurred while trying to create the file for
1393 a maildir delivery, the message "Mailbox is full" was not appended to the
1394 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1395 only to a quota excession during the actual writing of the file.
1397 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1398 characters?) which causes problems in log lines. The DN values are now
1399 passed through string_printing() before being added to log lines.
1401 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1402 and InterBase are left for another time.)
1404 PH/21 Added message_body_newlines option.
1406 PH/22 Guard against possible overflow in moan_check_errorcopy().
1408 PH/23 POSIX allows open() to be a macro; guard against that.
1410 PH/24 If the recipient of an error message contained an @ in the local part
1411 (suitably quoted, of course), incorrect values were put in $domain and
1412 $local_part during the evaluation of errors_copy.
1418 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1419 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1422 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1423 issue a MAIL command.
1425 PH/02 In an ACL statement such as
1427 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1429 if a client was not listed at all, or was listed with a value other than
1430 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1431 the condition was not true (as it should be), so access was not denied.
1432 The bug was that the ! inversion was incorrectly passed on to the second
1433 item. This has been fixed.
1435 PH/03 Added additional dnslists conditions == and =& which are different from
1436 = and & when the dns lookup returns more than one IP address.
1438 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1439 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1441 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1442 FSYNC, which compiles an option called disable_fsync that allows for
1443 bypassing fsync(). The documentation is heavily laced with warnings.
1445 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1447 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1448 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1449 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1450 including adding "make clean"; (3) Added -fPIC when compiling the test
1451 dynamically loaded module, to get rid of a warning.
1453 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1454 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1455 = 0s. The bug is otherwise harmless.
1457 PH/07 There was a bug in the dovecot authenticator such that the value of
1458 $auth1 could be overwritten, and so not correctly preserved, after a
1459 successful authentication. This usually meant that the value preserved by
1460 the server_setid option was incorrect.
1462 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1464 PH/09 Installed PCRE release 7.0.
1466 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1467 run for batched SMTP input. It is now run at the start of every message
1468 in the batch. While fixing this I discovered that the process information
1469 (output by running exiwhat) was not always getting set for -bs and -bS
1470 input. This is fixed, and it now also says "batched" for BSMTP.
1472 PH/11 Added control=no_pipelining.
1474 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1475 patch, slightly modified), and move the expansion of helo_data till after
1476 the connection is made in the smtp transport (so it can use these
1479 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1481 PH/14 Added log_selector = +pid.
1483 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1485 PH/16 Add ${if forany and ${if forall.
1487 PH/17 Added dsn_from option to vary the From: line in DSNs.
1489 PH/18 Flush SMTP output before performing a callout, unless control =
1490 no_callout_flush is set.
1492 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1493 was true (the default) a successful delivery failed to delete the retry
1494 item, thus causing premature timeout of the address. The bug is now
1497 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1499 PH/21 Long custom messages for fakedefer and fakereject are now split up
1500 into multiline reponses in the same way that messages for "deny" and
1501 other ACL rejections are.
1503 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1504 with slight modification.
1506 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1507 draft, changing the syntax and factoring some duplicate code".
1509 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1510 for deliveries of the second and subsequent messages over the same SMTP
1513 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1514 ${reduce, with only minor "tidies".
1516 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1518 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1519 expansion side effects.
1521 PH/27 When a message times out after an over-quota error from an Exim-imposed
1522 quota, the bounce message says "mailbox is full". This message was not
1523 being given when it was a system quota that was exceeded. It now should
1526 MH/03 Made $recipients available in local_scan(). local_scan() already has
1527 better access to the recipient list through recipients_list[], but
1528 $recipients can be useful in postmaster-provided expansion strings.
1530 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1531 in the case of a MAIL command with additional options following the
1532 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1533 were accidentally chopped off.
1535 PH/29 SMTP synchronization checks are implemented when a command is read -
1536 there is a check that no more input is waiting when there shouldn't be
1537 any. However, for some commands, a delay in an ACL can mean that it is
1538 some time before the response is written. In this time, more input might
1539 arrive, invalidly. So now there are extra checks after an ACL has run for
1540 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1541 pipelining has not been advertised.
1543 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1545 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1546 returned in the SMTP response when smtp_return_error_details was false.
1547 This has been fixed.
1549 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1550 instead of the C I/O that was originally supplied, because problems were
1551 reported on Solaris.
1553 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1554 Exim which did not show up earlier: it was assuming that a call to
1555 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1556 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1557 was a macro that became an assignment, so it seemed to work. This has
1558 changed to a proper function call with a void return, hence the compile
1559 error. Exim's code has been fixed.
1561 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1564 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1566 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1568 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1569 because it turns out to be too big or there is a timeout) there is no
1570 "Completed" line in the log. When some messages of this type were
1571 selected by exigrep, they were listed as "not completed". Others were
1572 picked up by some special patterns. I have improved the selection
1573 criteria to be more general.
1575 PH/38 The host_find_failed option in the manualroute router can now be set
1576 to "ignore", to completely ignore a host whose IP address cannot be
1577 found. If all hosts are ignored, the behaviour is controlled by the new
1578 host_all_ignored option.
1580 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1581 homing or because of multiple MX records with /mx) generated more than
1582 one IP address, and the following item turned out to be the local host,
1583 all the secondary addresses of the first item were incorrectly removed
1584 from the list, along with the local host and any following hosts (which
1585 is what is supposed to happen).
1587 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1588 whoever called Exim into the -H file. In the case of the daemon it was
1589 behaving confusingly. When first started, it used values for whoever
1590 started the daemon, but after a SIGHUP it used the Exim user (because it
1591 calls itself on a restart). I have changed the code so that it now always
1594 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1595 message are rejected with the same error (e.g. no authentication or bad
1596 sender address), and a DATA command is nevertheless sent (as can happen
1597 with PIPELINING or a stupid MUA), the error message that was given to the
1598 RCPT commands is included in the rejection of the DATA command. This is
1599 intended to be helpful for MUAs that show only the final error to their
1602 PH/42 Another patch from the Sieve maintainer.
1604 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1605 Eximstats - Fixed some broken HTML links and added missing column headers
1607 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1608 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1610 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1612 PH/43 Yet another patch from the Sieve maintainer.
1614 PH/44 I found a way to check for a TCP/IP connection going away before sending
1615 the response to the final '.' that terminates a message, but only in the
1616 case where the client has not sent further data following the '.'
1617 (unfortunately, this is allowed). However, in many cases there won't be
1618 any further data because there won't be any more messages to send. A call
1619 to select() can be used: if it shows that the input is "ready", there is
1620 either input waiting, or the socket has been closed. An attempt to read
1621 the next input character can distinguish the two cases. Previously, Exim
1622 would have sent an OK response which the client would never have see.
1623 This could lead to message repetition. This fix should cure that, at
1624 least in a lot of common cases.
1626 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1627 advertised in response to EHLO.
1633 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1634 fixed by 4.65/MH/01 (is this a record?) are fixed:
1636 (i) An empty string was always treated as zero by the numeric comparison
1637 operators. This behaviour has been restored.
1639 (ii) It is documented that the numeric comparison operators always treat
1640 their arguments as decimal numbers. This was broken in that numbers
1641 starting with 0 were being interpreted as octal.
1643 While fixing these problems I realized that there was another issue that
1644 hadn't been noticed. Values of message_size_limit (both the global option
1645 and the transport option) were treated as octal if they started with 0.
1646 The documentation was vague. These values are now always treated as
1647 decimal, and I will make that clear in the documentation.
1653 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1654 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1657 MH/01 Don't check that the operands of numeric comparison operators are
1658 integers when their expansion is in "skipping" mode (fixes bug
1659 introduced by 4.64-PH/07).
1661 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1662 child addresses, Exim now panics and dies. Previously, because the count
1663 is held in a short int, deliveries were likely to be lost. As such a
1664 large number of recipients for a single message is ridiculous
1665 (performance will be very, very poor), I have chosen to impose a limit
1666 rather than extend the field.
1672 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1673 leftover -K file (the existence of which was triggered by #402).
1674 While we were at it, introduced process PID as part of the -K
1675 filename. This should rule out race conditions when creating
1678 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1679 processing considerably. Previous code took too long for large mails,
1680 triggering a timeout which in turn triggers #401.
1682 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1683 in the DK code in transports.c. sendfile() is not really portable,
1684 hence the _LINUX specificness.
1686 TF/01 In the add_headers option to the mail command in an Exim filter,
1687 there was a bug that Exim would claim a syntax error in any
1688 header after the first one which had an odd number of characters
1691 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1692 callout verification, Exim cached a "reject" for the entire domain. This
1693 is correct for most verifications, but it is not correct for a recipient
1694 verification with use_sender or use_postmaster set, because in that case
1695 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1696 case of MAIL FROM:<> rejection from other early rejections (e.g.
1697 rejection of HELO). When verifying a recipient using a non-null MAIL
1698 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1699 Whatever the result of the callout, the value of the domain cache is
1700 left unchanged (for any other kind of callout, getting as far as trying
1701 RCPT means that the domain itself is ok).
1703 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1706 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1707 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1708 session, the daemon ignores ECONNECT errors and logs others; it now
1709 ignores EPIPE as well.
1711 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1712 (quoted-printable decoding).
1714 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1715 later the small subsequent patch to fix an introduced bug.
1717 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1719 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1721 PH/08 An error is now given if message_size_limit is specified negative.
1723 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1724 to be given (somewhat) arbitrary names.
1726 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1729 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1730 miscellaneous code fixes
1732 PH/10 Added the log_reject_target ACL modifier to specify where to log
1735 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1736 hostname. This is wrong, because it relates to the incoming message (and
1737 probably the interface on which it is arriving) and not to the outgoing
1738 callout (which could be using a different interface). This has been
1739 changed to use the value of the helo_data option from the smtp transport
1740 instead - this is what is used when a message is actually being sent. If
1741 there is no remote transport (possible with a router that sets up host
1742 addresses), $smtp_active_hostname is used.
1744 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1745 tweaks were necessary in order to get it to work (see also 21 below):
1746 (a) The code assumed that strncpy() returns a negative number on buffer
1747 overflow, which isn't the case. Replaced with Exim's string_format()
1749 (b) There were several signed/unsigned issues. I just did the minimum
1750 hacking in of casts. There is scope for a larger refactoring.
1751 (c) The code used strcasecmp() which is not a standard C function.
1752 Replaced with Exim's strcmpic() function.
1753 (d) The code set only $1; it now sets $auth1 as well.
1754 (e) A simple test gave the error "authentication client didn't specify
1755 service in request". It would seem that Dovecot has changed its
1756 interface. Fortunately there's a specification; I followed it and
1757 changed what the client sends and it appears to be working now.
1759 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1762 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1763 address A is aliased to B and C, where B exists and C does not. Without
1764 -v the output is "A verified" because verification stops after a
1765 successful redirection if more than one address is generated. However,
1766 with -v the child addresses are also verified. Exim was outputting "A
1767 failed to verify" and then showing the successful verification for C,
1768 with its parentage. It now outputs "B failed to verify", showing B's
1769 parentage before showing the successful verification of C.
1771 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1772 look up a TXT record in a specific list after matching in a combined
1775 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1776 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1777 they consult the DNS. I had assumed they would set it the way they
1778 wanted; and indeed my experiments on Linux seem to show that in some
1779 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1780 To be on the safe side, however, I have now made the interface to
1781 host_find_byname() similar to host_find_bydns(), with an argument
1782 containing the DNS resolver options. The host_find_byname() function now
1783 sets these options at its start, just as host_find_bydns() does. The smtp
1784 transport options dns_qualify_single and dns_search_parents are passed to
1785 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1786 of host_find_byname() use the default settings of RES_DEFNAMES
1787 (qualify_single) but not RES_DNSRCH (search_parents).
1789 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1790 spool_read_header() do less string testing, by means of a preliminary
1791 switch on the second character of optional "-foo" lines. (This is
1792 overdue, caused by the large number of possibilities that now exist.
1793 Originally there were few.) While I was there, I also converted the
1794 str(n)cmp tests so they don't re-test the leading "-" and the first
1795 character, in the hope this might squeeze out yet more improvement.
1797 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1798 flag allowing group syntax was set by the header_syntax check but not
1799 turned off, possible causing trouble later; (2) The flag was not being
1800 set at all for the header_verify test, causing "group"-style headers to
1801 be rejected. I have now set it in this case, and also caused header_
1802 verify to ignore an empty address taken from a group. While doing this, I
1803 came across some other cases where the code for allowing group syntax
1804 while scanning a header line wasn't quite right (mostly, not resetting
1805 the flag correctly in the right place). These bugs could have caused
1806 trouble for malformed header lines. I hope it is now all correct.
1808 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1809 with the "reply" argument non-NULL. The code, however (which originally
1810 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1811 but it didn't always do it. This confused somebody who was copying the
1812 code for some other use. I have removed all the tests.
1814 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1815 feature that was used to support insecure browsers during the U.S. crypto
1816 embargo. It requires special client support, and Exim is probably the
1817 only MTA that supported it -- and would never use it because real RSA is
1818 always available. This code has been removed, because it had the bad
1819 effect of slowing Exim down by computing (never used) parameters for the
1820 RSA_EXPORT functionality.
1822 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1823 authenticator to fail if there's a tab character in the incoming data
1824 (there should never be unless someone is messing about, as it's supposed
1825 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1826 if the connection is using TLS or if the remote IP is the same as the
1827 local IP, and the "valid-client-cert option" if a client certificate has
1830 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1831 authenticators. This can be used for authorization after authentication
1832 succeeds. (In the case of plaintext, it servers for both authentication
1835 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1836 if any retry times were supplied.
1838 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1839 connection, where there is no client IP address to check. In this
1840 situation, the verify now always succeeds.
1842 PH/25 Applied John Jetmore's -Mset patch.
1844 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1846 PH/27 In a string expansion for a processed (not raw) header when multiple
1847 headers of the same name were present, leading whitespace was being
1848 removed from all of them, but trailing whitespace was being removed only
1849 from the last one. Now trailing whitespace is removed from each header
1850 before concatenation. Completely empty headers in a concatenation (as
1851 before) are ignored.
1853 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1854 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1856 PH/29 [Removed. This was a change that I later backed out, and forgot to
1857 correct the ChangeLog entry (that I had efficiently created) before
1858 committing the later change.]
1860 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1861 address errors (4xx response to RCPT) over the same connection as other
1862 messages routed to the same hosts. Such deliveries are always "forced",
1863 so retry times are not inspected. This resulted in far too many retries
1864 for the affected addresses. The effect occurred only when there were more
1865 hosts than the hosts_max_try setting in the smtp transport when it had
1866 the 4xx errors. Those hosts that it had tried were not added to the list
1867 of hosts for which the message was waiting, so if all were tried, there
1868 was no problem. Two fixes have been applied:
1870 (i) If there are any address or message errors in an SMTP delivery, none
1871 of the hosts (tried or untried) are now added to the list of hosts
1872 for which the message is waiting, so the message should not be a
1873 candidate for sending over the same connection that was used for a
1874 successful delivery of some other message. This seems entirely
1875 reasonable: after all the message is NOT "waiting for some host".
1876 This is so "obvious" that I'm not sure why it wasn't done
1877 previously. Hope I haven't missed anything, but it can't do any
1878 harm, as the worst effect is to miss an optimization.
1880 (ii) If, despite (i), such a delivery is accidentally attempted, the
1881 routing retry time is respected, so at least it doesn't keep
1882 hammering the server.
1884 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1885 in ${readsocket because some servers need this prod.
1887 PH/32 Added some extra debug output when updating a wait-xxx database.
1889 PH/33 The hint "could be header name not terminated by colon", which has been
1890 given for certain expansion errors for a long time, was not being given
1891 for the ${if def:h_colon_omitted{... case.
1893 PH/34 The spec says: "With one important exception, whenever a domain list is
1894 being scanned, $domain contains the subject domain." There was at least
1895 one case where this was not true.
1897 PH/35 The error "getsockname() failed: connection reset by peer" was being
1898 written to the panic log as well as the main log, but it isn't really
1899 panic-worthy as it just means the connection died rather early on. I have
1900 removed the panic log writing for the ECONNRESET error when getsockname()
1903 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1904 runs only) independently of the message's sender address. This meant
1905 that, if the 4xx error was in fact related to the sender, a different
1906 message to the same recipient with a different sender could confuse
1907 things. In particualar, this can happen when sending to a greylisting
1908 server, but other circumstances could also provoke similar problems.
1909 I have changed the default so that the retry time for these errors is now
1910 based a combination of the sender and recipient addresses. This change
1911 can be overridden by setting address_retry_include_sender=false in the
1914 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1915 remote server are returned as part of bounce messages. This was not
1916 happening for LMTP over a pipe (the lmtp transport), but now it is the
1917 same for both kinds of LMTP.
1919 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1920 in header lines that were in fact CNAMEs. This is no longer the case.
1922 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1923 and queue runs started by the daemon processed all messages. This has
1924 been fixed so that -R and -S can now usefully be given with -q<time>.
1926 PH/40 Import PCRE release 6.7 (fixes some bugs).
1928 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1930 PH/42 Give an error if -q is specified more than once.
1932 PH/43 Renamed the variables $interface_address and $interface_port as
1933 $received_ip_address and $received_port, to make it clear that these
1934 values apply to message reception, and not to the outgoing interface when
1935 a message is delivered. (The old names remain recognized, of course.)
1937 PH/44 There was no timeout on the connect() call when using a Unix domain
1938 socket in the ${readsocket expansion. There now is.
1940 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1941 be meaningful with "accept".
1943 SC/01 Eximstats V1.43
1944 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1946 SC/02 Eximstats V1.44
1947 Use a glob alias rather than an array ref in the generated
1948 parser. This improves both readability and performance.
1950 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1951 Collect SpamAssassin and rejection statistics.
1952 Don't display local sender or destination tables unless
1953 there is data to show.
1954 Added average volumes into the top table text output.
1956 SC/04 Eximstats V1.46
1957 Collect data on the number of addresses (recipients)
1958 as well as the number of messages.
1960 SC/05 Eximstats V1.47
1961 Added 'Message too big' to the list of mail rejection
1962 reasons (thanks to Marco Gaiarin).
1964 SC/06 Eximstats V1.48
1965 Mainlog lines which have GMT offsets and are too short to
1966 have a flag are now skipped.
1968 SC/07 Eximstats V1.49 (Alain Williams)
1969 Added the -emptyok flag.
1971 SC/08 Eximstats V1.50
1972 Fixes for obtaining the IP address from reject messages.
1974 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1975 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1976 whitesspace changes from 4.64-PH/27
1978 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1981 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1982 are found, allow negative numbers in numeric criteria)
1984 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1986 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1987 to match changes made in 4.64-PH/43
1989 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1991 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1992 "domain" is actually an IP address, and force a failure. This locks out
1993 those revolvers/nameservers that support "A-for-A" lookups, in
1994 contravention of the specifications.
1996 PH/48 When a host name was looked up from an IP address, and the subsequent
1997 forward lookup of the name timed out, the host name was left in
1998 $sender_host_name, contrary to the specification.
2000 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2001 restricted to single-key lookups, Exim was not diagnosing an error if
2002 * or *@ was used with a query-style lookup.
2004 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2006 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2007 long ago, but noone interested enough thought of it. Let's just say that
2008 the "1.1" means that there are some new functions that weren't there at
2009 some point in the past.
2011 PH/51 Error processing for expansion failure of helo_data from an smtp
2012 transport during callout processing was broken.
2014 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2015 tested/used via the -bh/-bhc/-bs options.
2017 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2018 bug, fixed in subsequent PCRE releases).
2020 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2021 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2023 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2029 SC/01 Use a glob alias rather than an array ref in eximstats generated
2030 parser. This improves both readability and performance.
2032 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2033 Don't display local sender or destination tables in eximstats unless
2034 there is data to show.
2035 Added average volumes into the eximstats top table text output.
2037 SC/03 Collect data on the number of addresses (recipients) as well
2038 as the number of messages in eximstats.
2040 TF/01 Correct an error in the documentation for the redirect router. Exim
2041 does (usually) call initgroups() when daemonizing.
2043 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2044 with consistent privilege compared to when running as a daemon.
2046 TF/03 Note in the spec that $authenticated_id is not set for local
2047 submissions from trusted users.
2049 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2050 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2052 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2053 by adding some example configuration directives to the default
2054 configuration file. A little bit of work is required to uncomment the
2055 directives and define how usernames and passwords are checked, but
2056 there is now a framework to start from.
2058 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2059 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2060 without this. I don't know how relevant this is to other LDAP libraries.
2062 PH/02 Add the verb name to the "unknown ACL verb" error.
2064 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2066 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2068 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2069 directory not expanded when it should be if an expanded home directory
2070 was set for the address (which is overridden by the transport).
2072 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2075 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2076 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2077 because it is too late at that time, and has no effect.
2079 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2080 security issue with \' (bugzilla #107). I could not use the
2081 PQescapeStringConn() function, because it needs a PGconn value as one of
2084 PH/08 When testing addresses using -bt, indicate those final addresses that
2085 are duplicates that would not cause an additional delivery. At least one
2086 person was confused, thinking that -bt output corresponded to deliveries.
2087 (Suppressing duplicates isn't a good idea as you lose the information
2088 about possibly different redirections that led to the duplicates.)
2090 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2091 systems where poll() doesn't work, in particular OS X.
2093 PH/10 Added more information to debugging output for retry time not reached.
2095 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2096 operations in malware.c.
2098 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2101 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2102 syslog_duplication set false, log lines that would normally be written
2103 both the the main log and to the reject log were not written to syslog at
2106 PH/14 In the default configuration, change the use of "message" in ACL warn
2107 statements to "add_header".
2109 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2110 not followed by a command (e.g. "seen endif").
2112 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2113 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2116 PH/17 Added extra conditions to the default value of delay_warning_condition
2120 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2121 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2122 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2125 The Auto-Submitted: and various List- headers are standardised, whereas I
2126 don't think Precedence: ever was.
2128 PH/18 Refactored debugging code in route_finduser() to show more information,
2129 in particular, the error code if getpwnam() issues one.
2131 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2132 This is apparently needed in addition to the PH/07 change above to avoid
2133 any possible encoding problems.
2135 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2136 but not after initializing Perl.
2138 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2139 output them only if debugging. By default they are written stderr,
2140 apparently, which is not desirable.
2142 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2145 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2148 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2150 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2151 authenticated or an ident call has been made. Suppress the default
2152 values for $authenticated_id and $authenticated_sender (but permit -oMai
2153 and -oMas) when testing with -bh.
2155 PH/24 Re-jigged the order of the tests in the default configuration so that the
2156 tests for valid domains and recipients precede the DNS black list and CSA
2157 tests, on the grounds that those ones are more expensive.
2159 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2160 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2161 This bug exists in every version of Exim that I still have, right back to
2164 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2165 However, an attempt to turn on case-sensitivity in a regex key by
2166 including (?-i) didn't work because the subject string was already
2167 lowercased, and the effects were non-intuitive. It turns out that a
2168 one-line patch can be used to allow (?-i) to work as expected.
2174 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2175 other effects) broke the use of negated acl sub-conditions.
2177 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2180 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2181 "Deny" causes Exim to reject the incoming connection with a 554 error.
2182 Unfortunately, if there is a major crisis, such as a disk failure,
2183 tcp-wrappers gives "deny", whereas what one would like would be some
2184 kind of temporary error. A kludge has been added to help with this.
2185 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2186 554 error is used if errno is still zero or contains ENOENT (which occurs
2187 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2190 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2192 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2193 errors. Otherwise a message that provokes a temporary error (when other
2194 messages do not) can cause a whole host to time out.
2196 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2197 addresses were routed directly to files or pipes from a redirect router.
2198 File deliveries just didn't batch; pipe deliveries might have suffered
2201 PH/06 A failure to get a lock for a hints database would erroneously always say
2202 "Failed to get write lock", even when it was really a read lock.
2204 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2205 of 0600. This has been changed to use the value of the lockfile_mode
2206 option (which defaults to 0600).
2208 PH/08 Applied small patch from the Sieve maintainer.
2210 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2211 folder from quota calculations, a direct delivery into this folder messed
2212 up the contents of the maildirsize file. This was because the regex was
2213 used only to exclude .Trash (or whatever) when the size of the mailbox
2214 was calculated. There was no check that a delivery was happening into an
2215 excluded directory. This bug has been fixed by ignoring all quota
2216 processing for deliveries into excluded directories.
2218 PH/10 Added the maildirfolder_create_regex option to appendfile.
2224 PH/01 The code for finding all the local interface addresses on a FreeBSD
2225 system running IPv6 was broken. This may well have applied to all BSD
2226 systems, as well as to others that have similar system calls. The broken
2227 code found IPv4 interfaces correctly, but gave incorrect values for the
2228 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2229 that it would not match correctly against @[] and not recognize the IPv6
2232 PH/02 The ipliteral router was not recognizing addresses of the form user@
2233 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2235 PH/03 Added disable_ipv6.
2237 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2238 decoded form, because it is most often used to construct To: headers
2239 lines in autoreplies, and the decoded form may well be syntactically
2240 invalid. However, $reply_address has leading white space removed, and all
2241 newlines turned into spaces so that the autoreply transport does not
2244 PH/05 If group was specified without a user on a router, and no group or user
2245 was specified on a transport, the group from the router was ignored.
2247 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2248 for visible compile-time settings that can be used to change these
2249 numbers, for those that want even more. Backwards compatibility with old
2250 spool files has been maintained. However, going back to a previous Exim
2251 release will lost any variables that are in spool files.
2253 PH/07 Two small changes when running in the test harness: increase delay when
2254 passing a TCP/IP connection to a new process, in case the original
2255 process has to generate a bounce, and remove special handling of
2256 127.0.0.2 (sic), which is no longer necessary.
2258 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2259 be the same on different OS.
2261 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2264 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2265 whether --show-vars was specified or not
2267 JJ/02 exipick: Added support for new ACL variable spool format introduced
2270 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2271 syntactically invalid From: or Reply-to: line, and a filter used this to
2272 generate an autoreply, and therefore failed to obtain an address for the
2273 autoreply, Exim could try to deliver to a non-existent relative file
2274 name, causing unrelated and misleading errors. What now happens is that
2275 it logs this as a hard delivery error, but does not attempt to create a
2278 PH/11 The exinext utility has a -C option for testing purposes, but although
2279 the given file was scanned by exinext itself; it wasn't being passed on
2280 when Exim was called.
2282 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2283 an end-of-file indication when reading a command response.
2285 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2286 compiled. In many other places in Exim, IPv6 addresses are always
2287 recognized, so I have changed this. It also means that IPv4 domain
2288 literals of the form [IPV4:n.n.n.n] are now always recognized.
2290 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2291 used if the router is not running as root, for example, when verifying at
2292 ACL time, or when using -bh. The debugging output from this situation was
2293 non-existent - all you got was a failure to exec. I have made two
2296 (a) Failures to set uid/gid, the current directory, or a process leader
2297 in a subprocess such as that created by queryprogram now generate
2298 suitable debugging ouput when -d is set.
2300 (b) The queryprogram router detects when it is not running as root,
2301 outputs suitable debugging information if -d is set, and then runs
2302 the subprocess without attempting to change uid/gid.
2304 PH/15 Minor change to Makefile for building test_host (undocumented testing
2307 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2308 additional section of a DNS packet that returns MX or SRV records.
2309 Instead, it always explicitly searches for A/AAAA records. This avoids
2310 major problems that occur when a DNS server includes only records of one
2311 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2312 fixed another bug: if SRV records were looked up and the corresponding
2313 address records were *not* found in the additional section, the port
2314 values from the SRV records were lost.
2316 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2317 using the correct key (the original address) when searching the retry
2318 rules in order to find which one to use for generating the retry hint.
2320 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2321 adding the default one. Similarly, if it contains a Reply-To: header, the
2322 errors_reply_to option, if set, is not used.
2324 PH/19 When calculating a retry time, Exim used to measure the "time since
2325 failure" by looking at the "first failed" field in the retry record. Now
2326 it does not use this if it is later than than the arrival time of the
2327 message. Instead it uses the arrival time. This makes for better
2328 behaviour in cases where some deliveries succeed, thus re-setting the
2329 "first failed" field. An example is a quota failure for a huge message
2330 when small messages continue to be delivered. Without this change, the
2331 "time since failure" will always be short, possible causing more frequent
2332 delivery attempts for the huge message than are intended.
2333 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2335 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2336 $1, $2, $3) because the numerical variables can be reset during some
2337 expansion items (e.g. "match"), thereby losing the authentication data.
2339 PH/21 Make -bV show the size of off_t variables so that the test suite can
2340 decide whether to run tests for quotas > 2G.
2342 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2343 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2344 filecount value is greater than 2G or if a quota value is greater than 2G
2345 on a system where the size of off_t is not greater than 4, a panic error
2348 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2349 never match. The debug and -bh output now contains an explicit error
2350 message indicating a malformed IPv4 address or mask.
2352 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2353 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2354 PH/23 above applies.
2356 PH/25 Do not write to syslog when running in the test harness. The only
2357 occasion when this arises is a failure to open the main or panic logs
2358 (for which there is an explicit test).
2360 PH/26 Added the /no_tell option to "control=freeze".
2362 PH/27 If a host name lookup failed very early in a connection, for example, if
2363 the IP address matched host_lookup and the reverse lookup yielded a name
2364 that did not have a forward lookup, an error message of the form "no IP
2365 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2366 could be logged. Now it outputs the IP address instead of "NULL".
2368 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2369 allows the sender and the authenticated sender to be set when
2370 submitting a message from within Exim. Since child_open_exim() is
2371 documented for local_scan(), the new function should be too.
2373 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2374 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2375 results in an empty string is now treated as unset.
2377 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2379 PH/31 Added sender_verify_fail logging option.
2381 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2382 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2383 tidied the source and removed it altogether.
2385 PH/33 When a queue run was abandoned because the load average was too high, a
2386 log line was always written; now it is written only if the queue_run log
2387 selector is set. In addition, the log line for abandonment now contains
2388 information about the queue run such as the pid. This is always present
2389 in "start" and "stop" lines but was omitted from the "abandon" line.
2391 PH/34 Omit spaces between a header name and the colon in the error message that
2392 is given when verify = headers_syntax fails (if there are lots of them,
2393 the message gets confusing).
2395 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2396 names, as there are now some PTR records that contain slashes. This check
2397 is only to protect against broken name servers that fall over on strange
2398 characters, so the fact that it applies to all lookups doesn't matter.
2400 PH/36 Now that the new test suite is complete, we can remove some of the
2401 special code in Exim that was needed for the old test suite. For example,
2402 sorting DNS records because real resolvers return them in an arbitrary
2403 order. The new test suite's fake resolver always returns records in the
2406 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2407 bounces) except when queue_only is set, to avoid logging races between
2408 the different processes.
2410 PH/38 Panic-die if .include specifies a non-absolute path.
2412 PH/39 A tweak to the "H" retry rule from its user.
2414 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2415 a label. They prevented compilation on older perls.
2417 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2418 a warning to be raised on newish perls.
2420 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2421 on queue. Changes to match documented behaviour of showing count of
2422 messages matching specified criteria.
2424 PH/40 Changed the default ident timeout from 30s to 5s.
2426 PH/41 Added support for the use of login_cap features, on those BSD systems
2427 that have them, for controlling the resources used by pipe deliveries.
2429 PH/42 The content-scanning code uses fopen() to create files in which to put
2430 message data. Previously it was not paying any attention to the mode of
2431 the files. Exim runs with umask(0) because the rest of the code creates
2432 files with open(), and sets the required mode explicitly. Thus, these
2433 files were ending up world-writeable. This was not a big issue, because,
2434 being within the spool directory, they were not world-accessible. I have
2435 created a function called modefopen, which takes an additional mode
2436 argument. It sets umask(777), creates the file, chmods it to the required
2437 mode, then resets the umask. All the relevant calls to fopen() in the
2438 content scanning code have been changed to use this function.
2440 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2441 to 24 hours. This avoids potential overflow problems when processing G
2442 and H retry rules. I suspect nobody ever tinkers with this value.
2444 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2446 PH/45 When the plaintext authenticator is running as a client, the server's
2447 challenges are checked to ensure they are valid base64 strings. By
2448 default, the authentication attempt is cancelled if an invalid string is
2449 received. Setting client_ignore_invalid_base64 true ignores these errors.
2450 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2451 they are received. Thus, the responses can be made to depend on the
2452 challenges. If an invalid string is ignored, an empty string is placed in
2455 PH/46 Messages that are created by the autoreply transport now contains a
2456 References: header, in accordance with RFCs 2822 and 3834.
2458 PH/47 Added authenticated_sender_force to the smtp transport.
2460 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2462 PH/49 Installed latest patch from the Sieve maintainer.
2464 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2465 was also set, the appendfile transport was unnecessarily scanning a
2466 directory of message files (e.g. for maildir delivery) to find the count
2467 of files (along with the size), even though it did not need this
2468 information. It now does the scan only if it needs to find either the
2469 size of the count of files.
2471 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2473 PH/52 Two bugs concerned with error handling when the smtp transport is
2476 (i) Exim was not creating retry information for temporary errors given
2477 for individual recipients after the DATA command when the smtp transport
2478 was used in LMTP mode. This meant that they could be retried too
2479 frequently, and not timed out correctly.
2481 (ii) Exim was setting the flag that allows error details to be returned
2482 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2483 recipients that were returned after the DATA command.
2485 PH/53 This is related to PH/52, but is more general: for any failing address,
2486 when detailed error information was permitted to be returned to the
2487 sender, but the error was temporary, then after the final timeout, only
2488 "retry timeout exceeded" was returned. Now it returns the full error as
2489 well as "retry timeout exceeded".
2491 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2492 do this, and (what is worse) MTAs that accept it.
2494 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2495 will now be deprecated.
2497 PH/56 New os.c-cygwin from the Cygwin maintainer.
2499 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2500 formats (previously only available in exim formats via -bpr, -bpru,
2501 and -bpra. Now also available in native and exiqgrep formats)
2503 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2504 with very large, slow to parse queues
2506 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2508 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2510 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2511 responses to authentication challenges, though it was showing the
2512 challenges; (ii) I've removed the CR characters from the debug output for
2515 PH/58 Allow for the insertion of a newline as well as a space when a string
2516 is turned into more than one encoded-word during RFC 2047 encoding. The
2517 Sieve code now uses this.
2519 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2520 data_4xx, lost_connection, tls_required.
2522 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2523 message was being sent as an SMTP response.
2525 PH/61 Add -l and -k options to exicyclog.
2527 PH/62 When verifying, if an address was redirected to one new address, so that
2528 verification continued, and the new address failed or deferred after
2529 having set something in $address_data, the value of $address_data was not
2530 passed back to the ACL. This was different to the case when no
2531 redirection occurred. The value is now passed back in both cases.
2533 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2534 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2535 use login_cap.h, so on its own it isn't the distinguishing feature. The
2536 new name refers directly to the setclassresources() function.
2538 PH/65 Added configuration files for NetBSD3.
2540 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2542 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2543 is preferred over IPv4.
2545 PH/68 The bounce_return_message and bounce_return_body options were not being
2546 honoured for bounces generated during the reception of non-SMTP messages.
2547 In particular, this applied to messages rejected by the ACL. This bug has
2548 been fixed. However, if bounce_return_message is true and bounce_return_
2549 body is false, the headers that are returned for a non-SMTP message
2550 include only those that have been read before the error was detected.
2551 (In the case of an ACL rejection, they have all been read.)
2553 PH/69 The HTML version of the specification is now built in a directory called
2554 spec_html instead of spec.html, because the latter looks like a path with
2555 a MIME-type, and this confuses some software.
2557 PH/70 Catch two compiler warnings in sieve.c.
2559 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2560 function verify_get_ident() calls ip_connect() to connect a socket, but
2561 if the "connect()" function timed out, ip_connect() used to close the
2562 socket. However, verify_get_ident() also closes the socket later, and in
2563 between Exim writes to the log, which may get opened at this point. When
2564 the socket was closed in ip_connect(), the log could get the same file
2565 descriptor number as the socket. This naturally causes chaos. The fix is
2566 not to close the socket in ip_connect(); the socket should be closed by
2567 the function that creates it. There was only one place in the code where
2568 this was missing, in the iplookup router, which I don't think anybody now
2569 uses, but I've fixed it anyway.
2571 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2572 well as to direct DNS lookups. Otherwise the handling of names in host
2573 lists is inconsistent and therefore confusing.
2579 PH/01 Two changes to the default runtime configuration:
2581 (1) Move the checks for relay_from_hosts and authenticated clients from
2582 after to before the (commented out) DNS black list checks.
2584 (2) Add control=submission to the relay_from_hosts and authenticated
2585 clients checks, on the grounds that messages accepted by these
2586 statements are most likely to be submissions.
2588 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2590 (1) Generate an error if the third argument for the ${prvs expansion is
2593 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2596 (3) Reset the variables that are obtained from the first argument of
2597 ${prvscheck and used in the second argument before leaving the code,
2598 because their memory is reclaimed, so using them afterwards may do
2601 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2602 one (it's much easier than Tom thought :-).
2604 (5) Because of (4), we can now allow for the use of $prvscheck_result
2605 inside the third argument.
2607 PH/03 For some reason, the default setting of PATH when running a command from
2608 a pipe transport was just "/usr/bin". I have changed it to
2611 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2612 anything to be listed in the output from -bV.
2614 PH/05 When a filter generated an autoreply, the entire To: header line was
2615 quoted in the delivery log line, like this:
2617 => >A.N.Other <ano@some.domain> <original@ddress> ...
2619 This has been changed so that it extracts the operative address. There
2620 may be more than one such address. If so, they are comma-separated, like
2623 => >ano@some.domain,ona@other.domain <original@ddress> ...
2625 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2626 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2627 not being looked up in the rDNS to get a host name, Exim was showing the
2628 IP address twice in Received: lines, even though the IP addresses were
2629 identical. For example:
2631 Received: from [1.2.3.4] (helo=[1.2.3.4])
2633 However, if the real host name was known, it was omitting the HELO data
2634 if it matched the actual IP address. This has been tidied up so that it
2635 doesn't show the same IP address twice.
2637 PH/07 When both +timestamp and +memory debugging was on, the value given by
2638 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2639 called by the debug printing, thereby overwriting the timestamp buffer.
2640 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2642 PH/08 When the original message was included in an autoreply transport, it
2643 always said "this is a copy of the message, including all the headers",
2644 even if body_only or headers_only was set. It now gives an appropriate
2647 PH/09 Applied a patch from the Sieve maintainer which:
2649 o fixes some comments
2650 o adds the (disabled) notify extension core
2651 o adds some debug output for the result of if/elsif tests
2652 o points to the current vacation draft in the documentation
2653 and documents the missing references header update
2657 o fixes a bug in processing the envelope test (when testing
2658 multiple envelope elements, the last element determinted the
2661 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2662 Electronic Mail") by including:
2664 Auto-submitted: auto-generated
2666 in the messages that it generates (bounce messages and others, such as
2667 warnings). In the case of bounce messages for non-SMTP mesages, there was
2668 also a typo: it was using "Auto_submitted" (underscore instead of
2669 hyphen). Since every message generated by Exim is necessarily in response
2670 to another message, thes have all been changed to:
2672 Auto-Submitted: auto-replied
2674 in accordance with these statements in the RFC:
2676 The auto-replied keyword:
2678 - SHOULD be used on messages sent in direct response to another
2679 message by an automatic process,
2681 - MUST NOT be used on manually-generated messages,
2683 - MAY be used on Delivery Status Notifications (DSNs) and Message
2684 Disposition Notifications (MDNs),
2686 - MUST NOT be used on messages generated by automatic or periodic
2687 processes, except for messages which are automatic responses to
2690 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2691 to the default Received: header definition.
2693 PH/12 Added log selector acl_warn_skipped (default on).
2695 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2696 variables because (a) they are in the wrong storage pool and (b) even if
2697 they were copied, it wouldn't work properly because of the caching.
2699 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2700 checking when decoding. Apparently there are clients that generate
2701 overlong encoded strings. Why am I not surprised?
2703 PH/15 If the first argument of "${if match_address" was not empty, but did not
2704 contain an "@" character, Exim crashed. Now it writes a panic log message
2705 and treats the condition as false.
2707 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2709 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2710 extension "envelope-auth". The code is finished and in agreement with
2711 other implementations, but there is no documentation so far and in fact,
2712 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2713 not changing the active code.
2715 Print executed "if" and "elsif" statements when debugging is used. This
2716 helps a great deal to understand what a filter does.
2718 Document more things not specified clearly in RFC3028. I had all this
2719 sorted out, when out of a sudden new issues came to my mind. Oops."
2721 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2724 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2725 canonical form (as documented). However, after a host name lookup from
2726 the IP address, check_host() was doing a simple string comparison with
2727 addresses acquired from the DNS when checking that the found name did
2728 have the original IP as one of its addresses. Since any found IPv6
2729 addresses are likely to be in abbreviated form, the comparison could
2730 fail. Luckily, there already exists a function for doing the comparison
2731 by converting both addresses to binary, so now that is used instead of
2732 the text comparison.
2734 PH/20 There was another similar case to PH/19, when a complete host name was
2735 given in a host list; looking up its IP address could give an abbreviated
2736 form, whereas the current host's name might or might not be abbreviated.
2737 The same fix has been applied.
2743 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2744 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2747 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2748 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2750 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2752 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2753 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2754 Also, for "Auto-Submitted", treat anything other than "no" as
2755 non-personal, in accordance with RFC 3834. (Previously it treated
2756 anything starting "auto-" as non-personal.)
2758 TF/01 The control=submission/name=... option had a problem with syntax
2759 errors if the name included a slash character. The /name= option
2760 now slurps the rest of the string, so it can include any characters
2761 but it must come last in the list of options (after /sender_retain
2764 PH/05 Some modifications to the interface to the fake nameserver for the new
2772 TK/01 Added the "success_on_redirect" address verification option. See
2773 NewStuff for rationale and an example.
2775 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2777 PH/02 Patch to exigrep to allow it to work on syslog lines.
2779 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2780 fread() to skip over the body file's header line, because in Cygwin the
2781 header line is locked and is inaccessible.
2783 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2784 co-exist for some time) to make it clear that it is the Exim ID that is
2785 referenced, not the Message-ID: header line.
2787 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2788 string_format() function, because snprintf() does not exist on all
2791 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2792 ${stat: expansion item.
2794 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2795 protocol synchronization error", to keep the pedants happy.
2797 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2798 well as for IRIX systems, when gcc is being used. See the host.c source
2801 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2803 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2806 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2807 transport and to the smtp transport in LMTP mode.
2809 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2811 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2812 run a filter in a subprocess. This could lead to confusion in subsequent
2813 lookups in the parent process. There should also be a search_tidyup() at
2814 the end of the subprocess.
2816 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2817 only if the host matched helo_try_verify_hosts, which caused the
2818 verification to occur when the EHLO/HELO command was issued. The ACL just
2819 tested the remembered result. Now, if a previous verification attempt has
2820 not happened, "verify = helo" does it there and then.
2822 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2824 TK/03 Fix log output including CR from clamd.
2826 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2827 exist provoked a memory error which could cause a segfault.
2829 PH/15 Installed PCRE 6.2
2831 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2833 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2834 of the problem. Specifically, suggested +O2 rather than +O1 for the
2837 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2839 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2840 record, the port was not correctly logged when the outgoing_port log
2841 selector was set (it logged the transort's default port).
2843 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2844 fallback_hosts, and "hosts" in the smtp transport.
2846 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2847 host errors such as "Connection refused".
2849 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2850 authentication with radiusclient 0.4.9:
2852 - Error returned from rc_read_config was caught wrongly
2853 - Username/password not passed on to radius server due to wrong length.
2855 The presumption is that some radiusclient API changes for 4.51/PH/17
2856 were not taken care of correctly. The code is still untested by me (my
2857 Linux distribution still has 0.3.2 of radiusclient), but it was
2858 contributed by a Radius user.
2860 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2861 expanding the "port" option of the smtp transport.
2863 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2864 while reading a MIME header. Thanks to Tom Hughes for a patch.
2866 PH/24 Include config.h inside local_scan.h so that configuration settings are
2869 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2870 that in an ACL for RCPT (for example), you can examine exactly what was
2873 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2874 commands, but it was not correctly comparing the address with the actual
2875 client host address. Thus, it would show the EHLO address in Received:
2876 header lines when this was not necessary.
2878 PH/27 Added the % operator to ${eval:}.
2880 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2881 it should be ignoring failures (because with -C, for example, it has lost
2882 privilege). It wasn't ignoring creation failures other than "already
2885 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2888 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2889 because an input line was too long, either on its own, or by virtue of
2890 too many continuations, the temporary file was not being removed, and the
2891 return code was incorrect.
2893 PH/31 Missing "BOOL" in function definition in filtertest.c.
2895 PH/32 Applied Sieve patches from the maintainer.
2897 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2899 PH/33 Added "verify = not_blind".
2901 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2902 Local/Makefile (with some defaults set). These are used in built scripts
2903 such as exicyclog, but they have never been used in the exim_install
2904 script (though there are many overriding facilities there). I have
2905 arranged that the exim_install script now takes note of these two
2908 PH/35 Installed configuration files for Dragonfly.
2910 PH/36 When a locally submitted message by a trusted user did not contain a
2911 From: header, and the sender address was obtained from -f or from an SMTP
2912 MAIL command, and the trusted user did not use -F to supply a sender
2913 name, $originator_name was incorrectly used when constructing a From:
2914 header. Furthermore, $originator_name was used for submission mode
2915 messages from external hosts without From: headers in a similar way,
2916 which is clearly wrong.
2918 PH/37 Added control=suppress_local_fixups.
2920 PH/38 When log_selector = +received_sender was set, and the addition of the
2921 sender made the log line's construction buffer exactly full, or one byte
2922 less than full, an overflow happened when the terminating "\n" was
2925 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2926 when the result of a list match is failure because a DNS lookup failed.
2928 PH/40 RM_COMMAND is now used in the building process.
2930 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2931 the "build-* directories that it finds.
2933 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2934 address in a domain literal was a prefix of an interface address.
2936 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2937 when verifying a sender address, unless rewrite_headers is false.
2939 PH/44 Wrote a long comment about why errors_to addresses are verified as
2940 recipients, not senders.
2942 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2943 the ratelimit ACL was added.
2945 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2947 PH/46 Added extra information about PostgreSQL errors to the error string.
2949 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2950 avoiding the need to install special zones in a real server. This is
2951 backwards compatible; if it can't find the fake resolver, it drops back.
2952 Thus, both old and new test suites can be run.
2954 TF/02 Added util/ratelimit.pl
2956 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2957 clock is set back in time.
2959 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2960 Candler <B.Candler@pobox.com>.
2962 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2963 OK for addresses that are the result of redirections.
2965 PH/48 A number of further additions for the benefit of the new test suite,
2966 including a fake gethostbyname() that interfaces to the fake DNS resolver
2969 TF/06 The fix for widen_domains has also been applied to qualify_single and
2970 search_parents which are the other dnslookup options that can cause
2973 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2976 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2978 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2979 service" if it encountered an MX record with an empty target hostname.
2980 The message is now "an MX or SRV record indicated no SMTP service".
2982 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2983 if the DNS of the sending site is misconfigured. This is quite a
2984 common situation. This change restores the behaviour of treating a
2985 helo verification defer as a failure.
2987 PH/51 If self=fail was set on a router, the bounce message did not include the
2988 actual error message.
2994 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2996 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2997 command itself ended in error, the underlying message about the transport
2998 filter timeout was being overwritten with the pipe command error. Now the
2999 underlying error message should be appended to the second error message.
3001 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3003 PH/02 Reduce the amount of output that "make" produces by default. Full output
3004 can still be requested.
3006 PH/03 The warning log line about a condition test deferring for a "warn" verb
3007 was being output only once per connection, rather than after each
3008 occurrence (because it was using the same function as for successful
3009 "warn" verbs). This seems wrong, so I have changed it.
3011 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3012 it should not have, which might have caused a crash in the right
3013 circumstances, but probably never did.
3015 PH/04 Installed a modified version of Tony Finch's patch to make submission
3016 mode fix the return path as well as the Sender: header line, and to
3017 add a /name= option so that you can make the user's friendly name appear
3020 TF/03 Added the control = fakedefer ACL modifier.
3022 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3023 Mark Lowes for thorough testing.
3025 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3027 TK/03 Merged latest SRS patch from Miles Wilton.
3029 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3030 with the definition in sysexits.h (which is #included earlier).
3031 Fortunately, Exim does not actually use EX_OK. The code used to try to
3032 preserve the sysexits.h value, by assumimg that macro definitions were
3033 scanned for macro replacements. I have been disabused of this notion,
3034 so now the code just undefines EX_OK before #including unistd.h.
3036 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3037 in the smtp transport. When a block could not be written in a single
3038 write() function, the timeout was being re-applied to each part-write.
3039 This seems wrong - if the receiver was accepting one byte at a time it
3040 would take for ever. The timeout is now adjusted when this happens. It
3041 doesn't have to be particularly precise.
3043 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3044 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3046 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3047 without a domain if the check to <postmaster@domain> fails.
3049 SC/01 Eximstats: added -xls and the ability to specify output files
3050 (patch written by Frank Heydlauf).
3052 SC/02 Eximstats: use FileHandles for outputing results.
3054 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3056 SC/04 Eximstats: fixed display of large numbers with -nvr option
3058 SC/05 Eximstats: fixed merging of reports with empty tables.
3060 SC/06 Eximstats: added the -include_original_destination flag
3062 SC/07 Eximstats: removed tabs and trailing whitespace.
3064 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3066 TK/06 MBOX spool code: Add real "From " MBOX separator line
3067 so the .eml file is really in mbox format (even though
3068 most programs do not really care). Patch from Alex Miller.
3070 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3071 The latter is generated from $received_to and is only set if the
3072 message has one envelope recipient. SA can use these headers,
3073 obviously out-of-the-box. Patch from Alex Miller.
3075 PH/08 The ${def test on a variable was returning false if the variable's
3076 value was "0", contrary to what the specification has always said!
3077 The result should be true unless the variable is empty.
3079 PH/09 The syntax error of a character other than { following "${if
3080 def:variable_name" (after optional whitespace) was not being diagnosed.
3081 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3082 accidental colon was present, for example, could give incorrect results.
3084 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3085 result is used (not including appendfile, where other changes are about
3088 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3089 This involved changing a lot of size variables from int to off_t. It
3090 should work with maildirs and everything.
3092 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3093 spamd dying while we are connected to it.
3095 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3098 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3099 function for BDB 4.3.
3101 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3103 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3104 the sources for PCRE have been re-arranged and more files are now
3107 PH/15 The code I had for printing potentially long long variables in PH/11
3108 above was not the best (it lost precision). The length of off_t variables
3109 is now inspected at build time, and an appropriate printing format (%ld
3110 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3111 to be "long long int" or "long int". This is needed for the internal
3112 formatting function string_vformat().
3114 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3115 the configuration file to be ":syslog", then the script "guesses" where
3116 the logs files are, rather than using the compiled in default. In our
3117 case the guess is not the same as the compiled default, so the script
3118 suddenly stopped working when I started to use syslog. The patch checks
3119 to see if log_file_path is "". If so, it attempts to read it from exim
3120 with no configuration file to get the compiled in version, before it
3121 falls back to the previous guessing code."
3123 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3124 implementing BATV in an Exim configuration. See NewStuff for the gory
3127 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3128 Makefile that are specific to HP-UX.
3130 PH/18 If the "use_postmaster" option was set for a recipient callout together
3131 with the "random" option, the postmaster address was used as the MAIL
3132 FROM address for the random test, but not for the subsequent recipient
3133 test. It is now used for both.
3135 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3136 patch removes a few documentation additions to RFC 3028, because the
3137 latest draft now contains them. It adds the new en;ascii-case comparator
3138 and a new error check for 8bit text in MIME parts. Comparator and
3139 require names are now matched exactly. I enabled the subaddress
3140 extension, but it is not well tested yet (read: it works for me)."
3142 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3143 rework some of the code of TK/09 above to avoid the hardwired use of
3144 "%lld" and "long long". Replaced the call to snprintf() with a call to
3147 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3148 records point to non-existent hosts", "retry timeout exceeded", and
3149 "retry time not reached for any host after a long failure period".
3151 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3152 experimental DomainKeys support:
3154 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3155 (2) On an error such as an illegally used "control", the wrong name for
3156 the control was given.
3158 These problems did NOT occur unless DomainKeys support was compiled.
3160 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3162 PH/24 Added ${if match_ip condition.
3164 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3165 databases so that it will be absolutely obvious if a crash occurs in the
3166 DB library. This is a regular occurrence (often caused by mis-matched
3169 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3170 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3171 picked up on a user's system that detects such things. There doesn't seem
3172 to be a gcc warning option for this - only an attribute that has to be
3173 put on the function's prototype. It seems that in Fedora Core 4 they have
3174 set this on a number of new functions. No doubt there will be more in due
3177 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3178 specify a transport. However, if an address that was verified by such a
3179 router was the subject of a callout, Exim crashed because it tried to
3180 read the rcpt_include_affixes from the non-existent transport. Now it
3181 just assumes that the setting of that option is false. This bug was
3182 introduced by 4.51/PH/31.
3184 PH/28 Changed -d+all to exclude +memory, because that information is very
3185 rarely of interest, but it makes the output a lot bigger. People tend to
3186 do -d+all out of habit.
3188 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3189 code in os-type was giving problems when libc.so lives in lib64, like on
3192 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3193 aren't the modern standard, and it seems that some systems' include files
3194 don't always have them. Exim was already checking for some of the newer
3195 ones like T_AAAA, and defining it itself. I've added checks for all the
3196 record types that Exim uses.
3198 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3199 not automatically generating a new one, as it is supposed to. This
3200 prevented TLS from working. If the file did exist, but contained invalid
3201 data, a new version was generated, as expected. It was only the case of a
3202 non-existent file that was broken.
3204 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3205 with a change in libdomainkeys > 0.64.
3207 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3208 from DNS. If the selector record carries the flag, it now has
3209 precedence over the domain-wide flag.
3211 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3213 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3214 the use of a port name that isn't defined in /etc/services) occurred, the
3215 message was deferred as in a normal delivery, and thus remained on the
3216 spool, instead of being failed because of the mua_wrapper setting. This
3217 is now fixed, and I tidied up some of the mua_wrapper messages at the
3220 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3221 the messages in a hash of arrays rather than using individual hashes.
3222 This is a bit cleaner and results in dramatic memory savings, albeit
3223 at a slight CPU cost.
3225 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3226 as requested by Marc Sherman.
3228 SC/10 Eximstats: added histograms for user specified patterns as requested
3231 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3233 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3234 fopen() in the content-scanning modules that did not already have it.
3240 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3241 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3243 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3245 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3247 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3248 to test Sieve filters that use "vacation".
3250 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3251 that changes the way the GnuTLS parameters are stored in the cache file.
3252 The new format can be generated externally. For backward compatibility,
3253 if the data in the cache doesn't make sense, Exim assumes it has read an
3254 old-format file, and it generates new data and writes a new file. This
3255 means that you can't go back to an older release without removing the
3258 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3259 work if there are any delivery delays because "one_time" forces the
3260 parent to be marked "delivered", so its unseen clone is never tried
3261 again. For this reason, Exim now forbids the simultaneous setting of
3264 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3265 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3266 entry if you want to know the details. The fix, however, neglected to
3267 consider the case when local delivery batching is involved. The test for
3268 "previously delivered" was not happening when checking to see if an
3269 address could be batched with a previous (undelivered) one; under
3270 certain circumstances this could lead to multiple deliveries to the same
3273 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3274 in its include files, and this causes problems building Exim.
3276 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3277 header_syntax) but Exim was just ignoring anything given after a slash.
3278 In particular, this caused confusion with an attempt to use "verify =
3279 reverse_host_lookup/defer_ok". An error is now given when options are
3280 supplied for verify items that do not have them. (Maybe reverse_host_
3281 lookup should have a defer_ok option, but that's a different point.)
3283 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3284 defined by RFC 821) to 2048, because there were problems with some AUTH
3285 commands, and RFC 1869 says the size should be increased for extended
3286 SMTP commands that take arguments.
3288 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3291 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3292 "unknown" error; now it says that the functionality isn't in the binary.
3294 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3295 an address' error message when a string expansion fails (syntax or
3296 whatever). Otherwise the password may appear in the log. Following change
3297 PH/42 below, there is no longer a chance of it appearing in a bounce
3300 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3302 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3303 of its addresses was ever tried. (Bugzilla bug #2.)
3305 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3306 the result incorrectly in the debug output. (It correctly added a newline
3307 to what was transported.)
3309 TF/01 Added $received_time.
3311 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3312 commented out examples of how to interface to a virus scanner and to
3313 SpamAssassin. Also added commented examples of av_scanner and
3314 spamd_address settings.
3316 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3317 and controls are allowed in which ACLs. There were a couple of minor
3318 errors. Some of the entries in the conditions table (which is a table of
3319 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3320 negation of where the condition IS allowed.
3322 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3324 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3325 header file does not have a version number, so I've had to invent a new
3326 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3327 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3328 radiusclient), but it was contributed by a Radius user.
3330 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3331 files or directories, for OpenSSL.
3333 PH/19 When an Exim process that is running as root has to create an Exim log
3334 file, it does so in a subprocess that runs as exim:exim so as to get the
3335 ownership right at creation (otherwise, other Exim processes might see
3336 the file with the wrong ownership). There was no test for failure of this
3337 fork() call, which would lead to the process getting stuck as it waited
3338 for a non-existent subprocess. Forks do occasionally fail when resources
3339 run out. I reviewed all the other calls to fork(); they all seem to check
3342 PH/20 When checking for unexpected SMTP input at connect time (before writing
3343 the banner), Exim was not dealing correctly with a non-positive return
3344 from the read() function. If the client had disconnected by this time,
3345 the result was a log entry for a synchronization error with an empty
3346 string after "input=" when read() returned zero. If read() returned -1
3347 (an event I could not check), uninitialized data bytes were printed.
3348 There were reports of junk text (parts of files, etc) appearing after
3351 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3353 PH/22 Added support for macro redefinition, and (re)definition in between
3354 driver and ACL definitions.
3356 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3357 forgetting to use the resulting value; it was using the unexpanded value.
3359 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3360 hadn't been configured. The fix is from Juergen Kreileder, who
3361 understands it better than I do:
3363 "Here's what I see happening with three configured cyrus_sasl
3364 authenticators configured (plain, login, cram-md5):
3366 On startup auth_cyrus_sasl_init() gets called for each of these.
3367 This means three calls to sasl_listmech() without a specified mech_list.
3368 => SASL tests which mechs of all available mechs actually work
3369 => three warnings about OTP not working
3370 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3372 With the patch, sasl_listmech() also gets called three times. But now
3373 SASL's mech_list option is set to the server_mech specified in the the
3374 authenticator. Or in other words, the answer from sasl_listmech()
3375 gets limited to just the mech you're testing for (which is different
3377 => the return list contains just 'plain' or 'login', 'cram-md5' or
3378 nothing depending on the value of ob->server_mech.
3380 I've just tested the patch: Authentication still works fine,
3381 unavailable mechs specified in the exim configuration are still
3382 caught, and the auth.log warnings about OTP are gone."
3384 PH/25 When debugging is enabled, the contents of the command line are added
3385 to the debugging output, even when log_selector=+arguments is not
3388 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3389 answer is "GNU", and only if the return is "GNU/something" is the answer
3392 PH/27 $acl_verify_message is now set immediately after the failure of a
3393 verification in an ACL, and so is available in subsequent modifiers. In
3394 particular, the message can be preserved by coding like this:
3396 warn !verify = sender
3397 set acl_m0 = $acl_verify_message
3399 Previously, $acl_verify_message was set only while expanding "message"
3400 and "log_message" when a very denied access.
3402 PH/28 Modified OS/os.c-Linux with
3404 -#ifndef OS_LOAD_AVERAGE
3405 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3407 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3408 nomenclature these days.)
3410 PH/29 Installed patch from the Sieve maintainer that adds the options
3411 sieve_useraddress and sieve_subaddress to the redirect router.
3413 PH/30 In these circumstances:
3414 . Two addresses routed to the same list of hosts;
3415 . First host does not offer TLS;
3416 . First host accepts first address;
3417 . First host gives temporary error to second address;
3418 . Second host offers TLS and a TLS session is established;
3419 . Second host accepts second address.
3420 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3421 and peerdn, if requested) that were in fact used only for the second
3424 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3425 attention to any local part prefix or suffix that was matched by the
3426 router that accepted the address. It now behaves in the same way as it
3427 does for delivery: the affixes are removed from the local part unless
3428 rcpt_include_affixes is set on the transport.
3430 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3431 timeout during the DATA phase of an incoming message.
3433 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3434 applied a patch sanctioned by the Sieve maintainer.
3436 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3437 the uid or gid is negative. A case of a negative gid caused this to be
3438 noticed. The fix allows for either to be negative.
3440 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3441 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3444 PH/36 The change PH/12 above was broken. Fixed it.
3446 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3447 the grounds that routing the same address twice would always produce the
3448 same answer. This might have been true once, but it is certainly no
3449 longer true now. Routing a child address may depend on the previous
3450 routing that produced that child. Some complicated redirection strategies
3451 went wrong when messages had multiple recipients, and made Exim's
3452 behaviour dependent on the order in which the addresses were given.
3454 I have moved the duplicate checking until after the routing is complete.
3455 Exim scans the addresses that are assigned to local and remote
3456 transports, and removes any duplicates. This means that more work will be
3457 done, as duplicates will always all be routed, but duplicates are
3458 presumably rare, so I don't expect this is of any significance.
3460 For deliveries to pipes, files, and autoreplies, the duplicate checking
3461 still happens during the routing process, since they are not going to be
3464 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3465 It corrects a timeout issue with spamd. This is Ian's comment: "The
3466 background is that sometimes spamd either never reads data from a
3467 connection it has accepted, or it never writes response data. The exiscan
3468 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3469 blindly assumes that writes won't block so it may never time out."
3471 PH/39 Allow G after quota size as well as K and M.
3473 PH/40 The value set for $authenticated_id in an authenticator may not contain
3474 binary zeroes or newlines because the value is written to log lines and
3475 to spool files. There was no check on this. Now the value is run through
3476 the string_printing() function so that such characters are converted to
3477 printable escape sequences.
3479 PH/41 $message_linecount is a new variable that contains the total number of
3480 lines in the message. Compare $body_linecount, which is the count for the
3483 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3484 bounce and delay warning messages, except in certain special cases, which
3487 (a) An SMTP error message from a remote host;
3488 (b) A message specified in a :fail: redirection;
3489 (c) A message specified in a "fail" command in a system filter;
3490 (d) A message specified in a FAIL return from the queryprogram router;
3491 (e) A message specified by the cannot_route_message router option.
3493 In these cases only, Exim does include the error details in bounce and
3494 warning messages. There are also a few cases where bland messages such
3495 as "unrouteable address" or "local delivery error" are given.
3497 PH/43 $value is now also set for the "else" part of a ${run expansion.
3499 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3500 being worked on, but at least Exim now implements the latest version to
3503 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3504 process to complete was treated as a delivery failure, a timeout while
3505 writing the message to the pipe was logged, but erroneously treated as a
3506 successful delivery. Such timeouts include transport filter timeouts. For
3507 consistency with the overall process timeout, these timeouts are now
3508 treated as errors, giving rise to delivery failures by default. However,
3509 there is now a new Boolean option for the pipe transport called
3510 timeout_defer, which, if set TRUE, converts the failures into defers for
3511 both kinds of timeout. A transport filter timeout is now identified in
3514 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3515 systems where "make" and "gmake" are different, calling "gmake" at top
3516 level broke things. I've arranged for the value of $(MAKE) to be passed
3517 from the Makefile to this script so that it can call the same version of
3521 A note about Exim versions 4.44 and 4.50
3522 ----------------------------------------
3524 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3525 changes of various kinds. As a consequence, a big documentation update was
3526 needed. This delayed the release for rather longer than seemed good, especially
3527 in the light of a couple of (minor) security issues. Therefore, the changes
3528 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3529 release. So 4.44 and 4.50 are in effect two different branches that both start
3532 I have left the 4.50 change log unchanged; it contains all the changes since
3533 4.43. The change log for 4.44 is below; many of its items are identical to
3534 those for 4.50. This seems to be the most sensible way to preserve the
3535 historical information.
3541 1. Minor wording change to the doc/README.SIEVE file.
3543 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3544 computation of the current number of files was incorrect.
3546 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3547 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3550 4. Give more explanation in the error message when the command for a transport
3551 filter fails to execute.
3553 5. There are several places where Exim runs a non-Exim command in a
3554 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3555 was being done only for the command run by the queryprogram router. It is
3556 now done for all such subprocesses. The other cases are: ${run, transport
3557 filters, and the commands run by the lmtp and pipe transports.
3559 6. Added CONFIGURE_GROUP build-time option.
3561 7. Some older OS have a limit of 256 on the maximum number of file
3562 descriptors. Exim was using setrlimit() to set 1000 as a large value
3563 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3564 systems. I've change it so that if it can't get 1000, it tries for 256.
3566 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3567 was an oversight, and furthermore, ever since the addition of extra
3568 controls (e.g. 4.43/32), the checks on when to allow different forms of
3569 "control" were broken. There should now be diagnostics for all cases when a
3570 control that does not make sense is encountered.
3572 9. Added the /retain_sender option to "control=submission".
3574 10. $recipients is now available in the predata ACL (oversight).
3576 11. Tidy the search cache before the fork to do a delivery from a message
3577 received from the command line. Otherwise the child will trigger a lookup
3578 failure and thereby defer the delivery if it tries to use (for example) a
3579 cached ldap connection that the parent has called unbind on.
3581 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3582 of $address_data from the recipient verification was clobbered by the
3583 sender verification.
3585 13. The value of address_data from a sender verification is now available in
3586 $sender_address_data in subsequent conditions in the ACL statement.
3588 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3590 15. Added a new option "connect=<time>" to callout options, to set a different
3593 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3594 was its contents. (It was OK if the option was not defined at all.)
3596 17. A "Completed" log line is now written for messages that are removed from
3597 the spool by the -Mrm option.
3599 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3600 information about exactly what failed.
3602 19. Added -dd to debug only the daemon process.
3604 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3605 handles timeouts, both on the server side and network timeouts. Renamed the
3606 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3608 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3609 It is now set to "smtps".
3611 22. $host_address is now set to the target address during the checking of
3612 ignore_target_hosts.
3614 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3615 being passed; this would have caused $sender_host_name to have been used if
3616 matching the list had actually called for a host name (not very likely,
3617 since this list is usually IP addresses). A host name is now passed as
3620 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3621 code that specifies a non-restarting handler (typically sigaction() in
3622 modern systems) in an attempt to fix a rare and obscure crash bug.
3624 25. Narrowed the window for a race in the daemon that could cause it to ignore
3625 SIGCHLD signals. This is not a major problem, because they are used only to
3626 wake it up if nothing else does.
3628 26. A malformed maildirsize file could cause Exim to calculate negative values
3629 for the mailbox size or file count. Odd effects could occur as a result.
3630 The maildirsize information is now recalculated if the size or filecount
3633 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3634 support for a long time. Removed HAVE_SYS_VFS_H.
3636 28. Installed the latest version of exipick from John Jetmore.
3638 29. In an address list, if the pattern was not a regular expression, an empty
3639 subject address (from a bounce message) matched only if the pattern was an
3640 empty string. Non-empty patterns were not even tested. This was the wrong
3641 because it is perfectly reasonable to use an empty address as part of a
3642 database query. An empty address is now tested by patterns that are
3643 lookups. However, all the other forms of pattern expect the subject to
3644 contain a local part and a domain, and therefore, for them, an empty
3645 address still always fails if the pattern is not itself empty.
3647 30. Exim went into a mad DNS loop when attempting to do a callout where the
3648 host was specified on an smtp transport, and looking it up yielded more
3649 than one IP address.
3651 31. Re-factored the code for checking spool and log partition space into a
3652 function that finds that data and another that does the check. The former
3653 is then used to implement four new variables: $spool_space, $log_space,
3654 $spool_inodes, and $log_inodes.
3656 32. The RFC2047 encoding function was originally intended for short strings
3657 such as real names; it was not keeping to the 75-character limit for
3658 encoded words that the RFC imposes. It now respects the limit, and
3659 generates multiple encoded words if necessary. To be on the safe side, I
3660 have increased the buffer size for the ${rfc2047: expansion operator from
3663 33. It is now permitted to omit both strings after an "if" condition; if the
3664 condition is true, the result is "true". As before, when the second string
3665 is omitted, a false condition yields an empty string. This makes it less
3666 cumbersome to write custom ACL and router conditions.
3668 34. Failure to deliver a bounce message always caused it to be frozen, even if
3669 there was an errors_to setting on the router. The errors_to setting is now
3672 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3673 canonical form (fully expanded) before being placed in
3674 $sender_host_address.
3676 36. The table in the code that translates DNS record types into text (T_A to
3677 "A" for instance) was missing entries for NS and CNAME. It is just possible
3678 that this could have caused confusion if both these types were looked up
3679 for the same domain, because the text type is used as part of Exim's
3680 per-process caching. But the chance of anyone hitting this buglet seems
3683 37. The dnsdb lookup has been extended in a number of ways.
3685 (1) There is a new type, "zns", which walks up the domain tree until it
3686 finds some nameserver records. It should be used with care.
3688 (2) There is a new type, "mxh", which is like "mx" except that it returns
3689 just the host names, not the priorities.
3691 (3) It is now possible to give a list of domains (or IP addresses) to be
3692 looked up. The behaviour when one of the lookups defers can be
3693 controlled by a keyword.
3695 (4) It is now possible to specify the separator character for use when
3696 multiple records are returned.
3698 38. The dnslists ACL condition has been extended: it is now possible to supply
3699 a list of IP addresses and/or domains to be looked up in a particular DNS
3702 39. Added log_selector=+queue_time_overall.
3704 40. When running the queue in the test harness, wait just a tad after forking a
3705 delivery process, to get repeatability of debugging output.
3707 41. Include certificate and key file names in error message when GnuTLS fails
3708 to set them up, because the GnuTLS error message doesn't include the name
3709 of the failing file when there is a problem reading it.
3711 42. Allow both -bf and -bF in the same test run.
3713 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3715 44. The "Exiscan patch" is now merged into the mainline Exim source.
3717 45. Sometimes the final signoff response after QUIT could fail to get
3718 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3719 < 0 before doing a fflush(). This bug looks as though it goes back to the
3720 introduction of TLS in release 3.20, but "sometimes" must have been rare
3721 because the tests only now provoked it.
3723 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3724 (this can affect the format of dates).
3726 47. exim_tidydb, when checking for the continued existence of a message for
3727 which it has found a message-specific retry record, was not finding
3728 messages that were in split spool directories. Consequently, it was
3729 deleting retry records that should have stayed in existence.
3731 48. Steve fixed some bugs in eximstats.
3733 49. The SPA authentication driver was not abandoning authentication and moving
3734 on to the next authenticator when an expansion was forced to fail,
3735 contradicting the general specification for all authenticators. Instead it
3736 was generating a temporary error. It now behaves as specified.
3738 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3739 (the order specifies the preference for clients). The order is now AES256,
3740 AES128, 3DES, ARCFOUR128.
3742 51. Small patch to Sieve code - explicitly set From: when generating an
3745 52. Exim crashed if a remote delivery caused a very long error message to be
3746 recorded - for instance if somebody sent an entire SpamAssassin report back
3747 as a large number of 550 error lines. This bug was coincidentally fixed by
3748 increasing the size of one of Exim's internal buffers (big_buffer) that
3749 happened as part of the Exiscan merge. However, to be on the safe side, I
3750 have made the code more robust (and fixed the comments that describe what
3753 53. Now that there can be additional text after "Completed" in log lines (if
3754 the queue_time_overall log selector is set), a one-byte patch to exigrep
3755 was needed to allow it to recognize "Completed" as not the last thing in
3758 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3759 patch that reportedly fixes this has been added. I am not expert enough to
3760 create a test for it. This is what the patch creator wrote:
3762 "I found a little strange behaviour of ldap code when working with
3763 Windows 2003 AD Domain, where users was placed in more than one
3764 Organization Units. When I tried to give exim partial DN, the exit code
3765 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3766 But simultaneously result of request was absolutely normal ldap result,
3767 so I produce this patch..."
3769 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3770 REFERENCE, so I have modified the code to exclude the patch when that macro
3773 55. Some experimental protocols are using DNS PTR records for new purposes. The
3774 keys for these records are domain names, not reversed IP addresses. The
3775 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3776 leaves it alone. Component reversal etc. now happens only for IP addresses.
3779 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3781 57. Double the size of the debug message buffer (to 2048) so that more of very
3782 long debug lines gets shown.
3784 58. The exicyclog utility now does better if the number of log files to keep
3785 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3787 59. Two changes related to the smtp_active_hostname option:
3789 (1) $smtp_active_hostname is now available as a variable.
3790 (2) The default for smtp_banner uses $smtp_active_hostname instead
3791 of $primary_hostname.
3793 60. The host_aton() function is supposed to be passed a string that is known
3794 to be a valid IP address. However, in the case of IPv6 addresses, it was
3795 not checking this. This is a hostage to fortune. Exim now panics and dies
3796 if the condition is not met. A case was found where this could be provoked
3797 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3798 components; fortuitously, this particular loophole had already been fixed
3799 by change 4.50/55 above.
3801 If there are any other similar loopholes, the new check in host_aton()
3802 itself should stop them being exploited. The report I received stated that
3803 data on the command line could provoke the exploit when Exim was running as
3804 exim, but did not say which command line option was involved. All I could
3805 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3806 running as the user.
3809 61. There was a buffer overflow vulnerability in the SPA authentication code
3810 (which came originally from the Samba project). I have added a test to the
3811 spa_base64_to_bits() function which I hope fixes it.
3814 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3815 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3817 63. The daemon start-up calls getloadavg() while still root for those OS that
3818 need the first call to be done as root, but it missed one case: when
3819 deliver_queue_load_max is set with deliver_drop_privilege. This is
3820 necessary for the benefit of the queue runner, because there is no re-exec
3821 when deliver_drop_privilege is set.
3823 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3824 This has been fixed.
3826 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3827 was in use, was not putting the data itself into the right store pool;
3828 consequently, it could be overwritten for a subsequent message in the same
3829 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3832 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3834 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3835 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3836 the function were treating the return as a boolean value, which happened to
3837 work because 0=false and not-0=true, but is not correct code.
3839 68. The host_aton() function was not handling scoped IPv6 addresses (those
3840 with, for example, "%eth0" on the end) correctly.
3842 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3843 negated items (that is, ~something) in unsigned ints. Some compilers
3844 apparently mutter when there is no cast.
3846 70. If an address verification called from an ACL failed, and did not produce a
3847 user-specific message (i.e. there was only a "system" message), nothing was
3848 put in $acl_verify_message. In this situation, it now puts the system
3851 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3852 session; change 4.31/43 added the unwanted input to the log line - except
3853 that it did not do this in the start of session case. It now does.
3855 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3856 This is wrong and can cause the other end to generate a synchronization
3857 error if it is another Exim or anything else that does the synchronization
3858 check. A QUIT command is no longer sent after a timeout.
3860 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3861 during host lookups.
3863 74. The defer_ok option of callout verification was not working if it was used
3864 when verifying addresses in header lines, that is, for this case:
3866 verify = header_sender/callout=defer_ok
3868 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3869 those file descriptors could be used for SMTP connections. If anything
3870 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3871 could be sent to the SMTP client, causing chaos. The daemon now opens
3872 stdin, stdout, and stderr to /dev/null when it puts itself into the
3875 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3876 log by default. The user can override this with suitable Perl magic.
3878 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3879 the log message when discard triggers, was not working for the DATA ACL or
3880 for the non-SMTP ACL.
3882 78. Error message wording change in sieve.c.
3884 79. If smtp_accept_max_per_host was set, the number of connections could be
3885 restricted to fewer than expected, because the daemon was trying to set up
3886 a new connection before checking whether the processes handling previous
3887 connections had finished. The check for completed processes is now done
3888 earlier. On busy systems, this bug wouldn't be noticed because something
3889 else would have woken the daemon, and it would have reaped the completed
3892 80. If a message was submitted locally by a user whose login name contained one
3893 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3894 It caused a spool format error. I have fixed the spool reading code. A
3895 related problem was that the "from" clause in the Received: line became
3896 illegal because of the space(s). It is now covered by ${quote_local_part.
3898 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3901 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3903 83. Patch from Sieve maintainer to fix unterminated string problem in
3904 "vacation" handling.
3906 84. Some minor changes to the Linux configuration files to help with other
3907 OS variants using glibc.
3909 85. One more patch for Sieve to update vacation handling to latest spec.
3912 ----------------------------------------------------
3913 See the note above about the 4.44 and 4.50 releases.
3914 ----------------------------------------------------
3920 1. Change 4.43/35 introduced a bug that caused file counts to be
3921 incorrectly computed when quota_filecount was set in an appendfile
3924 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3925 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3928 3. Give more explanation in the error message when the command for a transport
3929 filter fails to execute.
3931 4. There are several places where Exim runs a non-Exim command in a
3932 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3933 was being done only for the command run by the queryprogram router. It is
3934 now done for all such subprocesses. The other cases are: ${run, transport
3935 filters, and the commands run by the lmtp and pipe transports.
3937 5. Some older OS have a limit of 256 on the maximum number of file
3938 descriptors. Exim was using setrlimit() to set 1000 as a large value
3939 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3940 systems. I've change it so that if it can't get 1000, it tries for 256.
3942 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3943 was an oversight, and furthermore, ever since the addition of extra
3944 controls (e.g. 4.43/32), the checks on when to allow different forms of
3945 "control" were broken. There should now be diagnostics for all cases when a
3946 control that does not make sense is encountered.
3948 7. $recipients is now available in the predata ACL (oversight).
3950 8. Tidy the search cache before the fork to do a delivery from a message
3951 received from the command line. Otherwise the child will trigger a lookup
3952 failure and thereby defer the delivery if it tries to use (for example) a
3953 cached ldap connection that the parent has called unbind on.
3955 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3956 of $address_data from the recipient verification was clobbered by the
3957 sender verification.
3959 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3960 was its contents. (It was OK if the option was not defined at all.)
3962 11. A "Completed" log line is now written for messages that are removed from
3963 the spool by the -Mrm option.
3965 12. $host_address is now set to the target address during the checking of
3966 ignore_target_hosts.
3968 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3969 being passed; this would have caused $sender_host_name to have been used if
3970 matching the list had actually called for a host name (not very likely,
3971 since this list is usually IP addresses). A host name is now passed as
3974 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3975 code that specifies a non-restarting handler (typically sigaction() in
3976 modern systems) in an attempt to fix a rare and obscure crash bug.
3978 15. Narrowed the window for a race in the daemon that could cause it to ignore
3979 SIGCHLD signals. This is not a major problem, because they are used only to
3980 wake it up if nothing else does.
3982 16. A malformed maildirsize file could cause Exim to calculate negative values
3983 for the mailbox size or file count. Odd effects could occur as a result.
3984 The maildirsize information is now recalculated if the size or filecount
3987 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3988 support for a long time. Removed HAVE_SYS_VFS_H.
3990 18. Updated exipick to current release from John Jetmore.
3992 19. Allow an empty sender to be matched against a lookup in an address list.
3993 Previously the only cases considered were a regular expression, or an
3996 20. Exim went into a mad DNS lookup loop when doing a callout where the
3997 host was specified on the transport, if the DNS lookup yielded more than
4000 21. The RFC2047 encoding function was originally intended for short strings
4001 such as real names; it was not keeping to the 75-character limit for
4002 encoded words that the RFC imposes. It now respects the limit, and
4003 generates multiple encoded words if necessary. To be on the safe side, I
4004 have increased the buffer size for the ${rfc2047: expansion operator from
4007 22. Failure to deliver a bounce message always caused it to be frozen, even if
4008 there was an errors_to setting on the router. The errors_to setting is now
4011 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4012 canonical form (fully expanded) before being placed in
4013 $sender_host_address.
4015 24. Updated eximstats to version 1.33
4017 25. Include certificate and key file names in error message when GnuTLS fails
4018 to set them up, because the GnuTLS error message doesn't include the name
4019 of the failing file when there is a problem reading it.
4021 26. Expand error message when OpenSSL has problems setting up cert/key files.
4024 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4025 (this can affect the format of dates).
4027 28. exim_tidydb, when checking for the continued existence of a message for
4028 which it has found a message-specific retry record, was not finding
4029 messages that were in split spool directories. Consequently, it was
4030 deleting retry records that should have stayed in existence.
4032 29. eximstats updated to version 1.35
4033 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4034 1.35 - bugfix such that pie charts by volume are generated correctly
4036 30. The SPA authentication driver was not abandoning authentication and moving
4037 on to the next authenticator when an expansion was forced to fail,
4038 contradicting the general specification for all authenticators. Instead it
4039 was generating a temporary error. It now behaves as specified.
4041 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4042 (the order specifies the preference for clients). The order is now AES256,
4043 AES128, 3DES, ARCFOUR128.
4045 31. Small patch to Sieve code - explicitly set From: when generating an
4048 32. Exim crashed if a remote delivery caused a very long error message to be
4049 recorded - for instance if somebody sent an entire SpamAssassin report back
4050 as a large number of 550 error lines. This bug was coincidentally fixed by
4051 increasing the size of one of Exim's internal buffers (big_buffer) that
4052 happened as part of the Exiscan merge. However, to be on the safe side, I
4053 have made the code more robust (and fixed the comments that describe what
4056 33. Some experimental protocols are using DNS PTR records for new purposes. The
4057 keys for these records are domain names, not reversed IP addresses. The
4058 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4059 leaves it alone. Component reversal etc. now happens only for IP addresses.
4062 34. The host_aton() function is supposed to be passed a string that is known
4063 to be a valid IP address. However, in the case of IPv6 addresses, it was
4064 not checking this. This is a hostage to fortune. Exim now panics and dies
4065 if the condition is not met. A case was found where this could be provoked
4066 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4067 components; fortuitously, this particular loophole had already been fixed
4068 by change 4.50/55 or 4.44/33 above.
4070 If there are any other similar loopholes, the new check in host_aton()
4071 itself should stop them being exploited. The report I received stated that
4072 data on the command line could provoke the exploit when Exim was running as
4073 exim, but did not say which command line option was involved. All I could
4074 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4075 running as the user.
4078 35. There was a buffer overflow vulnerability in the SPA authentication code
4079 (which came originally from the Samba project). I have added a test to the
4080 spa_base64_to_bits() function which I hope fixes it.
4083 36. The daemon start-up calls getloadavg() while still root for those OS that
4084 need the first call to be done as root, but it missed one case: when
4085 deliver_queue_load_max is set with deliver_drop_privilege. This is
4086 necessary for the benefit of the queue runner, because there is no re-exec
4087 when deliver_drop_privilege is set.
4089 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4090 was in use, was not putting the data itself into the right store pool;
4091 consequently, it could be overwritten for a subsequent message in the same
4092 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4095 38. Sometimes the final signoff response after QUIT could fail to get
4096 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4097 < 0 before doing a fflush(). This bug looks as though it goes back to the
4098 introduction of TLS in release 3.20, but "sometimes" must have been rare
4099 because the tests only now provoked it.
4105 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4106 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4107 Now it returns an integer. A number of calls were still expecting a T/F
4108 return. Fortuitously, in all cases, the tests worked in OK situations,
4109 which is the norm. However, things would have gone wrong on any write
4110 failures on the smtp file descriptor. This function is used when sending
4111 messages over SMTP and also when doing verify callouts.
4113 2. When Exim is called to do synchronous delivery of a locally submitted
4114 message (the -odf or -odi options), it no longer closes stderr before doing
4117 3. Implemented the mua_wrapper option.
4119 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4121 5. Implemented the functions header_remove(), header_testname(),
4122 header_add_at_position(), and receive_remove_recipient(), and exported them
4125 6. If an ACL "warn" statement specified the addition of headers, Exim already
4126 inserted X-ACL-Warn: at the start if there was no header name. However, it
4127 was not making this test for the second and subsequent header lines if
4128 there were newlines in the string. This meant that an invalid header could
4129 be inserted if Exim was badly configured.
4131 7. Allow an ACL "warn" statement to add header lines at the start or after all
4132 the Received: headers, as well as at the end.
4134 8. Added the rcpt_4xx retry error code.
4136 9. Added postmaster_mailfrom=xxx to callout verification option.
4138 10. Added mailfrom=xxxx to the callout verification option, for verify=
4141 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4142 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4144 12. Inserted some casts to stop certain compilers warning when using pointer
4145 differences as field lengths or precisions in printf-type calls (mostly
4146 affecting debugging statements).
4148 13. Added optional readline() support for -be (dynamically loaded).
4150 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4151 same clock tick as a message's arrival, so that its received time was the
4152 same as the "first fail" time on the retry record, and that message
4153 remained on the queue past the ultimate address timeout, every queue runner
4154 would try a delivery (because it was past the ultimate address timeout) but
4155 after another failure, the ultimate address timeout, which should have then
4156 bounced the address, did not kick in. This was a "< instead of <=" error;
4157 in most cases the first failure would have been in the next clock tick
4158 after the received time, and all would be well.
4160 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4161 being recognized when the domain list was tested by the match_domain
4162 condition in an expansion string.
4164 16. Added the ${str2b64: operator.
4166 17. Exim was always calling setrlimit() to set a large limit for the number of
4167 processes, without checking whether the existing limit was already
4168 adequate. (It did check for the limit on file descriptors.) Furthermore,
4169 errors from getrlimit() and setrlimit() were being ignored. Now they are
4170 logged to the main and panic logs, but Exim does carry on, to try to do its
4171 job under whatever limits there are.
4173 18. Imported PCRE 5.0.
4175 19. Trivial typo in log message " temporarily refused connection" (the leading
4178 20. If the log selector return_path_on_delivery was set and an address was
4179 redirected to /dev/null, the delivery process crashed because it assumed
4180 that a return path would always be set for a "successful" delivery. In this
4181 case, the whole delivery is bypassed as an optimization, and therefore no
4184 21. Internal re-arrangement: the function for sending a challenge and reading
4185 a response while authentication was assuming a zero-terminated challenge
4186 string. It's now changed to take a pointer and a length, to allow for
4187 binary data in such strings.
4189 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4191 23. Exim was not respecting finduser_retries when seeking the login of the
4192 uid under which it was called; it was always trying 10 times. (The default
4193 setting of finduser_retries is zero.) Also, it was sleeping after the final
4194 failure, which is pointless.
4196 24. Implemented tls_on_connect_ports.
4198 25. Implemented acl_smtp_predata.
4200 26. If the domain in control=submission is set empty, Exim assumes that the
4201 authenticated id is a complete email address when it generates From: or
4202 Sender: header lines.
4204 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4205 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4206 chown and chgrp in /bin and hostname in /usr/bin.
4208 28. Exim was keeping the "process log" file open after each use, just as it
4209 does for the main log. This opens the possibility of it remaining open for
4210 long periods when the USR1 signal hits a daemon. Occasional processlog
4211 errors were reported, that could have been caused by this. Anyway, it seems
4212 much more sensible not to leave this file open at all, so that is what now
4215 29. The long-running daemon process does not normally write to the log once it
4216 has entered its main loop, and it closes the log before doing so. This is
4217 so that log files can straightforwardly be renamed and moved. However,
4218 there are a couple of unusual error situations where the daemon does write
4219 log entries, and I had neglected to close the log afterwards.
4221 30. The text of an SMTP error response that was received during a remote
4222 delivery was being truncated at 512 bytes. This is too short for some of
4223 the long messages that one sometimes sees. I've increased the limit to
4226 31. It is now possible to make retry rules that apply only when a message has a
4227 specific sender, in particular, an empty sender.
4229 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4230 it possible to be selective about when SMTP synchronization is enforced.
4232 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4234 32. Implemented hosts_connection_nolog.
4236 33. Added an ACL for QUIT.
4238 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4241 35. Added mailbox_size and mailbox_filecount to appendfile.
4243 36. Added control = no_multiline_responses to ACLs.
4245 37. There was a bug in the logic of the code that waits for the clock to tick
4246 in the case where the clock went backwards by a substantial amount such
4247 that the microsecond fraction of "now" was more than the microsecond
4248 fraction of "then" (but the whole seconds number was less).
4250 38. Added support for the libradius Radius client library this is found on
4251 FreeBSD (previously only the radiusclient library was supported).
4257 1. When certain lookups returned multiple values in the form name=value, the
4258 quoting of the values was not always being done properly. Specifically:
4259 (a) If the value started with a double quote, but contained no whitespace,
4261 (b) If the value contained whitespace other than a space character (i.e.
4262 tabs or newlines or carriage returns) it was not quoted.
4263 This fix has been applied to the mysql and pgsql lookups by writing a
4264 separate quoting function and calling it from the lookup code. The fix
4265 should probably also be applied to nisplus, ibase and oracle lookups, but
4266 since I cannot test any of those, I have not disturbed their existing code.
4268 2. A hit in the callout cache for a specific address caused a log line with no
4269 reason for rejecting RCPT. Now it says "Previous (cached) callout
4270 verification failure".
4272 3. There was an off-by-one bug in the queryprogram router. An over-long
4273 return line was truncated at 256 instead of 255 characters, thereby
4274 overflowing its buffer with the terminating zero. As well as fixing this, I
4275 have increased the buffer size to 1024 (and made a note to document this).
4277 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4278 when Exim is waiting for an SMTP response from a remote server, Exim
4279 restarts its select() call on the socket, thereby resetting its timeout.
4280 This is not a problem when such interrupts are rare. Somebody set up a cron
4281 job to run exiwhat every 2 minutes, which is less than the normal select()
4282 timeout (5 or 10 minutes). This meant that the select() timeout never
4283 kicked in because it was always reset. I have fixed this by comparing the
4284 time when an interrupt arrives with the time at the start of the first call
4285 to select(). If more time than the timeout has elapsed, the interrupt is
4286 treated as a timeout.
4288 5. Some internal re-factoring in preparation for the addition of Sieve
4289 extensions (by MH). In particular, the "personal" test is moved to a
4290 separate function, and given an option for scanning Cc: and Bcc: (which is
4291 not set for Exim filters).
4293 6. When Exim created an email address using the login of the caller as the
4294 local part (e.g. when creating a From: or Sender: header line), it was not
4295 quoting the local part when it contained special characters such as @.
4297 7. Installed new OpenBSD configuration files.
4299 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4300 try to make them clearer.
4302 9. Callout options, other than the timeout value, were being ignored when
4303 verifying sender addresses in header lines. For example, when using
4305 verify = header_sender/callout=no_cache
4307 the cache was (incorrectly) being used.
4309 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4310 only the Cygwin environment.
4312 11. When return_path_on_delivery was set as a log selector, if different remote
4313 addresses in the same message used different return paths and parallel
4314 remote delivery occurred, the wrong values would sometimes be logged.
4315 (Whenever a remote delivery process finished, the return path value from
4316 the most recently started remote delivery process was logged.)
4318 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4319 lines when AUTH and/or TLS are in use. This is the "received protocol"
4320 field. Exim used to use "asmtp" for authenticated SMTP, without any
4321 indication (in the protocol name) for TLS use. Now it follows the RFC and
4322 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4323 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4324 names appear in log lines as well as in Received: header lines.
4326 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4327 extensions, and comparison tests, and to fix some bugs.
4329 14. Changes to the "personal" filter test:
4331 (1) The test was buggy in that it was just doing the equivalent of
4332 "contains" tests on header lines. For example, if a user's address was
4333 anne@some.where, the "personal" test would incorrectly be true for
4335 To: susanne@some.where
4337 This test is now done by extracting each address from the header in turn,
4338 and checking the entire address. Other tests that are part of "personal"
4339 are now done using regular expressions (for example, to check local parts
4340 of addresses in From: header lines).
4342 (2) The list of non-personal local parts in From: addresses has been
4343 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4344 taken from the Sieve specification recommendations.
4346 (3) If the message contains any header line starting with "List-" it is
4347 treated as non-personal.
4349 (4) The test for "circular" in the Subject: header line has been removed
4350 because it now seems ill-conceived.
4352 15. Minor typos in src/EDITME comments corrected.
4354 16. Installed latest exipick from John Jetmore.
4356 17. If headers_add on a router specified a text string that was too long for
4357 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4358 of string_sprintf() is now avoided.
4360 18. $message_body_size was not set (it was always zero) when running the DATA
4361 ACL and the local_scan() function.
4363 19. For the "mail" command in an Exim filter, no default was being set for
4364 the once_repeat time, causing a random time value to be used if "once" was
4365 specified. (If the value happened to be <= 0, no repeat happened.) The
4366 default is now 0s, meaning "never repeat". The "vacation" command was OK
4367 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4368 (I found it when inspecting the code).
4370 20. There is now an overall timeout for performing a callout verification. It
4371 defaults to 4 times the callout timeout, which applies to individual SMTP
4372 commands during the callout. The overall timeout applies when there is more
4373 than one host that can be tried. The timeout is checked before trying the
4374 next host. This prevents very long delays if there are a large number of
4375 hosts and all are timing out (e.g. when the network connections are timing
4376 out). The value of the overall timeout can be changed by specifying an
4377 additional sub-option for "callout", called "maxwait". For example:
4379 verify = sender/callout=5s,maxwait=20s
4381 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4382 to the end before writing, but this should make it even safer).
4384 22. Exim was forgetting that it had advertised PIPELINING for the second and
4385 subsequent messages on an SMTP connection. It was also not resetting its
4386 memory on STARTTLS and an internal HELO.
4388 23. When Exim logs an SMTP synchronization error within a session, it now
4389 records whether PIPELINING has been advertised or not.
4391 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4392 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4393 rather than long int.
4395 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4397 26. Added the never_mail option to autoreply.
4403 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4404 crash if the getsockname() call failed; this can happen if a connection is
4405 closed very soon after it is established. The problem was simply in the
4406 order in which certain operations were done, causing Exim to try to write
4407 to the SMTP stream before it had set up the file descriptor. The bug has
4408 been fixed by making things happen in the correct order.
4414 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4415 before the connection was closed, thus losing the rejection response.
4417 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4418 some early Solaris releases, but causes trouble in current releases where
4419 socklen_t is defined.
4421 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4424 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4427 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4428 that releasing the top of it at the end releases what was used for sub-
4429 expansions (unless the block got too big). However, discard this block if
4430 the first thing is a variable or header, so that we can use its block when
4431 it is dynamic (useful for very large $message_headers, for example).
4433 6. Lookups now cache *every* query, not just the most recent. A new, separate
4434 store pool is used for this. It can be recovered when all lookup caches are
4435 flushed. Lookups now release memory at the end of their result strings.
4436 This has involved some general refactoring of the lookup sources.
4438 7. Some code has been added to the store_xxx() functions to reduce the amount
4439 of flapping under certain conditions.
4441 8. log_incoming_interface used to affect only the <= reception log lines. Now
4442 it causes the local interface and port to be added to several more SMTP log
4443 lines, for example "SMTP connection from", and rejection lines.
4445 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4447 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4449 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4450 could be overwritten at the end of the current message (or the start of a
4451 new message if it was set in a HELO ACL). The value is now preserved for
4452 the duration of the SMTP connection.
4454 12. If a transport had a headers_rewrite setting, and a matching header line
4455 contained an unqualified address, that address was qualified, even if it
4456 did not match any rewriting rules. The underlying bug was that the values
4457 of the flags that permit the existence of unqualified sender and recipient
4458 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4459 non-local messages, and by -bnq for local messages) were not being
4460 preserved with the message after it was received.
4462 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4463 "next input=" as part of the text comprising the host identity instead of
4464 the correct text. The code was using the same buffer for two different
4465 strings. However, depending on which order the printing function evaluated
4466 its arguments, the bug did not always show up. Under Linux, for example, my
4467 test suite worked just fine.
4469 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4470 doesn't work with some older versions of Perl. It has been changed to "my",
4471 which in any case is probably the better facility to use.
4473 15. A really picky compiler found some instances of statements for creating
4474 error messages that either had too many or two few arguments for the format
4477 16. The size of the buffer for calls to the DNS resolver has been increased
4478 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4479 for addresses that have a lot of PTR records. This alleviates a problem; it
4480 does not fully solve it.
4482 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4483 buffer now truncates the list and logs the incident, which is the same
4484 action as happens when Exim is looking up a host name and its aliases.
4485 Previously in this situation something unpredictable would happen;
4486 sometimes it was "internal error: store_reset failed".
4488 18. If a server dropped the connection unexpectedly when an Exim client was
4489 using GnuTLS and trying to read a response, the client delivery process
4490 crashed while trying to generate an error log message.
4492 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4493 string, for example:
4495 warn message = H1: something\nH2: something
4497 the text was added as a single header line from Exim's point of view
4498 though it ended up OK in the delivered message. However, searching for the
4499 second and subsequent header lines using $h_h2: did not work. This has been
4500 fixed. Similarly, if a system filter added multiple headers in this way,
4501 the routers could not see them.
4503 20. Expanded the error message when iplsearch is called with an invalid key to
4504 suggest using net-iplsearch in a host list.
4506 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4507 ACLs are no longer actually imposed (and a message to that effect is
4510 22. If a "gecos" field in a passwd entry contained escaped characters, in
4511 particular, if it contained a \" sequence, Exim got it wrong when building
4512 a From: or a Sender: header from that name. A second bug also caused
4513 incorrect handling when an unquoted " was present following a character
4514 that needed quoting.
4516 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4517 was not being matched caselessly.
4519 24. Arranged for all hyphens in the exim.8 source to be escaped with
4522 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4523 properly. Recipient callout cache records were still being keyed to include
4524 the sender, even when use_sender was set false. This led to far more
4525 callouts that were necessary. The sender is no longer included in the key
4526 when use_sender is false.
4528 26. Added "control = submission" modifier to ACLs.
4530 27. Added the ${base62d: operator to decode base 62 numbers.
4532 28. dnsdb lookups can now access SRV records.
4534 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4535 the configuration file.
4537 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4538 (-v) mode. This makes the output for a verbose queue run more intelligible.
4540 31. Added a use_postmaster feature to recipient callouts.
4542 32. Added the $body_zerocount variable, containing the number of binary zero
4543 bytes in the message body.
4545 33. The time of last modification of the "new" subdirectory is now used as the
4546 "mailbox time last read" when there is a quota error for a maildir
4549 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4551 35. Added +ignore_unknown as a special item in host lists.
4553 36. Code for decoding IPv6 addresses in host lists is now included, even if
4554 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4555 address was recognized as an IP address, but was then not correctly decoded
4556 into binary, causing unexpected and incorrect effects when compared with
4563 1. Very minor rewording of debugging text in manualroute to say "list of
4564 hosts" instead of "hostlist".
4566 2. If verify=header_syntax was set, and a header line with an unqualified
4567 address (no domain) and a large number of spaces between the end of the
4568 name and the colon was received, the reception process suffered a buffer
4569 overflow, and (when I tested it) crashed. This was caused by some obsolete
4570 code that should have been removed. The fix is to remove it!
4572 3. When running in the test harness, delay a bit after writing a bounce
4573 message to get a bit more predictability in the log output.
4575 4. Added a call to search_tidyup() just before forking a reception process. In
4576 theory, someone could use a lookup in the expansion of smtp_accept_max_
4577 per_host which, without the tidyup, could leave open a database connection.
4579 5. Added the variables $recipient_data and $sender_data which get set from a
4580 lookup success in an ACL "recipients" or "senders" condition, or a router
4581 "senders" option, similar to $domain_data and $local_part_data.
4583 6. Moved the writing of debug_print from before to after the "senders" test
4586 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4587 problems for message scanning, either using a data ACL, or using
4588 local_scan() because the Received: header was not generated till after they
4589 were called (in order to set the time as the time of reception completion).
4590 I have revised the way this works. The header is now generated after the
4591 body is received, but before the ACL or local_scan() are called. After they
4592 are run, the timestamp in the header is updated.
4598 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4599 before starting a queue runner without re-exec. This happened only when
4600 deliver_drop_privilege was set or when the Exim user was set to root. The
4601 effect of the bug was that timeouts during subsequent deliveries caused
4602 crashes instead of being properly handled. The handler is now left at its
4603 default (and expected) setting.
4605 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4606 message, again when deliver_drop_privilege is set or Exim is run as root.
4607 The bug described in (1) was not present in this case, but the tidying up
4608 of the other signals was missing. I have made the two cases consistent.
4610 3. The ignore_target_hosts setting on a manualroute router was being ignored
4611 for hosts that were looked up using the /MX notation.
4613 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4616 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4617 operated on the sender address. After changing the $sender_address to <>
4618 for the sender address verify, Exim was re-instated it as the original
4619 (before rewriting) address, but remembering that it had rewritten it, so it
4620 wasn't rewriting it again. This bug also had the effect of breaking the
4621 sender address verification caching when the sender address was rewritten.
4623 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4624 This has been changed so that if the ip literal address matches
4625 ignore_target_hosts, the router declines.
4627 7. Added expansion conditions match_domain, match_address, and match_local_
4628 part (NOT match_host).
4630 8. The placeholder for the Received: header didn't have a length field set.
4632 9. Added code to Exim itself and to exim_lock to test for a specific race
4633 condition that could lead to file corruption when using MBX delivery. The
4634 issue is with the lockfile that is created in /tmp. If this file is removed
4635 after a process has opened it but before that process has acquired a lock,
4636 there is the potential for a second process to recreate the file and also
4637 acquire a lock. This could lead to two Exim processes writing to the file
4638 at the same time. The added code performs the same test as UW imapd; it
4639 checks after acquiring the lock that its file descriptor still refers to
4640 the same named file.
4642 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4643 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4646 11. Added the smtp_active_hostname option. If used, this will typically be made
4647 to depend on the incoming interface address. Because $interface_address is
4648 not set up until the daemon has forked a reception process, error responses
4649 that can happen earlier (such as "too many connections") no longer contain
4652 12. If an expansion in a condition on a "warn" statement fails because a lookup
4653 defers, the "warn" statement is abandoned, and the next ACL statement is
4654 processed. Previously this caused the whole ACL to be aborted.
4656 13. Added the iplsearch lookup type.
4658 14. Added ident_timeout as a log selector.
4660 15. Added tls_certificate_verified as a log selector.
4662 16. Added a global option tls_require_ciphers (compare the smtp transport
4663 option of the same name). This controls incoming TLS connections.
4665 17. I finally figured out how to make tls_require_ciphers do a similar thing
4666 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4667 before starting the TLS session.
4669 18. Tabs are now shown as \t in -bP output.
4671 19. If the log selector return_path_on_delivery was set, Exim crashed when
4672 bouncing a message because it had too many Received: header lines.
4674 20. If two routers both had headers_remove settings, and the first one included
4675 a superfluous trailing colon, the final name in the first list and the
4676 first name in the second list were incorrectly joined into one item (with a
4677 colon in the middle).
4683 1. Added -C and -D options to the exinext utility, mainly to make it easier
4684 to include in the automated testing, but these could be helpful when
4685 multiple configurations are in use.
4687 2. The exinext utility was not formatting the output nicely when there was
4688 an alternate port involved in the retry record key, nor when there was a
4689 message id as well (for retries that were specific to a specific message
4690 and a specific host). It was also confused by IPv6 addresses, because of
4691 the additional colons they contain. I have fixed the IPv4 problem, and
4692 patched it up to do a reasonable job for IPv6.
4694 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4695 delivery, the log line now contains "pipelined" if PIPELINING was used.
4697 4. An SMTP transport process used to panic and die if the bind() call to set
4698 an explicit outgoing interface failed. This has been changed; it is now
4699 treated in the same way as a connect() failure.
4701 5. A reference to $sender_host_name in the part of a conditional expansion
4702 that was being skipped was still causing a DNS lookup. This no longer
4705 6. The def: expansion condition was not recognizing references to header lines
4706 that used bh_ and bheader_.
4708 7. Added the _cache feature to named lists.
4710 8. The code for checking quota_filecount in the appendfile transport was
4711 allowing one more file than it should have been.
4713 9. For compatibility with Sendmail, the command line option
4721 and sets the incoming protocol and host name (for trusted callers). The
4722 host name and its colon can be omitted when only the protocol is to be set.
4723 Note the Exim already has two private options, -pd and -ps, that refer to
4724 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4725 "s", but I don't think that's a major issue.
4727 10. A number of refactoring changes to the code, none of which should affect
4730 (a) The number of logging options was getting close to filling up the
4731 32-bit word that was used as a bit map. I have split them into two classes:
4732 those that are passed in the argument to log_write(), and those that are
4733 only ever tested independently outside of that function. These are now in
4734 separate 32-bit words, so there is plenty of room for expansion again.
4735 There is no change in the user interface or the logging behaviour.
4737 (b) When building, for example, log lines, the code previously used a
4738 macro that called string_cat() twice, in order to add two strings. This is
4739 not really sufficiently general. Furthermore, there was one instance where
4740 it was actually wrong because one of the argument was used twice, and in
4741 one call a function was used. (As it happened, calling the function twice
4742 did not affect the overall behaviour.) The macro has been replaced by a
4743 function that can join an arbitrary number of extra strings onto a growing
4746 (c) The code for expansion conditions now uses a table and a binary chop
4747 instead of a serial search (which was left over from when there were very
4748 few conditions). Also, it now recognizes conditions like "pam" even when
4749 the relevant support is not compiled in: a suitably worded error message is
4750 given if an attempt is made to use such a condition.
4752 11. Added ${time_interval:xxxxx}.
4754 12. A bug was causing one of the ddress fields not to be passed back correctly
4755 from remote delivery subprocesses. The field in question was not being
4756 subsequently used, so this caused to problems in practice.
4758 13. Added new log selectors queue_time and deliver_time.
4760 14. Might have fixed a bug in maildirsizefile handling that threw up
4761 "unexpected character" debug warnings, and recalculated the data
4762 unnecessarily. In any case, I expanded the warning message to give more
4765 15. Added the message "Restricted characters in address" to the statements in
4766 the default ACL that block characters like @ and % in local parts.
4768 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4769 Three changes have been made:
4771 (a) There was a serious bug; a negative response to MAIL caused the whole
4772 recipient domain to be cached as invalid, thereby blocking all messages
4773 to all local parts at the same domain, from all senders. This bug has
4774 been fixed. The domain is no longer cached after a negative response to
4775 MAIL if the sender used is not empty.
4777 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4780 (c) A new callout option, "use_sender" has been added for people who want
4781 the modified behaviour.
4787 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4790 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4791 indeed breaks things for older releases.
4793 3. Added additional logging to the case where there is a problem reading data
4794 from a filter that is running in a subprocess using a pipe, in order to
4795 try to track down a specific problem.
4797 4. Testing facility fudge: when running in the test harness and attempting
4798 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4799 getting "No route to host". Convert this to a timeout.
4801 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4804 6. Some OS don't have socklen_t but use size_t instead. This affects the
4805 fifth argument of getsockopt() amongst other things. This is now
4806 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4807 can be set for individual OS. I have set it for SunOS5, OSF1, and
4808 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4809 some earlier ones do not.
4811 7. Change 4.30/15 was not doing the test caselessly.
4813 8. The standard form for an IPv6 address literal was being rejected by address
4814 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4815 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4816 this, as well as the form without the "IPv6" on the front (but only when
4817 address literals are enabled, of course).
4819 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4821 10. Exim crashed if a message with an empty sender address specified by -f
4822 encountered a router with an errors_to setting. This could be provoked only
4823 by a command such as
4827 where an empty string was supplied; "<>" did not hit this bug.
4829 11. Installed PCRE release 4.5.
4831 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4832 remained set. It is now erased.
4834 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4835 times from message ids (which are base 36 rather than the normal 62).
4837 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4838 were being counted as actual protocol errors, and logged if the log
4839 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4840 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4841 and DATA following a set of rejected RCPTs do not count as protocol errors.
4842 In other words, Exim assumes they were pipelined, though this may not
4843 actually be the case. Of course, in all cases the client gets an
4844 appropriate error code.
4846 15. If a lookup fails in an ACL condition, a message about the failure may
4847 be available; it is used if testing the ACL cannot continue, because most
4848 such messages specify what the cause of the deferral is. However, some
4849 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4850 that caused an old message to be retained and used if a later statement
4851 caused a defer, replacing the real cause of the deferral.
4853 16. If an IP address had so many PTR records that the DNS lookup buffer
4854 was not large enough to hold them, Exim could crash while trying to process
4855 the truncated data. It now detects and logs this case.
4857 17. Further to 4.21/58, another change has been made: if (and only if) the
4858 first line of a message (the first header line) ends with CRLF, a bare LF
4859 in a subsequent header line has a space inserted after it, so as not to
4860 terminate the header.
4862 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4863 unnecessarily, used atoi() instead of strtol(), and didn't check the
4864 termination when getting file sizes from file names by regex.
4866 19. Completely re-implemented the support for maildirsize files, in the light
4867 of a number of problems with the previous contributed implementation
4868 (4.30/29). In particular:
4870 . If the quota is zero, the maildirsize file is maintained, but no quota is
4873 . If the maildir directory does not exist, it is created before any attempt
4874 to write a maildirsize file.
4876 . The quota value in the file is just a cache; if the quota is changed in
4877 the transport, the new value overrides.
4879 . A regular expression is available for excluding directories from the
4882 20. The autoreply transport checks the characters in options that define the
4883 message's headers; it allows continued headers, but it was checking with
4884 isspace() after an embedded newline instead of explicitly looking for a
4887 21. If all the "regular" hosts to which an address was routed had passed their
4888 expiry times, and had not reached their retry times, the address was
4889 bounced, even if fallback hosts were defined. Now Exim should go on to try
4892 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4893 equivalent code in the SMTP transport. Some hosts send humungous responses
4894 to HELO/EHLO, more than 1024 it seems.
4896 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4897 gives compiler warnings in some environments. I've now done it "properly",
4900 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4901 (because of problems with the built-in one) was declared to return uschar *
4902 instead of char *, causing compiler failure.
4904 25. Fixed a file descriptor leak when processing alias/forward files.
4906 26. Fixed a minor format string issue in dbfn.c.
4908 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4910 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4911 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4914 29. When a sender address is verified, it is cached, to save repeating the test
4915 when there is more than one recipient in a message. However, when the
4916 verification involves a callout, it is possible for different callout
4917 options to be set for different recipients. It is too complicated to keep
4918 track of this in the cache, so now Exim always runs a verification when a
4919 callout is required, relying on the callout cache for the optimization.
4920 The overhead is duplication of the address routing, but this should not be
4923 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4924 to be verified with callout=postmaster, and the main callout worked but the
4925 postmaster check failed, the verification correctly failed. However, if a
4926 subsequent RCPT command asked for sender verification *without* the
4927 postmaster check, incorrect caching caused this verification also to fail,
4930 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4931 it was not caching the DNS options (qualify_single, search_parents) that
4932 were used when the lookup failed. A subsequent lookup with different
4933 options therefore always gave the same answer, though there were cases
4934 where it should not have. (Example: a "domains = !$mx_any" option on a
4935 dnslookup router: the "domains" option is always processed without any
4936 widening, but the router might have qualify_single set.) Now Exim uses the
4937 cached value only when the same options are set.
4939 32. Added John Jetmore's "exipick" utility to the distribution.
4941 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4942 than a timeout (e.g. a certificate is required, and is not provided), an
4943 Exim server now closes the connection immediately. Previously it waited for
4944 the client to close - but if the client is SSL, it seems that they each
4945 wait for each other, leading to a delay before one of them times out.
4947 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4948 maintained 0.8.x compatibility because I don't think many are using it, and
4949 it is clearly obsolete.
4951 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4954 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4955 client certificate was expired. A simple patch fixes this, though I don't
4956 understand the full logic of why the verify callback is called multiple
4959 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4960 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4961 which causes problems with some clients (such as the Certicom SSL Plus
4962 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4963 disables the coutermeasure allowing Eudora to connect."
4965 38. Exim was not checking that a write() to a log file succeeded. This could
4966 lead to Bad Things if a log got too big, in particular if it hit a file
4967 size limit. Exim now panics and dies if it cannot write to a log file, just
4968 as it does if it cannot open a log file.
4970 39. Modified OS/Makefile-Linux so that it now contains
4972 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4974 The two -D definitions ensure that Exim is compiled with large file
4975 support, which makes it possible to handle log files that are bigger than
4978 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4979 instance) a domain was checked against a named list that involved a lookup,
4980 causing $domain_data to be set, then another domain was checked against the
4981 same list, then the first domain was re-checked, the value of $domain_data
4982 after the final check could be wrong. In particular, if the second check
4983 failed, it could be set empty. This bug probably also applied to
4986 41. The strip_trailing_dot option was not being applied to the address given
4987 with the -f command-line option.
4989 42. The code for reading a message's header from the spool was incrementing
4990 $received_count, but never initializing it. This meant that the value was
4991 incorrect (doubled) while delivering a message in the same process in which
4992 it was received. In the most common configuration of Exim, this never
4993 happens - a fresh exec is done - but it can happen when
4994 deliver_drop_privilege is set.
4996 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4997 it now includes up to 150 characters of the unexpected data in the log
5000 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5001 and building data strings. The size of both of these buffers was 10 000
5002 bytes - far larger than anybody would *ever* want, thought I. Needless to
5003 say, somebody hit the limit. I have increased the maximum line length to
5004 20 000 and the maximum data length of concatenated lines to 100 000. I have
5005 also fixed two bugs, because there was no checking on these buffers. Tsk,
5006 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5007 buffer is too small.
5009 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5010 lsearch lookups. Now it does.
5012 46. When parsing a route_list item in a manualroute router, a fixed-length
5013 buffer was used for the list of hosts. I made this 1024 bytes long,
5014 thinking that nobody would ever have a list of hosts that long. Wrong.
5015 Somebody had a whole pile of complicated expansion conditions, and the
5016 string was silently truncated, leading to an expansion error. It turns out
5017 that it is easier to change to an unlimited length (owing to other changes
5018 that have happened since this code was originally written) than to build
5019 structure for giving a limitation error. The length of the item that
5020 expands into the list of hosts is now unlimited.
5022 47. The lsearch lookup could not handle data where the length of text line was
5023 more than 4095 characters. Such lines were truncated, leading to shortened
5024 data being returned. It should now handle lines of any length.
5026 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5027 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5030 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5031 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5032 no longer happens in comment lines. A list of replacements is now placed
5033 at the head of all of the source files, except those whose only change is
5034 to replace PERL_COMMAND in the very first #! line.
5036 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5037 messages on the queue, with a bottom-up merge sort, using code contributed
5038 by Michael Haardt. This should make operations like -bp somewhat faster on
5039 large queues. It won't affect queue runners, except when queue_run_in_order
5042 51. Installed eximstats 1.31 in the distribution.
5044 52. Added support for SRV lookups to the dnslookup router.
5046 53. If an ACL referred to $message_body or $message_body_end, the value was not
5047 reset for any messages that followed in the same SMTP session.
5049 54. The store-handling optimization for building very long strings was not
5050 differentiating between the different store pools. I don't think this
5051 actually made any difference in practice, but I've tidied it.
5053 55. While running the routers to verify a sender address, $sender_address
5054 was still set to the sender address. This is wrong, because when routing to
5055 send a bounce to the sender, it would be empty. Therefore, I have changed
5056 it so that, while verifying a sender address, $sender_address is set to <>.
5057 (There is no change to what happens when verifying a recipient address.)
5059 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5060 target A or AAAA records (if not already returned) without resetting the
5061 qualify_single or search_parents options of the DNS resolver. These are
5062 inappropriate in this case because the targets of MX and SRV records must
5063 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5064 target that, when qualified, matched something in the local domain. These
5065 two options are now turned off when doing these lookups.
5067 57. It seems that at least some releases of Reiserfs (which does not have the
5068 concept of a fixed number of inodes) returns zero and not -1 for the
5069 number of available inodes. This interacted badly with check_spool_inodes,
5070 which assumed that -1 was the "no such thing" setting. What I have done is
5071 to check that the total number of inodes is greater than zero before doing
5072 the test of how many are available.
5074 58. When a "warn" ACL statement has a log_message modifier, the message is
5075 remembered, and not repeated. This is to avoid a lot of repetition when a
5076 message has many recipients that cause the same warning to be written.
5077 Howewer, Exim was preserving the list of already written lines for an
5078 entire SMTP session, which doesn't seem right. The memory is now reset if a
5079 new message is started.
5081 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5082 debugging output unless log_rewrite was also set.
5084 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5085 of (int)(handle) when we know that handle contains (void *)(-1).
5087 61. The Exim daemon panic-logs an error return when it closes the incoming
5088 connection. However "connection reset by peer" seems to be common, and
5089 isn't really an error worthy of noting specially, so that particular error
5092 62. When Exim is trying to find all the local interfaces, it used to panic and
5093 die if the ioctl to get the interface flags failed. However, it seems that
5094 on at least one OS (Solaris 9) it is possible to have an interface that is
5095 included in the list of interfaces, but for which you get a failure error
5096 for this call. This happens when the interface is not "plumbed" into a
5097 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5098 failure of the "get flags" call assumes that the interface is down.
5100 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5101 makes life easier for people who are doing arithmetic on fields extracted
5102 from dates, where you often get leading zeros that should not be
5103 interpreted as octal.
5105 64. Added qualify_domain to the redirect router, to override the global
5108 65. If a pathologically long header line contained very many addresses (the
5109 report of this problem mentioned 10 000) and each of them was rewritten,
5110 Exim could use up a very large amount of memory. (It kept on making new
5111 copies of the header line as it rewrote, and never released the old ones.)
5112 At the expense of a bit more processing, the header rewriting function has
5113 been changed so that it no longer eats memory in this way.
5115 66. The generation of the Received: header has been moved from the time that a
5116 message starts to be received, to the time that it finishes. The timestamp
5117 in the Received: header should now be very close to that of the <= log
5118 line. There are two side-effects of this change:
5120 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5121 logged header lines no longer include the local Received: line, because
5122 it has not yet been created. The same applies to a copy of the message
5123 that is returned to a non-SMTP sender when a message is rejected.
5125 (b) When a filter file is tested using -bf, no additional Received: header
5126 is added to the test message. After some thought, I decided that this
5129 This change does not affect the value of $received_for. It is still set
5130 after address rewriting, but before local_scan() is called.
5132 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5134 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5135 gave an unhelpful panic error message, and a defer error. I have managed to
5136 change this behaviour so that it now rejects any supplied certificate,
5137 which seems right, as the list of acceptable certificates is empty.
5139 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5140 gave an unhelpful defer error. I have not managed to make this reject any
5141 supplied certificates, but the error message it gives is "no certificate
5142 supplied", which is not helpful.
5144 70. exigrep's output now also includes lines that are not associated with any
5145 message, but which match the given pattern. Implemented by a patch from
5146 Martin Sluka, which also tidied up the Perl a bit.
5148 71. Recipient callout verification, like sender verification, was using <> in
5149 the MAIL FROM command. This isn't really the right thing, since the actual
5150 sender may affect whether the remote host accepts the recipient or not. I
5151 have changed it to use the actual sender in the callout; this means that
5152 the cache record is now keyed on a recipient/sender pair, not just the
5153 recipient address. There doesn't seem to be a real danger of callout loops,
5154 since a callout by the remote host to check the sender would use <>.
5155 [SEE ABOVE: changed after hitting problems.]
5157 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5158 temporary errors. However, in the case of such a code being given after
5159 the end of a data transmission (i.e. after ".") Exim was failing to write
5160 a retry record for the message. (Yes, there was some broken host that was
5161 actually sending 8xx at this point.)
5163 73. An unknown lookup type in a host list could cause Exim to panic-die when
5164 the list was checked. (An example that provoked this was putting <; in the
5165 middle of a list instead of at the start.) If this happened during a DATA
5166 ACL check, a -D file could be left lying around. This kind of configuration
5167 error no longer causes Exim to die; instead it causes a defer errror. The
5168 incident is still logged to the main and panic logs.
5170 74. Buglet left over from Exim 3 conversion. The message "too many messages
5171 in one connection" was written to the rejectlog but not the mainlog, except
5172 when address rewriting (yes!) was being logged.
5174 75. Added write_rejectlog option.
5176 76. When a system filter was run not as root (that is, when system_filter_user
5177 was set), the values of the $n variables were not being returned to the
5178 main process; thus, they were not subsequently available in the $sn
5181 77. Added +return_path_on_delivery log selector.
5183 78. A connection timeout was being treated differently from recipients deferred
5184 when testing hosts_max_try with a message that was older than the host's
5185 retry timeout. (The host should not be counted, thus allowing all hosts to
5186 be tried at least once before bouncing.) This may have been the cause of an
5187 occasionally reported bug whereby a message would remain on the queue
5188 longer than the retry timeout, but would be bounced if a delivery was
5189 forced. I say "may" because I never totally pinned down the problem;
5190 setting up timeout/retry tests is difficult. See also the next item.
5192 79. The ultimate address timeout was not being applied to errors that involved
5193 a combination of host plus message (for example, a timeout on a MAIL
5194 command). When an address resolved to a number of possible hosts, and they
5195 were not all tried for each delivery (e.g. because of hosts_max_try), a
5196 message could remain on the queue longer than the retry timeout.
5198 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5201 81. Fixed an obscure SMTP outgoing bug which required at least the following
5202 conditions: (a) there was another message waiting for the same server;
5203 (b) the server returned 5xx to all RCPT commands in the first message so
5204 that the message was not completed; (c) the server dropped the connection
5205 or gave a negative response to the RSET that Exim sends to abort the
5206 transaction. The observed case was a dropped connection after DATA that had
5207 been sent in pipelining mode. That is, the server had advertised PIPELINING
5208 but was not implementing it correctly. The effect of the bug was incorrect
5209 behaviour, such as trying another host, and this could lead to a crash.
5215 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5216 and daemon.c were passed as pointers to ints; they should have been
5217 pointers to socklen_t variables (which are typically unsigned ints).
5219 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5222 3. Fixed a really odd bug that affected only the testing scheme; patching a
5223 certain fixed string in the binary changed the value of another string that
5224 happened to be identical to the end of the original first string.
5226 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5227 name", it returns that address as the IP address. On some operating
5228 systems (e.g. Solaris), it also passes back the IP address string as the
5229 "host name". However, on others (e.g. Linux), it passes back an empty
5230 string. Exim wasn't checking for this, and was changing the host name to an
5231 empty string, assuming it had been canonicized.
5233 5. Although rare, it is permitted to have more than one PTR record for a given
5234 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5235 all the names associated with an address, because they do in Solaris.
5236 However, it seems that they do not in Linux for data that comes from the
5237 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5238 I found this out when I moved to a new Linux workstation and tried to run
5239 the Exim test suite.
5241 To get round this problem I have changed the code so that it now does its
5242 own call to the DNS to look up PTR records when searching for a host name.
5243 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5244 addresses that are only in /etc/hosts are still found.
5246 This behaviour is, however, controlled by an option called host_lookup_
5247 order, which defaults to "bydns:byaddr". If people want to use the other
5248 order, or indeed, just use one or the other means of lookup, they can
5249 specify it in this variable.
5251 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5252 some operating systems, this comes back from gethostbyaddr() as an empty
5253 string, and this is what Exim used to test for. However, it seems that in
5254 other systems, "." is yielded. Exim now tests for this case too.
5256 7. The values of check_spool_space and check_log_space are now held internally
5257 as a number of kilobytes instead of an absolute number of bytes. If a
5258 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5259 kilobyte. This means that much larger values can be stored.
5261 8. Exim monitor: an attempt to get the action menu when not actually pointing
5262 at a message produces an empty menu entitled "No message selected". This
5263 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5264 no entries in it ("Shell widget menu has zero width and/or height"). So I
5265 have added a single, blank menu entry in this case.
5267 9. Added ${quote_local_part.
5269 10. MIME decoding is now applied to the contents of Subject: header lines when
5272 11. Now that a reference to $sender_host_address automatically causes a reverse
5273 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5274 host lookup before query-style lookups in lists that might use this
5275 variable. This has therefore been abolished, and the "net-" prefix is no
5276 longer necessary for query-style lookups.
5278 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5279 have been a typo for LFLAGS, so it has been changed.
5281 13. The install script calls Exim with "-C /dev/null" in order to find the
5282 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5283 to be output. Howeve, since Exim outputs its version number before the
5284 error, it didn't break the script. It just looked ugly. I fixed this by
5285 always allowing "-C /dev/null" if the caller is root.
5287 14. Ignore overlarge ACL variable number when reading spool file - insurance
5288 against a later release with more variables having written the file.
5290 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5291 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5292 this, as well as the form without the "IPv6" on the front.
5294 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5295 OS/Makefile-Darwin file.
5297 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5298 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5300 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5301 variable. However, some OS use ints for inodes. I've added cast to long int
5302 to get rid of the compiler warning.
5304 19. I had forgotten to lock out "/../" in configuration file names when
5305 ALT_CONFIG_PREFIX was set.
5307 20. Routers used for verification do not need to specify transports. However,
5308 if such a router generated a host list, and callout was configured, Exim
5309 crashed, because it could not find a port number from the (non-existent)
5310 transport. It now assumes port 25 in this circumstance.
5312 21. Added the -t option to exigrep.
5314 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5315 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5316 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5317 them is set, LOOKUP_LSEARCH is forced.
5319 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5321 24. Added sender and host information to the "rejected by local_scan()" log
5322 line; previously there was no indication of these.
5324 25. Added .include_if_exists.
5326 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5327 Linux. It turns out that not all file systems support this. Apparently some
5328 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5329 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5330 not supported on the file descriptor, is now ignored when Exim is trying to
5331 sync a directory. This applies only to Linux.
5333 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5335 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5336 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5339 29. Added support for maildirsize files from supplied patch (modified a bit).
5341 30. The use of :fail: followed by an empty string could lead Exim to respond to
5342 sender verification failures with (e.g.):
5344 550 Verification failed for <xxx>
5345 550 Sender verify failed
5347 where the first response line was missing the '-' that indicates it is not
5348 the final line of the response.
5350 31. The loop for finding the name of the user that called Exim had a hardwired
5351 limit of 10; it now uses the value of finduser_retries, which is used for
5352 all other user lookups.
5354 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5357 33. Exim was neglecting to zero errno before one call of strtol() when
5358 expanding a string and expecting an integer value. On some systems this
5359 resulted in spurious "integer overflow" errors. Also, it was casting the
5360 result into an int without checking.
5362 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5363 did not work. The code looks as if it has *never* worked, though it appears
5364 to have been documented since at least releast 1.62. I have made it work.
5366 35. The "timeout_DNS" error in retry rules, also documented since at least
5367 1.62, also never worked. As it isn't clear exactly what this means, and
5368 clearly it isn't a major issue, I have abolished the feature by treating it
5369 as "timeout", and writing a warning to the main and panic logs.
5371 36. The display of retry rules for -brt wasn't always showing the error code
5374 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5375 timeout_connect_A, timeout_connect_MX.
5377 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5378 to the empty sender.
5380 39. The daemon was not analysing the content of -oX till after it had closed
5381 stderr and disconnected from the controlling terminal. This meant that any
5382 syntax errors were only noted on the panic log, and the return code from
5383 the command was 0. By re-arranging the code a little, I've made the
5384 decoding happen first, so such errors now appear on stderr, and the return
5385 code is 1. However, the actual setting up of the sockets still happens in
5386 the disconnected process, so errors there are still only recorded on the
5389 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5390 connections (as happens on some IP stacks) was logged at start up time as
5391 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5392 it from "IPv6 and IPv4", which means that two separate sockets are being
5395 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5396 says whether AF_INET or AF_INET6 was passed as an argument.
5398 42. Exiwhat output was messed up when time zones were included in log
5401 43. Exiwhat now gives more information about the daemon's listening ports,
5402 and whether -tls-on-connect was used.
5404 44. The "port" option of the smtp transport is now expanded.
5406 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5407 silently ignored. Now an error message is written to the main and panic
5410 46. There's a new ACL modifier called "logwrite" which writes to a log file
5411 as soon as it is encountered.
5413 47. Added $local_user_uid and $local_user_gid at routing time.
5415 48. Exim crashed when trying to verify a sender address that was being
5418 49. Exim was recognizing only a space character after ".include". It now also
5419 recognizes a tab character.
5421 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5422 extracting the relevant information from the specification. The man page no
5423 longer contains scrambled data for the -d option, and I've added a section
5424 at the front about calling Exim under different names.
5426 51. Added "extra_headers" argument to the "mail" command in filter files.
5428 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5431 53. Installed eximstats 1.29.
5433 54. Added transport_filter_timeout as a generic transport option.
5435 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5436 Cc: header lines. This was required by RFC 822, but it not required by RFC
5439 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5440 incoming messages that did not have them. Now it does so only if the
5441 message originates locally, that is, if there is no associated remote host
5442 address. When Resent- header lines are present, this applies to the Resent-
5443 lines rather than the non-Resent- lines.
5445 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5446 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5448 58. Messages for configuration errors now include the name of the main
5449 configuration file - useful now that there may be more than one file in a
5450 list (.included file names were always shown).
5452 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5453 for those rare installations that do not start the daemon as root or run it
5454 setuid root. I've cut out the call to initgroups() if the daemon is not
5457 60. The Exim user and group can now be bound into the binary as text strings
5458 that are looked up at the start of Exim's processing.
5460 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5462 62. Added $mailstore_basename variable.
5464 63. Installed patch to sieve.c from Michael Haardt.
5466 64. When Exim failed to open the panic log after failing to open the main log,
5467 the original message it was trying to log was written to stderr and debug
5468 output, but if they were not available (the usual case in production), it
5469 was lost. Now it is written to syslog before the two lines that record the
5470 failures to open the logs.
5472 65. Users' Exim filters run in subprocesses under the user's uid. It is
5473 possible for a "deliver" command or an alias in a "personal" command to
5474 provoke an address rewrite. If logging of address rewriting is configured,
5475 this fails because the process is not running as root or exim. There may be
5476 a better way of dealing with this, but for the moment (because 4.30 needs
5477 to be released), I have disabled address rewrite logging when running a
5478 filter in a non-root, non-exim process.
5484 1. The buildconfig auxiliary program wasn't quoting the value set for
5485 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5486 not defined. This bug was masked in 4.22 by the effect that was fixed in
5489 2. Some messages that were rejected after a message id was allocated were
5490 shown as "incomplete" by exigrep. It no longer does this for messages that
5491 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5493 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5494 have allow_domain_literals set, the ID did not get logged in the <= line.
5495 Domain literals are now always recognized in Message-ID: header lines.
5497 4. The first argument for a ${extract expansion item is the key name or field
5498 number. Leading and trailing spaces in this item were not being ignored,
5499 causing some misleading effects.
5501 5. When deliver_drop_privilege was set, single queue runner processes started
5502 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5503 same command in the process it spins off) were not dropping privilege.
5505 6. When the daemon running as "exim" started a queue runner, it always
5506 re-executed Exim in the spun-off process. This is a waste of effort when
5507 deliver_drop_privilege is set. The new process now just calls the
5508 queue-runner function directly.
5514 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5517 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5518 ignored. Though the use of -oP was forcing the writing of a pid file, it
5519 was always written to the default place.
5521 3. If the message "no IP address found for host xxxx" is generated during
5522 incoming verification, it is now followed by identification of the incoming
5523 connection (so you can more easily find what provoked it).
5525 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5527 5. Added some features to "harden" Exim a bit more against certain attacks:
5529 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5530 be put in Local/Makefile. This is like the never_users runtime option,
5531 but it cannot be overridden. The default setting is "root".
5533 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5534 prefix string with which any file named in a -C command line option
5537 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5538 is retained for -C and -D only if the caller of Exim is root. Without
5539 it, the exim user may also use -C and -D and retain privilege.
5541 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5542 command line option is disabled.
5544 6. Macro names set by the -D option must start with an upper case letter, just
5545 like macro names defined in the configuration file.
5547 7. Added "dereference=" facility to LDAP.
5549 8. Two instances of the typo "uknown" in the source files are fixed.
5551 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5552 the Configure-Makefile script screwed up while processing it.
5554 10. Incorporated PCRE 4.4.
5556 11. The SMTP synchronization check was not operating right at the start of an
5557 SMTP session. For example, it could not catch a HELO sent before the client
5558 waited for the greeting. There is now a check for outstanding input at the
5559 point when the greeting is written. Because of the duplex, asynchronous
5560 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5561 way, but not yet received, when the check is performed.
5563 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5564 on TCP/IP sockets, because this apparently causes some broken clients to
5567 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5568 unchanged) from the Cygwin maintainer.
5570 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5571 was supported instead of testing for mbx. Effectively a typo.
5573 15. The spa authenticator server code was not checking that the input it
5574 received was valid base64.
5576 16. The debug output line for the "set" modifier in ACLs was not showing the
5577 name of the variable that was being set.
5579 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5581 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5582 lookup on its own. Something else was needed to trigger the lookup. For
5583 example, a match in host_lookup or the need for a host name in a host list.
5584 Now, if $sender_host_name is referenced and the host name has not yet been
5585 looked up, a lookup is performed. If the lookup fails, the variable remains
5586 empty, and $host_lookup_failed is set to "1".
5588 19. Added "eqi" as a case-independent comparison operator.
5590 20. The saslauthd authentication condition could segfault if neither service
5591 nor realm was specified.
5593 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5594 error message that was logged was misleading, and incoming SMTP
5595 connections were dropped. The message is now more accurate, and temporary
5596 errors are given to SMTP connections.
5598 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5599 (except RSET) with a 421 error, until QUIT is received. However, it was
5600 failing to send a response to QUIT.
5602 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5603 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5604 ACL is now run after the helo_try_verify_hosts code.
5606 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5607 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5608 case-independent, so other case variants are also recognized). Apparently
5609 some systems use these upper case variants.
5611 25. If more than two messages were waiting for the same host, and a transport
5612 filter was specified for the transport, Exim sent two messages over the
5613 same TCP/IP connection, and then failed with "socket operation on non-
5614 socket" when it tried to send the third.
5616 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5618 27. The extern definition of crypt16() in expand.c was not being excluded when
5619 the OS had its own crypt16() function.
5621 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5622 as a preferred synonym for return_size_limit, both as an option and as an
5625 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5627 30. Changed the default configuration ACL to relax the local part checking rule
5628 for addresses that are not in any local domains. For these addresses,
5629 slashes and pipe symbols are allowed within local parts, but the sequence
5630 /../ is explicitly forbidden.
5632 31. SPA server authentication was not clearing the challenge buffer before
5635 32. log_message in a "warn" ACL statement was writing to the reject log as
5636 well as to the main log, which contradicts the documentation and doesn't
5637 seem right (because no rejection is happening). So I have stopped it.
5639 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5640 However, I am unable to do any testing of this.
5642 34. Fixed an infelicity in the appendfile transport. When checking directories
5643 for a mailbox, to see if any needed to be created, it was accidentally
5644 using path names with one or more superfluous leading slashes; tracing
5645 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5647 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5648 contents are added to the log line that is written for every discarded
5649 recipient. (Previously a log_message setting was ignored.)
5651 36. The ${quote: operator now quotes the string if it is empty.
5653 37. The install script runs exim in order to find its version number. If for
5654 some reason other than non-existence or emptiness, which it checks, it
5655 could not run './exim', it was installing it with an empty version number,
5656 i.e. as "exim-". This error state is now caught, and the installation is
5659 38. An argument was missing from the function that creates an error message
5660 when Exim fails to connect to the socket for saslauthd authentication.
5661 This could cause Exim to crash, or give a corrupted message.
5663 39. Added isip, isip4, and isip6 to ${if conditions.
5665 40. The ACL variables $acl_xx are now saved with the message, and can be
5666 accessed later in routers, transports, and filters.
5668 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5669 strings in the file are not string-expanded.
5671 42. If a MAIL command specified a SIZE value that was too large to fit into an
5672 int variable, the check against message_size_limit failed. Such values are
5673 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5674 day this will have to be increased, but I don't think I want to be around
5675 when emails are that large.
5682 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5683 iconv() is not standard in FreeBSD.
5685 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5686 IPv6 enabled. The observed symptom was a segmentation fault on return from
5687 the function os_common_find_running_interfaces() in src/os.c.
5689 3. In the check_special_case() function in daemon.c I had used "errno" as an
5690 argument name, which causes warnings on some systems. This was basically a
5691 typo, since it was named "eno" in the comments!
5693 4. The code that waits for the clock to tick (at a resolution of some fraction
5694 of a second) so as to ensure message-id uniqueness was always waiting for
5695 at least one whole tick, when it could have waited for less. [This is
5696 almost certainly not relevant at current processor speeds, where it is
5697 unlikely to ever wait at all. But we try to future-proof.]
5699 5. The function that sleeps for a time interval that includes fractions of a
5700 second contained a race. It did not block SIGALRM between setting the
5701 timer, and suspending (a couple of lines later). If the interval was short
5702 and the sigsuspend() was delayed until after it had expired, the suspension
5703 never ended. On busy systems this could lead to processes getting stuck for
5706 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5707 process, before it forks any delivery processes. The open lookup caching
5708 mechanism meant that the open file or database connection was passed into
5709 the delivery process. The problem was that delivery processes always tidy
5710 up cached lookup data. This could cause a problem for the next delivery
5711 process started by the queue runner, because the external queue runner
5712 process does not know about the closure. So the next delivery process
5713 still has data in the lookup cache. In the case of a file lookup, there was
5714 no problem because closing a file descriptor in a subprocess doesn't affect
5715 the parent. However, if the lookup was caching a connection to a database,
5716 the connection was closed, and the second delivery process was likely to
5717 see errors such as "PGSQL: query failed: server closed the connection
5718 unexpectedly". The problem has been fixed by closing all cached lookups
5719 in a queue runner before running a delivery process.
5721 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5722 seem to have the "const" qualifier which it has on other OS. I've
5725 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5726 *of the same type* (client or server) with the same public name that an
5727 error should be diagnosed.
5729 9. When Exim looked up a host name for an IP address, but failed to find the
5730 original IP address when looking up the host name (a safety check), it
5731 output the message "<ip address> does not match any IP for NULL", which was
5732 confusing, to say the least. The bug was that the host name should have
5733 appeared instead of "NULL".
5735 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5736 user that is built into the binary, and the -C or -D options is used, root
5737 privilege is dropped before the configuration file is read. In addition,
5738 logging is switched to stderr instead of the normal log files. If the
5739 configuration then re-defines the Exim user, the unprivileged environment
5740 is probably not what is expected, so Exim logs a panic warning message (but
5743 However, if deliver_drop_privilege is set, the unprivileged state may well
5744 be exactly what is intended, so the warning has been cut out in that case,
5745 and Exim is allowed to try to write to its normal log files.
5751 1. smtp_return_error_details was not giving details for temporary sender
5752 or receiver verification errors.
5754 2. Diagnose a configuration error if two authenticators have the same public
5757 3. Exim used not to create the message log file for a message until the first
5758 delivery attempt. This could be confusing when incoming messages were held
5759 for policy or load reasons. The message log file is now created at the time
5760 the message is received, and an initial "Received" line is written to it.
5762 4. The automatically generated man page for command line options had a minor
5763 bug that caused no ill effects; however, a more serious problem was that
5764 the procedure for building the man page automatically didn't always
5765 operate. Consequently, release 4.20 contains an out-of-date version. This
5766 shouldn't happen again.
5768 5. When building Exim with embedded Perl support, the script that builds the
5769 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5770 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5772 6. The freeze_tell option was not being used for messages that were frozen on
5773 arrival, either by an ACL or by local_scan().
5775 7. Added the smtp_incomplete_transaction log selector.
5777 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5778 was accepting AUTH without a new EHLO.
5780 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5781 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5784 10. Logging of TCP/IP connections (when configured) now happens in the main
5785 daemon process instead of the child process, so that the TCP/IP connection
5786 count is more accurate (but it can never be perfect).
5788 11. The use of "drop" in a nested ACL was not being handled correctly in the
5789 outer ACL. Now, if condition failure induced by the nested "drop" causes
5790 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5791 or "require"), the connection is dropped.
5793 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5794 that yield "discard" can now be used with an "accept" or a "discard" verb,
5795 but an error is generated for any others (because I can't see a useful way
5796 to define what should happen).
5798 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5799 are now processed in the same way as lines in the Exim configuration file.
5800 In particular, continuation lines are supported.
5802 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5804 15. Added -ti meaning -t -i.
5806 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5807 domains, and warn by logging if others are found.
5809 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5810 returned from GIFCONF when Exim is trying to find the list of interfaces on
5811 a host. The code in os.c has been modified to copy each ifreq to an aligned
5812 structure in all cases.
5814 Also, in some cases, the returned ifreq's were being copied to a 'struct
5815 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5816 means the last couple of bytes of an IPv6 address could be chopped if the
5817 ifreq contained only a normal sockaddr (14 bytes storage).
5819 18. Named domain lists were not supported in the hosts_treat_as_local option.
5820 An entry such as +xxxx was not recognized, and was treated as a literal
5823 19. Ensure that header lines added by a DATA ACL are included in the reject log
5824 if the ACL subsequently rejects the message.
5826 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5827 MD5 (which is deprecated).
5829 21. When testing a filter file using -bf, Exim was writing a message when it
5830 took the sender from a "From " line in the message, but it was not doing so
5831 when it took $return_path from a Return-Path: header line. It now does.
5833 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5834 with a valid header line field name (a series of printing characters
5835 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5837 23. Changed "disc" in the source to "disk" to conform to the documentation and
5838 the book and for uniformity.
5840 24. Ignore Sendmail's -Ooption=value command line item.
5842 25. When execve() failed while trying to run a command in a pipe transport,
5843 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5844 could be confused with a return value of 69 from the command itself. This
5845 has been changed to 127, the value the shell returns if it is asked to run
5846 a non-existent command. The wording for the related log line suggests a
5847 non-existent command as the problem.
5849 26. If received_header_text expands to an empty string, do not add a Received:
5850 header line to the message. (Well, it adds a token one on the spool, but
5851 marks it "old" so that it doesn't get used or transmitted.)
5853 27. Installed eximstats 1.28 (addition of -nt option).
5855 28. There was no check for failure on the call to getsockname() in the daemon
5856 code. This can fail if there is a shortage of resources on the system, with
5857 ENOMEM, for example. A temporary error is now given on failure.
5859 29. Contrary to the C standard, it seems that in some environments, the
5860 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5861 program. Exim now does this explicitly; it affects the formatting of
5862 timestamps using strftime().
5864 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5865 complaints. I've now initialized all the variables, to avoid this.
5867 32. Header lines added by a system filter were not being "seen" during
5868 transport-time rewrites.
5870 33. The info_callback() function passed to OpenSSL is set up with type void
5871 (*)(SSL *, int, int), as described somewhere. However, when calling the
5872 function (actually a macro) that sets it up, the type void(*)() is
5873 expected. I've put in a cast to prevent warnings from picky compilers.
5875 34. If a DNS black list lookup found a CNAME record, but there were no A
5876 records associated with the domain it pointed at, Exim crashed.
5878 35. If a DNS black list lookup returned more than one A record, Exim ignored
5879 all but the first. It now scans all returned addresses if a particular IP
5880 value is being sought. In this situation, the contents of the
5881 $dnslist_value variable are a list of all the addresses, separated by a
5884 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5885 to accept a host name and all its aliases if the forward lookup for any of
5886 them yielded the IP address of the incoming connection. Now it accepts only
5887 those names whose forward lookup yields the correct IP address. Any other
5888 names are discarded. This closes a loophole whereby a rogue DNS
5889 administrator could create reverse DNS records to break through a
5890 wildcarded host restriction in an ACL.
5892 37. If a user filter or a system filter that ran in a subprocess used any of
5893 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5894 the wrong values were passed to the pipe command ($thisaddress had the
5895 value of $0, $0 had the value of $1, etc). This bug was introduced by
5896 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5898 38. Improved the line breaking for long SMTP error messages from ACLs.
5899 Previously, if there was no break point between 40 and 75 characters, Exim
5900 left the rest of the message alone. Two changes have been made: (a) I've
5901 reduced the minimum length to 35 characters; (b) if it can't find a break
5902 point between 35 and 75 characters, it looks ahead and uses the first one
5903 that it finds. This may give the occasional overlong line, but at least the
5904 remaining text gets split now.
5906 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5907 file descriptors might be low, and that setting 1000 would always raise it.
5908 It turns out that in some environments, the limit is already over 1000 and
5909 that lowering it causes trouble. So now Exim takes care not to decrease it.
5911 40. When delivering a message, the value of $return_path is set to $sender_
5912 address at the start of routing (routers may change the value). By an
5913 oversight, this default was not being set up when an address was tested by
5914 -bt or -bv, which affected the outcome if any router or filter referred to
5917 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5918 log only when "message" or "log_message" are set. However, if one of the
5919 conditions was an address verification, or a call to a nested ACL, the
5920 messages generated by the underlying test were being passed through. This
5921 no longer happens. The underlying message is available in $acl_verify_
5922 message for both "message" and "log_message" expansions, so it can be
5923 passed through if needed.
5925 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5926 new expansion $bh_ to give the encoded byte string without charset
5927 translation. Translation happens only if iconv() is available; HAVE_ICONV
5928 indicates this at build time. HEADERS_CHARSET gives the charset to
5929 translate to; headers_charset can change it in the configuration, and
5930 "headers charset" can change it in an individual filter file.
5932 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5933 that creates RFC 2047 encoded "words" labels them as that charset instead
5934 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5935 expansion operator; (ii) when Exim creates a From: line for a local
5936 message; (iii) when a header line is rewritten to include a "phrase" part.
5938 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5939 buggy, causing it to skip the first lines of messages whose message ID
5940 ended in 'D'. This would not have bitten before Exim release 4.14, because
5941 message IDs were unlikely to end in 'D' before then. The effect was to have
5942 incorrect size information for certain domains.
5944 45. #include "config.h" was missing at the start of the crypt16.c module. This
5945 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5948 46. If there was a timeout during a "random" callout check, Exim treated it as
5949 a failure of the random address, and carried on sending RSET and the real
5950 address. If the delay was just some slowness somewhere, the response to the
5951 original RCPT would be taken as a response to RSET and so on, causing
5952 mayhem of various kinds.
5954 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5955 when I implemented it. It didn't allow for the fact that some option values
5956 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5957 the right test for positive values.
5959 48. Domain names in DNS records are case-independent. Exim always looks them up
5960 in lower case. Some resolvers return domain names in exactly the case they
5961 appear in the zone file, that is, they may contain uppercase letters. Not
5962 all resolvers do this - some return always lower case. Exim was treating a
5963 change of case by a resolver as a change of domain, similar to a widening
5964 of a domain abbreviation. This triggered its re-routing code and so it was
5965 trying to route what was effectively the same domain again. This normally
5966 caused routing to fail (because the router wouldn't handle the domain
5967 twice). Now Exim checks for this case specially, and just changes the
5968 casing of the domain that it ultimately uses when it transmits the message
5971 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5974 50. If a filter generated a file delivery with a non-absolute name (possible if
5975 no home directory exists for the router), the forbid_file option was not
5978 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5979 the existing equality matching.
5981 52. Exim was using ints instead of ino_t variables in some places where it was
5982 dealing with inode numbers.
5984 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5985 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5986 called TMPDIR, and if it finds it is different, it changes its value.
5988 54. The smtp_printf() function is now made available to local_scan() so
5989 additional output lines can be written before returning. There is also an
5990 smtp_fflush() function to enable the detection of a dropped connection.
5991 The variables smtp_input and smtp_batched_input are exported to
5994 55. Changed the default runtime configuration: the message "Unknown user"
5995 has been removed from the ACL, and instead placed on the localuser router,
5996 using the cannot_route_message feature. This means that any verification
5997 failures that generate their own messages won't get overridden. Similarly,
5998 the "Unrouteable address" message that was in the ACL for unverifiable
5999 relay addresses has also been removed.
6001 56. Added hosts_avoid_esmtp to the smtp transport.
6003 57. The exicyclog script was not checking for the esoteric option
6004 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6005 will work only if exicyclog is run under the appropriate euid.
6007 58. Following a discussion on the list, the rules by which Exim recognises line
6008 endings on incoming messages have been changed. The -dropcr and drop_cr
6009 options are now no-ops, retained only for backwards compatibility. The
6010 following line terminators are recognized: LF CRLF CR. However, special
6011 processing applies to CR:
6013 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6014 nor a local message in the state where . is a terminator.
6016 (ii) If a bare CR is encountered in a header line, an extra space is added
6017 after the line terminator so as not to end the header. The reasoning
6018 behind this is that bare CRs in header lines are most likely either
6019 to be mistakes, or people trying to play silly games.
6021 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6022 was being incorrectly given as 18 bytes larger than it should have been.
6023 This is a VOB (very old bug).
6025 60. This may never have affected anything current, but just in case it has:
6026 When the local host is found other than at the start of a list of hosts,
6027 the local host, those with the same MX, and any that follow, are discarded.
6028 When the list in question was part of a longer list of hosts, the following
6029 hosts (not currently being processed) were also being discarded. This no
6030 longer happens. I'm not sure if this situation could ever has previously
6033 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6036 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6037 header. This is something that is recommended in a new Internet Draft, and
6038 is something that is documented as being done by Sendmail. There are two
6039 possible values. For messages generated by the autoreply transport, Exim
6042 Auto-Submitted: auto-replied
6044 whereas for all other generated messages (e.g. bounces) it adds
6046 Auto-Submitted: auto-generated
6048 63. The "personal" condition in filters now includes a test for the
6049 Auto-Submitted: header. If it contains the string "auto-" the message it
6050 not considered personal.
6052 64. Added rcpt_include_affixes as a generic transport option.
6054 65. Added queue_only_override (default true).
6056 66. Added the syslog_duplication option.
6058 67. If what should have been the first header line of a message consisted of
6059 a space followed by a colon, Exim was mis-interpreting it as a header line.
6060 It isn't of course - it is syntactically invalid and should therefore be
6061 treated as the start of the message body. The misbehaviour could have
6062 caused a number of strange effects, including loss of data in subsequent
6063 header lines, and spool format errors.
6065 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6066 client host had authenticated. This control can now be exercised by an ACL
6067 for more flexibility.
6069 69. By default, callouts do not happen when testing with -bh. There is now a
6070 variant, -bhc, which does actually run the callout code, including
6071 consulting and updating the callout cache.
6073 70. Added support for saslauthd authentication, courtesy of Alexander
6076 71. If statvfs() failed on the spool or log directories while checking their
6077 size for availability, Exim confusingly gave the error "space shortage".
6078 Furthermore, in debugging mode it crashed with a floating point exception.
6079 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6080 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6081 serious problem, Exim now writes to the main and panic logs when this
6082 happens, with details of the failure. It then refuses to accept the
6083 incoming message, giving the message "spool directory problem" or "log
6084 directory problem" with a 421 code for SMTP messages.
6086 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6087 0, 1, and 2 exist, because some OS complain for execs without them (see
6088 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6089 descriptors. However, the code omitted to check that the open succeeded,
6090 causing mysterious errors if for some reason the permissions on /dev/null
6091 got screwed. Now Exim writes a message to the main and panic logs, and
6092 bombs out if it can't open /dev/null.
6094 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6095 interact so that it is all more flexible. It is supposed to remain
6096 backwards compatible. Also added extra_local_interfaces.
6098 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6099 to bomb out with an assertion failure - to the client this appears as a
6100 connection drop. This problem occurs in the part of the code that was taken
6101 from the Samba project. Fortunately, the assertion is in a very simple
6102 function, so I have fixed this by reproducing the function inline in the
6103 one place where it is called, and arranging for authentication to fail
6104 instead of killing the process with assert().
6106 75. The SPA client code was not working when the server requested OEM rather
6107 than Unicode encoding.
6109 76. Added code to make require_files with a specific uid setting more usable in
6110 the case where statting the file as root fails - usually a non-root-mounted
6111 NFS file system. When this happens and the failure is EACCES, Exim now
6112 forks a subprocess and does the per-uid checking as the relevant uid.
6114 77. Added process_log_path.
6116 78. If log_file_path was not explicitly set, a setting of check_log_space or
6117 check_log_inodes was ignored.
6119 79. If a space check for the spool or log partitions fails, the incident is now
6120 logged. Of course, in the latter case the data may get lost...
6122 80. Added the %p formatting code to string_format() so that it can be used to
6123 print addresses in debug_print(). Adjusted all the address printing in the
6124 debugging in store.c to use %p rather than %d.
6126 81. There was a concern that a line of code in smtp_in.c could overflow a
6127 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6128 initially expressed, the concern was not well-founded, because trailing
6129 spaces are removed early. However, if the trailing spaces were followed by
6130 a NULL, they did not get removed, so the overflow was possible. Two fixes
6133 (a) I re-wrote the offending code in a cleaner fashion.
6134 (b) If an incoming SMTP command contains a NULL character, it is rejected
6137 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6138 runs initgroups(), so that if the Exim user is in any additional groups,
6139 they will be used during message reception.
6145 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob