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.
49 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
50 checkers that were complaining about end of non-void function with no
53 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
54 This was a regression intruduced in 4.83 by another bugfix.
56 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
58 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
59 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
65 TF/01 Correctly close the server side of TLS when forking for delivery.
67 When a message was received over SMTP with TLS, Exim failed to clear up
68 the incoming connection properly after forking off the child process to
69 deliver the message. In some situations the subsequent outgoing
70 delivery connection happened to have the same fd number as the incoming
71 connection previously had. Exim would try to use TLS and fail, logging
72 a "Bad file descriptor" error.
74 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
75 utilities have not been installed.
77 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
78 temporary space as the ACL may create new global variables.
80 TL/01 LDAP support uses per connection or global context settings, depending
81 upon the detected version of the libraries at build time.
83 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
84 to extract and use the src ip:port in logging and expansions as if it
85 were a direct connection from the outside internet. PPv2 support was
86 updated based on HAProxy spec change in May 2014.
88 JH/02 Add ${listextract {number}{list}{success}{fail}}.
90 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
91 Properly escape header and check for NULL return.
93 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
96 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
98 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
99 characters in header names, implemented as a verify condition.
100 Contributed by Michael Fischer v. Mollard.
102 TL/05 Rename SPF condition results err_perm and err_temp to standardized
103 results permerror and temperror. Previous values are deprecated but
104 still accepted. In a future release, err_perm and err_temp will be
105 completely removed, which will be a backward incompatibility if the
106 ACL tests for either of these two old results. Patch contributed by
107 user bes-internal on the mailing list.
109 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
111 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
112 selectors, in both main and reject logs.
114 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
117 JH/07 Add malware type "sock" for talking to simple daemon.
119 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
121 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
122 routers/transports under cutthrough routing.
124 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
125 numbers. Touch up "bool" conditional to keep the same definition.
127 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
129 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
131 JH/12 Expand items in router/transport headers_add or headers_remove lists
132 individually rather than the list as a whole. Bug 1452.
134 Required for reasonable handling of multiple headers_ options when
135 they may be empty; requires that headers_remove items with embedded
136 colons must have them doubled (or the list-separator changed).
138 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
139 view the policy declared in the DMARC record. Currently, $dmarc_status
140 is a combined value of both the record presence and the result of the
143 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
145 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
146 dnslookup router and the smtp transport (applying to the forward
149 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
150 of ldap servers used for a specific lookup. Patch provided by Heiko
153 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
154 New variable $lookup_dnssec_authenticated for observability.
156 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
157 Patch submitted by Lars Timman.
159 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
161 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
162 Requires trusted mode and valid format message id, aborts otherwise.
163 Patch contributed by Heiko Schlichting.
165 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
166 certextract with support for various fields. Bug 1358.
168 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
169 is requested by default, modifiable by smtp transport option
172 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
173 operate on certificate variables to give certificate fingerprints
174 Also new ${sha256:cert_variable}.
176 JH/23 The PRDR feature is moved from being Experimental into the mainline.
178 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
179 Christian Aistleitner.
181 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
183 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
184 file. Patch from Wolfgang Breyha.
186 JH/25 Expand the coverage of the delivery $host and $host_address to
187 client authenticators run in verify callout. Bug 1476.
189 JH/26 Port service names are now accepted for tls_on_connect_ports, to
190 align with daemon_smtp_ports. Bug 72.
192 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
193 support and error reporting did not work properly.
195 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
196 and is readable. Patch from Andrew Colin Kissa.
198 TL/14 Enhance documentation of ${run expansion and how it parses the
199 commandline after expansion, particularly in the case when an
200 unquoted variable expansion results in an empty value.
202 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
204 PP/02 Fix internal collision of T_APL on systems which support RFC3123
205 by renaming away from it. Addresses GH issue 15, reported by
208 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
210 TL/15 SECURITY: prevent double expansion in math comparison functions
211 (can expand unsanitized data). Not remotely exploitable.
218 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
220 PP/02 Make -n do something, by making it not do something.
221 When combined with -bP, the name of an option is not output.
223 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
226 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
227 $sender_host_name and config options to manage this, and basic check
230 PP/05 DSCP support for outbound connections and control modifier for inbound.
232 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
233 (Only plugin which currently uses this is kerberos4, which nobody should
234 be using, but we should make it available and other future plugins might
235 conceivably use it, even though it would break NAT; stuff *should* be
236 using channel bindings instead).
238 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
239 name; added for Sendmail compatibility; requires admin caller.
240 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
241 just ignore it); requires trusted caller.
242 Also parse but ignore: -Ac -Am -X<logfile>
245 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
247 TL/02 Add +smtp_confirmation as a default logging option.
249 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
250 Patch by Magnus Holmgren from 2007-02-20.
252 TL/04 Bugzilla 1281 - Spec typo.
253 Bugzilla 1283 - Spec typo.
254 Bugzilla 1290 - Spec grammar fixes.
256 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
258 TL/06 Add Experimental DMARC support using libopendmarc libraries.
260 TL/07 Fix an out of order global option causing a segfault. Reported to dev
261 mailing list by by Dmitry Isaikin.
263 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
265 JH/02 Support "G" suffix to numbers in ${if comparisons.
267 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
269 NM/01 Bugzilla 1197 - Spec typo
270 Bugzilla 1196 - Spec examples corrections
272 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
274 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
275 gnutls_enable_pkcs11, but renamed to more accurately indicate its
278 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
279 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
281 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
282 "acl {{name}{arg}...}", and optional args on acl condition
285 JH/05 Permit multiple router/transport headers_add/remove lines.
287 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
289 JH/07 Avoid using a waiting database for a single-message-only transport.
290 Performance patch from Paul Fisher. Bugzilla 1262.
292 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
295 JH/09 Add $headers_added variable, with content from use of ACL modifier
296 add_header (but not yet added to the message). Bugzilla 199.
298 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
299 Pulled from Bugzilla 817 by Wolfgang Breyha.
301 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
303 (nb: this is the same fix as in Exim 4.80.1)
305 JH/11 Add A= logging on delivery lines, and a client_set_id option on
308 JH/12 Add optional authenticated_sender logging to A= and a log_selector
311 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
313 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
314 advertise SMTP AUTH mechanism to us, instead of a generic
315 protocol violation error. Also, make Exim more robust to bad
316 data from the Dovecot auth socket.
318 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
320 When a queue runner is handling a message, Exim first routes the
321 recipient addresses, during which it prunes them based on the retry
322 hints database. After that it attempts to deliver the message to
323 any remaining recipients. It then updates the hints database using
326 So if a recipient address works intermittently, it can get repeatedly
327 deferred at routing time. The retry hints record remains fresh so the
328 address never reaches the final cutoff time.
330 This is a fairly common occurrence when a user is bumping up against
331 their storage quota. Exim had some logic in its local delivery code
332 to deal with this. However it did not apply to per-recipient defers
333 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
335 This change adds a proper retry rule check during routing so that the
336 final cutoff time is checked against the message's age. We only do
337 this check if there is an address retry record and there is not a
338 domain retry record; this implies that previous attempts to handle
339 the address had the retry_use_local_parts option turned on. We use
340 this as an approximation for the destination being like a local
341 delivery, as in LMTP.
343 I suspect this new check makes the old local delivery cutoff check
344 redundant, but I have not verified this so I left the code in place.
346 TF/02 Correct gecos expansion when From: is a prefix of the username.
348 Test 0254 submits a message to Exim with the header
352 When I ran the test suite under the user fanf2, Exim expanded
353 the header to contain my full name, whereas it should have added
354 a Resent-Sender: header. It erroneously treats any prefix of the
355 username as equal to the username.
357 This change corrects that bug.
359 GF/01 DCC debug and logging tidyup
360 Error conditions log to paniclog rather than rejectlog.
361 Debug lines prefixed by "DCC: " to remove any ambiguity.
363 TF/03 Avoid unnecessary rebuilds of lookup-related code.
365 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
366 Bug spotted by Jeremy Harris; was flawed since initial commit.
367 Would have resulted in OCSP responses post-SNI triggering an Exim
368 NULL dereference and crash.
370 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
372 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
373 Bug detection, analysis and fix by Samuel Thibault.
374 Bugzilla 1331, Debian bug #698092.
376 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
378 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
379 Server implementation by Todd Lyons, client by JH.
380 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
381 config variable "prdr_enable" controls whether the server
382 advertises the facility. If the client requests PRDR a new
383 acl_data_smtp_prdr ACL is called once for each recipient, after
384 the body content is received and before the acl_smtp_data ACL.
385 The client is controlled by bolth of: a hosts_try_prdr option
386 on the smtp transport, and the server advertisement.
387 Default client logging of deliveries and rejections involving
388 PRDR are flagged with the string "PRDR".
390 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
391 fclose(). Diagnosis by Todd Lyons.
393 PP/17 Update configure.default to handle IPv6 localhost better.
394 Patch by Alain Williams (plus minor tweaks).
397 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
398 This is now consistent with GnuTLS, and is now documented: the
399 previous undocumented portable approach to treating the option as
400 unset was to force an expansion failure. That still works, and
401 an empty string is now equivalent.
403 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
404 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
405 not performing validation itself.
407 PP/20 Added force_command boolean option to pipe transport.
408 Patch from Nick Koston, of cPanel Inc.
410 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
413 TF/04 Added udpsend ACL modifer and hexquote expansion operator
415 PP/21 Fix eximon continuous updating with timestamped log-files.
416 Broken in a format-string cleanup in 4.80, missed when I repaired the
417 other false fix of the same issue.
418 Report and fix from Heiko Schlichting.
421 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
422 Report from Prashanth Katuri.
424 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
425 It's SecureTransport, so affects any MacOS clients which use the
426 system-integrated TLS libraries, including email clients.
428 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
429 using a MIME ACL for non-SMTP local injection.
430 Report and assistance in diagnosis by Warren Baker.
432 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
434 JH/16 Fix comparisons for 64b. Bugzilla 1385.
436 TL/09 Add expansion variable $authenticated_fail_id to keep track of
437 last id that failed so it may be referenced in subsequent ACL's.
439 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
442 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
443 ldap library initialization, allowing self-signed CA's to be
444 used. Also properly sets require_cert option later in code by
445 using NULL (global ldap config) instead of ldap handle (per
446 session). Bug diagnosis and testing by alxgomz.
448 TL/12 Enhanced documentation in the ratelimit.pl script provided in
449 the src/util/ subdirectory.
451 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
452 renamed to Transport Post Delivery Action by Jeremy Harris, as
455 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
456 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
457 redis_servers = needs to be configured which will be used by the redis
458 lookup. Patch from Warren Baker, of The Packet Hub.
460 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
462 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
463 hostname or reverse DNS when processing a host list. Used suggestions
464 from multiple comments on this bug.
466 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
468 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
469 Missed a few lines, added it to make the runtest require no keyboard
472 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
473 contains upper case chars. Make router use caseful_local_part.
475 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
476 support when GnuTLS has been built with p11-kit.
482 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
484 This, or similar/improved, will also be change PP/11 of 4.82.
490 PP/01 Handle short writes when writing local log-files.
491 In practice, only affects FreeBSD (8 onwards).
492 Bugzilla 1053, with thanks to Dmitry Isaikin.
494 NM/01 Bugzilla 949 - Documentation tweak
496 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
499 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
501 PP/02 Implemented gsasl authenticator.
503 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
505 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
506 `pkg-config foo` for cflags/libs.
508 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
509 with rest of GSASL and with heimdal_gssapi.
511 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
512 `pkg-config foo` for cflags/libs for the TLS implementation.
514 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
515 properties get this fed in as external SSF. A number of robustness
516 and debugging improvements to the cyrus_sasl authenticator.
518 PP/08 cyrus_sasl server now expands the server_realm option.
520 PP/09 Bugzilla 1214 - Log authentication information in reject log.
521 Patch by Jeremy Harris.
523 PP/10 Added dbmjz lookup type.
525 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
527 PP/12 MAIL args handles TAB as well as SP, for better interop with
528 non-compliant senders.
529 Analysis and variant patch by Todd Lyons.
531 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
532 Bug report from Lars Müller <lars@samba.org> (via SUSE),
533 Patch from Dirk Mueller <dmueller@suse.com>
535 PP/13 tls_peerdn now print-escaped for spool files.
536 Observed some $tls_peerdn in wild which contained \n, which resulted
537 in spool file corruption.
539 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
540 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
541 or write after TLS renegotiation, which otherwise led to messages
544 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
545 as a tracking header (ie: a signed header comes before the signature).
546 Patch from Wolfgang Breyha.
548 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
549 comma-sep list; embedded commas doubled.
551 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
553 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
555 Report and patch from Dmitry Banschikov.
557 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
558 Removed SSL_clear() after SSL_new() which led to protocol negotiation
559 failures. We appear to now support TLS1.1+ with Exim.
561 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
562 lets Exim select keys and certificates based upon TLS SNI from client.
563 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
564 before an outbound SMTP session. New log_selector, +tls_sni.
566 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
567 NULL dereference. Report and patch from Alun Jones.
569 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
570 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
571 Not seeing resolver debug output on NetBSD, but suspect this is a
572 resolver implementation change.
574 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
575 Left warnings. Added "eximon gdb" invocation mode.
577 PP/21 Defaulting "accept_8bitmime" to true, not false.
579 PP/22 Added -bw for inetd wait mode support.
581 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
582 locate the relevant includes and libraries. Made this the default.
584 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
585 Bugzilla 1246, report and most of solution from Tomasz Kusy.
587 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
588 This may cause build issues on older platforms.
590 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
591 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
592 gnutls_require_mac & gnutls_require_protocols (no longer supported).
593 Added SNI support via GnuTLS too.
594 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
596 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
598 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
599 Applied second patch from Janne, implementing suggestion to default
600 multiple-strings-in-record handling to match SPF spec.
602 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
604 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
605 read-only, out of scope).
606 Patch from Wolfgang Breyha, report from Stuart Northfield.
608 PP/29 Fix three issues highlighted by clang analyser static analysis.
609 Only crash-plausible issue would require the Cambridge-specific
610 iplookup router and a misconfiguration.
611 Report from Marcin Mirosław.
613 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
615 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
616 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
617 As part of this, removing so much warning spew let me fix some minor
618 real issues in debug logging.
620 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
621 assignment on my part. Fixed.
623 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
624 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
625 Janne Snabb (who went above and beyond: thank you).
627 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
628 string otherwise requires a connection and a bunch more work and it's
629 relatively easy to get wrong. Should also expose TLS library linkage
632 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
633 64-bit ${eval} (JH/03).
635 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
636 GNU libc to support some of the 64-bit stuff, should not lead to
637 conflicts. Defined before os.h is pulled in, so if a given platform
638 needs to override this, it can.
640 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
641 protection layer was required, which is not implemented.
642 Bugzilla 1254, patch from Wolfgang Breyha.
644 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
645 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
646 tls_dhparam take prime identifiers. Also unbreak combination of
647 OpenSSL+DH_params+TLSSNI.
649 PP/39 Disable SSLv2 by default in OpenSSL support.
655 PP/01 Solaris build fix for Oracle's LDAP libraries.
656 Bugzilla 1109, patch from Stephen Usher.
658 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
660 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
663 TF/02 Fix a couple more cases where we did not log the error message
664 when unlink() failed. See also change 4.74-TF/03.
666 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
667 lock up or crash if it happened to be inside a call to libc when it
668 got a SIGUSR1 from exiwhat.
670 The SIGUSR1 handler appends the current process status to the process
671 log which is later printed by exiwhat. It used to use the general
672 purpose logging code to do this, but several functions it calls are
673 not safe for signals.
675 The new output code in the SIGUSR1 handler is specific to the process
676 log, and simple enough that it's easy to inspect for signal safety.
677 Removing some special cases also simplifies the general logging code.
678 Removing the spurious timestamps from the process log simplifies
681 TF/04 Improved ratelimit ACL condition.
683 The /noupdate option has been deprecated in favour of /readonly which
684 has clearer semantics. The /leaky, /strict, and /readonly update modes
685 are mutually exclusive. The update mode is no longer included in the
686 database key; it just determines when the database is updated. (This
687 means that when you upgrde Exim will forget old rate measurements.)
689 Exim now checks that the per_* options are used with an update mode that
690 makes sense for the current ACL. For example, when Exim is processing a
691 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
692 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
693 must specify per_mail/readonly. If you omit the update mode it defaults to
694 /leaky where that makes sense (as before) or /readonly where required.
696 The /noupdate option is now undocumented but still supported for
697 backwards compatibility. It is equivalent to /readonly except that in
698 ACLs where /readonly is required you may specify /leaky/noupdate or
699 /strict/noupdate which are treated the same as /readonly.
701 A useful new feature is the /count= option. This is a generalization
702 of the per_byte option, so that you can measure the throughput of other
703 aggregate values. For example, the per_byte option is now equivalent
704 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
706 The per_rcpt option has been generalized using the /count= mechanism
707 (though it's more complicated than the per_byte equivalence). When it is
708 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
709 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
710 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
711 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
712 Note that using per_rcpt with a non-readonly update mode in more than
713 one ACL will cause the recipients to be double-counted. (The per_mail
714 and per_byte options don't have this problem.)
716 The handling of very low rates has changed slightly. If the computed rate
717 is less than the event's count (usually one) then this event is the first
718 after a long gap. In this case the rate is set to the same as this event's
719 count, so that the first message of a spam run is counted properly.
721 The major new feature is a mechanism for counting the rate of unique
722 events. The new per_addr option counts the number of different
723 recipients that someone has sent messages to in the last time period. It
724 behaves like per_rcpt if all the recipient addresses are different, but
725 duplicate recipient addresses do not increase the measured rate. Like
726 the /count= option this is a general mechanism, so the per_addr option
727 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
728 example, measure the rate that a client uses different sender addresses
729 with the options per_mail/unique=$sender_address. There are further
730 details in the main documentation.
732 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
734 TF/06 Removed a few PCRE remnants.
736 TF/07 Automatically extract Exim's version number from tags in the git
737 repository when doing development or release builds.
739 PP/02 Raise smtp_cmd_buffer_size to 16kB.
740 Bugzilla 879. Patch from Paul Fisher.
742 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
743 Heavily based on revision 40f9a89a from Simon Arlott's tree.
746 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
748 PP/05 Variable $av_failed, true if the AV scanner deferred.
749 Bugzilla 1078. Patch from John Horne.
751 PP/06 Stop make process more reliably on build failure.
752 Bugzilla 1087. Patch from Heiko Schlittermann.
754 PP/07 Make maildir_use_size_file an _expandable_ boolean.
755 Bugzilla 1089. Patch from Heiko Schlittermann.
757 PP/08 Handle ${run} returning more data than OS pipe buffer size.
758 Bugzilla 1131. Patch from Holger Weiß.
760 PP/09 Handle IPv6 addresses with SPF.
761 Bugzilla 860. Patch from Wolfgang Breyha.
763 PP/10 GnuTLS: support TLS 1.2 & 1.1.
765 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
768 PP/11 match_* no longer expand right-hand-side by default.
769 New compile-time build option, EXPAND_LISTMATCH_RHS.
770 New expansion conditions, "inlist", "inlisti".
772 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
774 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
776 PP/14 fix log_write() format string regression from TF/03.
777 Bugzilla 1152. Patch from Dmitry Isaikin.
783 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
785 PP/02 Harmonised TLS library version reporting; only show if debugging.
786 Layout now matches that introduced for other libraries in 4.74 PP/03.
788 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
790 PP/04 New "dns_use_edns0" global option.
792 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
795 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
796 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
798 TK/01 Updated PolarSSL code to 0.14.2.
799 Bugzilla 1097. Patch from Andreas Metzler.
801 PP/07 Catch divide-by-zero in ${eval:...}.
804 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
807 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
808 format-string attack -- SECURITY: remote arbitrary code execution.
810 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
811 time unintentionally subject to list matching rules, letting the header
812 cause arbitrary Exim lookups (of items which can occur in lists, *not*
813 arbitrary string expansion). This allowed for information disclosure.
815 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
816 INT_MIN/-1 -- value coerced to INT_MAX.
822 NM/01 Workround for PCRE version dependency in version reporting
825 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
826 This fixes portability to compilers other than gcc, notably
827 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
829 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
830 makefiles for portability to HP-UX and POSIX correctness.
832 PP/01 Permit LOOKUP_foo enabling on the make command-line.
833 Also via indented variable definition in the Makefile.
834 (Debugging by Oliver Heesakkers).
836 PP/02 Restore caching of spamd results with expanded spamd_address.
837 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
839 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
840 Improves build reliability. Fix from: Frank Elsner
842 NM/02 Fix wide character breakage in the rfc2047 coding
843 Fixes bug 1064. Patch from Andrey N. Oktyabrski
845 NM/03 Allow underscore in dnslist lookups
846 Fixes bug 1026. Patch from Graeme Fowler
848 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
849 Code patches from Adam Ciarcinski of NetBSD.
851 NM/04 Fixed exiqgrep to cope with mailq missing size issue
854 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
855 is logged, to avoid truncation. Patch from John Horne.
857 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
858 Patch from Jakob Hirsch.
860 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
861 SQL string expansion failure details.
862 Patch from Andrey Oktyabrski.
864 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
865 Patch from Simon Arlott.
867 PP/09 New lookups functionality failed to compile on old gcc which rejects
868 extern declarations in function scope.
869 Patch from Oliver Fleischmann
871 PP/10 Use sig_atomic_t for flags set from signal handlers.
872 Check getgroups() return and improve debugging.
873 Fixed developed for diagnosis in bug 927 (which turned out to be
876 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
877 Patch from Mark Zealey.
879 PP/12 Bugzilla 1056: Improved spamd server selection.
880 Patch from Mark Zealey.
882 PP/13 Bugzilla 1086: Deal with maildir quota file races.
883 Based on patch from Heiko Schlittermann.
885 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
886 Patch from Uwe Doering, sign-off by Michael Haardt.
888 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
889 variable declaration deep within a block. Bug and patch from
892 PP/15 lookups-Makefile IRIX compatibilty coercion.
894 PP/16 Make DISABLE_DKIM build knob functional.
896 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
897 Patch by Simon Arlott
899 TF/03 Fix valgrind.h portability to C89 compilers that do not support
900 variable argument macros. Our copy now differs from upstream.
906 TF/01 Failure to get a lock on a hints database can have serious
907 consequences so log it to the panic log.
909 TF/02 Log LMTP confirmation messages in the same way as SMTP,
910 controlled using the smtp_confirmation log selector.
912 TF/03 Include the error message when we fail to unlink a spool file.
914 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
915 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
916 for maintaining out-of-tree patches for some time.
918 PP/01 Bugzilla 139: Documentation and portability issues.
919 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
920 Handle per-OS dynamic-module compilation flags.
922 PP/02 Let /dev/null have normal permissions.
923 The 4.73 fixes were a little too stringent and complained about the
924 permissions on /dev/null. Exempt it from some checks.
925 Reported by Andreas M. Kirchwitz.
927 PP/03 Report version information for many libraries, including
928 Exim version information for dynamically loaded libraries. Created
929 version.h, now support a version extension string for distributors
930 who patch heavily. Dynamic module ABI change.
932 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
933 privilege escalation vulnerability whereby the Exim run-time user
934 can cause root to append content of the attacker's choosing to
937 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
940 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
941 If dropping privileges for untrusted macros, we disabled normal logging
942 on the basis that it would fail; for the Exim run-time user, this is not
943 the case, and it resulted in successful deliveries going unlogged.
944 Fixed. Reported by Andreas Metzler.
950 PP/01 Date: & Message-Id: revert to normally being appended to a message,
951 only prepend for the Resent-* case. Fixes regression introduced in
952 Exim 4.70 by NM/22 for Bugzilla 607.
954 PP/02 Include check_rfc2047_length in configure.default because we're seeing
955 increasing numbers of administrators be bitten by this.
957 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
959 PP/03 Bugzilla 994: added openssl_options main configuration option.
961 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
963 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
965 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
967 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
968 without a peer certificate, leading to a segfault because of an
969 assumption that peers always have certificates. Be a little more
970 paranoid. Problem reported by Martin Tscholak.
972 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
973 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
974 NB: ClamAV planning to remove STREAM in "middle of 2010".
975 CL also introduces -bmalware, various -d+acl logging additions and
976 more caution in buffer sizes.
978 PP/09 Implemented reverse_ip expansion operator.
980 PP/10 Bugzilla 937: provide a "debug" ACL control.
982 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
984 PP/12 Bugzilla 973: Implement --version.
986 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
988 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
990 PP/15 Bugzilla 816: support multiple condition rules on Routers.
992 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
993 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
994 ignore trailing whitespace.
996 JJ/02 prevent non-panic DKIM error from being sent to paniclog
998 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1001 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1002 Notification from Dr Andrew Aitchison.
1004 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1005 ExtendedDetectionInfo response format.
1006 Notification from John Horne.
1008 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1011 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1012 XSL and documented dependency on system catalogs, with examples of how
1015 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1018 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1019 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1020 configuration file which is writeable by the Exim user or group.
1022 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1023 of configuration files to cover files specified with the -C option if
1024 they are going to be used with root privileges, not just the default
1027 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1028 option (effectively making it always true).
1030 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1031 files to be used while preserving root privileges.
1033 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1034 that rogue child processes cannot use them.
1036 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1037 run-time user, instead of root.
1039 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1040 Exim run-time user without dropping privileges.
1042 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1043 result string, instead of calling string_vformat() twice with the same
1046 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1047 for other users. Others should always drop root privileges if they use
1048 -C on the command line, even for a whitelisted configure file.
1050 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1052 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1058 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1059 $data_path, and $header_path variables; fixed documentation bugs and
1062 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1063 exipick to access non-standard spools, including the "frozen" queue
1066 NM/01 Bugzilla 965: Support mysql stored procedures.
1067 Patch from Alain Williams
1069 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1071 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1072 Patch from Andreas Metzler
1074 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1075 Patch from Kirill Miazine
1077 NM/05 Bugzilla 671: Added umask to procmail example.
1079 JJ/03 installed exipick 20100323.0, fixing doc bug
1081 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1082 directory. Notification and patch from Dan Rosenberg.
1084 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1086 TK/02 Improve log output when DKIM signing operation fails.
1088 MH/01 Treat the transport option dkim_domain as a colon separated
1089 list, not as a single string, and sign the message with each element,
1090 omitting multiple occurences of the same signer.
1092 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1093 Bugzilla 985, 986. Patch by Simon Arlott
1095 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1096 Patch by Simon Arlott
1098 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1099 MBX locking. Notification from Dan Rosenberg.
1105 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1107 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1109 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1111 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1113 NM/04 Bugzilla 915: Flag broken perl installation during build.
1119 TK/01 Added patch by Johannes Berg that expands the main option
1120 "spamd_address" if it starts with a dollar sign.
1122 TK/02 Write list of recipients to X-Envelope-Sender header when building
1123 the mbox-format spool file for content scanning (suggested by Jakob
1126 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1127 (http://www.dcc-servers.net/) support via dccifd. Activated by
1128 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1130 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1131 by Mark Daniel Reidel <mr@df.eu>.
1133 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1134 When building exim an external PCRE library is now needed -
1135 PCRE is a system library on the majority of modern systems.
1136 See entry on PCRE_LIBS in EDITME file.
1138 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1139 conversation. Added nologin parameter to request.
1140 Patch contributed by Kirill Miazine.
1142 TF/01 Do not log submission mode rewrites if they do not change the address.
1144 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1146 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1147 log files in place. Contributed by Roberto Lima.
1149 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1151 TF/03 Bugzilla 615: When checking the local_parts router precondition
1152 after a local_part_suffix or local_part_prefix option, Exim now
1153 does not use the address's named list lookup cache, since this
1154 contains cached lookups for the whole local part.
1156 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1157 Robert Millan. Documentation is in experimental-spec.txt.
1159 TF/04 Bugzilla 668: Fix parallel build (make -j).
1161 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1163 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1164 Patch provided by Jan Srzednicki.
1166 TF/05 Leading white space used to be stripped from $spam_report which
1167 wrecked the formatting. Now it is preserved.
1169 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1170 that they are available at delivery time.
1172 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1174 TF/08 TLS error reporting now respects the incoming_interface and
1175 incoming_port log selectors.
1177 TF/09 Produce a more useful error message if an SMTP transport's hosts
1178 setting expands to an empty string.
1180 NM/06 Bugzilla 744: EXPN did not work under TLS.
1181 Patch provided by Phil Pennock.
1183 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1184 Patch provided by Richard Godbee.
1186 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1187 acl_smtp_notquit, added index entry.
1189 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1190 Patch provided by Eugene Bujak.
1192 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1193 accept(). Patch provided by Maxim Dounin.
1195 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1196 Patch provided by Phil Pennock.
1198 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1200 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1201 Patch provided by Brad "anomie" Jorsch.
1203 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1204 Patch provided by Dean Brooks.
1206 TK/05 Add native DKIM support (does not depend on external libraries).
1208 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1209 Patch provided by Graeme Fowler.
1211 NM/16 Bugzilla 851: Documentation example syntax fix.
1213 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1215 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1218 NM/19 Bugzilla 745: TLS version reporting.
1219 Patch provided by Phil Pennock.
1221 NM/20 Bugzilla 167: bool: condition support.
1222 Patch provided by Phil Pennock.
1224 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1225 clients. Patch provided by Phil Pennock.
1227 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1228 Patch provided by Brad "anomie" Jorsch.
1230 NM/23 Bugzilla 687: Fix misparses in eximstats.
1231 Patch provided by Heiko Schlittermann.
1233 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1234 Patch provided by Heiko Schlittermann.
1236 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1237 plus update to original patch.
1239 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1241 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1242 Patch provided by David Brownlee.
1244 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1246 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1248 NM/30 Bugzilla 888: TLS documentation bugfixes.
1250 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1252 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1253 Unlike the original bugzilla I have changed all shell scripts in src tree.
1255 NM/33 Bugzilla 898: Transport filter timeout fix.
1256 Patch by Todd Rinaldo.
1258 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1259 Patch by Serge Demonchaux.
1261 NM/35 Bugzilla 39: Base64 decode bug fixes.
1262 Patch by Jakob Hirsch.
1264 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1266 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1268 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1270 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1276 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1277 ALT-N's libdkim that I have put here:
1278 http://duncanthrax.net/exim-experimental/
1280 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1281 were called 'true' and it seems that C99 defines that as a reserved
1282 keyword to be used with 'bool' variable types. That means you could
1283 not include C99-style headers which use bools without triggering
1284 build errors in sieve.c.
1286 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1287 as mailq or other aliases. Changed the --help handling significantly
1288 to do whats expected. exim_usage() emits usage/help information.
1290 SC/01 Added the -bylocaldomain option to eximstats.
1292 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1294 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1296 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1302 PH/01 Another patch from the Sieve maintainer.
1304 PH/02 When an IPv6 address is converted to a string for single-key lookup
1305 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1306 dots are used instead of colons so that keys in lsearch files need not
1307 contain colons. This was done some time before quoting was made available
1308 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1309 (notated using the quote facility) so as to distinguish them from IPv4
1310 keys. This meant that lookups for IP addresses in host lists did not work
1311 for iplsearch lookups.
1313 This has been fixed by arranging for IPv6 addresses to be expressed with
1314 colons if the lookup type is iplsearch. This is not incompatible, because
1315 previously such lookups could never work.
1317 The situation is now rather anomolous, since one *can* have colons in
1318 ordinary lsearch keys. However, making the change in all cases is
1319 incompatible and would probably break a number of configurations.
1321 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1324 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1325 conversion specification without a maximum field width, thereby enabling
1326 a rogue spamd server to cause a buffer overflow. While nobody in their
1327 right mind would setup Exim to query an untrusted spamd server, an
1328 attacker that gains access to a server running spamd could potentially
1329 exploit this vulnerability to run arbitrary code as the Exim user.
1331 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1332 $primary_hostname instead of what libspf2 thinks the hosts name is.
1334 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1335 a directory entry by the name of the lookup key. Previously, if a
1336 symlink pointed to a non-existing file or a file in a directory that
1337 Exim lacked permissions to read, a lookup for a key matching that
1338 symlink would fail. Now it is enough that a matching directory entry
1339 exists, symlink or not. (Bugzilla 503.)
1341 PH/03 The body_linecount and body_zerocount variables are now exported in the
1344 PH/04 Added the $dnslist_matched variable.
1346 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1347 This means they are set thereafter only if the connection becomes
1350 PH/06 Added the client_condition to authenticators so that some can be skipped
1351 by clients under certain conditions.
1353 PH/07 The error message for a badly-placed control=no_multiline_responses left
1354 "_responses" off the end of the name.
1356 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1358 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1359 (without spaces) instead of just copying the configuration text.
1361 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1363 PH/11 Added $max_received_linelength.
1365 PH/12 Added +ignore_defer and +include_defer to host lists.
1367 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1368 way in which PCRE > 7.0 is built.
1370 PH/14 Implemented queue_only_load_latch.
1372 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1373 MAIL command. The effect was to mangle the value on 64-bit systems.
1375 PH/16 Another patch from the Sieve maintainer.
1377 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1379 PH/18 If a system quota error occurred while trying to create the file for
1380 a maildir delivery, the message "Mailbox is full" was not appended to the
1381 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1382 only to a quota excession during the actual writing of the file.
1384 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1385 characters?) which causes problems in log lines. The DN values are now
1386 passed through string_printing() before being added to log lines.
1388 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1389 and InterBase are left for another time.)
1391 PH/21 Added message_body_newlines option.
1393 PH/22 Guard against possible overflow in moan_check_errorcopy().
1395 PH/23 POSIX allows open() to be a macro; guard against that.
1397 PH/24 If the recipient of an error message contained an @ in the local part
1398 (suitably quoted, of course), incorrect values were put in $domain and
1399 $local_part during the evaluation of errors_copy.
1405 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1406 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1409 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1410 issue a MAIL command.
1412 PH/02 In an ACL statement such as
1414 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1416 if a client was not listed at all, or was listed with a value other than
1417 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1418 the condition was not true (as it should be), so access was not denied.
1419 The bug was that the ! inversion was incorrectly passed on to the second
1420 item. This has been fixed.
1422 PH/03 Added additional dnslists conditions == and =& which are different from
1423 = and & when the dns lookup returns more than one IP address.
1425 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1426 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1428 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1429 FSYNC, which compiles an option called disable_fsync that allows for
1430 bypassing fsync(). The documentation is heavily laced with warnings.
1432 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1434 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1435 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1436 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1437 including adding "make clean"; (3) Added -fPIC when compiling the test
1438 dynamically loaded module, to get rid of a warning.
1440 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1441 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1442 = 0s. The bug is otherwise harmless.
1444 PH/07 There was a bug in the dovecot authenticator such that the value of
1445 $auth1 could be overwritten, and so not correctly preserved, after a
1446 successful authentication. This usually meant that the value preserved by
1447 the server_setid option was incorrect.
1449 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1451 PH/09 Installed PCRE release 7.0.
1453 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1454 run for batched SMTP input. It is now run at the start of every message
1455 in the batch. While fixing this I discovered that the process information
1456 (output by running exiwhat) was not always getting set for -bs and -bS
1457 input. This is fixed, and it now also says "batched" for BSMTP.
1459 PH/11 Added control=no_pipelining.
1461 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1462 patch, slightly modified), and move the expansion of helo_data till after
1463 the connection is made in the smtp transport (so it can use these
1466 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1468 PH/14 Added log_selector = +pid.
1470 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1472 PH/16 Add ${if forany and ${if forall.
1474 PH/17 Added dsn_from option to vary the From: line in DSNs.
1476 PH/18 Flush SMTP output before performing a callout, unless control =
1477 no_callout_flush is set.
1479 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1480 was true (the default) a successful delivery failed to delete the retry
1481 item, thus causing premature timeout of the address. The bug is now
1484 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1486 PH/21 Long custom messages for fakedefer and fakereject are now split up
1487 into multiline reponses in the same way that messages for "deny" and
1488 other ACL rejections are.
1490 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1491 with slight modification.
1493 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1494 draft, changing the syntax and factoring some duplicate code".
1496 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1497 for deliveries of the second and subsequent messages over the same SMTP
1500 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1501 ${reduce, with only minor "tidies".
1503 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1505 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1506 expansion side effects.
1508 PH/27 When a message times out after an over-quota error from an Exim-imposed
1509 quota, the bounce message says "mailbox is full". This message was not
1510 being given when it was a system quota that was exceeded. It now should
1513 MH/03 Made $recipients available in local_scan(). local_scan() already has
1514 better access to the recipient list through recipients_list[], but
1515 $recipients can be useful in postmaster-provided expansion strings.
1517 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1518 in the case of a MAIL command with additional options following the
1519 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1520 were accidentally chopped off.
1522 PH/29 SMTP synchronization checks are implemented when a command is read -
1523 there is a check that no more input is waiting when there shouldn't be
1524 any. However, for some commands, a delay in an ACL can mean that it is
1525 some time before the response is written. In this time, more input might
1526 arrive, invalidly. So now there are extra checks after an ACL has run for
1527 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1528 pipelining has not been advertised.
1530 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1532 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1533 returned in the SMTP response when smtp_return_error_details was false.
1534 This has been fixed.
1536 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1537 instead of the C I/O that was originally supplied, because problems were
1538 reported on Solaris.
1540 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1541 Exim which did not show up earlier: it was assuming that a call to
1542 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1543 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1544 was a macro that became an assignment, so it seemed to work. This has
1545 changed to a proper function call with a void return, hence the compile
1546 error. Exim's code has been fixed.
1548 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1551 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1553 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1555 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1556 because it turns out to be too big or there is a timeout) there is no
1557 "Completed" line in the log. When some messages of this type were
1558 selected by exigrep, they were listed as "not completed". Others were
1559 picked up by some special patterns. I have improved the selection
1560 criteria to be more general.
1562 PH/38 The host_find_failed option in the manualroute router can now be set
1563 to "ignore", to completely ignore a host whose IP address cannot be
1564 found. If all hosts are ignored, the behaviour is controlled by the new
1565 host_all_ignored option.
1567 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1568 homing or because of multiple MX records with /mx) generated more than
1569 one IP address, and the following item turned out to be the local host,
1570 all the secondary addresses of the first item were incorrectly removed
1571 from the list, along with the local host and any following hosts (which
1572 is what is supposed to happen).
1574 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1575 whoever called Exim into the -H file. In the case of the daemon it was
1576 behaving confusingly. When first started, it used values for whoever
1577 started the daemon, but after a SIGHUP it used the Exim user (because it
1578 calls itself on a restart). I have changed the code so that it now always
1581 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1582 message are rejected with the same error (e.g. no authentication or bad
1583 sender address), and a DATA command is nevertheless sent (as can happen
1584 with PIPELINING or a stupid MUA), the error message that was given to the
1585 RCPT commands is included in the rejection of the DATA command. This is
1586 intended to be helpful for MUAs that show only the final error to their
1589 PH/42 Another patch from the Sieve maintainer.
1591 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1592 Eximstats - Fixed some broken HTML links and added missing column headers
1594 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1595 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1597 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1599 PH/43 Yet another patch from the Sieve maintainer.
1601 PH/44 I found a way to check for a TCP/IP connection going away before sending
1602 the response to the final '.' that terminates a message, but only in the
1603 case where the client has not sent further data following the '.'
1604 (unfortunately, this is allowed). However, in many cases there won't be
1605 any further data because there won't be any more messages to send. A call
1606 to select() can be used: if it shows that the input is "ready", there is
1607 either input waiting, or the socket has been closed. An attempt to read
1608 the next input character can distinguish the two cases. Previously, Exim
1609 would have sent an OK response which the client would never have see.
1610 This could lead to message repetition. This fix should cure that, at
1611 least in a lot of common cases.
1613 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1614 advertised in response to EHLO.
1620 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1621 fixed by 4.65/MH/01 (is this a record?) are fixed:
1623 (i) An empty string was always treated as zero by the numeric comparison
1624 operators. This behaviour has been restored.
1626 (ii) It is documented that the numeric comparison operators always treat
1627 their arguments as decimal numbers. This was broken in that numbers
1628 starting with 0 were being interpreted as octal.
1630 While fixing these problems I realized that there was another issue that
1631 hadn't been noticed. Values of message_size_limit (both the global option
1632 and the transport option) were treated as octal if they started with 0.
1633 The documentation was vague. These values are now always treated as
1634 decimal, and I will make that clear in the documentation.
1640 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1641 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1644 MH/01 Don't check that the operands of numeric comparison operators are
1645 integers when their expansion is in "skipping" mode (fixes bug
1646 introduced by 4.64-PH/07).
1648 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1649 child addresses, Exim now panics and dies. Previously, because the count
1650 is held in a short int, deliveries were likely to be lost. As such a
1651 large number of recipients for a single message is ridiculous
1652 (performance will be very, very poor), I have chosen to impose a limit
1653 rather than extend the field.
1659 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1660 leftover -K file (the existence of which was triggered by #402).
1661 While we were at it, introduced process PID as part of the -K
1662 filename. This should rule out race conditions when creating
1665 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1666 processing considerably. Previous code took too long for large mails,
1667 triggering a timeout which in turn triggers #401.
1669 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1670 in the DK code in transports.c. sendfile() is not really portable,
1671 hence the _LINUX specificness.
1673 TF/01 In the add_headers option to the mail command in an Exim filter,
1674 there was a bug that Exim would claim a syntax error in any
1675 header after the first one which had an odd number of characters
1678 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1679 callout verification, Exim cached a "reject" for the entire domain. This
1680 is correct for most verifications, but it is not correct for a recipient
1681 verification with use_sender or use_postmaster set, because in that case
1682 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1683 case of MAIL FROM:<> rejection from other early rejections (e.g.
1684 rejection of HELO). When verifying a recipient using a non-null MAIL
1685 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1686 Whatever the result of the callout, the value of the domain cache is
1687 left unchanged (for any other kind of callout, getting as far as trying
1688 RCPT means that the domain itself is ok).
1690 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1693 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1694 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1695 session, the daemon ignores ECONNECT errors and logs others; it now
1696 ignores EPIPE as well.
1698 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1699 (quoted-printable decoding).
1701 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1702 later the small subsequent patch to fix an introduced bug.
1704 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1706 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1708 PH/08 An error is now given if message_size_limit is specified negative.
1710 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1711 to be given (somewhat) arbitrary names.
1713 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1716 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1717 miscellaneous code fixes
1719 PH/10 Added the log_reject_target ACL modifier to specify where to log
1722 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1723 hostname. This is wrong, because it relates to the incoming message (and
1724 probably the interface on which it is arriving) and not to the outgoing
1725 callout (which could be using a different interface). This has been
1726 changed to use the value of the helo_data option from the smtp transport
1727 instead - this is what is used when a message is actually being sent. If
1728 there is no remote transport (possible with a router that sets up host
1729 addresses), $smtp_active_hostname is used.
1731 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1732 tweaks were necessary in order to get it to work (see also 21 below):
1733 (a) The code assumed that strncpy() returns a negative number on buffer
1734 overflow, which isn't the case. Replaced with Exim's string_format()
1736 (b) There were several signed/unsigned issues. I just did the minimum
1737 hacking in of casts. There is scope for a larger refactoring.
1738 (c) The code used strcasecmp() which is not a standard C function.
1739 Replaced with Exim's strcmpic() function.
1740 (d) The code set only $1; it now sets $auth1 as well.
1741 (e) A simple test gave the error "authentication client didn't specify
1742 service in request". It would seem that Dovecot has changed its
1743 interface. Fortunately there's a specification; I followed it and
1744 changed what the client sends and it appears to be working now.
1746 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1749 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1750 address A is aliased to B and C, where B exists and C does not. Without
1751 -v the output is "A verified" because verification stops after a
1752 successful redirection if more than one address is generated. However,
1753 with -v the child addresses are also verified. Exim was outputting "A
1754 failed to verify" and then showing the successful verification for C,
1755 with its parentage. It now outputs "B failed to verify", showing B's
1756 parentage before showing the successful verification of C.
1758 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1759 look up a TXT record in a specific list after matching in a combined
1762 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1763 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1764 they consult the DNS. I had assumed they would set it the way they
1765 wanted; and indeed my experiments on Linux seem to show that in some
1766 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1767 To be on the safe side, however, I have now made the interface to
1768 host_find_byname() similar to host_find_bydns(), with an argument
1769 containing the DNS resolver options. The host_find_byname() function now
1770 sets these options at its start, just as host_find_bydns() does. The smtp
1771 transport options dns_qualify_single and dns_search_parents are passed to
1772 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1773 of host_find_byname() use the default settings of RES_DEFNAMES
1774 (qualify_single) but not RES_DNSRCH (search_parents).
1776 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1777 spool_read_header() do less string testing, by means of a preliminary
1778 switch on the second character of optional "-foo" lines. (This is
1779 overdue, caused by the large number of possibilities that now exist.
1780 Originally there were few.) While I was there, I also converted the
1781 str(n)cmp tests so they don't re-test the leading "-" and the first
1782 character, in the hope this might squeeze out yet more improvement.
1784 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1785 flag allowing group syntax was set by the header_syntax check but not
1786 turned off, possible causing trouble later; (2) The flag was not being
1787 set at all for the header_verify test, causing "group"-style headers to
1788 be rejected. I have now set it in this case, and also caused header_
1789 verify to ignore an empty address taken from a group. While doing this, I
1790 came across some other cases where the code for allowing group syntax
1791 while scanning a header line wasn't quite right (mostly, not resetting
1792 the flag correctly in the right place). These bugs could have caused
1793 trouble for malformed header lines. I hope it is now all correct.
1795 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1796 with the "reply" argument non-NULL. The code, however (which originally
1797 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1798 but it didn't always do it. This confused somebody who was copying the
1799 code for some other use. I have removed all the tests.
1801 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1802 feature that was used to support insecure browsers during the U.S. crypto
1803 embargo. It requires special client support, and Exim is probably the
1804 only MTA that supported it -- and would never use it because real RSA is
1805 always available. This code has been removed, because it had the bad
1806 effect of slowing Exim down by computing (never used) parameters for the
1807 RSA_EXPORT functionality.
1809 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1810 authenticator to fail if there's a tab character in the incoming data
1811 (there should never be unless someone is messing about, as it's supposed
1812 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1813 if the connection is using TLS or if the remote IP is the same as the
1814 local IP, and the "valid-client-cert option" if a client certificate has
1817 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1818 authenticators. This can be used for authorization after authentication
1819 succeeds. (In the case of plaintext, it servers for both authentication
1822 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1823 if any retry times were supplied.
1825 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1826 connection, where there is no client IP address to check. In this
1827 situation, the verify now always succeeds.
1829 PH/25 Applied John Jetmore's -Mset patch.
1831 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1833 PH/27 In a string expansion for a processed (not raw) header when multiple
1834 headers of the same name were present, leading whitespace was being
1835 removed from all of them, but trailing whitespace was being removed only
1836 from the last one. Now trailing whitespace is removed from each header
1837 before concatenation. Completely empty headers in a concatenation (as
1838 before) are ignored.
1840 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1841 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1843 PH/29 [Removed. This was a change that I later backed out, and forgot to
1844 correct the ChangeLog entry (that I had efficiently created) before
1845 committing the later change.]
1847 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1848 address errors (4xx response to RCPT) over the same connection as other
1849 messages routed to the same hosts. Such deliveries are always "forced",
1850 so retry times are not inspected. This resulted in far too many retries
1851 for the affected addresses. The effect occurred only when there were more
1852 hosts than the hosts_max_try setting in the smtp transport when it had
1853 the 4xx errors. Those hosts that it had tried were not added to the list
1854 of hosts for which the message was waiting, so if all were tried, there
1855 was no problem. Two fixes have been applied:
1857 (i) If there are any address or message errors in an SMTP delivery, none
1858 of the hosts (tried or untried) are now added to the list of hosts
1859 for which the message is waiting, so the message should not be a
1860 candidate for sending over the same connection that was used for a
1861 successful delivery of some other message. This seems entirely
1862 reasonable: after all the message is NOT "waiting for some host".
1863 This is so "obvious" that I'm not sure why it wasn't done
1864 previously. Hope I haven't missed anything, but it can't do any
1865 harm, as the worst effect is to miss an optimization.
1867 (ii) If, despite (i), such a delivery is accidentally attempted, the
1868 routing retry time is respected, so at least it doesn't keep
1869 hammering the server.
1871 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1872 in ${readsocket because some servers need this prod.
1874 PH/32 Added some extra debug output when updating a wait-xxx database.
1876 PH/33 The hint "could be header name not terminated by colon", which has been
1877 given for certain expansion errors for a long time, was not being given
1878 for the ${if def:h_colon_omitted{... case.
1880 PH/34 The spec says: "With one important exception, whenever a domain list is
1881 being scanned, $domain contains the subject domain." There was at least
1882 one case where this was not true.
1884 PH/35 The error "getsockname() failed: connection reset by peer" was being
1885 written to the panic log as well as the main log, but it isn't really
1886 panic-worthy as it just means the connection died rather early on. I have
1887 removed the panic log writing for the ECONNRESET error when getsockname()
1890 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1891 runs only) independently of the message's sender address. This meant
1892 that, if the 4xx error was in fact related to the sender, a different
1893 message to the same recipient with a different sender could confuse
1894 things. In particualar, this can happen when sending to a greylisting
1895 server, but other circumstances could also provoke similar problems.
1896 I have changed the default so that the retry time for these errors is now
1897 based a combination of the sender and recipient addresses. This change
1898 can be overridden by setting address_retry_include_sender=false in the
1901 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1902 remote server are returned as part of bounce messages. This was not
1903 happening for LMTP over a pipe (the lmtp transport), but now it is the
1904 same for both kinds of LMTP.
1906 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1907 in header lines that were in fact CNAMEs. This is no longer the case.
1909 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1910 and queue runs started by the daemon processed all messages. This has
1911 been fixed so that -R and -S can now usefully be given with -q<time>.
1913 PH/40 Import PCRE release 6.7 (fixes some bugs).
1915 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1917 PH/42 Give an error if -q is specified more than once.
1919 PH/43 Renamed the variables $interface_address and $interface_port as
1920 $received_ip_address and $received_port, to make it clear that these
1921 values apply to message reception, and not to the outgoing interface when
1922 a message is delivered. (The old names remain recognized, of course.)
1924 PH/44 There was no timeout on the connect() call when using a Unix domain
1925 socket in the ${readsocket expansion. There now is.
1927 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1928 be meaningful with "accept".
1930 SC/01 Eximstats V1.43
1931 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1933 SC/02 Eximstats V1.44
1934 Use a glob alias rather than an array ref in the generated
1935 parser. This improves both readability and performance.
1937 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1938 Collect SpamAssassin and rejection statistics.
1939 Don't display local sender or destination tables unless
1940 there is data to show.
1941 Added average volumes into the top table text output.
1943 SC/04 Eximstats V1.46
1944 Collect data on the number of addresses (recipients)
1945 as well as the number of messages.
1947 SC/05 Eximstats V1.47
1948 Added 'Message too big' to the list of mail rejection
1949 reasons (thanks to Marco Gaiarin).
1951 SC/06 Eximstats V1.48
1952 Mainlog lines which have GMT offsets and are too short to
1953 have a flag are now skipped.
1955 SC/07 Eximstats V1.49 (Alain Williams)
1956 Added the -emptyok flag.
1958 SC/08 Eximstats V1.50
1959 Fixes for obtaining the IP address from reject messages.
1961 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1962 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1963 whitesspace changes from 4.64-PH/27
1965 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1968 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1969 are found, allow negative numbers in numeric criteria)
1971 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1973 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1974 to match changes made in 4.64-PH/43
1976 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1978 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1979 "domain" is actually an IP address, and force a failure. This locks out
1980 those revolvers/nameservers that support "A-for-A" lookups, in
1981 contravention of the specifications.
1983 PH/48 When a host name was looked up from an IP address, and the subsequent
1984 forward lookup of the name timed out, the host name was left in
1985 $sender_host_name, contrary to the specification.
1987 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1988 restricted to single-key lookups, Exim was not diagnosing an error if
1989 * or *@ was used with a query-style lookup.
1991 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1993 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1994 long ago, but noone interested enough thought of it. Let's just say that
1995 the "1.1" means that there are some new functions that weren't there at
1996 some point in the past.
1998 PH/51 Error processing for expansion failure of helo_data from an smtp
1999 transport during callout processing was broken.
2001 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2002 tested/used via the -bh/-bhc/-bs options.
2004 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2005 bug, fixed in subsequent PCRE releases).
2007 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2008 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2010 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2016 SC/01 Use a glob alias rather than an array ref in eximstats generated
2017 parser. This improves both readability and performance.
2019 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2020 Don't display local sender or destination tables in eximstats unless
2021 there is data to show.
2022 Added average volumes into the eximstats top table text output.
2024 SC/03 Collect data on the number of addresses (recipients) as well
2025 as the number of messages in eximstats.
2027 TF/01 Correct an error in the documentation for the redirect router. Exim
2028 does (usually) call initgroups() when daemonizing.
2030 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2031 with consistent privilege compared to when running as a daemon.
2033 TF/03 Note in the spec that $authenticated_id is not set for local
2034 submissions from trusted users.
2036 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2037 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2039 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2040 by adding some example configuration directives to the default
2041 configuration file. A little bit of work is required to uncomment the
2042 directives and define how usernames and passwords are checked, but
2043 there is now a framework to start from.
2045 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2046 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2047 without this. I don't know how relevant this is to other LDAP libraries.
2049 PH/02 Add the verb name to the "unknown ACL verb" error.
2051 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2053 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2055 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2056 directory not expanded when it should be if an expanded home directory
2057 was set for the address (which is overridden by the transport).
2059 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2062 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2063 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2064 because it is too late at that time, and has no effect.
2066 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2067 security issue with \' (bugzilla #107). I could not use the
2068 PQescapeStringConn() function, because it needs a PGconn value as one of
2071 PH/08 When testing addresses using -bt, indicate those final addresses that
2072 are duplicates that would not cause an additional delivery. At least one
2073 person was confused, thinking that -bt output corresponded to deliveries.
2074 (Suppressing duplicates isn't a good idea as you lose the information
2075 about possibly different redirections that led to the duplicates.)
2077 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2078 systems where poll() doesn't work, in particular OS X.
2080 PH/10 Added more information to debugging output for retry time not reached.
2082 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2083 operations in malware.c.
2085 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2088 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2089 syslog_duplication set false, log lines that would normally be written
2090 both the the main log and to the reject log were not written to syslog at
2093 PH/14 In the default configuration, change the use of "message" in ACL warn
2094 statements to "add_header".
2096 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2097 not followed by a command (e.g. "seen endif").
2099 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2100 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2103 PH/17 Added extra conditions to the default value of delay_warning_condition
2107 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2108 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2109 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2112 The Auto-Submitted: and various List- headers are standardised, whereas I
2113 don't think Precedence: ever was.
2115 PH/18 Refactored debugging code in route_finduser() to show more information,
2116 in particular, the error code if getpwnam() issues one.
2118 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2119 This is apparently needed in addition to the PH/07 change above to avoid
2120 any possible encoding problems.
2122 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2123 but not after initializing Perl.
2125 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2126 output them only if debugging. By default they are written stderr,
2127 apparently, which is not desirable.
2129 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2132 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2135 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2137 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2138 authenticated or an ident call has been made. Suppress the default
2139 values for $authenticated_id and $authenticated_sender (but permit -oMai
2140 and -oMas) when testing with -bh.
2142 PH/24 Re-jigged the order of the tests in the default configuration so that the
2143 tests for valid domains and recipients precede the DNS black list and CSA
2144 tests, on the grounds that those ones are more expensive.
2146 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2147 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2148 This bug exists in every version of Exim that I still have, right back to
2151 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2152 However, an attempt to turn on case-sensitivity in a regex key by
2153 including (?-i) didn't work because the subject string was already
2154 lowercased, and the effects were non-intuitive. It turns out that a
2155 one-line patch can be used to allow (?-i) to work as expected.
2161 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2162 other effects) broke the use of negated acl sub-conditions.
2164 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2167 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2168 "Deny" causes Exim to reject the incoming connection with a 554 error.
2169 Unfortunately, if there is a major crisis, such as a disk failure,
2170 tcp-wrappers gives "deny", whereas what one would like would be some
2171 kind of temporary error. A kludge has been added to help with this.
2172 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2173 554 error is used if errno is still zero or contains ENOENT (which occurs
2174 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2177 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2179 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2180 errors. Otherwise a message that provokes a temporary error (when other
2181 messages do not) can cause a whole host to time out.
2183 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2184 addresses were routed directly to files or pipes from a redirect router.
2185 File deliveries just didn't batch; pipe deliveries might have suffered
2188 PH/06 A failure to get a lock for a hints database would erroneously always say
2189 "Failed to get write lock", even when it was really a read lock.
2191 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2192 of 0600. This has been changed to use the value of the lockfile_mode
2193 option (which defaults to 0600).
2195 PH/08 Applied small patch from the Sieve maintainer.
2197 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2198 folder from quota calculations, a direct delivery into this folder messed
2199 up the contents of the maildirsize file. This was because the regex was
2200 used only to exclude .Trash (or whatever) when the size of the mailbox
2201 was calculated. There was no check that a delivery was happening into an
2202 excluded directory. This bug has been fixed by ignoring all quota
2203 processing for deliveries into excluded directories.
2205 PH/10 Added the maildirfolder_create_regex option to appendfile.
2211 PH/01 The code for finding all the local interface addresses on a FreeBSD
2212 system running IPv6 was broken. This may well have applied to all BSD
2213 systems, as well as to others that have similar system calls. The broken
2214 code found IPv4 interfaces correctly, but gave incorrect values for the
2215 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2216 that it would not match correctly against @[] and not recognize the IPv6
2219 PH/02 The ipliteral router was not recognizing addresses of the form user@
2220 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2222 PH/03 Added disable_ipv6.
2224 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2225 decoded form, because it is most often used to construct To: headers
2226 lines in autoreplies, and the decoded form may well be syntactically
2227 invalid. However, $reply_address has leading white space removed, and all
2228 newlines turned into spaces so that the autoreply transport does not
2231 PH/05 If group was specified without a user on a router, and no group or user
2232 was specified on a transport, the group from the router was ignored.
2234 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2235 for visible compile-time settings that can be used to change these
2236 numbers, for those that want even more. Backwards compatibility with old
2237 spool files has been maintained. However, going back to a previous Exim
2238 release will lost any variables that are in spool files.
2240 PH/07 Two small changes when running in the test harness: increase delay when
2241 passing a TCP/IP connection to a new process, in case the original
2242 process has to generate a bounce, and remove special handling of
2243 127.0.0.2 (sic), which is no longer necessary.
2245 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2246 be the same on different OS.
2248 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2251 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2252 whether --show-vars was specified or not
2254 JJ/02 exipick: Added support for new ACL variable spool format introduced
2257 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2258 syntactically invalid From: or Reply-to: line, and a filter used this to
2259 generate an autoreply, and therefore failed to obtain an address for the
2260 autoreply, Exim could try to deliver to a non-existent relative file
2261 name, causing unrelated and misleading errors. What now happens is that
2262 it logs this as a hard delivery error, but does not attempt to create a
2265 PH/11 The exinext utility has a -C option for testing purposes, but although
2266 the given file was scanned by exinext itself; it wasn't being passed on
2267 when Exim was called.
2269 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2270 an end-of-file indication when reading a command response.
2272 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2273 compiled. In many other places in Exim, IPv6 addresses are always
2274 recognized, so I have changed this. It also means that IPv4 domain
2275 literals of the form [IPV4:n.n.n.n] are now always recognized.
2277 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2278 used if the router is not running as root, for example, when verifying at
2279 ACL time, or when using -bh. The debugging output from this situation was
2280 non-existent - all you got was a failure to exec. I have made two
2283 (a) Failures to set uid/gid, the current directory, or a process leader
2284 in a subprocess such as that created by queryprogram now generate
2285 suitable debugging ouput when -d is set.
2287 (b) The queryprogram router detects when it is not running as root,
2288 outputs suitable debugging information if -d is set, and then runs
2289 the subprocess without attempting to change uid/gid.
2291 PH/15 Minor change to Makefile for building test_host (undocumented testing
2294 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2295 additional section of a DNS packet that returns MX or SRV records.
2296 Instead, it always explicitly searches for A/AAAA records. This avoids
2297 major problems that occur when a DNS server includes only records of one
2298 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2299 fixed another bug: if SRV records were looked up and the corresponding
2300 address records were *not* found in the additional section, the port
2301 values from the SRV records were lost.
2303 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2304 using the correct key (the original address) when searching the retry
2305 rules in order to find which one to use for generating the retry hint.
2307 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2308 adding the default one. Similarly, if it contains a Reply-To: header, the
2309 errors_reply_to option, if set, is not used.
2311 PH/19 When calculating a retry time, Exim used to measure the "time since
2312 failure" by looking at the "first failed" field in the retry record. Now
2313 it does not use this if it is later than than the arrival time of the
2314 message. Instead it uses the arrival time. This makes for better
2315 behaviour in cases where some deliveries succeed, thus re-setting the
2316 "first failed" field. An example is a quota failure for a huge message
2317 when small messages continue to be delivered. Without this change, the
2318 "time since failure" will always be short, possible causing more frequent
2319 delivery attempts for the huge message than are intended.
2320 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2322 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2323 $1, $2, $3) because the numerical variables can be reset during some
2324 expansion items (e.g. "match"), thereby losing the authentication data.
2326 PH/21 Make -bV show the size of off_t variables so that the test suite can
2327 decide whether to run tests for quotas > 2G.
2329 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2330 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2331 filecount value is greater than 2G or if a quota value is greater than 2G
2332 on a system where the size of off_t is not greater than 4, a panic error
2335 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2336 never match. The debug and -bh output now contains an explicit error
2337 message indicating a malformed IPv4 address or mask.
2339 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2340 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2341 PH/23 above applies.
2343 PH/25 Do not write to syslog when running in the test harness. The only
2344 occasion when this arises is a failure to open the main or panic logs
2345 (for which there is an explicit test).
2347 PH/26 Added the /no_tell option to "control=freeze".
2349 PH/27 If a host name lookup failed very early in a connection, for example, if
2350 the IP address matched host_lookup and the reverse lookup yielded a name
2351 that did not have a forward lookup, an error message of the form "no IP
2352 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2353 could be logged. Now it outputs the IP address instead of "NULL".
2355 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2356 allows the sender and the authenticated sender to be set when
2357 submitting a message from within Exim. Since child_open_exim() is
2358 documented for local_scan(), the new function should be too.
2360 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2361 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2362 results in an empty string is now treated as unset.
2364 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2366 PH/31 Added sender_verify_fail logging option.
2368 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2369 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2370 tidied the source and removed it altogether.
2372 PH/33 When a queue run was abandoned because the load average was too high, a
2373 log line was always written; now it is written only if the queue_run log
2374 selector is set. In addition, the log line for abandonment now contains
2375 information about the queue run such as the pid. This is always present
2376 in "start" and "stop" lines but was omitted from the "abandon" line.
2378 PH/34 Omit spaces between a header name and the colon in the error message that
2379 is given when verify = headers_syntax fails (if there are lots of them,
2380 the message gets confusing).
2382 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2383 names, as there are now some PTR records that contain slashes. This check
2384 is only to protect against broken name servers that fall over on strange
2385 characters, so the fact that it applies to all lookups doesn't matter.
2387 PH/36 Now that the new test suite is complete, we can remove some of the
2388 special code in Exim that was needed for the old test suite. For example,
2389 sorting DNS records because real resolvers return them in an arbitrary
2390 order. The new test suite's fake resolver always returns records in the
2393 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2394 bounces) except when queue_only is set, to avoid logging races between
2395 the different processes.
2397 PH/38 Panic-die if .include specifies a non-absolute path.
2399 PH/39 A tweak to the "H" retry rule from its user.
2401 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2402 a label. They prevented compilation on older perls.
2404 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2405 a warning to be raised on newish perls.
2407 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2408 on queue. Changes to match documented behaviour of showing count of
2409 messages matching specified criteria.
2411 PH/40 Changed the default ident timeout from 30s to 5s.
2413 PH/41 Added support for the use of login_cap features, on those BSD systems
2414 that have them, for controlling the resources used by pipe deliveries.
2416 PH/42 The content-scanning code uses fopen() to create files in which to put
2417 message data. Previously it was not paying any attention to the mode of
2418 the files. Exim runs with umask(0) because the rest of the code creates
2419 files with open(), and sets the required mode explicitly. Thus, these
2420 files were ending up world-writeable. This was not a big issue, because,
2421 being within the spool directory, they were not world-accessible. I have
2422 created a function called modefopen, which takes an additional mode
2423 argument. It sets umask(777), creates the file, chmods it to the required
2424 mode, then resets the umask. All the relevant calls to fopen() in the
2425 content scanning code have been changed to use this function.
2427 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2428 to 24 hours. This avoids potential overflow problems when processing G
2429 and H retry rules. I suspect nobody ever tinkers with this value.
2431 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2433 PH/45 When the plaintext authenticator is running as a client, the server's
2434 challenges are checked to ensure they are valid base64 strings. By
2435 default, the authentication attempt is cancelled if an invalid string is
2436 received. Setting client_ignore_invalid_base64 true ignores these errors.
2437 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2438 they are received. Thus, the responses can be made to depend on the
2439 challenges. If an invalid string is ignored, an empty string is placed in
2442 PH/46 Messages that are created by the autoreply transport now contains a
2443 References: header, in accordance with RFCs 2822 and 3834.
2445 PH/47 Added authenticated_sender_force to the smtp transport.
2447 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2449 PH/49 Installed latest patch from the Sieve maintainer.
2451 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2452 was also set, the appendfile transport was unnecessarily scanning a
2453 directory of message files (e.g. for maildir delivery) to find the count
2454 of files (along with the size), even though it did not need this
2455 information. It now does the scan only if it needs to find either the
2456 size of the count of files.
2458 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2460 PH/52 Two bugs concerned with error handling when the smtp transport is
2463 (i) Exim was not creating retry information for temporary errors given
2464 for individual recipients after the DATA command when the smtp transport
2465 was used in LMTP mode. This meant that they could be retried too
2466 frequently, and not timed out correctly.
2468 (ii) Exim was setting the flag that allows error details to be returned
2469 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2470 recipients that were returned after the DATA command.
2472 PH/53 This is related to PH/52, but is more general: for any failing address,
2473 when detailed error information was permitted to be returned to the
2474 sender, but the error was temporary, then after the final timeout, only
2475 "retry timeout exceeded" was returned. Now it returns the full error as
2476 well as "retry timeout exceeded".
2478 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2479 do this, and (what is worse) MTAs that accept it.
2481 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2482 will now be deprecated.
2484 PH/56 New os.c-cygwin from the Cygwin maintainer.
2486 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2487 formats (previously only available in exim formats via -bpr, -bpru,
2488 and -bpra. Now also available in native and exiqgrep formats)
2490 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2491 with very large, slow to parse queues
2493 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2495 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2497 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2498 responses to authentication challenges, though it was showing the
2499 challenges; (ii) I've removed the CR characters from the debug output for
2502 PH/58 Allow for the insertion of a newline as well as a space when a string
2503 is turned into more than one encoded-word during RFC 2047 encoding. The
2504 Sieve code now uses this.
2506 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2507 data_4xx, lost_connection, tls_required.
2509 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2510 message was being sent as an SMTP response.
2512 PH/61 Add -l and -k options to exicyclog.
2514 PH/62 When verifying, if an address was redirected to one new address, so that
2515 verification continued, and the new address failed or deferred after
2516 having set something in $address_data, the value of $address_data was not
2517 passed back to the ACL. This was different to the case when no
2518 redirection occurred. The value is now passed back in both cases.
2520 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2521 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2522 use login_cap.h, so on its own it isn't the distinguishing feature. The
2523 new name refers directly to the setclassresources() function.
2525 PH/65 Added configuration files for NetBSD3.
2527 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2529 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2530 is preferred over IPv4.
2532 PH/68 The bounce_return_message and bounce_return_body options were not being
2533 honoured for bounces generated during the reception of non-SMTP messages.
2534 In particular, this applied to messages rejected by the ACL. This bug has
2535 been fixed. However, if bounce_return_message is true and bounce_return_
2536 body is false, the headers that are returned for a non-SMTP message
2537 include only those that have been read before the error was detected.
2538 (In the case of an ACL rejection, they have all been read.)
2540 PH/69 The HTML version of the specification is now built in a directory called
2541 spec_html instead of spec.html, because the latter looks like a path with
2542 a MIME-type, and this confuses some software.
2544 PH/70 Catch two compiler warnings in sieve.c.
2546 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2547 function verify_get_ident() calls ip_connect() to connect a socket, but
2548 if the "connect()" function timed out, ip_connect() used to close the
2549 socket. However, verify_get_ident() also closes the socket later, and in
2550 between Exim writes to the log, which may get opened at this point. When
2551 the socket was closed in ip_connect(), the log could get the same file
2552 descriptor number as the socket. This naturally causes chaos. The fix is
2553 not to close the socket in ip_connect(); the socket should be closed by
2554 the function that creates it. There was only one place in the code where
2555 this was missing, in the iplookup router, which I don't think anybody now
2556 uses, but I've fixed it anyway.
2558 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2559 well as to direct DNS lookups. Otherwise the handling of names in host
2560 lists is inconsistent and therefore confusing.
2566 PH/01 Two changes to the default runtime configuration:
2568 (1) Move the checks for relay_from_hosts and authenticated clients from
2569 after to before the (commented out) DNS black list checks.
2571 (2) Add control=submission to the relay_from_hosts and authenticated
2572 clients checks, on the grounds that messages accepted by these
2573 statements are most likely to be submissions.
2575 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2577 (1) Generate an error if the third argument for the ${prvs expansion is
2580 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2583 (3) Reset the variables that are obtained from the first argument of
2584 ${prvscheck and used in the second argument before leaving the code,
2585 because their memory is reclaimed, so using them afterwards may do
2588 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2589 one (it's much easier than Tom thought :-).
2591 (5) Because of (4), we can now allow for the use of $prvscheck_result
2592 inside the third argument.
2594 PH/03 For some reason, the default setting of PATH when running a command from
2595 a pipe transport was just "/usr/bin". I have changed it to
2598 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2599 anything to be listed in the output from -bV.
2601 PH/05 When a filter generated an autoreply, the entire To: header line was
2602 quoted in the delivery log line, like this:
2604 => >A.N.Other <ano@some.domain> <original@ddress> ...
2606 This has been changed so that it extracts the operative address. There
2607 may be more than one such address. If so, they are comma-separated, like
2610 => >ano@some.domain,ona@other.domain <original@ddress> ...
2612 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2613 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2614 not being looked up in the rDNS to get a host name, Exim was showing the
2615 IP address twice in Received: lines, even though the IP addresses were
2616 identical. For example:
2618 Received: from [1.2.3.4] (helo=[1.2.3.4])
2620 However, if the real host name was known, it was omitting the HELO data
2621 if it matched the actual IP address. This has been tidied up so that it
2622 doesn't show the same IP address twice.
2624 PH/07 When both +timestamp and +memory debugging was on, the value given by
2625 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2626 called by the debug printing, thereby overwriting the timestamp buffer.
2627 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2629 PH/08 When the original message was included in an autoreply transport, it
2630 always said "this is a copy of the message, including all the headers",
2631 even if body_only or headers_only was set. It now gives an appropriate
2634 PH/09 Applied a patch from the Sieve maintainer which:
2636 o fixes some comments
2637 o adds the (disabled) notify extension core
2638 o adds some debug output for the result of if/elsif tests
2639 o points to the current vacation draft in the documentation
2640 and documents the missing references header update
2644 o fixes a bug in processing the envelope test (when testing
2645 multiple envelope elements, the last element determinted the
2648 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2649 Electronic Mail") by including:
2651 Auto-submitted: auto-generated
2653 in the messages that it generates (bounce messages and others, such as
2654 warnings). In the case of bounce messages for non-SMTP mesages, there was
2655 also a typo: it was using "Auto_submitted" (underscore instead of
2656 hyphen). Since every message generated by Exim is necessarily in response
2657 to another message, thes have all been changed to:
2659 Auto-Submitted: auto-replied
2661 in accordance with these statements in the RFC:
2663 The auto-replied keyword:
2665 - SHOULD be used on messages sent in direct response to another
2666 message by an automatic process,
2668 - MUST NOT be used on manually-generated messages,
2670 - MAY be used on Delivery Status Notifications (DSNs) and Message
2671 Disposition Notifications (MDNs),
2673 - MUST NOT be used on messages generated by automatic or periodic
2674 processes, except for messages which are automatic responses to
2677 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2678 to the default Received: header definition.
2680 PH/12 Added log selector acl_warn_skipped (default on).
2682 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2683 variables because (a) they are in the wrong storage pool and (b) even if
2684 they were copied, it wouldn't work properly because of the caching.
2686 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2687 checking when decoding. Apparently there are clients that generate
2688 overlong encoded strings. Why am I not surprised?
2690 PH/15 If the first argument of "${if match_address" was not empty, but did not
2691 contain an "@" character, Exim crashed. Now it writes a panic log message
2692 and treats the condition as false.
2694 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2696 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2697 extension "envelope-auth". The code is finished and in agreement with
2698 other implementations, but there is no documentation so far and in fact,
2699 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2700 not changing the active code.
2702 Print executed "if" and "elsif" statements when debugging is used. This
2703 helps a great deal to understand what a filter does.
2705 Document more things not specified clearly in RFC3028. I had all this
2706 sorted out, when out of a sudden new issues came to my mind. Oops."
2708 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2711 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2712 canonical form (as documented). However, after a host name lookup from
2713 the IP address, check_host() was doing a simple string comparison with
2714 addresses acquired from the DNS when checking that the found name did
2715 have the original IP as one of its addresses. Since any found IPv6
2716 addresses are likely to be in abbreviated form, the comparison could
2717 fail. Luckily, there already exists a function for doing the comparison
2718 by converting both addresses to binary, so now that is used instead of
2719 the text comparison.
2721 PH/20 There was another similar case to PH/19, when a complete host name was
2722 given in a host list; looking up its IP address could give an abbreviated
2723 form, whereas the current host's name might or might not be abbreviated.
2724 The same fix has been applied.
2730 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2731 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2734 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2735 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2737 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2739 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2740 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2741 Also, for "Auto-Submitted", treat anything other than "no" as
2742 non-personal, in accordance with RFC 3834. (Previously it treated
2743 anything starting "auto-" as non-personal.)
2745 TF/01 The control=submission/name=... option had a problem with syntax
2746 errors if the name included a slash character. The /name= option
2747 now slurps the rest of the string, so it can include any characters
2748 but it must come last in the list of options (after /sender_retain
2751 PH/05 Some modifications to the interface to the fake nameserver for the new
2759 TK/01 Added the "success_on_redirect" address verification option. See
2760 NewStuff for rationale and an example.
2762 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2764 PH/02 Patch to exigrep to allow it to work on syslog lines.
2766 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2767 fread() to skip over the body file's header line, because in Cygwin the
2768 header line is locked and is inaccessible.
2770 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2771 co-exist for some time) to make it clear that it is the Exim ID that is
2772 referenced, not the Message-ID: header line.
2774 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2775 string_format() function, because snprintf() does not exist on all
2778 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2779 ${stat: expansion item.
2781 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2782 protocol synchronization error", to keep the pedants happy.
2784 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2785 well as for IRIX systems, when gcc is being used. See the host.c source
2788 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2790 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2793 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2794 transport and to the smtp transport in LMTP mode.
2796 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2798 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2799 run a filter in a subprocess. This could lead to confusion in subsequent
2800 lookups in the parent process. There should also be a search_tidyup() at
2801 the end of the subprocess.
2803 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2804 only if the host matched helo_try_verify_hosts, which caused the
2805 verification to occur when the EHLO/HELO command was issued. The ACL just
2806 tested the remembered result. Now, if a previous verification attempt has
2807 not happened, "verify = helo" does it there and then.
2809 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2811 TK/03 Fix log output including CR from clamd.
2813 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2814 exist provoked a memory error which could cause a segfault.
2816 PH/15 Installed PCRE 6.2
2818 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2820 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2821 of the problem. Specifically, suggested +O2 rather than +O1 for the
2824 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2826 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2827 record, the port was not correctly logged when the outgoing_port log
2828 selector was set (it logged the transort's default port).
2830 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2831 fallback_hosts, and "hosts" in the smtp transport.
2833 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2834 host errors such as "Connection refused".
2836 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2837 authentication with radiusclient 0.4.9:
2839 - Error returned from rc_read_config was caught wrongly
2840 - Username/password not passed on to radius server due to wrong length.
2842 The presumption is that some radiusclient API changes for 4.51/PH/17
2843 were not taken care of correctly. The code is still untested by me (my
2844 Linux distribution still has 0.3.2 of radiusclient), but it was
2845 contributed by a Radius user.
2847 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2848 expanding the "port" option of the smtp transport.
2850 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2851 while reading a MIME header. Thanks to Tom Hughes for a patch.
2853 PH/24 Include config.h inside local_scan.h so that configuration settings are
2856 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2857 that in an ACL for RCPT (for example), you can examine exactly what was
2860 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2861 commands, but it was not correctly comparing the address with the actual
2862 client host address. Thus, it would show the EHLO address in Received:
2863 header lines when this was not necessary.
2865 PH/27 Added the % operator to ${eval:}.
2867 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2868 it should be ignoring failures (because with -C, for example, it has lost
2869 privilege). It wasn't ignoring creation failures other than "already
2872 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2875 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2876 because an input line was too long, either on its own, or by virtue of
2877 too many continuations, the temporary file was not being removed, and the
2878 return code was incorrect.
2880 PH/31 Missing "BOOL" in function definition in filtertest.c.
2882 PH/32 Applied Sieve patches from the maintainer.
2884 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2886 PH/33 Added "verify = not_blind".
2888 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2889 Local/Makefile (with some defaults set). These are used in built scripts
2890 such as exicyclog, but they have never been used in the exim_install
2891 script (though there are many overriding facilities there). I have
2892 arranged that the exim_install script now takes note of these two
2895 PH/35 Installed configuration files for Dragonfly.
2897 PH/36 When a locally submitted message by a trusted user did not contain a
2898 From: header, and the sender address was obtained from -f or from an SMTP
2899 MAIL command, and the trusted user did not use -F to supply a sender
2900 name, $originator_name was incorrectly used when constructing a From:
2901 header. Furthermore, $originator_name was used for submission mode
2902 messages from external hosts without From: headers in a similar way,
2903 which is clearly wrong.
2905 PH/37 Added control=suppress_local_fixups.
2907 PH/38 When log_selector = +received_sender was set, and the addition of the
2908 sender made the log line's construction buffer exactly full, or one byte
2909 less than full, an overflow happened when the terminating "\n" was
2912 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2913 when the result of a list match is failure because a DNS lookup failed.
2915 PH/40 RM_COMMAND is now used in the building process.
2917 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2918 the "build-* directories that it finds.
2920 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2921 address in a domain literal was a prefix of an interface address.
2923 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2924 when verifying a sender address, unless rewrite_headers is false.
2926 PH/44 Wrote a long comment about why errors_to addresses are verified as
2927 recipients, not senders.
2929 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2930 the ratelimit ACL was added.
2932 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2934 PH/46 Added extra information about PostgreSQL errors to the error string.
2936 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2937 avoiding the need to install special zones in a real server. This is
2938 backwards compatible; if it can't find the fake resolver, it drops back.
2939 Thus, both old and new test suites can be run.
2941 TF/02 Added util/ratelimit.pl
2943 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2944 clock is set back in time.
2946 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2947 Candler <B.Candler@pobox.com>.
2949 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2950 OK for addresses that are the result of redirections.
2952 PH/48 A number of further additions for the benefit of the new test suite,
2953 including a fake gethostbyname() that interfaces to the fake DNS resolver
2956 TF/06 The fix for widen_domains has also been applied to qualify_single and
2957 search_parents which are the other dnslookup options that can cause
2960 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2963 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2965 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2966 service" if it encountered an MX record with an empty target hostname.
2967 The message is now "an MX or SRV record indicated no SMTP service".
2969 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2970 if the DNS of the sending site is misconfigured. This is quite a
2971 common situation. This change restores the behaviour of treating a
2972 helo verification defer as a failure.
2974 PH/51 If self=fail was set on a router, the bounce message did not include the
2975 actual error message.
2981 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2983 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2984 command itself ended in error, the underlying message about the transport
2985 filter timeout was being overwritten with the pipe command error. Now the
2986 underlying error message should be appended to the second error message.
2988 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2990 PH/02 Reduce the amount of output that "make" produces by default. Full output
2991 can still be requested.
2993 PH/03 The warning log line about a condition test deferring for a "warn" verb
2994 was being output only once per connection, rather than after each
2995 occurrence (because it was using the same function as for successful
2996 "warn" verbs). This seems wrong, so I have changed it.
2998 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2999 it should not have, which might have caused a crash in the right
3000 circumstances, but probably never did.
3002 PH/04 Installed a modified version of Tony Finch's patch to make submission
3003 mode fix the return path as well as the Sender: header line, and to
3004 add a /name= option so that you can make the user's friendly name appear
3007 TF/03 Added the control = fakedefer ACL modifier.
3009 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3010 Mark Lowes for thorough testing.
3012 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3014 TK/03 Merged latest SRS patch from Miles Wilton.
3016 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3017 with the definition in sysexits.h (which is #included earlier).
3018 Fortunately, Exim does not actually use EX_OK. The code used to try to
3019 preserve the sysexits.h value, by assumimg that macro definitions were
3020 scanned for macro replacements. I have been disabused of this notion,
3021 so now the code just undefines EX_OK before #including unistd.h.
3023 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3024 in the smtp transport. When a block could not be written in a single
3025 write() function, the timeout was being re-applied to each part-write.
3026 This seems wrong - if the receiver was accepting one byte at a time it
3027 would take for ever. The timeout is now adjusted when this happens. It
3028 doesn't have to be particularly precise.
3030 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3031 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3033 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3034 without a domain if the check to <postmaster@domain> fails.
3036 SC/01 Eximstats: added -xls and the ability to specify output files
3037 (patch written by Frank Heydlauf).
3039 SC/02 Eximstats: use FileHandles for outputing results.
3041 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3043 SC/04 Eximstats: fixed display of large numbers with -nvr option
3045 SC/05 Eximstats: fixed merging of reports with empty tables.
3047 SC/06 Eximstats: added the -include_original_destination flag
3049 SC/07 Eximstats: removed tabs and trailing whitespace.
3051 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3053 TK/06 MBOX spool code: Add real "From " MBOX separator line
3054 so the .eml file is really in mbox format (even though
3055 most programs do not really care). Patch from Alex Miller.
3057 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3058 The latter is generated from $received_to and is only set if the
3059 message has one envelope recipient. SA can use these headers,
3060 obviously out-of-the-box. Patch from Alex Miller.
3062 PH/08 The ${def test on a variable was returning false if the variable's
3063 value was "0", contrary to what the specification has always said!
3064 The result should be true unless the variable is empty.
3066 PH/09 The syntax error of a character other than { following "${if
3067 def:variable_name" (after optional whitespace) was not being diagnosed.
3068 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3069 accidental colon was present, for example, could give incorrect results.
3071 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3072 result is used (not including appendfile, where other changes are about
3075 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3076 This involved changing a lot of size variables from int to off_t. It
3077 should work with maildirs and everything.
3079 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3080 spamd dying while we are connected to it.
3082 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3085 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3086 function for BDB 4.3.
3088 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3090 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3091 the sources for PCRE have been re-arranged and more files are now
3094 PH/15 The code I had for printing potentially long long variables in PH/11
3095 above was not the best (it lost precision). The length of off_t variables
3096 is now inspected at build time, and an appropriate printing format (%ld
3097 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3098 to be "long long int" or "long int". This is needed for the internal
3099 formatting function string_vformat().
3101 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3102 the configuration file to be ":syslog", then the script "guesses" where
3103 the logs files are, rather than using the compiled in default. In our
3104 case the guess is not the same as the compiled default, so the script
3105 suddenly stopped working when I started to use syslog. The patch checks
3106 to see if log_file_path is "". If so, it attempts to read it from exim
3107 with no configuration file to get the compiled in version, before it
3108 falls back to the previous guessing code."
3110 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3111 implementing BATV in an Exim configuration. See NewStuff for the gory
3114 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3115 Makefile that are specific to HP-UX.
3117 PH/18 If the "use_postmaster" option was set for a recipient callout together
3118 with the "random" option, the postmaster address was used as the MAIL
3119 FROM address for the random test, but not for the subsequent recipient
3120 test. It is now used for both.
3122 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3123 patch removes a few documentation additions to RFC 3028, because the
3124 latest draft now contains them. It adds the new en;ascii-case comparator
3125 and a new error check for 8bit text in MIME parts. Comparator and
3126 require names are now matched exactly. I enabled the subaddress
3127 extension, but it is not well tested yet (read: it works for me)."
3129 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3130 rework some of the code of TK/09 above to avoid the hardwired use of
3131 "%lld" and "long long". Replaced the call to snprintf() with a call to
3134 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3135 records point to non-existent hosts", "retry timeout exceeded", and
3136 "retry time not reached for any host after a long failure period".
3138 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3139 experimental DomainKeys support:
3141 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3142 (2) On an error such as an illegally used "control", the wrong name for
3143 the control was given.
3145 These problems did NOT occur unless DomainKeys support was compiled.
3147 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3149 PH/24 Added ${if match_ip condition.
3151 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3152 databases so that it will be absolutely obvious if a crash occurs in the
3153 DB library. This is a regular occurrence (often caused by mis-matched
3156 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3157 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3158 picked up on a user's system that detects such things. There doesn't seem
3159 to be a gcc warning option for this - only an attribute that has to be
3160 put on the function's prototype. It seems that in Fedora Core 4 they have
3161 set this on a number of new functions. No doubt there will be more in due
3164 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3165 specify a transport. However, if an address that was verified by such a
3166 router was the subject of a callout, Exim crashed because it tried to
3167 read the rcpt_include_affixes from the non-existent transport. Now it
3168 just assumes that the setting of that option is false. This bug was
3169 introduced by 4.51/PH/31.
3171 PH/28 Changed -d+all to exclude +memory, because that information is very
3172 rarely of interest, but it makes the output a lot bigger. People tend to
3173 do -d+all out of habit.
3175 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3176 code in os-type was giving problems when libc.so lives in lib64, like on
3179 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3180 aren't the modern standard, and it seems that some systems' include files
3181 don't always have them. Exim was already checking for some of the newer
3182 ones like T_AAAA, and defining it itself. I've added checks for all the
3183 record types that Exim uses.
3185 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3186 not automatically generating a new one, as it is supposed to. This
3187 prevented TLS from working. If the file did exist, but contained invalid
3188 data, a new version was generated, as expected. It was only the case of a
3189 non-existent file that was broken.
3191 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3192 with a change in libdomainkeys > 0.64.
3194 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3195 from DNS. If the selector record carries the flag, it now has
3196 precedence over the domain-wide flag.
3198 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3200 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3201 the use of a port name that isn't defined in /etc/services) occurred, the
3202 message was deferred as in a normal delivery, and thus remained on the
3203 spool, instead of being failed because of the mua_wrapper setting. This
3204 is now fixed, and I tidied up some of the mua_wrapper messages at the
3207 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3208 the messages in a hash of arrays rather than using individual hashes.
3209 This is a bit cleaner and results in dramatic memory savings, albeit
3210 at a slight CPU cost.
3212 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3213 as requested by Marc Sherman.
3215 SC/10 Eximstats: added histograms for user specified patterns as requested
3218 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3220 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3221 fopen() in the content-scanning modules that did not already have it.
3227 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3228 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3230 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3232 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3234 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3235 to test Sieve filters that use "vacation".
3237 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3238 that changes the way the GnuTLS parameters are stored in the cache file.
3239 The new format can be generated externally. For backward compatibility,
3240 if the data in the cache doesn't make sense, Exim assumes it has read an
3241 old-format file, and it generates new data and writes a new file. This
3242 means that you can't go back to an older release without removing the
3245 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3246 work if there are any delivery delays because "one_time" forces the
3247 parent to be marked "delivered", so its unseen clone is never tried
3248 again. For this reason, Exim now forbids the simultaneous setting of
3251 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3252 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3253 entry if you want to know the details. The fix, however, neglected to
3254 consider the case when local delivery batching is involved. The test for
3255 "previously delivered" was not happening when checking to see if an
3256 address could be batched with a previous (undelivered) one; under
3257 certain circumstances this could lead to multiple deliveries to the same
3260 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3261 in its include files, and this causes problems building Exim.
3263 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3264 header_syntax) but Exim was just ignoring anything given after a slash.
3265 In particular, this caused confusion with an attempt to use "verify =
3266 reverse_host_lookup/defer_ok". An error is now given when options are
3267 supplied for verify items that do not have them. (Maybe reverse_host_
3268 lookup should have a defer_ok option, but that's a different point.)
3270 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3271 defined by RFC 821) to 2048, because there were problems with some AUTH
3272 commands, and RFC 1869 says the size should be increased for extended
3273 SMTP commands that take arguments.
3275 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3278 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3279 "unknown" error; now it says that the functionality isn't in the binary.
3281 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3282 an address' error message when a string expansion fails (syntax or
3283 whatever). Otherwise the password may appear in the log. Following change
3284 PH/42 below, there is no longer a chance of it appearing in a bounce
3287 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3289 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3290 of its addresses was ever tried. (Bugzilla bug #2.)
3292 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3293 the result incorrectly in the debug output. (It correctly added a newline
3294 to what was transported.)
3296 TF/01 Added $received_time.
3298 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3299 commented out examples of how to interface to a virus scanner and to
3300 SpamAssassin. Also added commented examples of av_scanner and
3301 spamd_address settings.
3303 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3304 and controls are allowed in which ACLs. There were a couple of minor
3305 errors. Some of the entries in the conditions table (which is a table of
3306 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3307 negation of where the condition IS allowed.
3309 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3311 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3312 header file does not have a version number, so I've had to invent a new
3313 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3314 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3315 radiusclient), but it was contributed by a Radius user.
3317 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3318 files or directories, for OpenSSL.
3320 PH/19 When an Exim process that is running as root has to create an Exim log
3321 file, it does so in a subprocess that runs as exim:exim so as to get the
3322 ownership right at creation (otherwise, other Exim processes might see
3323 the file with the wrong ownership). There was no test for failure of this
3324 fork() call, which would lead to the process getting stuck as it waited
3325 for a non-existent subprocess. Forks do occasionally fail when resources
3326 run out. I reviewed all the other calls to fork(); they all seem to check
3329 PH/20 When checking for unexpected SMTP input at connect time (before writing
3330 the banner), Exim was not dealing correctly with a non-positive return
3331 from the read() function. If the client had disconnected by this time,
3332 the result was a log entry for a synchronization error with an empty
3333 string after "input=" when read() returned zero. If read() returned -1
3334 (an event I could not check), uninitialized data bytes were printed.
3335 There were reports of junk text (parts of files, etc) appearing after
3338 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3340 PH/22 Added support for macro redefinition, and (re)definition in between
3341 driver and ACL definitions.
3343 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3344 forgetting to use the resulting value; it was using the unexpanded value.
3346 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3347 hadn't been configured. The fix is from Juergen Kreileder, who
3348 understands it better than I do:
3350 "Here's what I see happening with three configured cyrus_sasl
3351 authenticators configured (plain, login, cram-md5):
3353 On startup auth_cyrus_sasl_init() gets called for each of these.
3354 This means three calls to sasl_listmech() without a specified mech_list.
3355 => SASL tests which mechs of all available mechs actually work
3356 => three warnings about OTP not working
3357 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3359 With the patch, sasl_listmech() also gets called three times. But now
3360 SASL's mech_list option is set to the server_mech specified in the the
3361 authenticator. Or in other words, the answer from sasl_listmech()
3362 gets limited to just the mech you're testing for (which is different
3364 => the return list contains just 'plain' or 'login', 'cram-md5' or
3365 nothing depending on the value of ob->server_mech.
3367 I've just tested the patch: Authentication still works fine,
3368 unavailable mechs specified in the exim configuration are still
3369 caught, and the auth.log warnings about OTP are gone."
3371 PH/25 When debugging is enabled, the contents of the command line are added
3372 to the debugging output, even when log_selector=+arguments is not
3375 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3376 answer is "GNU", and only if the return is "GNU/something" is the answer
3379 PH/27 $acl_verify_message is now set immediately after the failure of a
3380 verification in an ACL, and so is available in subsequent modifiers. In
3381 particular, the message can be preserved by coding like this:
3383 warn !verify = sender
3384 set acl_m0 = $acl_verify_message
3386 Previously, $acl_verify_message was set only while expanding "message"
3387 and "log_message" when a very denied access.
3389 PH/28 Modified OS/os.c-Linux with
3391 -#ifndef OS_LOAD_AVERAGE
3392 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3394 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3395 nomenclature these days.)
3397 PH/29 Installed patch from the Sieve maintainer that adds the options
3398 sieve_useraddress and sieve_subaddress to the redirect router.
3400 PH/30 In these circumstances:
3401 . Two addresses routed to the same list of hosts;
3402 . First host does not offer TLS;
3403 . First host accepts first address;
3404 . First host gives temporary error to second address;
3405 . Second host offers TLS and a TLS session is established;
3406 . Second host accepts second address.
3407 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3408 and peerdn, if requested) that were in fact used only for the second
3411 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3412 attention to any local part prefix or suffix that was matched by the
3413 router that accepted the address. It now behaves in the same way as it
3414 does for delivery: the affixes are removed from the local part unless
3415 rcpt_include_affixes is set on the transport.
3417 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3418 timeout during the DATA phase of an incoming message.
3420 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3421 applied a patch sanctioned by the Sieve maintainer.
3423 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3424 the uid or gid is negative. A case of a negative gid caused this to be
3425 noticed. The fix allows for either to be negative.
3427 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3428 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3431 PH/36 The change PH/12 above was broken. Fixed it.
3433 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3434 the grounds that routing the same address twice would always produce the
3435 same answer. This might have been true once, but it is certainly no
3436 longer true now. Routing a child address may depend on the previous
3437 routing that produced that child. Some complicated redirection strategies
3438 went wrong when messages had multiple recipients, and made Exim's
3439 behaviour dependent on the order in which the addresses were given.
3441 I have moved the duplicate checking until after the routing is complete.
3442 Exim scans the addresses that are assigned to local and remote
3443 transports, and removes any duplicates. This means that more work will be
3444 done, as duplicates will always all be routed, but duplicates are
3445 presumably rare, so I don't expect this is of any significance.
3447 For deliveries to pipes, files, and autoreplies, the duplicate checking
3448 still happens during the routing process, since they are not going to be
3451 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3452 It corrects a timeout issue with spamd. This is Ian's comment: "The
3453 background is that sometimes spamd either never reads data from a
3454 connection it has accepted, or it never writes response data. The exiscan
3455 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3456 blindly assumes that writes won't block so it may never time out."
3458 PH/39 Allow G after quota size as well as K and M.
3460 PH/40 The value set for $authenticated_id in an authenticator may not contain
3461 binary zeroes or newlines because the value is written to log lines and
3462 to spool files. There was no check on this. Now the value is run through
3463 the string_printing() function so that such characters are converted to
3464 printable escape sequences.
3466 PH/41 $message_linecount is a new variable that contains the total number of
3467 lines in the message. Compare $body_linecount, which is the count for the
3470 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3471 bounce and delay warning messages, except in certain special cases, which
3474 (a) An SMTP error message from a remote host;
3475 (b) A message specified in a :fail: redirection;
3476 (c) A message specified in a "fail" command in a system filter;
3477 (d) A message specified in a FAIL return from the queryprogram router;
3478 (e) A message specified by the cannot_route_message router option.
3480 In these cases only, Exim does include the error details in bounce and
3481 warning messages. There are also a few cases where bland messages such
3482 as "unrouteable address" or "local delivery error" are given.
3484 PH/43 $value is now also set for the "else" part of a ${run expansion.
3486 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3487 being worked on, but at least Exim now implements the latest version to
3490 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3491 process to complete was treated as a delivery failure, a timeout while
3492 writing the message to the pipe was logged, but erroneously treated as a
3493 successful delivery. Such timeouts include transport filter timeouts. For
3494 consistency with the overall process timeout, these timeouts are now
3495 treated as errors, giving rise to delivery failures by default. However,
3496 there is now a new Boolean option for the pipe transport called
3497 timeout_defer, which, if set TRUE, converts the failures into defers for
3498 both kinds of timeout. A transport filter timeout is now identified in
3501 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3502 systems where "make" and "gmake" are different, calling "gmake" at top
3503 level broke things. I've arranged for the value of $(MAKE) to be passed
3504 from the Makefile to this script so that it can call the same version of
3508 A note about Exim versions 4.44 and 4.50
3509 ----------------------------------------
3511 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3512 changes of various kinds. As a consequence, a big documentation update was
3513 needed. This delayed the release for rather longer than seemed good, especially
3514 in the light of a couple of (minor) security issues. Therefore, the changes
3515 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3516 release. So 4.44 and 4.50 are in effect two different branches that both start
3519 I have left the 4.50 change log unchanged; it contains all the changes since
3520 4.43. The change log for 4.44 is below; many of its items are identical to
3521 those for 4.50. This seems to be the most sensible way to preserve the
3522 historical information.
3528 1. Minor wording change to the doc/README.SIEVE file.
3530 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3531 computation of the current number of files was incorrect.
3533 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3534 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3537 4. Give more explanation in the error message when the command for a transport
3538 filter fails to execute.
3540 5. There are several places where Exim runs a non-Exim command in a
3541 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3542 was being done only for the command run by the queryprogram router. It is
3543 now done for all such subprocesses. The other cases are: ${run, transport
3544 filters, and the commands run by the lmtp and pipe transports.
3546 6. Added CONFIGURE_GROUP build-time option.
3548 7. Some older OS have a limit of 256 on the maximum number of file
3549 descriptors. Exim was using setrlimit() to set 1000 as a large value
3550 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3551 systems. I've change it so that if it can't get 1000, it tries for 256.
3553 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3554 was an oversight, and furthermore, ever since the addition of extra
3555 controls (e.g. 4.43/32), the checks on when to allow different forms of
3556 "control" were broken. There should now be diagnostics for all cases when a
3557 control that does not make sense is encountered.
3559 9. Added the /retain_sender option to "control=submission".
3561 10. $recipients is now available in the predata ACL (oversight).
3563 11. Tidy the search cache before the fork to do a delivery from a message
3564 received from the command line. Otherwise the child will trigger a lookup
3565 failure and thereby defer the delivery if it tries to use (for example) a
3566 cached ldap connection that the parent has called unbind on.
3568 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3569 of $address_data from the recipient verification was clobbered by the
3570 sender verification.
3572 13. The value of address_data from a sender verification is now available in
3573 $sender_address_data in subsequent conditions in the ACL statement.
3575 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3577 15. Added a new option "connect=<time>" to callout options, to set a different
3580 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3581 was its contents. (It was OK if the option was not defined at all.)
3583 17. A "Completed" log line is now written for messages that are removed from
3584 the spool by the -Mrm option.
3586 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3587 information about exactly what failed.
3589 19. Added -dd to debug only the daemon process.
3591 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3592 handles timeouts, both on the server side and network timeouts. Renamed the
3593 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3595 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3596 It is now set to "smtps".
3598 22. $host_address is now set to the target address during the checking of
3599 ignore_target_hosts.
3601 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3602 being passed; this would have caused $sender_host_name to have been used if
3603 matching the list had actually called for a host name (not very likely,
3604 since this list is usually IP addresses). A host name is now passed as
3607 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3608 code that specifies a non-restarting handler (typically sigaction() in
3609 modern systems) in an attempt to fix a rare and obscure crash bug.
3611 25. Narrowed the window for a race in the daemon that could cause it to ignore
3612 SIGCHLD signals. This is not a major problem, because they are used only to
3613 wake it up if nothing else does.
3615 26. A malformed maildirsize file could cause Exim to calculate negative values
3616 for the mailbox size or file count. Odd effects could occur as a result.
3617 The maildirsize information is now recalculated if the size or filecount
3620 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3621 support for a long time. Removed HAVE_SYS_VFS_H.
3623 28. Installed the latest version of exipick from John Jetmore.
3625 29. In an address list, if the pattern was not a regular expression, an empty
3626 subject address (from a bounce message) matched only if the pattern was an
3627 empty string. Non-empty patterns were not even tested. This was the wrong
3628 because it is perfectly reasonable to use an empty address as part of a
3629 database query. An empty address is now tested by patterns that are
3630 lookups. However, all the other forms of pattern expect the subject to
3631 contain a local part and a domain, and therefore, for them, an empty
3632 address still always fails if the pattern is not itself empty.
3634 30. Exim went into a mad DNS loop when attempting to do a callout where the
3635 host was specified on an smtp transport, and looking it up yielded more
3636 than one IP address.
3638 31. Re-factored the code for checking spool and log partition space into a
3639 function that finds that data and another that does the check. The former
3640 is then used to implement four new variables: $spool_space, $log_space,
3641 $spool_inodes, and $log_inodes.
3643 32. The RFC2047 encoding function was originally intended for short strings
3644 such as real names; it was not keeping to the 75-character limit for
3645 encoded words that the RFC imposes. It now respects the limit, and
3646 generates multiple encoded words if necessary. To be on the safe side, I
3647 have increased the buffer size for the ${rfc2047: expansion operator from
3650 33. It is now permitted to omit both strings after an "if" condition; if the
3651 condition is true, the result is "true". As before, when the second string
3652 is omitted, a false condition yields an empty string. This makes it less
3653 cumbersome to write custom ACL and router conditions.
3655 34. Failure to deliver a bounce message always caused it to be frozen, even if
3656 there was an errors_to setting on the router. The errors_to setting is now
3659 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3660 canonical form (fully expanded) before being placed in
3661 $sender_host_address.
3663 36. The table in the code that translates DNS record types into text (T_A to
3664 "A" for instance) was missing entries for NS and CNAME. It is just possible
3665 that this could have caused confusion if both these types were looked up
3666 for the same domain, because the text type is used as part of Exim's
3667 per-process caching. But the chance of anyone hitting this buglet seems
3670 37. The dnsdb lookup has been extended in a number of ways.
3672 (1) There is a new type, "zns", which walks up the domain tree until it
3673 finds some nameserver records. It should be used with care.
3675 (2) There is a new type, "mxh", which is like "mx" except that it returns
3676 just the host names, not the priorities.
3678 (3) It is now possible to give a list of domains (or IP addresses) to be
3679 looked up. The behaviour when one of the lookups defers can be
3680 controlled by a keyword.
3682 (4) It is now possible to specify the separator character for use when
3683 multiple records are returned.
3685 38. The dnslists ACL condition has been extended: it is now possible to supply
3686 a list of IP addresses and/or domains to be looked up in a particular DNS
3689 39. Added log_selector=+queue_time_overall.
3691 40. When running the queue in the test harness, wait just a tad after forking a
3692 delivery process, to get repeatability of debugging output.
3694 41. Include certificate and key file names in error message when GnuTLS fails
3695 to set them up, because the GnuTLS error message doesn't include the name
3696 of the failing file when there is a problem reading it.
3698 42. Allow both -bf and -bF in the same test run.
3700 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3702 44. The "Exiscan patch" is now merged into the mainline Exim source.
3704 45. Sometimes the final signoff response after QUIT could fail to get
3705 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3706 < 0 before doing a fflush(). This bug looks as though it goes back to the
3707 introduction of TLS in release 3.20, but "sometimes" must have been rare
3708 because the tests only now provoked it.
3710 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3711 (this can affect the format of dates).
3713 47. exim_tidydb, when checking for the continued existence of a message for
3714 which it has found a message-specific retry record, was not finding
3715 messages that were in split spool directories. Consequently, it was
3716 deleting retry records that should have stayed in existence.
3718 48. Steve fixed some bugs in eximstats.
3720 49. The SPA authentication driver was not abandoning authentication and moving
3721 on to the next authenticator when an expansion was forced to fail,
3722 contradicting the general specification for all authenticators. Instead it
3723 was generating a temporary error. It now behaves as specified.
3725 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3726 (the order specifies the preference for clients). The order is now AES256,
3727 AES128, 3DES, ARCFOUR128.
3729 51. Small patch to Sieve code - explicitly set From: when generating an
3732 52. Exim crashed if a remote delivery caused a very long error message to be
3733 recorded - for instance if somebody sent an entire SpamAssassin report back
3734 as a large number of 550 error lines. This bug was coincidentally fixed by
3735 increasing the size of one of Exim's internal buffers (big_buffer) that
3736 happened as part of the Exiscan merge. However, to be on the safe side, I
3737 have made the code more robust (and fixed the comments that describe what
3740 53. Now that there can be additional text after "Completed" in log lines (if
3741 the queue_time_overall log selector is set), a one-byte patch to exigrep
3742 was needed to allow it to recognize "Completed" as not the last thing in
3745 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3746 patch that reportedly fixes this has been added. I am not expert enough to
3747 create a test for it. This is what the patch creator wrote:
3749 "I found a little strange behaviour of ldap code when working with
3750 Windows 2003 AD Domain, where users was placed in more than one
3751 Organization Units. When I tried to give exim partial DN, the exit code
3752 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3753 But simultaneously result of request was absolutely normal ldap result,
3754 so I produce this patch..."
3756 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3757 REFERENCE, so I have modified the code to exclude the patch when that macro
3760 55. Some experimental protocols are using DNS PTR records for new purposes. The
3761 keys for these records are domain names, not reversed IP addresses. The
3762 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3763 leaves it alone. Component reversal etc. now happens only for IP addresses.
3766 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3768 57. Double the size of the debug message buffer (to 2048) so that more of very
3769 long debug lines gets shown.
3771 58. The exicyclog utility now does better if the number of log files to keep
3772 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3774 59. Two changes related to the smtp_active_hostname option:
3776 (1) $smtp_active_hostname is now available as a variable.
3777 (2) The default for smtp_banner uses $smtp_active_hostname instead
3778 of $primary_hostname.
3780 60. The host_aton() function is supposed to be passed a string that is known
3781 to be a valid IP address. However, in the case of IPv6 addresses, it was
3782 not checking this. This is a hostage to fortune. Exim now panics and dies
3783 if the condition is not met. A case was found where this could be provoked
3784 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3785 components; fortuitously, this particular loophole had already been fixed
3786 by change 4.50/55 above.
3788 If there are any other similar loopholes, the new check in host_aton()
3789 itself should stop them being exploited. The report I received stated that
3790 data on the command line could provoke the exploit when Exim was running as
3791 exim, but did not say which command line option was involved. All I could
3792 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3793 running as the user.
3796 61. There was a buffer overflow vulnerability in the SPA authentication code
3797 (which came originally from the Samba project). I have added a test to the
3798 spa_base64_to_bits() function which I hope fixes it.
3801 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3802 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3804 63. The daemon start-up calls getloadavg() while still root for those OS that
3805 need the first call to be done as root, but it missed one case: when
3806 deliver_queue_load_max is set with deliver_drop_privilege. This is
3807 necessary for the benefit of the queue runner, because there is no re-exec
3808 when deliver_drop_privilege is set.
3810 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3811 This has been fixed.
3813 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3814 was in use, was not putting the data itself into the right store pool;
3815 consequently, it could be overwritten for a subsequent message in the same
3816 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3819 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3821 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3822 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3823 the function were treating the return as a boolean value, which happened to
3824 work because 0=false and not-0=true, but is not correct code.
3826 68. The host_aton() function was not handling scoped IPv6 addresses (those
3827 with, for example, "%eth0" on the end) correctly.
3829 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3830 negated items (that is, ~something) in unsigned ints. Some compilers
3831 apparently mutter when there is no cast.
3833 70. If an address verification called from an ACL failed, and did not produce a
3834 user-specific message (i.e. there was only a "system" message), nothing was
3835 put in $acl_verify_message. In this situation, it now puts the system
3838 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3839 session; change 4.31/43 added the unwanted input to the log line - except
3840 that it did not do this in the start of session case. It now does.
3842 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3843 This is wrong and can cause the other end to generate a synchronization
3844 error if it is another Exim or anything else that does the synchronization
3845 check. A QUIT command is no longer sent after a timeout.
3847 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3848 during host lookups.
3850 74. The defer_ok option of callout verification was not working if it was used
3851 when verifying addresses in header lines, that is, for this case:
3853 verify = header_sender/callout=defer_ok
3855 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3856 those file descriptors could be used for SMTP connections. If anything
3857 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3858 could be sent to the SMTP client, causing chaos. The daemon now opens
3859 stdin, stdout, and stderr to /dev/null when it puts itself into the
3862 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3863 log by default. The user can override this with suitable Perl magic.
3865 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3866 the log message when discard triggers, was not working for the DATA ACL or
3867 for the non-SMTP ACL.
3869 78. Error message wording change in sieve.c.
3871 79. If smtp_accept_max_per_host was set, the number of connections could be
3872 restricted to fewer than expected, because the daemon was trying to set up
3873 a new connection before checking whether the processes handling previous
3874 connections had finished. The check for completed processes is now done
3875 earlier. On busy systems, this bug wouldn't be noticed because something
3876 else would have woken the daemon, and it would have reaped the completed
3879 80. If a message was submitted locally by a user whose login name contained one
3880 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3881 It caused a spool format error. I have fixed the spool reading code. A
3882 related problem was that the "from" clause in the Received: line became
3883 illegal because of the space(s). It is now covered by ${quote_local_part.
3885 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3888 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3890 83. Patch from Sieve maintainer to fix unterminated string problem in
3891 "vacation" handling.
3893 84. Some minor changes to the Linux configuration files to help with other
3894 OS variants using glibc.
3896 85. One more patch for Sieve to update vacation handling to latest spec.
3899 ----------------------------------------------------
3900 See the note above about the 4.44 and 4.50 releases.
3901 ----------------------------------------------------
3907 1. Change 4.43/35 introduced a bug that caused file counts to be
3908 incorrectly computed when quota_filecount was set in an appendfile
3911 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3912 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3915 3. Give more explanation in the error message when the command for a transport
3916 filter fails to execute.
3918 4. There are several places where Exim runs a non-Exim command in a
3919 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3920 was being done only for the command run by the queryprogram router. It is
3921 now done for all such subprocesses. The other cases are: ${run, transport
3922 filters, and the commands run by the lmtp and pipe transports.
3924 5. Some older OS have a limit of 256 on the maximum number of file
3925 descriptors. Exim was using setrlimit() to set 1000 as a large value
3926 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3927 systems. I've change it so that if it can't get 1000, it tries for 256.
3929 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3930 was an oversight, and furthermore, ever since the addition of extra
3931 controls (e.g. 4.43/32), the checks on when to allow different forms of
3932 "control" were broken. There should now be diagnostics for all cases when a
3933 control that does not make sense is encountered.
3935 7. $recipients is now available in the predata ACL (oversight).
3937 8. Tidy the search cache before the fork to do a delivery from a message
3938 received from the command line. Otherwise the child will trigger a lookup
3939 failure and thereby defer the delivery if it tries to use (for example) a
3940 cached ldap connection that the parent has called unbind on.
3942 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3943 of $address_data from the recipient verification was clobbered by the
3944 sender verification.
3946 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3947 was its contents. (It was OK if the option was not defined at all.)
3949 11. A "Completed" log line is now written for messages that are removed from
3950 the spool by the -Mrm option.
3952 12. $host_address is now set to the target address during the checking of
3953 ignore_target_hosts.
3955 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3956 being passed; this would have caused $sender_host_name to have been used if
3957 matching the list had actually called for a host name (not very likely,
3958 since this list is usually IP addresses). A host name is now passed as
3961 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3962 code that specifies a non-restarting handler (typically sigaction() in
3963 modern systems) in an attempt to fix a rare and obscure crash bug.
3965 15. Narrowed the window for a race in the daemon that could cause it to ignore
3966 SIGCHLD signals. This is not a major problem, because they are used only to
3967 wake it up if nothing else does.
3969 16. A malformed maildirsize file could cause Exim to calculate negative values
3970 for the mailbox size or file count. Odd effects could occur as a result.
3971 The maildirsize information is now recalculated if the size or filecount
3974 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3975 support for a long time. Removed HAVE_SYS_VFS_H.
3977 18. Updated exipick to current release from John Jetmore.
3979 19. Allow an empty sender to be matched against a lookup in an address list.
3980 Previously the only cases considered were a regular expression, or an
3983 20. Exim went into a mad DNS lookup loop when doing a callout where the
3984 host was specified on the transport, if the DNS lookup yielded more than
3987 21. The RFC2047 encoding function was originally intended for short strings
3988 such as real names; it was not keeping to the 75-character limit for
3989 encoded words that the RFC imposes. It now respects the limit, and
3990 generates multiple encoded words if necessary. To be on the safe side, I
3991 have increased the buffer size for the ${rfc2047: expansion operator from
3994 22. Failure to deliver a bounce message always caused it to be frozen, even if
3995 there was an errors_to setting on the router. The errors_to setting is now
3998 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3999 canonical form (fully expanded) before being placed in
4000 $sender_host_address.
4002 24. Updated eximstats to version 1.33
4004 25. Include certificate and key file names in error message when GnuTLS fails
4005 to set them up, because the GnuTLS error message doesn't include the name
4006 of the failing file when there is a problem reading it.
4008 26. Expand error message when OpenSSL has problems setting up cert/key files.
4011 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4012 (this can affect the format of dates).
4014 28. exim_tidydb, when checking for the continued existence of a message for
4015 which it has found a message-specific retry record, was not finding
4016 messages that were in split spool directories. Consequently, it was
4017 deleting retry records that should have stayed in existence.
4019 29. eximstats updated to version 1.35
4020 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4021 1.35 - bugfix such that pie charts by volume are generated correctly
4023 30. The SPA authentication driver was not abandoning authentication and moving
4024 on to the next authenticator when an expansion was forced to fail,
4025 contradicting the general specification for all authenticators. Instead it
4026 was generating a temporary error. It now behaves as specified.
4028 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4029 (the order specifies the preference for clients). The order is now AES256,
4030 AES128, 3DES, ARCFOUR128.
4032 31. Small patch to Sieve code - explicitly set From: when generating an
4035 32. Exim crashed if a remote delivery caused a very long error message to be
4036 recorded - for instance if somebody sent an entire SpamAssassin report back
4037 as a large number of 550 error lines. This bug was coincidentally fixed by
4038 increasing the size of one of Exim's internal buffers (big_buffer) that
4039 happened as part of the Exiscan merge. However, to be on the safe side, I
4040 have made the code more robust (and fixed the comments that describe what
4043 33. Some experimental protocols are using DNS PTR records for new purposes. The
4044 keys for these records are domain names, not reversed IP addresses. The
4045 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4046 leaves it alone. Component reversal etc. now happens only for IP addresses.
4049 34. The host_aton() function is supposed to be passed a string that is known
4050 to be a valid IP address. However, in the case of IPv6 addresses, it was
4051 not checking this. This is a hostage to fortune. Exim now panics and dies
4052 if the condition is not met. A case was found where this could be provoked
4053 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4054 components; fortuitously, this particular loophole had already been fixed
4055 by change 4.50/55 or 4.44/33 above.
4057 If there are any other similar loopholes, the new check in host_aton()
4058 itself should stop them being exploited. The report I received stated that
4059 data on the command line could provoke the exploit when Exim was running as
4060 exim, but did not say which command line option was involved. All I could
4061 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4062 running as the user.
4065 35. There was a buffer overflow vulnerability in the SPA authentication code
4066 (which came originally from the Samba project). I have added a test to the
4067 spa_base64_to_bits() function which I hope fixes it.
4070 36. The daemon start-up calls getloadavg() while still root for those OS that
4071 need the first call to be done as root, but it missed one case: when
4072 deliver_queue_load_max is set with deliver_drop_privilege. This is
4073 necessary for the benefit of the queue runner, because there is no re-exec
4074 when deliver_drop_privilege is set.
4076 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4077 was in use, was not putting the data itself into the right store pool;
4078 consequently, it could be overwritten for a subsequent message in the same
4079 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4082 38. Sometimes the final signoff response after QUIT could fail to get
4083 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4084 < 0 before doing a fflush(). This bug looks as though it goes back to the
4085 introduction of TLS in release 3.20, but "sometimes" must have been rare
4086 because the tests only now provoked it.
4092 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4093 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4094 Now it returns an integer. A number of calls were still expecting a T/F
4095 return. Fortuitously, in all cases, the tests worked in OK situations,
4096 which is the norm. However, things would have gone wrong on any write
4097 failures on the smtp file descriptor. This function is used when sending
4098 messages over SMTP and also when doing verify callouts.
4100 2. When Exim is called to do synchronous delivery of a locally submitted
4101 message (the -odf or -odi options), it no longer closes stderr before doing
4104 3. Implemented the mua_wrapper option.
4106 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4108 5. Implemented the functions header_remove(), header_testname(),
4109 header_add_at_position(), and receive_remove_recipient(), and exported them
4112 6. If an ACL "warn" statement specified the addition of headers, Exim already
4113 inserted X-ACL-Warn: at the start if there was no header name. However, it
4114 was not making this test for the second and subsequent header lines if
4115 there were newlines in the string. This meant that an invalid header could
4116 be inserted if Exim was badly configured.
4118 7. Allow an ACL "warn" statement to add header lines at the start or after all
4119 the Received: headers, as well as at the end.
4121 8. Added the rcpt_4xx retry error code.
4123 9. Added postmaster_mailfrom=xxx to callout verification option.
4125 10. Added mailfrom=xxxx to the callout verification option, for verify=
4128 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4129 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4131 12. Inserted some casts to stop certain compilers warning when using pointer
4132 differences as field lengths or precisions in printf-type calls (mostly
4133 affecting debugging statements).
4135 13. Added optional readline() support for -be (dynamically loaded).
4137 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4138 same clock tick as a message's arrival, so that its received time was the
4139 same as the "first fail" time on the retry record, and that message
4140 remained on the queue past the ultimate address timeout, every queue runner
4141 would try a delivery (because it was past the ultimate address timeout) but
4142 after another failure, the ultimate address timeout, which should have then
4143 bounced the address, did not kick in. This was a "< instead of <=" error;
4144 in most cases the first failure would have been in the next clock tick
4145 after the received time, and all would be well.
4147 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4148 being recognized when the domain list was tested by the match_domain
4149 condition in an expansion string.
4151 16. Added the ${str2b64: operator.
4153 17. Exim was always calling setrlimit() to set a large limit for the number of
4154 processes, without checking whether the existing limit was already
4155 adequate. (It did check for the limit on file descriptors.) Furthermore,
4156 errors from getrlimit() and setrlimit() were being ignored. Now they are
4157 logged to the main and panic logs, but Exim does carry on, to try to do its
4158 job under whatever limits there are.
4160 18. Imported PCRE 5.0.
4162 19. Trivial typo in log message " temporarily refused connection" (the leading
4165 20. If the log selector return_path_on_delivery was set and an address was
4166 redirected to /dev/null, the delivery process crashed because it assumed
4167 that a return path would always be set for a "successful" delivery. In this
4168 case, the whole delivery is bypassed as an optimization, and therefore no
4171 21. Internal re-arrangement: the function for sending a challenge and reading
4172 a response while authentication was assuming a zero-terminated challenge
4173 string. It's now changed to take a pointer and a length, to allow for
4174 binary data in such strings.
4176 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4178 23. Exim was not respecting finduser_retries when seeking the login of the
4179 uid under which it was called; it was always trying 10 times. (The default
4180 setting of finduser_retries is zero.) Also, it was sleeping after the final
4181 failure, which is pointless.
4183 24. Implemented tls_on_connect_ports.
4185 25. Implemented acl_smtp_predata.
4187 26. If the domain in control=submission is set empty, Exim assumes that the
4188 authenticated id is a complete email address when it generates From: or
4189 Sender: header lines.
4191 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4192 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4193 chown and chgrp in /bin and hostname in /usr/bin.
4195 28. Exim was keeping the "process log" file open after each use, just as it
4196 does for the main log. This opens the possibility of it remaining open for
4197 long periods when the USR1 signal hits a daemon. Occasional processlog
4198 errors were reported, that could have been caused by this. Anyway, it seems
4199 much more sensible not to leave this file open at all, so that is what now
4202 29. The long-running daemon process does not normally write to the log once it
4203 has entered its main loop, and it closes the log before doing so. This is
4204 so that log files can straightforwardly be renamed and moved. However,
4205 there are a couple of unusual error situations where the daemon does write
4206 log entries, and I had neglected to close the log afterwards.
4208 30. The text of an SMTP error response that was received during a remote
4209 delivery was being truncated at 512 bytes. This is too short for some of
4210 the long messages that one sometimes sees. I've increased the limit to
4213 31. It is now possible to make retry rules that apply only when a message has a
4214 specific sender, in particular, an empty sender.
4216 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4217 it possible to be selective about when SMTP synchronization is enforced.
4219 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4221 32. Implemented hosts_connection_nolog.
4223 33. Added an ACL for QUIT.
4225 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4228 35. Added mailbox_size and mailbox_filecount to appendfile.
4230 36. Added control = no_multiline_responses to ACLs.
4232 37. There was a bug in the logic of the code that waits for the clock to tick
4233 in the case where the clock went backwards by a substantial amount such
4234 that the microsecond fraction of "now" was more than the microsecond
4235 fraction of "then" (but the whole seconds number was less).
4237 38. Added support for the libradius Radius client library this is found on
4238 FreeBSD (previously only the radiusclient library was supported).
4244 1. When certain lookups returned multiple values in the form name=value, the
4245 quoting of the values was not always being done properly. Specifically:
4246 (a) If the value started with a double quote, but contained no whitespace,
4248 (b) If the value contained whitespace other than a space character (i.e.
4249 tabs or newlines or carriage returns) it was not quoted.
4250 This fix has been applied to the mysql and pgsql lookups by writing a
4251 separate quoting function and calling it from the lookup code. The fix
4252 should probably also be applied to nisplus, ibase and oracle lookups, but
4253 since I cannot test any of those, I have not disturbed their existing code.
4255 2. A hit in the callout cache for a specific address caused a log line with no
4256 reason for rejecting RCPT. Now it says "Previous (cached) callout
4257 verification failure".
4259 3. There was an off-by-one bug in the queryprogram router. An over-long
4260 return line was truncated at 256 instead of 255 characters, thereby
4261 overflowing its buffer with the terminating zero. As well as fixing this, I
4262 have increased the buffer size to 1024 (and made a note to document this).
4264 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4265 when Exim is waiting for an SMTP response from a remote server, Exim
4266 restarts its select() call on the socket, thereby resetting its timeout.
4267 This is not a problem when such interrupts are rare. Somebody set up a cron
4268 job to run exiwhat every 2 minutes, which is less than the normal select()
4269 timeout (5 or 10 minutes). This meant that the select() timeout never
4270 kicked in because it was always reset. I have fixed this by comparing the
4271 time when an interrupt arrives with the time at the start of the first call
4272 to select(). If more time than the timeout has elapsed, the interrupt is
4273 treated as a timeout.
4275 5. Some internal re-factoring in preparation for the addition of Sieve
4276 extensions (by MH). In particular, the "personal" test is moved to a
4277 separate function, and given an option for scanning Cc: and Bcc: (which is
4278 not set for Exim filters).
4280 6. When Exim created an email address using the login of the caller as the
4281 local part (e.g. when creating a From: or Sender: header line), it was not
4282 quoting the local part when it contained special characters such as @.
4284 7. Installed new OpenBSD configuration files.
4286 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4287 try to make them clearer.
4289 9. Callout options, other than the timeout value, were being ignored when
4290 verifying sender addresses in header lines. For example, when using
4292 verify = header_sender/callout=no_cache
4294 the cache was (incorrectly) being used.
4296 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4297 only the Cygwin environment.
4299 11. When return_path_on_delivery was set as a log selector, if different remote
4300 addresses in the same message used different return paths and parallel
4301 remote delivery occurred, the wrong values would sometimes be logged.
4302 (Whenever a remote delivery process finished, the return path value from
4303 the most recently started remote delivery process was logged.)
4305 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4306 lines when AUTH and/or TLS are in use. This is the "received protocol"
4307 field. Exim used to use "asmtp" for authenticated SMTP, without any
4308 indication (in the protocol name) for TLS use. Now it follows the RFC and
4309 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4310 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4311 names appear in log lines as well as in Received: header lines.
4313 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4314 extensions, and comparison tests, and to fix some bugs.
4316 14. Changes to the "personal" filter test:
4318 (1) The test was buggy in that it was just doing the equivalent of
4319 "contains" tests on header lines. For example, if a user's address was
4320 anne@some.where, the "personal" test would incorrectly be true for
4322 To: susanne@some.where
4324 This test is now done by extracting each address from the header in turn,
4325 and checking the entire address. Other tests that are part of "personal"
4326 are now done using regular expressions (for example, to check local parts
4327 of addresses in From: header lines).
4329 (2) The list of non-personal local parts in From: addresses has been
4330 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4331 taken from the Sieve specification recommendations.
4333 (3) If the message contains any header line starting with "List-" it is
4334 treated as non-personal.
4336 (4) The test for "circular" in the Subject: header line has been removed
4337 because it now seems ill-conceived.
4339 15. Minor typos in src/EDITME comments corrected.
4341 16. Installed latest exipick from John Jetmore.
4343 17. If headers_add on a router specified a text string that was too long for
4344 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4345 of string_sprintf() is now avoided.
4347 18. $message_body_size was not set (it was always zero) when running the DATA
4348 ACL and the local_scan() function.
4350 19. For the "mail" command in an Exim filter, no default was being set for
4351 the once_repeat time, causing a random time value to be used if "once" was
4352 specified. (If the value happened to be <= 0, no repeat happened.) The
4353 default is now 0s, meaning "never repeat". The "vacation" command was OK
4354 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4355 (I found it when inspecting the code).
4357 20. There is now an overall timeout for performing a callout verification. It
4358 defaults to 4 times the callout timeout, which applies to individual SMTP
4359 commands during the callout. The overall timeout applies when there is more
4360 than one host that can be tried. The timeout is checked before trying the
4361 next host. This prevents very long delays if there are a large number of
4362 hosts and all are timing out (e.g. when the network connections are timing
4363 out). The value of the overall timeout can be changed by specifying an
4364 additional sub-option for "callout", called "maxwait". For example:
4366 verify = sender/callout=5s,maxwait=20s
4368 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4369 to the end before writing, but this should make it even safer).
4371 22. Exim was forgetting that it had advertised PIPELINING for the second and
4372 subsequent messages on an SMTP connection. It was also not resetting its
4373 memory on STARTTLS and an internal HELO.
4375 23. When Exim logs an SMTP synchronization error within a session, it now
4376 records whether PIPELINING has been advertised or not.
4378 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4379 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4380 rather than long int.
4382 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4384 26. Added the never_mail option to autoreply.
4390 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4391 crash if the getsockname() call failed; this can happen if a connection is
4392 closed very soon after it is established. The problem was simply in the
4393 order in which certain operations were done, causing Exim to try to write
4394 to the SMTP stream before it had set up the file descriptor. The bug has
4395 been fixed by making things happen in the correct order.
4401 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4402 before the connection was closed, thus losing the rejection response.
4404 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4405 some early Solaris releases, but causes trouble in current releases where
4406 socklen_t is defined.
4408 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4411 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4414 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4415 that releasing the top of it at the end releases what was used for sub-
4416 expansions (unless the block got too big). However, discard this block if
4417 the first thing is a variable or header, so that we can use its block when
4418 it is dynamic (useful for very large $message_headers, for example).
4420 6. Lookups now cache *every* query, not just the most recent. A new, separate
4421 store pool is used for this. It can be recovered when all lookup caches are
4422 flushed. Lookups now release memory at the end of their result strings.
4423 This has involved some general refactoring of the lookup sources.
4425 7. Some code has been added to the store_xxx() functions to reduce the amount
4426 of flapping under certain conditions.
4428 8. log_incoming_interface used to affect only the <= reception log lines. Now
4429 it causes the local interface and port to be added to several more SMTP log
4430 lines, for example "SMTP connection from", and rejection lines.
4432 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4434 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4436 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4437 could be overwritten at the end of the current message (or the start of a
4438 new message if it was set in a HELO ACL). The value is now preserved for
4439 the duration of the SMTP connection.
4441 12. If a transport had a headers_rewrite setting, and a matching header line
4442 contained an unqualified address, that address was qualified, even if it
4443 did not match any rewriting rules. The underlying bug was that the values
4444 of the flags that permit the existence of unqualified sender and recipient
4445 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4446 non-local messages, and by -bnq for local messages) were not being
4447 preserved with the message after it was received.
4449 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4450 "next input=" as part of the text comprising the host identity instead of
4451 the correct text. The code was using the same buffer for two different
4452 strings. However, depending on which order the printing function evaluated
4453 its arguments, the bug did not always show up. Under Linux, for example, my
4454 test suite worked just fine.
4456 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4457 doesn't work with some older versions of Perl. It has been changed to "my",
4458 which in any case is probably the better facility to use.
4460 15. A really picky compiler found some instances of statements for creating
4461 error messages that either had too many or two few arguments for the format
4464 16. The size of the buffer for calls to the DNS resolver has been increased
4465 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4466 for addresses that have a lot of PTR records. This alleviates a problem; it
4467 does not fully solve it.
4469 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4470 buffer now truncates the list and logs the incident, which is the same
4471 action as happens when Exim is looking up a host name and its aliases.
4472 Previously in this situation something unpredictable would happen;
4473 sometimes it was "internal error: store_reset failed".
4475 18. If a server dropped the connection unexpectedly when an Exim client was
4476 using GnuTLS and trying to read a response, the client delivery process
4477 crashed while trying to generate an error log message.
4479 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4480 string, for example:
4482 warn message = H1: something\nH2: something
4484 the text was added as a single header line from Exim's point of view
4485 though it ended up OK in the delivered message. However, searching for the
4486 second and subsequent header lines using $h_h2: did not work. This has been
4487 fixed. Similarly, if a system filter added multiple headers in this way,
4488 the routers could not see them.
4490 20. Expanded the error message when iplsearch is called with an invalid key to
4491 suggest using net-iplsearch in a host list.
4493 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4494 ACLs are no longer actually imposed (and a message to that effect is
4497 22. If a "gecos" field in a passwd entry contained escaped characters, in
4498 particular, if it contained a \" sequence, Exim got it wrong when building
4499 a From: or a Sender: header from that name. A second bug also caused
4500 incorrect handling when an unquoted " was present following a character
4501 that needed quoting.
4503 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4504 was not being matched caselessly.
4506 24. Arranged for all hyphens in the exim.8 source to be escaped with
4509 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4510 properly. Recipient callout cache records were still being keyed to include
4511 the sender, even when use_sender was set false. This led to far more
4512 callouts that were necessary. The sender is no longer included in the key
4513 when use_sender is false.
4515 26. Added "control = submission" modifier to ACLs.
4517 27. Added the ${base62d: operator to decode base 62 numbers.
4519 28. dnsdb lookups can now access SRV records.
4521 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4522 the configuration file.
4524 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4525 (-v) mode. This makes the output for a verbose queue run more intelligible.
4527 31. Added a use_postmaster feature to recipient callouts.
4529 32. Added the $body_zerocount variable, containing the number of binary zero
4530 bytes in the message body.
4532 33. The time of last modification of the "new" subdirectory is now used as the
4533 "mailbox time last read" when there is a quota error for a maildir
4536 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4538 35. Added +ignore_unknown as a special item in host lists.
4540 36. Code for decoding IPv6 addresses in host lists is now included, even if
4541 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4542 address was recognized as an IP address, but was then not correctly decoded
4543 into binary, causing unexpected and incorrect effects when compared with
4550 1. Very minor rewording of debugging text in manualroute to say "list of
4551 hosts" instead of "hostlist".
4553 2. If verify=header_syntax was set, and a header line with an unqualified
4554 address (no domain) and a large number of spaces between the end of the
4555 name and the colon was received, the reception process suffered a buffer
4556 overflow, and (when I tested it) crashed. This was caused by some obsolete
4557 code that should have been removed. The fix is to remove it!
4559 3. When running in the test harness, delay a bit after writing a bounce
4560 message to get a bit more predictability in the log output.
4562 4. Added a call to search_tidyup() just before forking a reception process. In
4563 theory, someone could use a lookup in the expansion of smtp_accept_max_
4564 per_host which, without the tidyup, could leave open a database connection.
4566 5. Added the variables $recipient_data and $sender_data which get set from a
4567 lookup success in an ACL "recipients" or "senders" condition, or a router
4568 "senders" option, similar to $domain_data and $local_part_data.
4570 6. Moved the writing of debug_print from before to after the "senders" test
4573 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4574 problems for message scanning, either using a data ACL, or using
4575 local_scan() because the Received: header was not generated till after they
4576 were called (in order to set the time as the time of reception completion).
4577 I have revised the way this works. The header is now generated after the
4578 body is received, but before the ACL or local_scan() are called. After they
4579 are run, the timestamp in the header is updated.
4585 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4586 before starting a queue runner without re-exec. This happened only when
4587 deliver_drop_privilege was set or when the Exim user was set to root. The
4588 effect of the bug was that timeouts during subsequent deliveries caused
4589 crashes instead of being properly handled. The handler is now left at its
4590 default (and expected) setting.
4592 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4593 message, again when deliver_drop_privilege is set or Exim is run as root.
4594 The bug described in (1) was not present in this case, but the tidying up
4595 of the other signals was missing. I have made the two cases consistent.
4597 3. The ignore_target_hosts setting on a manualroute router was being ignored
4598 for hosts that were looked up using the /MX notation.
4600 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4603 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4604 operated on the sender address. After changing the $sender_address to <>
4605 for the sender address verify, Exim was re-instated it as the original
4606 (before rewriting) address, but remembering that it had rewritten it, so it
4607 wasn't rewriting it again. This bug also had the effect of breaking the
4608 sender address verification caching when the sender address was rewritten.
4610 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4611 This has been changed so that if the ip literal address matches
4612 ignore_target_hosts, the router declines.
4614 7. Added expansion conditions match_domain, match_address, and match_local_
4615 part (NOT match_host).
4617 8. The placeholder for the Received: header didn't have a length field set.
4619 9. Added code to Exim itself and to exim_lock to test for a specific race
4620 condition that could lead to file corruption when using MBX delivery. The
4621 issue is with the lockfile that is created in /tmp. If this file is removed
4622 after a process has opened it but before that process has acquired a lock,
4623 there is the potential for a second process to recreate the file and also
4624 acquire a lock. This could lead to two Exim processes writing to the file
4625 at the same time. The added code performs the same test as UW imapd; it
4626 checks after acquiring the lock that its file descriptor still refers to
4627 the same named file.
4629 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4630 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4633 11. Added the smtp_active_hostname option. If used, this will typically be made
4634 to depend on the incoming interface address. Because $interface_address is
4635 not set up until the daemon has forked a reception process, error responses
4636 that can happen earlier (such as "too many connections") no longer contain
4639 12. If an expansion in a condition on a "warn" statement fails because a lookup
4640 defers, the "warn" statement is abandoned, and the next ACL statement is
4641 processed. Previously this caused the whole ACL to be aborted.
4643 13. Added the iplsearch lookup type.
4645 14. Added ident_timeout as a log selector.
4647 15. Added tls_certificate_verified as a log selector.
4649 16. Added a global option tls_require_ciphers (compare the smtp transport
4650 option of the same name). This controls incoming TLS connections.
4652 17. I finally figured out how to make tls_require_ciphers do a similar thing
4653 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4654 before starting the TLS session.
4656 18. Tabs are now shown as \t in -bP output.
4658 19. If the log selector return_path_on_delivery was set, Exim crashed when
4659 bouncing a message because it had too many Received: header lines.
4661 20. If two routers both had headers_remove settings, and the first one included
4662 a superfluous trailing colon, the final name in the first list and the
4663 first name in the second list were incorrectly joined into one item (with a
4664 colon in the middle).
4670 1. Added -C and -D options to the exinext utility, mainly to make it easier
4671 to include in the automated testing, but these could be helpful when
4672 multiple configurations are in use.
4674 2. The exinext utility was not formatting the output nicely when there was
4675 an alternate port involved in the retry record key, nor when there was a
4676 message id as well (for retries that were specific to a specific message
4677 and a specific host). It was also confused by IPv6 addresses, because of
4678 the additional colons they contain. I have fixed the IPv4 problem, and
4679 patched it up to do a reasonable job for IPv6.
4681 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4682 delivery, the log line now contains "pipelined" if PIPELINING was used.
4684 4. An SMTP transport process used to panic and die if the bind() call to set
4685 an explicit outgoing interface failed. This has been changed; it is now
4686 treated in the same way as a connect() failure.
4688 5. A reference to $sender_host_name in the part of a conditional expansion
4689 that was being skipped was still causing a DNS lookup. This no longer
4692 6. The def: expansion condition was not recognizing references to header lines
4693 that used bh_ and bheader_.
4695 7. Added the _cache feature to named lists.
4697 8. The code for checking quota_filecount in the appendfile transport was
4698 allowing one more file than it should have been.
4700 9. For compatibility with Sendmail, the command line option
4708 and sets the incoming protocol and host name (for trusted callers). The
4709 host name and its colon can be omitted when only the protocol is to be set.
4710 Note the Exim already has two private options, -pd and -ps, that refer to
4711 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4712 "s", but I don't think that's a major issue.
4714 10. A number of refactoring changes to the code, none of which should affect
4717 (a) The number of logging options was getting close to filling up the
4718 32-bit word that was used as a bit map. I have split them into two classes:
4719 those that are passed in the argument to log_write(), and those that are
4720 only ever tested independently outside of that function. These are now in
4721 separate 32-bit words, so there is plenty of room for expansion again.
4722 There is no change in the user interface or the logging behaviour.
4724 (b) When building, for example, log lines, the code previously used a
4725 macro that called string_cat() twice, in order to add two strings. This is
4726 not really sufficiently general. Furthermore, there was one instance where
4727 it was actually wrong because one of the argument was used twice, and in
4728 one call a function was used. (As it happened, calling the function twice
4729 did not affect the overall behaviour.) The macro has been replaced by a
4730 function that can join an arbitrary number of extra strings onto a growing
4733 (c) The code for expansion conditions now uses a table and a binary chop
4734 instead of a serial search (which was left over from when there were very
4735 few conditions). Also, it now recognizes conditions like "pam" even when
4736 the relevant support is not compiled in: a suitably worded error message is
4737 given if an attempt is made to use such a condition.
4739 11. Added ${time_interval:xxxxx}.
4741 12. A bug was causing one of the ddress fields not to be passed back correctly
4742 from remote delivery subprocesses. The field in question was not being
4743 subsequently used, so this caused to problems in practice.
4745 13. Added new log selectors queue_time and deliver_time.
4747 14. Might have fixed a bug in maildirsizefile handling that threw up
4748 "unexpected character" debug warnings, and recalculated the data
4749 unnecessarily. In any case, I expanded the warning message to give more
4752 15. Added the message "Restricted characters in address" to the statements in
4753 the default ACL that block characters like @ and % in local parts.
4755 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4756 Three changes have been made:
4758 (a) There was a serious bug; a negative response to MAIL caused the whole
4759 recipient domain to be cached as invalid, thereby blocking all messages
4760 to all local parts at the same domain, from all senders. This bug has
4761 been fixed. The domain is no longer cached after a negative response to
4762 MAIL if the sender used is not empty.
4764 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4767 (c) A new callout option, "use_sender" has been added for people who want
4768 the modified behaviour.
4774 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4777 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4778 indeed breaks things for older releases.
4780 3. Added additional logging to the case where there is a problem reading data
4781 from a filter that is running in a subprocess using a pipe, in order to
4782 try to track down a specific problem.
4784 4. Testing facility fudge: when running in the test harness and attempting
4785 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4786 getting "No route to host". Convert this to a timeout.
4788 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4791 6. Some OS don't have socklen_t but use size_t instead. This affects the
4792 fifth argument of getsockopt() amongst other things. This is now
4793 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4794 can be set for individual OS. I have set it for SunOS5, OSF1, and
4795 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4796 some earlier ones do not.
4798 7. Change 4.30/15 was not doing the test caselessly.
4800 8. The standard form for an IPv6 address literal was being rejected by address
4801 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4802 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4803 this, as well as the form without the "IPv6" on the front (but only when
4804 address literals are enabled, of course).
4806 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4808 10. Exim crashed if a message with an empty sender address specified by -f
4809 encountered a router with an errors_to setting. This could be provoked only
4810 by a command such as
4814 where an empty string was supplied; "<>" did not hit this bug.
4816 11. Installed PCRE release 4.5.
4818 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4819 remained set. It is now erased.
4821 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4822 times from message ids (which are base 36 rather than the normal 62).
4824 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4825 were being counted as actual protocol errors, and logged if the log
4826 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4827 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4828 and DATA following a set of rejected RCPTs do not count as protocol errors.
4829 In other words, Exim assumes they were pipelined, though this may not
4830 actually be the case. Of course, in all cases the client gets an
4831 appropriate error code.
4833 15. If a lookup fails in an ACL condition, a message about the failure may
4834 be available; it is used if testing the ACL cannot continue, because most
4835 such messages specify what the cause of the deferral is. However, some
4836 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4837 that caused an old message to be retained and used if a later statement
4838 caused a defer, replacing the real cause of the deferral.
4840 16. If an IP address had so many PTR records that the DNS lookup buffer
4841 was not large enough to hold them, Exim could crash while trying to process
4842 the truncated data. It now detects and logs this case.
4844 17. Further to 4.21/58, another change has been made: if (and only if) the
4845 first line of a message (the first header line) ends with CRLF, a bare LF
4846 in a subsequent header line has a space inserted after it, so as not to
4847 terminate the header.
4849 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4850 unnecessarily, used atoi() instead of strtol(), and didn't check the
4851 termination when getting file sizes from file names by regex.
4853 19. Completely re-implemented the support for maildirsize files, in the light
4854 of a number of problems with the previous contributed implementation
4855 (4.30/29). In particular:
4857 . If the quota is zero, the maildirsize file is maintained, but no quota is
4860 . If the maildir directory does not exist, it is created before any attempt
4861 to write a maildirsize file.
4863 . The quota value in the file is just a cache; if the quota is changed in
4864 the transport, the new value overrides.
4866 . A regular expression is available for excluding directories from the
4869 20. The autoreply transport checks the characters in options that define the
4870 message's headers; it allows continued headers, but it was checking with
4871 isspace() after an embedded newline instead of explicitly looking for a
4874 21. If all the "regular" hosts to which an address was routed had passed their
4875 expiry times, and had not reached their retry times, the address was
4876 bounced, even if fallback hosts were defined. Now Exim should go on to try
4879 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4880 equivalent code in the SMTP transport. Some hosts send humungous responses
4881 to HELO/EHLO, more than 1024 it seems.
4883 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4884 gives compiler warnings in some environments. I've now done it "properly",
4887 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4888 (because of problems with the built-in one) was declared to return uschar *
4889 instead of char *, causing compiler failure.
4891 25. Fixed a file descriptor leak when processing alias/forward files.
4893 26. Fixed a minor format string issue in dbfn.c.
4895 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4897 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4898 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4901 29. When a sender address is verified, it is cached, to save repeating the test
4902 when there is more than one recipient in a message. However, when the
4903 verification involves a callout, it is possible for different callout
4904 options to be set for different recipients. It is too complicated to keep
4905 track of this in the cache, so now Exim always runs a verification when a
4906 callout is required, relying on the callout cache for the optimization.
4907 The overhead is duplication of the address routing, but this should not be
4910 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4911 to be verified with callout=postmaster, and the main callout worked but the
4912 postmaster check failed, the verification correctly failed. However, if a
4913 subsequent RCPT command asked for sender verification *without* the
4914 postmaster check, incorrect caching caused this verification also to fail,
4917 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4918 it was not caching the DNS options (qualify_single, search_parents) that
4919 were used when the lookup failed. A subsequent lookup with different
4920 options therefore always gave the same answer, though there were cases
4921 where it should not have. (Example: a "domains = !$mx_any" option on a
4922 dnslookup router: the "domains" option is always processed without any
4923 widening, but the router might have qualify_single set.) Now Exim uses the
4924 cached value only when the same options are set.
4926 32. Added John Jetmore's "exipick" utility to the distribution.
4928 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4929 than a timeout (e.g. a certificate is required, and is not provided), an
4930 Exim server now closes the connection immediately. Previously it waited for
4931 the client to close - but if the client is SSL, it seems that they each
4932 wait for each other, leading to a delay before one of them times out.
4934 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4935 maintained 0.8.x compatibility because I don't think many are using it, and
4936 it is clearly obsolete.
4938 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4941 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4942 client certificate was expired. A simple patch fixes this, though I don't
4943 understand the full logic of why the verify callback is called multiple
4946 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4947 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4948 which causes problems with some clients (such as the Certicom SSL Plus
4949 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4950 disables the coutermeasure allowing Eudora to connect."
4952 38. Exim was not checking that a write() to a log file succeeded. This could
4953 lead to Bad Things if a log got too big, in particular if it hit a file
4954 size limit. Exim now panics and dies if it cannot write to a log file, just
4955 as it does if it cannot open a log file.
4957 39. Modified OS/Makefile-Linux so that it now contains
4959 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4961 The two -D definitions ensure that Exim is compiled with large file
4962 support, which makes it possible to handle log files that are bigger than
4965 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4966 instance) a domain was checked against a named list that involved a lookup,
4967 causing $domain_data to be set, then another domain was checked against the
4968 same list, then the first domain was re-checked, the value of $domain_data
4969 after the final check could be wrong. In particular, if the second check
4970 failed, it could be set empty. This bug probably also applied to
4973 41. The strip_trailing_dot option was not being applied to the address given
4974 with the -f command-line option.
4976 42. The code for reading a message's header from the spool was incrementing
4977 $received_count, but never initializing it. This meant that the value was
4978 incorrect (doubled) while delivering a message in the same process in which
4979 it was received. In the most common configuration of Exim, this never
4980 happens - a fresh exec is done - but it can happen when
4981 deliver_drop_privilege is set.
4983 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4984 it now includes up to 150 characters of the unexpected data in the log
4987 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4988 and building data strings. The size of both of these buffers was 10 000
4989 bytes - far larger than anybody would *ever* want, thought I. Needless to
4990 say, somebody hit the limit. I have increased the maximum line length to
4991 20 000 and the maximum data length of concatenated lines to 100 000. I have
4992 also fixed two bugs, because there was no checking on these buffers. Tsk,
4993 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4994 buffer is too small.
4996 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4997 lsearch lookups. Now it does.
4999 46. When parsing a route_list item in a manualroute router, a fixed-length
5000 buffer was used for the list of hosts. I made this 1024 bytes long,
5001 thinking that nobody would ever have a list of hosts that long. Wrong.
5002 Somebody had a whole pile of complicated expansion conditions, and the
5003 string was silently truncated, leading to an expansion error. It turns out
5004 that it is easier to change to an unlimited length (owing to other changes
5005 that have happened since this code was originally written) than to build
5006 structure for giving a limitation error. The length of the item that
5007 expands into the list of hosts is now unlimited.
5009 47. The lsearch lookup could not handle data where the length of text line was
5010 more than 4095 characters. Such lines were truncated, leading to shortened
5011 data being returned. It should now handle lines of any length.
5013 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5014 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5017 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5018 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5019 no longer happens in comment lines. A list of replacements is now placed
5020 at the head of all of the source files, except those whose only change is
5021 to replace PERL_COMMAND in the very first #! line.
5023 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5024 messages on the queue, with a bottom-up merge sort, using code contributed
5025 by Michael Haardt. This should make operations like -bp somewhat faster on
5026 large queues. It won't affect queue runners, except when queue_run_in_order
5029 51. Installed eximstats 1.31 in the distribution.
5031 52. Added support for SRV lookups to the dnslookup router.
5033 53. If an ACL referred to $message_body or $message_body_end, the value was not
5034 reset for any messages that followed in the same SMTP session.
5036 54. The store-handling optimization for building very long strings was not
5037 differentiating between the different store pools. I don't think this
5038 actually made any difference in practice, but I've tidied it.
5040 55. While running the routers to verify a sender address, $sender_address
5041 was still set to the sender address. This is wrong, because when routing to
5042 send a bounce to the sender, it would be empty. Therefore, I have changed
5043 it so that, while verifying a sender address, $sender_address is set to <>.
5044 (There is no change to what happens when verifying a recipient address.)
5046 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5047 target A or AAAA records (if not already returned) without resetting the
5048 qualify_single or search_parents options of the DNS resolver. These are
5049 inappropriate in this case because the targets of MX and SRV records must
5050 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5051 target that, when qualified, matched something in the local domain. These
5052 two options are now turned off when doing these lookups.
5054 57. It seems that at least some releases of Reiserfs (which does not have the
5055 concept of a fixed number of inodes) returns zero and not -1 for the
5056 number of available inodes. This interacted badly with check_spool_inodes,
5057 which assumed that -1 was the "no such thing" setting. What I have done is
5058 to check that the total number of inodes is greater than zero before doing
5059 the test of how many are available.
5061 58. When a "warn" ACL statement has a log_message modifier, the message is
5062 remembered, and not repeated. This is to avoid a lot of repetition when a
5063 message has many recipients that cause the same warning to be written.
5064 Howewer, Exim was preserving the list of already written lines for an
5065 entire SMTP session, which doesn't seem right. The memory is now reset if a
5066 new message is started.
5068 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5069 debugging output unless log_rewrite was also set.
5071 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5072 of (int)(handle) when we know that handle contains (void *)(-1).
5074 61. The Exim daemon panic-logs an error return when it closes the incoming
5075 connection. However "connection reset by peer" seems to be common, and
5076 isn't really an error worthy of noting specially, so that particular error
5079 62. When Exim is trying to find all the local interfaces, it used to panic and
5080 die if the ioctl to get the interface flags failed. However, it seems that
5081 on at least one OS (Solaris 9) it is possible to have an interface that is
5082 included in the list of interfaces, but for which you get a failure error
5083 for this call. This happens when the interface is not "plumbed" into a
5084 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5085 failure of the "get flags" call assumes that the interface is down.
5087 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5088 makes life easier for people who are doing arithmetic on fields extracted
5089 from dates, where you often get leading zeros that should not be
5090 interpreted as octal.
5092 64. Added qualify_domain to the redirect router, to override the global
5095 65. If a pathologically long header line contained very many addresses (the
5096 report of this problem mentioned 10 000) and each of them was rewritten,
5097 Exim could use up a very large amount of memory. (It kept on making new
5098 copies of the header line as it rewrote, and never released the old ones.)
5099 At the expense of a bit more processing, the header rewriting function has
5100 been changed so that it no longer eats memory in this way.
5102 66. The generation of the Received: header has been moved from the time that a
5103 message starts to be received, to the time that it finishes. The timestamp
5104 in the Received: header should now be very close to that of the <= log
5105 line. There are two side-effects of this change:
5107 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5108 logged header lines no longer include the local Received: line, because
5109 it has not yet been created. The same applies to a copy of the message
5110 that is returned to a non-SMTP sender when a message is rejected.
5112 (b) When a filter file is tested using -bf, no additional Received: header
5113 is added to the test message. After some thought, I decided that this
5116 This change does not affect the value of $received_for. It is still set
5117 after address rewriting, but before local_scan() is called.
5119 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5121 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5122 gave an unhelpful panic error message, and a defer error. I have managed to
5123 change this behaviour so that it now rejects any supplied certificate,
5124 which seems right, as the list of acceptable certificates is empty.
5126 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5127 gave an unhelpful defer error. I have not managed to make this reject any
5128 supplied certificates, but the error message it gives is "no certificate
5129 supplied", which is not helpful.
5131 70. exigrep's output now also includes lines that are not associated with any
5132 message, but which match the given pattern. Implemented by a patch from
5133 Martin Sluka, which also tidied up the Perl a bit.
5135 71. Recipient callout verification, like sender verification, was using <> in
5136 the MAIL FROM command. This isn't really the right thing, since the actual
5137 sender may affect whether the remote host accepts the recipient or not. I
5138 have changed it to use the actual sender in the callout; this means that
5139 the cache record is now keyed on a recipient/sender pair, not just the
5140 recipient address. There doesn't seem to be a real danger of callout loops,
5141 since a callout by the remote host to check the sender would use <>.
5142 [SEE ABOVE: changed after hitting problems.]
5144 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5145 temporary errors. However, in the case of such a code being given after
5146 the end of a data transmission (i.e. after ".") Exim was failing to write
5147 a retry record for the message. (Yes, there was some broken host that was
5148 actually sending 8xx at this point.)
5150 73. An unknown lookup type in a host list could cause Exim to panic-die when
5151 the list was checked. (An example that provoked this was putting <; in the
5152 middle of a list instead of at the start.) If this happened during a DATA
5153 ACL check, a -D file could be left lying around. This kind of configuration
5154 error no longer causes Exim to die; instead it causes a defer errror. The
5155 incident is still logged to the main and panic logs.
5157 74. Buglet left over from Exim 3 conversion. The message "too many messages
5158 in one connection" was written to the rejectlog but not the mainlog, except
5159 when address rewriting (yes!) was being logged.
5161 75. Added write_rejectlog option.
5163 76. When a system filter was run not as root (that is, when system_filter_user
5164 was set), the values of the $n variables were not being returned to the
5165 main process; thus, they were not subsequently available in the $sn
5168 77. Added +return_path_on_delivery log selector.
5170 78. A connection timeout was being treated differently from recipients deferred
5171 when testing hosts_max_try with a message that was older than the host's
5172 retry timeout. (The host should not be counted, thus allowing all hosts to
5173 be tried at least once before bouncing.) This may have been the cause of an
5174 occasionally reported bug whereby a message would remain on the queue
5175 longer than the retry timeout, but would be bounced if a delivery was
5176 forced. I say "may" because I never totally pinned down the problem;
5177 setting up timeout/retry tests is difficult. See also the next item.
5179 79. The ultimate address timeout was not being applied to errors that involved
5180 a combination of host plus message (for example, a timeout on a MAIL
5181 command). When an address resolved to a number of possible hosts, and they
5182 were not all tried for each delivery (e.g. because of hosts_max_try), a
5183 message could remain on the queue longer than the retry timeout.
5185 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5188 81. Fixed an obscure SMTP outgoing bug which required at least the following
5189 conditions: (a) there was another message waiting for the same server;
5190 (b) the server returned 5xx to all RCPT commands in the first message so
5191 that the message was not completed; (c) the server dropped the connection
5192 or gave a negative response to the RSET that Exim sends to abort the
5193 transaction. The observed case was a dropped connection after DATA that had
5194 been sent in pipelining mode. That is, the server had advertised PIPELINING
5195 but was not implementing it correctly. The effect of the bug was incorrect
5196 behaviour, such as trying another host, and this could lead to a crash.
5202 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5203 and daemon.c were passed as pointers to ints; they should have been
5204 pointers to socklen_t variables (which are typically unsigned ints).
5206 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5209 3. Fixed a really odd bug that affected only the testing scheme; patching a
5210 certain fixed string in the binary changed the value of another string that
5211 happened to be identical to the end of the original first string.
5213 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5214 name", it returns that address as the IP address. On some operating
5215 systems (e.g. Solaris), it also passes back the IP address string as the
5216 "host name". However, on others (e.g. Linux), it passes back an empty
5217 string. Exim wasn't checking for this, and was changing the host name to an
5218 empty string, assuming it had been canonicized.
5220 5. Although rare, it is permitted to have more than one PTR record for a given
5221 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5222 all the names associated with an address, because they do in Solaris.
5223 However, it seems that they do not in Linux for data that comes from the
5224 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5225 I found this out when I moved to a new Linux workstation and tried to run
5226 the Exim test suite.
5228 To get round this problem I have changed the code so that it now does its
5229 own call to the DNS to look up PTR records when searching for a host name.
5230 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5231 addresses that are only in /etc/hosts are still found.
5233 This behaviour is, however, controlled by an option called host_lookup_
5234 order, which defaults to "bydns:byaddr". If people want to use the other
5235 order, or indeed, just use one or the other means of lookup, they can
5236 specify it in this variable.
5238 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5239 some operating systems, this comes back from gethostbyaddr() as an empty
5240 string, and this is what Exim used to test for. However, it seems that in
5241 other systems, "." is yielded. Exim now tests for this case too.
5243 7. The values of check_spool_space and check_log_space are now held internally
5244 as a number of kilobytes instead of an absolute number of bytes. If a
5245 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5246 kilobyte. This means that much larger values can be stored.
5248 8. Exim monitor: an attempt to get the action menu when not actually pointing
5249 at a message produces an empty menu entitled "No message selected". This
5250 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5251 no entries in it ("Shell widget menu has zero width and/or height"). So I
5252 have added a single, blank menu entry in this case.
5254 9. Added ${quote_local_part.
5256 10. MIME decoding is now applied to the contents of Subject: header lines when
5259 11. Now that a reference to $sender_host_address automatically causes a reverse
5260 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5261 host lookup before query-style lookups in lists that might use this
5262 variable. This has therefore been abolished, and the "net-" prefix is no
5263 longer necessary for query-style lookups.
5265 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5266 have been a typo for LFLAGS, so it has been changed.
5268 13. The install script calls Exim with "-C /dev/null" in order to find the
5269 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5270 to be output. Howeve, since Exim outputs its version number before the
5271 error, it didn't break the script. It just looked ugly. I fixed this by
5272 always allowing "-C /dev/null" if the caller is root.
5274 14. Ignore overlarge ACL variable number when reading spool file - insurance
5275 against a later release with more variables having written the file.
5277 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5278 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5279 this, as well as the form without the "IPv6" on the front.
5281 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5282 OS/Makefile-Darwin file.
5284 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5285 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5287 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5288 variable. However, some OS use ints for inodes. I've added cast to long int
5289 to get rid of the compiler warning.
5291 19. I had forgotten to lock out "/../" in configuration file names when
5292 ALT_CONFIG_PREFIX was set.
5294 20. Routers used for verification do not need to specify transports. However,
5295 if such a router generated a host list, and callout was configured, Exim
5296 crashed, because it could not find a port number from the (non-existent)
5297 transport. It now assumes port 25 in this circumstance.
5299 21. Added the -t option to exigrep.
5301 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5302 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5303 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5304 them is set, LOOKUP_LSEARCH is forced.
5306 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5308 24. Added sender and host information to the "rejected by local_scan()" log
5309 line; previously there was no indication of these.
5311 25. Added .include_if_exists.
5313 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5314 Linux. It turns out that not all file systems support this. Apparently some
5315 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5316 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5317 not supported on the file descriptor, is now ignored when Exim is trying to
5318 sync a directory. This applies only to Linux.
5320 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5322 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5323 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5326 29. Added support for maildirsize files from supplied patch (modified a bit).
5328 30. The use of :fail: followed by an empty string could lead Exim to respond to
5329 sender verification failures with (e.g.):
5331 550 Verification failed for <xxx>
5332 550 Sender verify failed
5334 where the first response line was missing the '-' that indicates it is not
5335 the final line of the response.
5337 31. The loop for finding the name of the user that called Exim had a hardwired
5338 limit of 10; it now uses the value of finduser_retries, which is used for
5339 all other user lookups.
5341 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5344 33. Exim was neglecting to zero errno before one call of strtol() when
5345 expanding a string and expecting an integer value. On some systems this
5346 resulted in spurious "integer overflow" errors. Also, it was casting the
5347 result into an int without checking.
5349 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5350 did not work. The code looks as if it has *never* worked, though it appears
5351 to have been documented since at least releast 1.62. I have made it work.
5353 35. The "timeout_DNS" error in retry rules, also documented since at least
5354 1.62, also never worked. As it isn't clear exactly what this means, and
5355 clearly it isn't a major issue, I have abolished the feature by treating it
5356 as "timeout", and writing a warning to the main and panic logs.
5358 36. The display of retry rules for -brt wasn't always showing the error code
5361 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5362 timeout_connect_A, timeout_connect_MX.
5364 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5365 to the empty sender.
5367 39. The daemon was not analysing the content of -oX till after it had closed
5368 stderr and disconnected from the controlling terminal. This meant that any
5369 syntax errors were only noted on the panic log, and the return code from
5370 the command was 0. By re-arranging the code a little, I've made the
5371 decoding happen first, so such errors now appear on stderr, and the return
5372 code is 1. However, the actual setting up of the sockets still happens in
5373 the disconnected process, so errors there are still only recorded on the
5376 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5377 connections (as happens on some IP stacks) was logged at start up time as
5378 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5379 it from "IPv6 and IPv4", which means that two separate sockets are being
5382 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5383 says whether AF_INET or AF_INET6 was passed as an argument.
5385 42. Exiwhat output was messed up when time zones were included in log
5388 43. Exiwhat now gives more information about the daemon's listening ports,
5389 and whether -tls-on-connect was used.
5391 44. The "port" option of the smtp transport is now expanded.
5393 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5394 silently ignored. Now an error message is written to the main and panic
5397 46. There's a new ACL modifier called "logwrite" which writes to a log file
5398 as soon as it is encountered.
5400 47. Added $local_user_uid and $local_user_gid at routing time.
5402 48. Exim crashed when trying to verify a sender address that was being
5405 49. Exim was recognizing only a space character after ".include". It now also
5406 recognizes a tab character.
5408 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5409 extracting the relevant information from the specification. The man page no
5410 longer contains scrambled data for the -d option, and I've added a section
5411 at the front about calling Exim under different names.
5413 51. Added "extra_headers" argument to the "mail" command in filter files.
5415 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5418 53. Installed eximstats 1.29.
5420 54. Added transport_filter_timeout as a generic transport option.
5422 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5423 Cc: header lines. This was required by RFC 822, but it not required by RFC
5426 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5427 incoming messages that did not have them. Now it does so only if the
5428 message originates locally, that is, if there is no associated remote host
5429 address. When Resent- header lines are present, this applies to the Resent-
5430 lines rather than the non-Resent- lines.
5432 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5433 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5435 58. Messages for configuration errors now include the name of the main
5436 configuration file - useful now that there may be more than one file in a
5437 list (.included file names were always shown).
5439 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5440 for those rare installations that do not start the daemon as root or run it
5441 setuid root. I've cut out the call to initgroups() if the daemon is not
5444 60. The Exim user and group can now be bound into the binary as text strings
5445 that are looked up at the start of Exim's processing.
5447 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5449 62. Added $mailstore_basename variable.
5451 63. Installed patch to sieve.c from Michael Haardt.
5453 64. When Exim failed to open the panic log after failing to open the main log,
5454 the original message it was trying to log was written to stderr and debug
5455 output, but if they were not available (the usual case in production), it
5456 was lost. Now it is written to syslog before the two lines that record the
5457 failures to open the logs.
5459 65. Users' Exim filters run in subprocesses under the user's uid. It is
5460 possible for a "deliver" command or an alias in a "personal" command to
5461 provoke an address rewrite. If logging of address rewriting is configured,
5462 this fails because the process is not running as root or exim. There may be
5463 a better way of dealing with this, but for the moment (because 4.30 needs
5464 to be released), I have disabled address rewrite logging when running a
5465 filter in a non-root, non-exim process.
5471 1. The buildconfig auxiliary program wasn't quoting the value set for
5472 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5473 not defined. This bug was masked in 4.22 by the effect that was fixed in
5476 2. Some messages that were rejected after a message id was allocated were
5477 shown as "incomplete" by exigrep. It no longer does this for messages that
5478 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5480 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5481 have allow_domain_literals set, the ID did not get logged in the <= line.
5482 Domain literals are now always recognized in Message-ID: header lines.
5484 4. The first argument for a ${extract expansion item is the key name or field
5485 number. Leading and trailing spaces in this item were not being ignored,
5486 causing some misleading effects.
5488 5. When deliver_drop_privilege was set, single queue runner processes started
5489 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5490 same command in the process it spins off) were not dropping privilege.
5492 6. When the daemon running as "exim" started a queue runner, it always
5493 re-executed Exim in the spun-off process. This is a waste of effort when
5494 deliver_drop_privilege is set. The new process now just calls the
5495 queue-runner function directly.
5501 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5504 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5505 ignored. Though the use of -oP was forcing the writing of a pid file, it
5506 was always written to the default place.
5508 3. If the message "no IP address found for host xxxx" is generated during
5509 incoming verification, it is now followed by identification of the incoming
5510 connection (so you can more easily find what provoked it).
5512 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5514 5. Added some features to "harden" Exim a bit more against certain attacks:
5516 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5517 be put in Local/Makefile. This is like the never_users runtime option,
5518 but it cannot be overridden. The default setting is "root".
5520 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5521 prefix string with which any file named in a -C command line option
5524 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5525 is retained for -C and -D only if the caller of Exim is root. Without
5526 it, the exim user may also use -C and -D and retain privilege.
5528 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5529 command line option is disabled.
5531 6. Macro names set by the -D option must start with an upper case letter, just
5532 like macro names defined in the configuration file.
5534 7. Added "dereference=" facility to LDAP.
5536 8. Two instances of the typo "uknown" in the source files are fixed.
5538 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5539 the Configure-Makefile script screwed up while processing it.
5541 10. Incorporated PCRE 4.4.
5543 11. The SMTP synchronization check was not operating right at the start of an
5544 SMTP session. For example, it could not catch a HELO sent before the client
5545 waited for the greeting. There is now a check for outstanding input at the
5546 point when the greeting is written. Because of the duplex, asynchronous
5547 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5548 way, but not yet received, when the check is performed.
5550 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5551 on TCP/IP sockets, because this apparently causes some broken clients to
5554 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5555 unchanged) from the Cygwin maintainer.
5557 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5558 was supported instead of testing for mbx. Effectively a typo.
5560 15. The spa authenticator server code was not checking that the input it
5561 received was valid base64.
5563 16. The debug output line for the "set" modifier in ACLs was not showing the
5564 name of the variable that was being set.
5566 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5568 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5569 lookup on its own. Something else was needed to trigger the lookup. For
5570 example, a match in host_lookup or the need for a host name in a host list.
5571 Now, if $sender_host_name is referenced and the host name has not yet been
5572 looked up, a lookup is performed. If the lookup fails, the variable remains
5573 empty, and $host_lookup_failed is set to "1".
5575 19. Added "eqi" as a case-independent comparison operator.
5577 20. The saslauthd authentication condition could segfault if neither service
5578 nor realm was specified.
5580 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5581 error message that was logged was misleading, and incoming SMTP
5582 connections were dropped. The message is now more accurate, and temporary
5583 errors are given to SMTP connections.
5585 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5586 (except RSET) with a 421 error, until QUIT is received. However, it was
5587 failing to send a response to QUIT.
5589 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5590 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5591 ACL is now run after the helo_try_verify_hosts code.
5593 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5594 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5595 case-independent, so other case variants are also recognized). Apparently
5596 some systems use these upper case variants.
5598 25. If more than two messages were waiting for the same host, and a transport
5599 filter was specified for the transport, Exim sent two messages over the
5600 same TCP/IP connection, and then failed with "socket operation on non-
5601 socket" when it tried to send the third.
5603 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5605 27. The extern definition of crypt16() in expand.c was not being excluded when
5606 the OS had its own crypt16() function.
5608 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5609 as a preferred synonym for return_size_limit, both as an option and as an
5612 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5614 30. Changed the default configuration ACL to relax the local part checking rule
5615 for addresses that are not in any local domains. For these addresses,
5616 slashes and pipe symbols are allowed within local parts, but the sequence
5617 /../ is explicitly forbidden.
5619 31. SPA server authentication was not clearing the challenge buffer before
5622 32. log_message in a "warn" ACL statement was writing to the reject log as
5623 well as to the main log, which contradicts the documentation and doesn't
5624 seem right (because no rejection is happening). So I have stopped it.
5626 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5627 However, I am unable to do any testing of this.
5629 34. Fixed an infelicity in the appendfile transport. When checking directories
5630 for a mailbox, to see if any needed to be created, it was accidentally
5631 using path names with one or more superfluous leading slashes; tracing
5632 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5634 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5635 contents are added to the log line that is written for every discarded
5636 recipient. (Previously a log_message setting was ignored.)
5638 36. The ${quote: operator now quotes the string if it is empty.
5640 37. The install script runs exim in order to find its version number. If for
5641 some reason other than non-existence or emptiness, which it checks, it
5642 could not run './exim', it was installing it with an empty version number,
5643 i.e. as "exim-". This error state is now caught, and the installation is
5646 38. An argument was missing from the function that creates an error message
5647 when Exim fails to connect to the socket for saslauthd authentication.
5648 This could cause Exim to crash, or give a corrupted message.
5650 39. Added isip, isip4, and isip6 to ${if conditions.
5652 40. The ACL variables $acl_xx are now saved with the message, and can be
5653 accessed later in routers, transports, and filters.
5655 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5656 strings in the file are not string-expanded.
5658 42. If a MAIL command specified a SIZE value that was too large to fit into an
5659 int variable, the check against message_size_limit failed. Such values are
5660 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5661 day this will have to be increased, but I don't think I want to be around
5662 when emails are that large.
5669 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5670 iconv() is not standard in FreeBSD.
5672 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5673 IPv6 enabled. The observed symptom was a segmentation fault on return from
5674 the function os_common_find_running_interfaces() in src/os.c.
5676 3. In the check_special_case() function in daemon.c I had used "errno" as an
5677 argument name, which causes warnings on some systems. This was basically a
5678 typo, since it was named "eno" in the comments!
5680 4. The code that waits for the clock to tick (at a resolution of some fraction
5681 of a second) so as to ensure message-id uniqueness was always waiting for
5682 at least one whole tick, when it could have waited for less. [This is
5683 almost certainly not relevant at current processor speeds, where it is
5684 unlikely to ever wait at all. But we try to future-proof.]
5686 5. The function that sleeps for a time interval that includes fractions of a
5687 second contained a race. It did not block SIGALRM between setting the
5688 timer, and suspending (a couple of lines later). If the interval was short
5689 and the sigsuspend() was delayed until after it had expired, the suspension
5690 never ended. On busy systems this could lead to processes getting stuck for
5693 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5694 process, before it forks any delivery processes. The open lookup caching
5695 mechanism meant that the open file or database connection was passed into
5696 the delivery process. The problem was that delivery processes always tidy
5697 up cached lookup data. This could cause a problem for the next delivery
5698 process started by the queue runner, because the external queue runner
5699 process does not know about the closure. So the next delivery process
5700 still has data in the lookup cache. In the case of a file lookup, there was
5701 no problem because closing a file descriptor in a subprocess doesn't affect
5702 the parent. However, if the lookup was caching a connection to a database,
5703 the connection was closed, and the second delivery process was likely to
5704 see errors such as "PGSQL: query failed: server closed the connection
5705 unexpectedly". The problem has been fixed by closing all cached lookups
5706 in a queue runner before running a delivery process.
5708 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5709 seem to have the "const" qualifier which it has on other OS. I've
5712 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5713 *of the same type* (client or server) with the same public name that an
5714 error should be diagnosed.
5716 9. When Exim looked up a host name for an IP address, but failed to find the
5717 original IP address when looking up the host name (a safety check), it
5718 output the message "<ip address> does not match any IP for NULL", which was
5719 confusing, to say the least. The bug was that the host name should have
5720 appeared instead of "NULL".
5722 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5723 user that is built into the binary, and the -C or -D options is used, root
5724 privilege is dropped before the configuration file is read. In addition,
5725 logging is switched to stderr instead of the normal log files. If the
5726 configuration then re-defines the Exim user, the unprivileged environment
5727 is probably not what is expected, so Exim logs a panic warning message (but
5730 However, if deliver_drop_privilege is set, the unprivileged state may well
5731 be exactly what is intended, so the warning has been cut out in that case,
5732 and Exim is allowed to try to write to its normal log files.
5738 1. smtp_return_error_details was not giving details for temporary sender
5739 or receiver verification errors.
5741 2. Diagnose a configuration error if two authenticators have the same public
5744 3. Exim used not to create the message log file for a message until the first
5745 delivery attempt. This could be confusing when incoming messages were held
5746 for policy or load reasons. The message log file is now created at the time
5747 the message is received, and an initial "Received" line is written to it.
5749 4. The automatically generated man page for command line options had a minor
5750 bug that caused no ill effects; however, a more serious problem was that
5751 the procedure for building the man page automatically didn't always
5752 operate. Consequently, release 4.20 contains an out-of-date version. This
5753 shouldn't happen again.
5755 5. When building Exim with embedded Perl support, the script that builds the
5756 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5757 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5759 6. The freeze_tell option was not being used for messages that were frozen on
5760 arrival, either by an ACL or by local_scan().
5762 7. Added the smtp_incomplete_transaction log selector.
5764 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5765 was accepting AUTH without a new EHLO.
5767 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5768 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5771 10. Logging of TCP/IP connections (when configured) now happens in the main
5772 daemon process instead of the child process, so that the TCP/IP connection
5773 count is more accurate (but it can never be perfect).
5775 11. The use of "drop" in a nested ACL was not being handled correctly in the
5776 outer ACL. Now, if condition failure induced by the nested "drop" causes
5777 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5778 or "require"), the connection is dropped.
5780 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5781 that yield "discard" can now be used with an "accept" or a "discard" verb,
5782 but an error is generated for any others (because I can't see a useful way
5783 to define what should happen).
5785 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5786 are now processed in the same way as lines in the Exim configuration file.
5787 In particular, continuation lines are supported.
5789 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5791 15. Added -ti meaning -t -i.
5793 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5794 domains, and warn by logging if others are found.
5796 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5797 returned from GIFCONF when Exim is trying to find the list of interfaces on
5798 a host. The code in os.c has been modified to copy each ifreq to an aligned
5799 structure in all cases.
5801 Also, in some cases, the returned ifreq's were being copied to a 'struct
5802 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5803 means the last couple of bytes of an IPv6 address could be chopped if the
5804 ifreq contained only a normal sockaddr (14 bytes storage).
5806 18. Named domain lists were not supported in the hosts_treat_as_local option.
5807 An entry such as +xxxx was not recognized, and was treated as a literal
5810 19. Ensure that header lines added by a DATA ACL are included in the reject log
5811 if the ACL subsequently rejects the message.
5813 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5814 MD5 (which is deprecated).
5816 21. When testing a filter file using -bf, Exim was writing a message when it
5817 took the sender from a "From " line in the message, but it was not doing so
5818 when it took $return_path from a Return-Path: header line. It now does.
5820 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5821 with a valid header line field name (a series of printing characters
5822 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5824 23. Changed "disc" in the source to "disk" to conform to the documentation and
5825 the book and for uniformity.
5827 24. Ignore Sendmail's -Ooption=value command line item.
5829 25. When execve() failed while trying to run a command in a pipe transport,
5830 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5831 could be confused with a return value of 69 from the command itself. This
5832 has been changed to 127, the value the shell returns if it is asked to run
5833 a non-existent command. The wording for the related log line suggests a
5834 non-existent command as the problem.
5836 26. If received_header_text expands to an empty string, do not add a Received:
5837 header line to the message. (Well, it adds a token one on the spool, but
5838 marks it "old" so that it doesn't get used or transmitted.)
5840 27. Installed eximstats 1.28 (addition of -nt option).
5842 28. There was no check for failure on the call to getsockname() in the daemon
5843 code. This can fail if there is a shortage of resources on the system, with
5844 ENOMEM, for example. A temporary error is now given on failure.
5846 29. Contrary to the C standard, it seems that in some environments, the
5847 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5848 program. Exim now does this explicitly; it affects the formatting of
5849 timestamps using strftime().
5851 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5852 complaints. I've now initialized all the variables, to avoid this.
5854 32. Header lines added by a system filter were not being "seen" during
5855 transport-time rewrites.
5857 33. The info_callback() function passed to OpenSSL is set up with type void
5858 (*)(SSL *, int, int), as described somewhere. However, when calling the
5859 function (actually a macro) that sets it up, the type void(*)() is
5860 expected. I've put in a cast to prevent warnings from picky compilers.
5862 34. If a DNS black list lookup found a CNAME record, but there were no A
5863 records associated with the domain it pointed at, Exim crashed.
5865 35. If a DNS black list lookup returned more than one A record, Exim ignored
5866 all but the first. It now scans all returned addresses if a particular IP
5867 value is being sought. In this situation, the contents of the
5868 $dnslist_value variable are a list of all the addresses, separated by a
5871 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5872 to accept a host name and all its aliases if the forward lookup for any of
5873 them yielded the IP address of the incoming connection. Now it accepts only
5874 those names whose forward lookup yields the correct IP address. Any other
5875 names are discarded. This closes a loophole whereby a rogue DNS
5876 administrator could create reverse DNS records to break through a
5877 wildcarded host restriction in an ACL.
5879 37. If a user filter or a system filter that ran in a subprocess used any of
5880 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5881 the wrong values were passed to the pipe command ($thisaddress had the
5882 value of $0, $0 had the value of $1, etc). This bug was introduced by
5883 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5885 38. Improved the line breaking for long SMTP error messages from ACLs.
5886 Previously, if there was no break point between 40 and 75 characters, Exim
5887 left the rest of the message alone. Two changes have been made: (a) I've
5888 reduced the minimum length to 35 characters; (b) if it can't find a break
5889 point between 35 and 75 characters, it looks ahead and uses the first one
5890 that it finds. This may give the occasional overlong line, but at least the
5891 remaining text gets split now.
5893 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5894 file descriptors might be low, and that setting 1000 would always raise it.
5895 It turns out that in some environments, the limit is already over 1000 and
5896 that lowering it causes trouble. So now Exim takes care not to decrease it.
5898 40. When delivering a message, the value of $return_path is set to $sender_
5899 address at the start of routing (routers may change the value). By an
5900 oversight, this default was not being set up when an address was tested by
5901 -bt or -bv, which affected the outcome if any router or filter referred to
5904 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5905 log only when "message" or "log_message" are set. However, if one of the
5906 conditions was an address verification, or a call to a nested ACL, the
5907 messages generated by the underlying test were being passed through. This
5908 no longer happens. The underlying message is available in $acl_verify_
5909 message for both "message" and "log_message" expansions, so it can be
5910 passed through if needed.
5912 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5913 new expansion $bh_ to give the encoded byte string without charset
5914 translation. Translation happens only if iconv() is available; HAVE_ICONV
5915 indicates this at build time. HEADERS_CHARSET gives the charset to
5916 translate to; headers_charset can change it in the configuration, and
5917 "headers charset" can change it in an individual filter file.
5919 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5920 that creates RFC 2047 encoded "words" labels them as that charset instead
5921 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5922 expansion operator; (ii) when Exim creates a From: line for a local
5923 message; (iii) when a header line is rewritten to include a "phrase" part.
5925 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5926 buggy, causing it to skip the first lines of messages whose message ID
5927 ended in 'D'. This would not have bitten before Exim release 4.14, because
5928 message IDs were unlikely to end in 'D' before then. The effect was to have
5929 incorrect size information for certain domains.
5931 45. #include "config.h" was missing at the start of the crypt16.c module. This
5932 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5935 46. If there was a timeout during a "random" callout check, Exim treated it as
5936 a failure of the random address, and carried on sending RSET and the real
5937 address. If the delay was just some slowness somewhere, the response to the
5938 original RCPT would be taken as a response to RSET and so on, causing
5939 mayhem of various kinds.
5941 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5942 when I implemented it. It didn't allow for the fact that some option values
5943 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5944 the right test for positive values.
5946 48. Domain names in DNS records are case-independent. Exim always looks them up
5947 in lower case. Some resolvers return domain names in exactly the case they
5948 appear in the zone file, that is, they may contain uppercase letters. Not
5949 all resolvers do this - some return always lower case. Exim was treating a
5950 change of case by a resolver as a change of domain, similar to a widening
5951 of a domain abbreviation. This triggered its re-routing code and so it was
5952 trying to route what was effectively the same domain again. This normally
5953 caused routing to fail (because the router wouldn't handle the domain
5954 twice). Now Exim checks for this case specially, and just changes the
5955 casing of the domain that it ultimately uses when it transmits the message
5958 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5961 50. If a filter generated a file delivery with a non-absolute name (possible if
5962 no home directory exists for the router), the forbid_file option was not
5965 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5966 the existing equality matching.
5968 52. Exim was using ints instead of ino_t variables in some places where it was
5969 dealing with inode numbers.
5971 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5972 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5973 called TMPDIR, and if it finds it is different, it changes its value.
5975 54. The smtp_printf() function is now made available to local_scan() so
5976 additional output lines can be written before returning. There is also an
5977 smtp_fflush() function to enable the detection of a dropped connection.
5978 The variables smtp_input and smtp_batched_input are exported to
5981 55. Changed the default runtime configuration: the message "Unknown user"
5982 has been removed from the ACL, and instead placed on the localuser router,
5983 using the cannot_route_message feature. This means that any verification
5984 failures that generate their own messages won't get overridden. Similarly,
5985 the "Unrouteable address" message that was in the ACL for unverifiable
5986 relay addresses has also been removed.
5988 56. Added hosts_avoid_esmtp to the smtp transport.
5990 57. The exicyclog script was not checking for the esoteric option
5991 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5992 will work only if exicyclog is run under the appropriate euid.
5994 58. Following a discussion on the list, the rules by which Exim recognises line
5995 endings on incoming messages have been changed. The -dropcr and drop_cr
5996 options are now no-ops, retained only for backwards compatibility. The
5997 following line terminators are recognized: LF CRLF CR. However, special
5998 processing applies to CR:
6000 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6001 nor a local message in the state where . is a terminator.
6003 (ii) If a bare CR is encountered in a header line, an extra space is added
6004 after the line terminator so as not to end the header. The reasoning
6005 behind this is that bare CRs in header lines are most likely either
6006 to be mistakes, or people trying to play silly games.
6008 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6009 was being incorrectly given as 18 bytes larger than it should have been.
6010 This is a VOB (very old bug).
6012 60. This may never have affected anything current, but just in case it has:
6013 When the local host is found other than at the start of a list of hosts,
6014 the local host, those with the same MX, and any that follow, are discarded.
6015 When the list in question was part of a longer list of hosts, the following
6016 hosts (not currently being processed) were also being discarded. This no
6017 longer happens. I'm not sure if this situation could ever has previously
6020 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6023 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6024 header. This is something that is recommended in a new Internet Draft, and
6025 is something that is documented as being done by Sendmail. There are two
6026 possible values. For messages generated by the autoreply transport, Exim
6029 Auto-Submitted: auto-replied
6031 whereas for all other generated messages (e.g. bounces) it adds
6033 Auto-Submitted: auto-generated
6035 63. The "personal" condition in filters now includes a test for the
6036 Auto-Submitted: header. If it contains the string "auto-" the message it
6037 not considered personal.
6039 64. Added rcpt_include_affixes as a generic transport option.
6041 65. Added queue_only_override (default true).
6043 66. Added the syslog_duplication option.
6045 67. If what should have been the first header line of a message consisted of
6046 a space followed by a colon, Exim was mis-interpreting it as a header line.
6047 It isn't of course - it is syntactically invalid and should therefore be
6048 treated as the start of the message body. The misbehaviour could have
6049 caused a number of strange effects, including loss of data in subsequent
6050 header lines, and spool format errors.
6052 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6053 client host had authenticated. This control can now be exercised by an ACL
6054 for more flexibility.
6056 69. By default, callouts do not happen when testing with -bh. There is now a
6057 variant, -bhc, which does actually run the callout code, including
6058 consulting and updating the callout cache.
6060 70. Added support for saslauthd authentication, courtesy of Alexander
6063 71. If statvfs() failed on the spool or log directories while checking their
6064 size for availability, Exim confusingly gave the error "space shortage".
6065 Furthermore, in debugging mode it crashed with a floating point exception.
6066 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6067 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6068 serious problem, Exim now writes to the main and panic logs when this
6069 happens, with details of the failure. It then refuses to accept the
6070 incoming message, giving the message "spool directory problem" or "log
6071 directory problem" with a 421 code for SMTP messages.
6073 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6074 0, 1, and 2 exist, because some OS complain for execs without them (see
6075 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6076 descriptors. However, the code omitted to check that the open succeeded,
6077 causing mysterious errors if for some reason the permissions on /dev/null
6078 got screwed. Now Exim writes a message to the main and panic logs, and
6079 bombs out if it can't open /dev/null.
6081 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6082 interact so that it is all more flexible. It is supposed to remain
6083 backwards compatible. Also added extra_local_interfaces.
6085 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6086 to bomb out with an assertion failure - to the client this appears as a
6087 connection drop. This problem occurs in the part of the code that was taken
6088 from the Samba project. Fortunately, the assertion is in a very simple
6089 function, so I have fixed this by reproducing the function inline in the
6090 one place where it is called, and arranging for authentication to fail
6091 instead of killing the process with assert().
6093 75. The SPA client code was not working when the server requested OEM rather
6094 than Unicode encoding.
6096 76. Added code to make require_files with a specific uid setting more usable in
6097 the case where statting the file as root fails - usually a non-root-mounted
6098 NFS file system. When this happens and the failure is EACCES, Exim now
6099 forks a subprocess and does the per-uid checking as the relevant uid.
6101 77. Added process_log_path.
6103 78. If log_file_path was not explicitly set, a setting of check_log_space or
6104 check_log_inodes was ignored.
6106 79. If a space check for the spool or log partitions fails, the incident is now
6107 logged. Of course, in the latter case the data may get lost...
6109 80. Added the %p formatting code to string_format() so that it can be used to
6110 print addresses in debug_print(). Adjusted all the address printing in the
6111 debugging in store.c to use %p rather than %d.
6113 81. There was a concern that a line of code in smtp_in.c could overflow a
6114 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6115 initially expressed, the concern was not well-founded, because trailing
6116 spaces are removed early. However, if the trailing spaces were followed by
6117 a NULL, they did not get removed, so the overflow was possible. Two fixes
6120 (a) I re-wrote the offending code in a cleaner fashion.
6121 (b) If an incoming SMTP command contains a NULL character, it is rejected
6124 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6125 runs initgroups(), so that if the Exim user is in any additional groups,
6126 they will be used during message reception.
6132 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob