1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
126 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
127 file. Patch from Wolfgang Breyha.
129 JH/25 Expand the coverage of the delivery $host and $host_address to
130 client authenticators run in verify callout. Bug 1476.
132 JH/26 Port service names are now accepted for tls_on_connect_ports, to
133 align with daemon_smtp_ports. Bug 72.
139 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
141 PP/02 Make -n do something, by making it not do something.
142 When combined with -bP, the name of an option is not output.
144 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
147 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
148 $sender_host_name and config options to manage this, and basic check
151 PP/05 DSCP support for outbound connections and control modifier for inbound.
153 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
154 (Only plugin which currently uses this is kerberos4, which nobody should
155 be using, but we should make it available and other future plugins might
156 conceivably use it, even though it would break NAT; stuff *should* be
157 using channel bindings instead).
159 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
160 name; added for Sendmail compatibility; requires admin caller.
161 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
162 just ignore it); requires trusted caller.
163 Also parse but ignore: -Ac -Am -X<logfile>
166 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
168 TL/02 Add +smtp_confirmation as a default logging option.
170 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
171 Patch by Magnus Holmgren from 2007-02-20.
173 TL/04 Bugzilla 1281 - Spec typo.
174 Bugzilla 1283 - Spec typo.
175 Bugzilla 1290 - Spec grammar fixes.
177 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
179 TL/06 Add Experimental DMARC support using libopendmarc libraries.
181 TL/07 Fix an out of order global option causing a segfault. Reported to dev
182 mailing list by by Dmitry Isaikin.
184 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
186 JH/02 Support "G" suffix to numbers in ${if comparisons.
188 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
190 NM/01 Bugzilla 1197 - Spec typo
191 Bugzilla 1196 - Spec examples corrections
193 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
195 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
196 gnutls_enable_pkcs11, but renamed to more accurately indicate its
199 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
200 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
202 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
203 "acl {{name}{arg}...}", and optional args on acl condition
206 JH/05 Permit multiple router/transport headers_add/remove lines.
208 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
210 JH/07 Avoid using a waiting database for a single-message-only transport.
211 Performance patch from Paul Fisher. Bugzilla 1262.
213 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
216 JH/09 Add $headers_added variable, with content from use of ACL modifier
217 add_header (but not yet added to the message). Bugzilla 199.
219 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
220 Pulled from Bugzilla 817 by Wolfgang Breyha.
222 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
224 (nb: this is the same fix as in Exim 4.80.1)
226 JH/11 Add A= logging on delivery lines, and a client_set_id option on
229 JH/12 Add optional authenticated_sender logging to A= and a log_selector
232 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
234 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
235 advertise SMTP AUTH mechanism to us, instead of a generic
236 protocol violation error. Also, make Exim more robust to bad
237 data from the Dovecot auth socket.
239 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
241 When a queue runner is handling a message, Exim first routes the
242 recipient addresses, during which it prunes them based on the retry
243 hints database. After that it attempts to deliver the message to
244 any remaining recipients. It then updates the hints database using
247 So if a recipient address works intermittently, it can get repeatedly
248 deferred at routing time. The retry hints record remains fresh so the
249 address never reaches the final cutoff time.
251 This is a fairly common occurrence when a user is bumping up against
252 their storage quota. Exim had some logic in its local delivery code
253 to deal with this. However it did not apply to per-recipient defers
254 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
256 This change adds a proper retry rule check during routing so that the
257 final cutoff time is checked against the message's age. We only do
258 this check if there is an address retry record and there is not a
259 domain retry record; this implies that previous attempts to handle
260 the address had the retry_use_local_parts option turned on. We use
261 this as an approximation for the destination being like a local
262 delivery, as in LMTP.
264 I suspect this new check makes the old local delivery cutoff check
265 redundant, but I have not verified this so I left the code in place.
267 TF/02 Correct gecos expansion when From: is a prefix of the username.
269 Test 0254 submits a message to Exim with the header
273 When I ran the test suite under the user fanf2, Exim expanded
274 the header to contain my full name, whereas it should have added
275 a Resent-Sender: header. It erroneously treats any prefix of the
276 username as equal to the username.
278 This change corrects that bug.
280 GF/01 DCC debug and logging tidyup
281 Error conditions log to paniclog rather than rejectlog.
282 Debug lines prefixed by "DCC: " to remove any ambiguity.
284 TF/03 Avoid unnecessary rebuilds of lookup-related code.
286 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
287 Bug spotted by Jeremy Harris; was flawed since initial commit.
288 Would have resulted in OCSP responses post-SNI triggering an Exim
289 NULL dereference and crash.
291 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
293 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
294 Bug detection, analysis and fix by Samuel Thibault.
295 Bugzilla 1331, Debian bug #698092.
297 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
299 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
300 Server implementation by Todd Lyons, client by JH.
301 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
302 config variable "prdr_enable" controls whether the server
303 advertises the facility. If the client requests PRDR a new
304 acl_data_smtp_prdr ACL is called once for each recipient, after
305 the body content is received and before the acl_smtp_data ACL.
306 The client is controlled by bolth of: a hosts_try_prdr option
307 on the smtp transport, and the server advertisement.
308 Default client logging of deliveries and rejections involving
309 PRDR are flagged with the string "PRDR".
311 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
312 fclose(). Diagnosis by Todd Lyons.
314 PP/17 Update configure.default to handle IPv6 localhost better.
315 Patch by Alain Williams (plus minor tweaks).
318 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
319 This is now consistent with GnuTLS, and is now documented: the
320 previous undocumented portable approach to treating the option as
321 unset was to force an expansion failure. That still works, and
322 an empty string is now equivalent.
324 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
325 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
326 not performing validation itself.
328 PP/20 Added force_command boolean option to pipe transport.
329 Patch from Nick Koston, of cPanel Inc.
331 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
334 TF/04 Added udpsend ACL modifer and hexquote expansion operator
336 PP/21 Fix eximon continuous updating with timestamped log-files.
337 Broken in a format-string cleanup in 4.80, missed when I repaired the
338 other false fix of the same issue.
339 Report and fix from Heiko Schlichting.
342 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
343 Report from Prashanth Katuri.
345 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
346 It's SecureTransport, so affects any MacOS clients which use the
347 system-integrated TLS libraries, including email clients.
349 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
350 using a MIME ACL for non-SMTP local injection.
351 Report and assistance in diagnosis by Warren Baker.
353 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
355 JH/16 Fix comparisons for 64b. Bugzilla 1385.
357 TL/09 Add expansion variable $authenticated_fail_id to keep track of
358 last id that failed so it may be referenced in subsequent ACL's.
360 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
363 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
364 ldap library initialization, allowing self-signed CA's to be
365 used. Also properly sets require_cert option later in code by
366 using NULL (global ldap config) instead of ldap handle (per
367 session). Bug diagnosis and testing by alxgomz.
369 TL/12 Enhanced documentation in the ratelimit.pl script provided in
370 the src/util/ subdirectory.
372 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
373 renamed to Transport Post Delivery Action by Jeremy Harris, as
376 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
377 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
378 redis_servers = needs to be configured which will be used by the redis
379 lookup. Patch from Warren Baker, of The Packet Hub.
381 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
383 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
384 hostname or reverse DNS when processing a host list. Used suggestions
385 from multiple comments on this bug.
387 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
389 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
390 Missed a few lines, added it to make the runtest require no keyboard
393 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
394 contains upper case chars. Make router use caseful_local_part.
396 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
397 support when GnuTLS has been built with p11-kit.
403 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
405 This, or similar/improved, will also be change PP/11 of 4.82.
411 PP/01 Handle short writes when writing local log-files.
412 In practice, only affects FreeBSD (8 onwards).
413 Bugzilla 1053, with thanks to Dmitry Isaikin.
415 NM/01 Bugzilla 949 - Documentation tweak
417 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
420 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
422 PP/02 Implemented gsasl authenticator.
424 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
426 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
427 `pkg-config foo` for cflags/libs.
429 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
430 with rest of GSASL and with heimdal_gssapi.
432 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
433 `pkg-config foo` for cflags/libs for the TLS implementation.
435 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
436 properties get this fed in as external SSF. A number of robustness
437 and debugging improvements to the cyrus_sasl authenticator.
439 PP/08 cyrus_sasl server now expands the server_realm option.
441 PP/09 Bugzilla 1214 - Log authentication information in reject log.
442 Patch by Jeremy Harris.
444 PP/10 Added dbmjz lookup type.
446 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
448 PP/12 MAIL args handles TAB as well as SP, for better interop with
449 non-compliant senders.
450 Analysis and variant patch by Todd Lyons.
452 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
453 Bug report from Lars Müller <lars@samba.org> (via SUSE),
454 Patch from Dirk Mueller <dmueller@suse.com>
456 PP/13 tls_peerdn now print-escaped for spool files.
457 Observed some $tls_peerdn in wild which contained \n, which resulted
458 in spool file corruption.
460 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
461 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
462 or write after TLS renegotiation, which otherwise led to messages
465 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
466 as a tracking header (ie: a signed header comes before the signature).
467 Patch from Wolfgang Breyha.
469 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
470 comma-sep list; embedded commas doubled.
472 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
474 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
476 Report and patch from Dmitry Banschikov.
478 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
479 Removed SSL_clear() after SSL_new() which led to protocol negotiation
480 failures. We appear to now support TLS1.1+ with Exim.
482 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
483 lets Exim select keys and certificates based upon TLS SNI from client.
484 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
485 before an outbound SMTP session. New log_selector, +tls_sni.
487 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
488 NULL dereference. Report and patch from Alun Jones.
490 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
491 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
492 Not seeing resolver debug output on NetBSD, but suspect this is a
493 resolver implementation change.
495 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
496 Left warnings. Added "eximon gdb" invocation mode.
498 PP/21 Defaulting "accept_8bitmime" to true, not false.
500 PP/22 Added -bw for inetd wait mode support.
502 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
503 locate the relevant includes and libraries. Made this the default.
505 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
506 Bugzilla 1246, report and most of solution from Tomasz Kusy.
508 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
509 This may cause build issues on older platforms.
511 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
512 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
513 gnutls_require_mac & gnutls_require_protocols (no longer supported).
514 Added SNI support via GnuTLS too.
515 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
517 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
519 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
520 Applied second patch from Janne, implementing suggestion to default
521 multiple-strings-in-record handling to match SPF spec.
523 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
525 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
526 read-only, out of scope).
527 Patch from Wolfgang Breyha, report from Stuart Northfield.
529 PP/29 Fix three issues highlighted by clang analyser static analysis.
530 Only crash-plausible issue would require the Cambridge-specific
531 iplookup router and a misconfiguration.
532 Report from Marcin Mirosław.
534 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
536 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
537 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
538 As part of this, removing so much warning spew let me fix some minor
539 real issues in debug logging.
541 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
542 assignment on my part. Fixed.
544 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
545 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
546 Janne Snabb (who went above and beyond: thank you).
548 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
549 string otherwise requires a connection and a bunch more work and it's
550 relatively easy to get wrong. Should also expose TLS library linkage
553 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
554 64-bit ${eval} (JH/03).
556 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
557 GNU libc to support some of the 64-bit stuff, should not lead to
558 conflicts. Defined before os.h is pulled in, so if a given platform
559 needs to override this, it can.
561 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
562 protection layer was required, which is not implemented.
563 Bugzilla 1254, patch from Wolfgang Breyha.
565 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
566 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
567 tls_dhparam take prime identifiers. Also unbreak combination of
568 OpenSSL+DH_params+TLSSNI.
570 PP/39 Disable SSLv2 by default in OpenSSL support.
576 PP/01 Solaris build fix for Oracle's LDAP libraries.
577 Bugzilla 1109, patch from Stephen Usher.
579 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
581 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
584 TF/02 Fix a couple more cases where we did not log the error message
585 when unlink() failed. See also change 4.74-TF/03.
587 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
588 lock up or crash if it happened to be inside a call to libc when it
589 got a SIGUSR1 from exiwhat.
591 The SIGUSR1 handler appends the current process status to the process
592 log which is later printed by exiwhat. It used to use the general
593 purpose logging code to do this, but several functions it calls are
594 not safe for signals.
596 The new output code in the SIGUSR1 handler is specific to the process
597 log, and simple enough that it's easy to inspect for signal safety.
598 Removing some special cases also simplifies the general logging code.
599 Removing the spurious timestamps from the process log simplifies
602 TF/04 Improved ratelimit ACL condition.
604 The /noupdate option has been deprecated in favour of /readonly which
605 has clearer semantics. The /leaky, /strict, and /readonly update modes
606 are mutually exclusive. The update mode is no longer included in the
607 database key; it just determines when the database is updated. (This
608 means that when you upgrde Exim will forget old rate measurements.)
610 Exim now checks that the per_* options are used with an update mode that
611 makes sense for the current ACL. For example, when Exim is processing a
612 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
613 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
614 must specify per_mail/readonly. If you omit the update mode it defaults to
615 /leaky where that makes sense (as before) or /readonly where required.
617 The /noupdate option is now undocumented but still supported for
618 backwards compatibility. It is equivalent to /readonly except that in
619 ACLs where /readonly is required you may specify /leaky/noupdate or
620 /strict/noupdate which are treated the same as /readonly.
622 A useful new feature is the /count= option. This is a generalization
623 of the per_byte option, so that you can measure the throughput of other
624 aggregate values. For example, the per_byte option is now equivalent
625 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
627 The per_rcpt option has been generalized using the /count= mechanism
628 (though it's more complicated than the per_byte equivalence). When it is
629 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
630 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
631 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
632 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
633 Note that using per_rcpt with a non-readonly update mode in more than
634 one ACL will cause the recipients to be double-counted. (The per_mail
635 and per_byte options don't have this problem.)
637 The handling of very low rates has changed slightly. If the computed rate
638 is less than the event's count (usually one) then this event is the first
639 after a long gap. In this case the rate is set to the same as this event's
640 count, so that the first message of a spam run is counted properly.
642 The major new feature is a mechanism for counting the rate of unique
643 events. The new per_addr option counts the number of different
644 recipients that someone has sent messages to in the last time period. It
645 behaves like per_rcpt if all the recipient addresses are different, but
646 duplicate recipient addresses do not increase the measured rate. Like
647 the /count= option this is a general mechanism, so the per_addr option
648 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
649 example, measure the rate that a client uses different sender addresses
650 with the options per_mail/unique=$sender_address. There are further
651 details in the main documentation.
653 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
655 TF/06 Removed a few PCRE remnants.
657 TF/07 Automatically extract Exim's version number from tags in the git
658 repository when doing development or release builds.
660 PP/02 Raise smtp_cmd_buffer_size to 16kB.
661 Bugzilla 879. Patch from Paul Fisher.
663 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
664 Heavily based on revision 40f9a89a from Simon Arlott's tree.
667 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
669 PP/05 Variable $av_failed, true if the AV scanner deferred.
670 Bugzilla 1078. Patch from John Horne.
672 PP/06 Stop make process more reliably on build failure.
673 Bugzilla 1087. Patch from Heiko Schlittermann.
675 PP/07 Make maildir_use_size_file an _expandable_ boolean.
676 Bugzilla 1089. Patch from Heiko Schlittermann.
678 PP/08 Handle ${run} returning more data than OS pipe buffer size.
679 Bugzilla 1131. Patch from Holger Weiß.
681 PP/09 Handle IPv6 addresses with SPF.
682 Bugzilla 860. Patch from Wolfgang Breyha.
684 PP/10 GnuTLS: support TLS 1.2 & 1.1.
686 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
689 PP/11 match_* no longer expand right-hand-side by default.
690 New compile-time build option, EXPAND_LISTMATCH_RHS.
691 New expansion conditions, "inlist", "inlisti".
693 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
695 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
697 PP/14 fix log_write() format string regression from TF/03.
698 Bugzilla 1152. Patch from Dmitry Isaikin.
704 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
706 PP/02 Harmonised TLS library version reporting; only show if debugging.
707 Layout now matches that introduced for other libraries in 4.74 PP/03.
709 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
711 PP/04 New "dns_use_edns0" global option.
713 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
716 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
717 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
719 TK/01 Updated PolarSSL code to 0.14.2.
720 Bugzilla 1097. Patch from Andreas Metzler.
722 PP/07 Catch divide-by-zero in ${eval:...}.
725 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
728 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
729 format-string attack -- SECURITY: remote arbitrary code execution.
731 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
732 time unintentionally subject to list matching rules, letting the header
733 cause arbitrary Exim lookups (of items which can occur in lists, *not*
734 arbitrary string expansion). This allowed for information disclosure.
736 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
737 INT_MIN/-1 -- value coerced to INT_MAX.
743 NM/01 Workround for PCRE version dependency in version reporting
746 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
747 This fixes portability to compilers other than gcc, notably
748 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
750 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
751 makefiles for portability to HP-UX and POSIX correctness.
753 PP/01 Permit LOOKUP_foo enabling on the make command-line.
754 Also via indented variable definition in the Makefile.
755 (Debugging by Oliver Heesakkers).
757 PP/02 Restore caching of spamd results with expanded spamd_address.
758 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
760 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
761 Improves build reliability. Fix from: Frank Elsner
763 NM/02 Fix wide character breakage in the rfc2047 coding
764 Fixes bug 1064. Patch from Andrey N. Oktyabrski
766 NM/03 Allow underscore in dnslist lookups
767 Fixes bug 1026. Patch from Graeme Fowler
769 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
770 Code patches from Adam Ciarcinski of NetBSD.
772 NM/04 Fixed exiqgrep to cope with mailq missing size issue
775 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
776 is logged, to avoid truncation. Patch from John Horne.
778 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
779 Patch from Jakob Hirsch.
781 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
782 SQL string expansion failure details.
783 Patch from Andrey Oktyabrski.
785 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
786 Patch from Simon Arlott.
788 PP/09 New lookups functionality failed to compile on old gcc which rejects
789 extern declarations in function scope.
790 Patch from Oliver Fleischmann
792 PP/10 Use sig_atomic_t for flags set from signal handlers.
793 Check getgroups() return and improve debugging.
794 Fixed developed for diagnosis in bug 927 (which turned out to be
797 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
798 Patch from Mark Zealey.
800 PP/12 Bugzilla 1056: Improved spamd server selection.
801 Patch from Mark Zealey.
803 PP/13 Bugzilla 1086: Deal with maildir quota file races.
804 Based on patch from Heiko Schlittermann.
806 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
807 Patch from Uwe Doering, sign-off by Michael Haardt.
809 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
810 variable declaration deep within a block. Bug and patch from
813 PP/15 lookups-Makefile IRIX compatibilty coercion.
815 PP/16 Make DISABLE_DKIM build knob functional.
817 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
818 Patch by Simon Arlott
820 TF/03 Fix valgrind.h portability to C89 compilers that do not support
821 variable argument macros. Our copy now differs from upstream.
827 TF/01 Failure to get a lock on a hints database can have serious
828 consequences so log it to the panic log.
830 TF/02 Log LMTP confirmation messages in the same way as SMTP,
831 controlled using the smtp_confirmation log selector.
833 TF/03 Include the error message when we fail to unlink a spool file.
835 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
836 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
837 for maintaining out-of-tree patches for some time.
839 PP/01 Bugzilla 139: Documentation and portability issues.
840 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
841 Handle per-OS dynamic-module compilation flags.
843 PP/02 Let /dev/null have normal permissions.
844 The 4.73 fixes were a little too stringent and complained about the
845 permissions on /dev/null. Exempt it from some checks.
846 Reported by Andreas M. Kirchwitz.
848 PP/03 Report version information for many libraries, including
849 Exim version information for dynamically loaded libraries. Created
850 version.h, now support a version extension string for distributors
851 who patch heavily. Dynamic module ABI change.
853 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
854 privilege escalation vulnerability whereby the Exim run-time user
855 can cause root to append content of the attacker's choosing to
858 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
861 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
862 If dropping privileges for untrusted macros, we disabled normal logging
863 on the basis that it would fail; for the Exim run-time user, this is not
864 the case, and it resulted in successful deliveries going unlogged.
865 Fixed. Reported by Andreas Metzler.
871 PP/01 Date: & Message-Id: revert to normally being appended to a message,
872 only prepend for the Resent-* case. Fixes regression introduced in
873 Exim 4.70 by NM/22 for Bugzilla 607.
875 PP/02 Include check_rfc2047_length in configure.default because we're seeing
876 increasing numbers of administrators be bitten by this.
878 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
880 PP/03 Bugzilla 994: added openssl_options main configuration option.
882 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
884 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
886 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
888 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
889 without a peer certificate, leading to a segfault because of an
890 assumption that peers always have certificates. Be a little more
891 paranoid. Problem reported by Martin Tscholak.
893 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
894 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
895 NB: ClamAV planning to remove STREAM in "middle of 2010".
896 CL also introduces -bmalware, various -d+acl logging additions and
897 more caution in buffer sizes.
899 PP/09 Implemented reverse_ip expansion operator.
901 PP/10 Bugzilla 937: provide a "debug" ACL control.
903 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
905 PP/12 Bugzilla 973: Implement --version.
907 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
909 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
911 PP/15 Bugzilla 816: support multiple condition rules on Routers.
913 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
914 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
915 ignore trailing whitespace.
917 JJ/02 prevent non-panic DKIM error from being sent to paniclog
919 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
922 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
923 Notification from Dr Andrew Aitchison.
925 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
926 ExtendedDetectionInfo response format.
927 Notification from John Horne.
929 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
932 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
933 XSL and documented dependency on system catalogs, with examples of how
936 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
939 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
940 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
941 configuration file which is writeable by the Exim user or group.
943 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
944 of configuration files to cover files specified with the -C option if
945 they are going to be used with root privileges, not just the default
948 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
949 option (effectively making it always true).
951 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
952 files to be used while preserving root privileges.
954 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
955 that rogue child processes cannot use them.
957 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
958 run-time user, instead of root.
960 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
961 Exim run-time user without dropping privileges.
963 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
964 result string, instead of calling string_vformat() twice with the same
967 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
968 for other users. Others should always drop root privileges if they use
969 -C on the command line, even for a whitelisted configure file.
971 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
973 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
979 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
980 $data_path, and $header_path variables; fixed documentation bugs and
983 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
984 exipick to access non-standard spools, including the "frozen" queue
987 NM/01 Bugzilla 965: Support mysql stored procedures.
988 Patch from Alain Williams
990 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
992 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
993 Patch from Andreas Metzler
995 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
996 Patch from Kirill Miazine
998 NM/05 Bugzilla 671: Added umask to procmail example.
1000 JJ/03 installed exipick 20100323.0, fixing doc bug
1002 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1003 directory. Notification and patch from Dan Rosenberg.
1005 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1007 TK/02 Improve log output when DKIM signing operation fails.
1009 MH/01 Treat the transport option dkim_domain as a colon separated
1010 list, not as a single string, and sign the message with each element,
1011 omitting multiple occurences of the same signer.
1013 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1014 Bugzilla 985, 986. Patch by Simon Arlott
1016 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1017 Patch by Simon Arlott
1019 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1020 MBX locking. Notification from Dan Rosenberg.
1026 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1028 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1030 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1032 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1034 NM/04 Bugzilla 915: Flag broken perl installation during build.
1040 TK/01 Added patch by Johannes Berg that expands the main option
1041 "spamd_address" if it starts with a dollar sign.
1043 TK/02 Write list of recipients to X-Envelope-Sender header when building
1044 the mbox-format spool file for content scanning (suggested by Jakob
1047 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1048 (http://www.dcc-servers.net/) support via dccifd. Activated by
1049 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1051 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1052 by Mark Daniel Reidel <mr@df.eu>.
1054 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1055 When building exim an external PCRE library is now needed -
1056 PCRE is a system library on the majority of modern systems.
1057 See entry on PCRE_LIBS in EDITME file.
1059 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1060 conversation. Added nologin parameter to request.
1061 Patch contributed by Kirill Miazine.
1063 TF/01 Do not log submission mode rewrites if they do not change the address.
1065 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1067 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1068 log files in place. Contributed by Roberto Lima.
1070 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1072 TF/03 Bugzilla 615: When checking the local_parts router precondition
1073 after a local_part_suffix or local_part_prefix option, Exim now
1074 does not use the address's named list lookup cache, since this
1075 contains cached lookups for the whole local part.
1077 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1078 Robert Millan. Documentation is in experimental-spec.txt.
1080 TF/04 Bugzilla 668: Fix parallel build (make -j).
1082 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1084 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1085 Patch provided by Jan Srzednicki.
1087 TF/05 Leading white space used to be stripped from $spam_report which
1088 wrecked the formatting. Now it is preserved.
1090 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1091 that they are available at delivery time.
1093 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1095 TF/08 TLS error reporting now respects the incoming_interface and
1096 incoming_port log selectors.
1098 TF/09 Produce a more useful error message if an SMTP transport's hosts
1099 setting expands to an empty string.
1101 NM/06 Bugzilla 744: EXPN did not work under TLS.
1102 Patch provided by Phil Pennock.
1104 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1105 Patch provided by Richard Godbee.
1107 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1108 acl_smtp_notquit, added index entry.
1110 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1111 Patch provided by Eugene Bujak.
1113 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1114 accept(). Patch provided by Maxim Dounin.
1116 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1117 Patch provided by Phil Pennock.
1119 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1121 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1122 Patch provided by Brad "anomie" Jorsch.
1124 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1125 Patch provided by Dean Brooks.
1127 TK/05 Add native DKIM support (does not depend on external libraries).
1129 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1130 Patch provided by Graeme Fowler.
1132 NM/16 Bugzilla 851: Documentation example syntax fix.
1134 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1136 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1139 NM/19 Bugzilla 745: TLS version reporting.
1140 Patch provided by Phil Pennock.
1142 NM/20 Bugzilla 167: bool: condition support.
1143 Patch provided by Phil Pennock.
1145 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1146 clients. Patch provided by Phil Pennock.
1148 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1149 Patch provided by Brad "anomie" Jorsch.
1151 NM/23 Bugzilla 687: Fix misparses in eximstats.
1152 Patch provided by Heiko Schlittermann.
1154 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1155 Patch provided by Heiko Schlittermann.
1157 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1158 plus update to original patch.
1160 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1162 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1163 Patch provided by David Brownlee.
1165 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1167 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1169 NM/30 Bugzilla 888: TLS documentation bugfixes.
1171 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1173 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1174 Unlike the original bugzilla I have changed all shell scripts in src tree.
1176 NM/33 Bugzilla 898: Transport filter timeout fix.
1177 Patch by Todd Rinaldo.
1179 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1180 Patch by Serge Demonchaux.
1182 NM/35 Bugzilla 39: Base64 decode bug fixes.
1183 Patch by Jakob Hirsch.
1185 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1187 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1189 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1191 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1197 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1198 ALT-N's libdkim that I have put here:
1199 http://duncanthrax.net/exim-experimental/
1201 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1202 were called 'true' and it seems that C99 defines that as a reserved
1203 keyword to be used with 'bool' variable types. That means you could
1204 not include C99-style headers which use bools without triggering
1205 build errors in sieve.c.
1207 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1208 as mailq or other aliases. Changed the --help handling significantly
1209 to do whats expected. exim_usage() emits usage/help information.
1211 SC/01 Added the -bylocaldomain option to eximstats.
1213 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1215 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1217 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1223 PH/01 Another patch from the Sieve maintainer.
1225 PH/02 When an IPv6 address is converted to a string for single-key lookup
1226 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1227 dots are used instead of colons so that keys in lsearch files need not
1228 contain colons. This was done some time before quoting was made available
1229 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1230 (notated using the quote facility) so as to distinguish them from IPv4
1231 keys. This meant that lookups for IP addresses in host lists did not work
1232 for iplsearch lookups.
1234 This has been fixed by arranging for IPv6 addresses to be expressed with
1235 colons if the lookup type is iplsearch. This is not incompatible, because
1236 previously such lookups could never work.
1238 The situation is now rather anomolous, since one *can* have colons in
1239 ordinary lsearch keys. However, making the change in all cases is
1240 incompatible and would probably break a number of configurations.
1242 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1245 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1246 conversion specification without a maximum field width, thereby enabling
1247 a rogue spamd server to cause a buffer overflow. While nobody in their
1248 right mind would setup Exim to query an untrusted spamd server, an
1249 attacker that gains access to a server running spamd could potentially
1250 exploit this vulnerability to run arbitrary code as the Exim user.
1252 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1253 $primary_hostname instead of what libspf2 thinks the hosts name is.
1255 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1256 a directory entry by the name of the lookup key. Previously, if a
1257 symlink pointed to a non-existing file or a file in a directory that
1258 Exim lacked permissions to read, a lookup for a key matching that
1259 symlink would fail. Now it is enough that a matching directory entry
1260 exists, symlink or not. (Bugzilla 503.)
1262 PH/03 The body_linecount and body_zerocount variables are now exported in the
1265 PH/04 Added the $dnslist_matched variable.
1267 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1268 This means they are set thereafter only if the connection becomes
1271 PH/06 Added the client_condition to authenticators so that some can be skipped
1272 by clients under certain conditions.
1274 PH/07 The error message for a badly-placed control=no_multiline_responses left
1275 "_responses" off the end of the name.
1277 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1279 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1280 (without spaces) instead of just copying the configuration text.
1282 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1284 PH/11 Added $max_received_linelength.
1286 PH/12 Added +ignore_defer and +include_defer to host lists.
1288 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1289 way in which PCRE > 7.0 is built.
1291 PH/14 Implemented queue_only_load_latch.
1293 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1294 MAIL command. The effect was to mangle the value on 64-bit systems.
1296 PH/16 Another patch from the Sieve maintainer.
1298 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1300 PH/18 If a system quota error occurred while trying to create the file for
1301 a maildir delivery, the message "Mailbox is full" was not appended to the
1302 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1303 only to a quota excession during the actual writing of the file.
1305 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1306 characters?) which causes problems in log lines. The DN values are now
1307 passed through string_printing() before being added to log lines.
1309 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1310 and InterBase are left for another time.)
1312 PH/21 Added message_body_newlines option.
1314 PH/22 Guard against possible overflow in moan_check_errorcopy().
1316 PH/23 POSIX allows open() to be a macro; guard against that.
1318 PH/24 If the recipient of an error message contained an @ in the local part
1319 (suitably quoted, of course), incorrect values were put in $domain and
1320 $local_part during the evaluation of errors_copy.
1326 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1327 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1330 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1331 issue a MAIL command.
1333 PH/02 In an ACL statement such as
1335 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1337 if a client was not listed at all, or was listed with a value other than
1338 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1339 the condition was not true (as it should be), so access was not denied.
1340 The bug was that the ! inversion was incorrectly passed on to the second
1341 item. This has been fixed.
1343 PH/03 Added additional dnslists conditions == and =& which are different from
1344 = and & when the dns lookup returns more than one IP address.
1346 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1347 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1349 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1350 FSYNC, which compiles an option called disable_fsync that allows for
1351 bypassing fsync(). The documentation is heavily laced with warnings.
1353 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1355 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1356 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1357 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1358 including adding "make clean"; (3) Added -fPIC when compiling the test
1359 dynamically loaded module, to get rid of a warning.
1361 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1362 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1363 = 0s. The bug is otherwise harmless.
1365 PH/07 There was a bug in the dovecot authenticator such that the value of
1366 $auth1 could be overwritten, and so not correctly preserved, after a
1367 successful authentication. This usually meant that the value preserved by
1368 the server_setid option was incorrect.
1370 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1372 PH/09 Installed PCRE release 7.0.
1374 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1375 run for batched SMTP input. It is now run at the start of every message
1376 in the batch. While fixing this I discovered that the process information
1377 (output by running exiwhat) was not always getting set for -bs and -bS
1378 input. This is fixed, and it now also says "batched" for BSMTP.
1380 PH/11 Added control=no_pipelining.
1382 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1383 patch, slightly modified), and move the expansion of helo_data till after
1384 the connection is made in the smtp transport (so it can use these
1387 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1389 PH/14 Added log_selector = +pid.
1391 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1393 PH/16 Add ${if forany and ${if forall.
1395 PH/17 Added dsn_from option to vary the From: line in DSNs.
1397 PH/18 Flush SMTP output before performing a callout, unless control =
1398 no_callout_flush is set.
1400 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1401 was true (the default) a successful delivery failed to delete the retry
1402 item, thus causing premature timeout of the address. The bug is now
1405 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1407 PH/21 Long custom messages for fakedefer and fakereject are now split up
1408 into multiline reponses in the same way that messages for "deny" and
1409 other ACL rejections are.
1411 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1412 with slight modification.
1414 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1415 draft, changing the syntax and factoring some duplicate code".
1417 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1418 for deliveries of the second and subsequent messages over the same SMTP
1421 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1422 ${reduce, with only minor "tidies".
1424 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1426 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1427 expansion side effects.
1429 PH/27 When a message times out after an over-quota error from an Exim-imposed
1430 quota, the bounce message says "mailbox is full". This message was not
1431 being given when it was a system quota that was exceeded. It now should
1434 MH/03 Made $recipients available in local_scan(). local_scan() already has
1435 better access to the recipient list through recipients_list[], but
1436 $recipients can be useful in postmaster-provided expansion strings.
1438 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1439 in the case of a MAIL command with additional options following the
1440 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1441 were accidentally chopped off.
1443 PH/29 SMTP synchronization checks are implemented when a command is read -
1444 there is a check that no more input is waiting when there shouldn't be
1445 any. However, for some commands, a delay in an ACL can mean that it is
1446 some time before the response is written. In this time, more input might
1447 arrive, invalidly. So now there are extra checks after an ACL has run for
1448 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1449 pipelining has not been advertised.
1451 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1453 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1454 returned in the SMTP response when smtp_return_error_details was false.
1455 This has been fixed.
1457 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1458 instead of the C I/O that was originally supplied, because problems were
1459 reported on Solaris.
1461 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1462 Exim which did not show up earlier: it was assuming that a call to
1463 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1464 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1465 was a macro that became an assignment, so it seemed to work. This has
1466 changed to a proper function call with a void return, hence the compile
1467 error. Exim's code has been fixed.
1469 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1472 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1474 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1476 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1477 because it turns out to be too big or there is a timeout) there is no
1478 "Completed" line in the log. When some messages of this type were
1479 selected by exigrep, they were listed as "not completed". Others were
1480 picked up by some special patterns. I have improved the selection
1481 criteria to be more general.
1483 PH/38 The host_find_failed option in the manualroute router can now be set
1484 to "ignore", to completely ignore a host whose IP address cannot be
1485 found. If all hosts are ignored, the behaviour is controlled by the new
1486 host_all_ignored option.
1488 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1489 homing or because of multiple MX records with /mx) generated more than
1490 one IP address, and the following item turned out to be the local host,
1491 all the secondary addresses of the first item were incorrectly removed
1492 from the list, along with the local host and any following hosts (which
1493 is what is supposed to happen).
1495 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1496 whoever called Exim into the -H file. In the case of the daemon it was
1497 behaving confusingly. When first started, it used values for whoever
1498 started the daemon, but after a SIGHUP it used the Exim user (because it
1499 calls itself on a restart). I have changed the code so that it now always
1502 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1503 message are rejected with the same error (e.g. no authentication or bad
1504 sender address), and a DATA command is nevertheless sent (as can happen
1505 with PIPELINING or a stupid MUA), the error message that was given to the
1506 RCPT commands is included in the rejection of the DATA command. This is
1507 intended to be helpful for MUAs that show only the final error to their
1510 PH/42 Another patch from the Sieve maintainer.
1512 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1513 Eximstats - Fixed some broken HTML links and added missing column headers
1515 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1516 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1518 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1520 PH/43 Yet another patch from the Sieve maintainer.
1522 PH/44 I found a way to check for a TCP/IP connection going away before sending
1523 the response to the final '.' that terminates a message, but only in the
1524 case where the client has not sent further data following the '.'
1525 (unfortunately, this is allowed). However, in many cases there won't be
1526 any further data because there won't be any more messages to send. A call
1527 to select() can be used: if it shows that the input is "ready", there is
1528 either input waiting, or the socket has been closed. An attempt to read
1529 the next input character can distinguish the two cases. Previously, Exim
1530 would have sent an OK response which the client would never have see.
1531 This could lead to message repetition. This fix should cure that, at
1532 least in a lot of common cases.
1534 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1535 advertised in response to EHLO.
1541 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1542 fixed by 4.65/MH/01 (is this a record?) are fixed:
1544 (i) An empty string was always treated as zero by the numeric comparison
1545 operators. This behaviour has been restored.
1547 (ii) It is documented that the numeric comparison operators always treat
1548 their arguments as decimal numbers. This was broken in that numbers
1549 starting with 0 were being interpreted as octal.
1551 While fixing these problems I realized that there was another issue that
1552 hadn't been noticed. Values of message_size_limit (both the global option
1553 and the transport option) were treated as octal if they started with 0.
1554 The documentation was vague. These values are now always treated as
1555 decimal, and I will make that clear in the documentation.
1561 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1562 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1565 MH/01 Don't check that the operands of numeric comparison operators are
1566 integers when their expansion is in "skipping" mode (fixes bug
1567 introduced by 4.64-PH/07).
1569 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1570 child addresses, Exim now panics and dies. Previously, because the count
1571 is held in a short int, deliveries were likely to be lost. As such a
1572 large number of recipients for a single message is ridiculous
1573 (performance will be very, very poor), I have chosen to impose a limit
1574 rather than extend the field.
1580 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1581 leftover -K file (the existence of which was triggered by #402).
1582 While we were at it, introduced process PID as part of the -K
1583 filename. This should rule out race conditions when creating
1586 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1587 processing considerably. Previous code took too long for large mails,
1588 triggering a timeout which in turn triggers #401.
1590 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1591 in the DK code in transports.c. sendfile() is not really portable,
1592 hence the _LINUX specificness.
1594 TF/01 In the add_headers option to the mail command in an Exim filter,
1595 there was a bug that Exim would claim a syntax error in any
1596 header after the first one which had an odd number of characters
1599 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1600 callout verification, Exim cached a "reject" for the entire domain. This
1601 is correct for most verifications, but it is not correct for a recipient
1602 verification with use_sender or use_postmaster set, because in that case
1603 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1604 case of MAIL FROM:<> rejection from other early rejections (e.g.
1605 rejection of HELO). When verifying a recipient using a non-null MAIL
1606 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1607 Whatever the result of the callout, the value of the domain cache is
1608 left unchanged (for any other kind of callout, getting as far as trying
1609 RCPT means that the domain itself is ok).
1611 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1614 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1615 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1616 session, the daemon ignores ECONNECT errors and logs others; it now
1617 ignores EPIPE as well.
1619 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1620 (quoted-printable decoding).
1622 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1623 later the small subsequent patch to fix an introduced bug.
1625 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1627 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1629 PH/08 An error is now given if message_size_limit is specified negative.
1631 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1632 to be given (somewhat) arbitrary names.
1634 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1637 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1638 miscellaneous code fixes
1640 PH/10 Added the log_reject_target ACL modifier to specify where to log
1643 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1644 hostname. This is wrong, because it relates to the incoming message (and
1645 probably the interface on which it is arriving) and not to the outgoing
1646 callout (which could be using a different interface). This has been
1647 changed to use the value of the helo_data option from the smtp transport
1648 instead - this is what is used when a message is actually being sent. If
1649 there is no remote transport (possible with a router that sets up host
1650 addresses), $smtp_active_hostname is used.
1652 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1653 tweaks were necessary in order to get it to work (see also 21 below):
1654 (a) The code assumed that strncpy() returns a negative number on buffer
1655 overflow, which isn't the case. Replaced with Exim's string_format()
1657 (b) There were several signed/unsigned issues. I just did the minimum
1658 hacking in of casts. There is scope for a larger refactoring.
1659 (c) The code used strcasecmp() which is not a standard C function.
1660 Replaced with Exim's strcmpic() function.
1661 (d) The code set only $1; it now sets $auth1 as well.
1662 (e) A simple test gave the error "authentication client didn't specify
1663 service in request". It would seem that Dovecot has changed its
1664 interface. Fortunately there's a specification; I followed it and
1665 changed what the client sends and it appears to be working now.
1667 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1670 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1671 address A is aliased to B and C, where B exists and C does not. Without
1672 -v the output is "A verified" because verification stops after a
1673 successful redirection if more than one address is generated. However,
1674 with -v the child addresses are also verified. Exim was outputting "A
1675 failed to verify" and then showing the successful verification for C,
1676 with its parentage. It now outputs "B failed to verify", showing B's
1677 parentage before showing the successful verification of C.
1679 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1680 look up a TXT record in a specific list after matching in a combined
1683 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1684 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1685 they consult the DNS. I had assumed they would set it the way they
1686 wanted; and indeed my experiments on Linux seem to show that in some
1687 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1688 To be on the safe side, however, I have now made the interface to
1689 host_find_byname() similar to host_find_bydns(), with an argument
1690 containing the DNS resolver options. The host_find_byname() function now
1691 sets these options at its start, just as host_find_bydns() does. The smtp
1692 transport options dns_qualify_single and dns_search_parents are passed to
1693 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1694 of host_find_byname() use the default settings of RES_DEFNAMES
1695 (qualify_single) but not RES_DNSRCH (search_parents).
1697 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1698 spool_read_header() do less string testing, by means of a preliminary
1699 switch on the second character of optional "-foo" lines. (This is
1700 overdue, caused by the large number of possibilities that now exist.
1701 Originally there were few.) While I was there, I also converted the
1702 str(n)cmp tests so they don't re-test the leading "-" and the first
1703 character, in the hope this might squeeze out yet more improvement.
1705 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1706 flag allowing group syntax was set by the header_syntax check but not
1707 turned off, possible causing trouble later; (2) The flag was not being
1708 set at all for the header_verify test, causing "group"-style headers to
1709 be rejected. I have now set it in this case, and also caused header_
1710 verify to ignore an empty address taken from a group. While doing this, I
1711 came across some other cases where the code for allowing group syntax
1712 while scanning a header line wasn't quite right (mostly, not resetting
1713 the flag correctly in the right place). These bugs could have caused
1714 trouble for malformed header lines. I hope it is now all correct.
1716 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1717 with the "reply" argument non-NULL. The code, however (which originally
1718 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1719 but it didn't always do it. This confused somebody who was copying the
1720 code for some other use. I have removed all the tests.
1722 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1723 feature that was used to support insecure browsers during the U.S. crypto
1724 embargo. It requires special client support, and Exim is probably the
1725 only MTA that supported it -- and would never use it because real RSA is
1726 always available. This code has been removed, because it had the bad
1727 effect of slowing Exim down by computing (never used) parameters for the
1728 RSA_EXPORT functionality.
1730 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1731 authenticator to fail if there's a tab character in the incoming data
1732 (there should never be unless someone is messing about, as it's supposed
1733 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1734 if the connection is using TLS or if the remote IP is the same as the
1735 local IP, and the "valid-client-cert option" if a client certificate has
1738 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1739 authenticators. This can be used for authorization after authentication
1740 succeeds. (In the case of plaintext, it servers for both authentication
1743 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1744 if any retry times were supplied.
1746 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1747 connection, where there is no client IP address to check. In this
1748 situation, the verify now always succeeds.
1750 PH/25 Applied John Jetmore's -Mset patch.
1752 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1754 PH/27 In a string expansion for a processed (not raw) header when multiple
1755 headers of the same name were present, leading whitespace was being
1756 removed from all of them, but trailing whitespace was being removed only
1757 from the last one. Now trailing whitespace is removed from each header
1758 before concatenation. Completely empty headers in a concatenation (as
1759 before) are ignored.
1761 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1762 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1764 PH/29 [Removed. This was a change that I later backed out, and forgot to
1765 correct the ChangeLog entry (that I had efficiently created) before
1766 committing the later change.]
1768 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1769 address errors (4xx response to RCPT) over the same connection as other
1770 messages routed to the same hosts. Such deliveries are always "forced",
1771 so retry times are not inspected. This resulted in far too many retries
1772 for the affected addresses. The effect occurred only when there were more
1773 hosts than the hosts_max_try setting in the smtp transport when it had
1774 the 4xx errors. Those hosts that it had tried were not added to the list
1775 of hosts for which the message was waiting, so if all were tried, there
1776 was no problem. Two fixes have been applied:
1778 (i) If there are any address or message errors in an SMTP delivery, none
1779 of the hosts (tried or untried) are now added to the list of hosts
1780 for which the message is waiting, so the message should not be a
1781 candidate for sending over the same connection that was used for a
1782 successful delivery of some other message. This seems entirely
1783 reasonable: after all the message is NOT "waiting for some host".
1784 This is so "obvious" that I'm not sure why it wasn't done
1785 previously. Hope I haven't missed anything, but it can't do any
1786 harm, as the worst effect is to miss an optimization.
1788 (ii) If, despite (i), such a delivery is accidentally attempted, the
1789 routing retry time is respected, so at least it doesn't keep
1790 hammering the server.
1792 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1793 in ${readsocket because some servers need this prod.
1795 PH/32 Added some extra debug output when updating a wait-xxx database.
1797 PH/33 The hint "could be header name not terminated by colon", which has been
1798 given for certain expansion errors for a long time, was not being given
1799 for the ${if def:h_colon_omitted{... case.
1801 PH/34 The spec says: "With one important exception, whenever a domain list is
1802 being scanned, $domain contains the subject domain." There was at least
1803 one case where this was not true.
1805 PH/35 The error "getsockname() failed: connection reset by peer" was being
1806 written to the panic log as well as the main log, but it isn't really
1807 panic-worthy as it just means the connection died rather early on. I have
1808 removed the panic log writing for the ECONNRESET error when getsockname()
1811 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1812 runs only) independently of the message's sender address. This meant
1813 that, if the 4xx error was in fact related to the sender, a different
1814 message to the same recipient with a different sender could confuse
1815 things. In particualar, this can happen when sending to a greylisting
1816 server, but other circumstances could also provoke similar problems.
1817 I have changed the default so that the retry time for these errors is now
1818 based a combination of the sender and recipient addresses. This change
1819 can be overridden by setting address_retry_include_sender=false in the
1822 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1823 remote server are returned as part of bounce messages. This was not
1824 happening for LMTP over a pipe (the lmtp transport), but now it is the
1825 same for both kinds of LMTP.
1827 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1828 in header lines that were in fact CNAMEs. This is no longer the case.
1830 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1831 and queue runs started by the daemon processed all messages. This has
1832 been fixed so that -R and -S can now usefully be given with -q<time>.
1834 PH/40 Import PCRE release 6.7 (fixes some bugs).
1836 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1838 PH/42 Give an error if -q is specified more than once.
1840 PH/43 Renamed the variables $interface_address and $interface_port as
1841 $received_ip_address and $received_port, to make it clear that these
1842 values apply to message reception, and not to the outgoing interface when
1843 a message is delivered. (The old names remain recognized, of course.)
1845 PH/44 There was no timeout on the connect() call when using a Unix domain
1846 socket in the ${readsocket expansion. There now is.
1848 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1849 be meaningful with "accept".
1851 SC/01 Eximstats V1.43
1852 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1854 SC/02 Eximstats V1.44
1855 Use a glob alias rather than an array ref in the generated
1856 parser. This improves both readability and performance.
1858 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1859 Collect SpamAssassin and rejection statistics.
1860 Don't display local sender or destination tables unless
1861 there is data to show.
1862 Added average volumes into the top table text output.
1864 SC/04 Eximstats V1.46
1865 Collect data on the number of addresses (recipients)
1866 as well as the number of messages.
1868 SC/05 Eximstats V1.47
1869 Added 'Message too big' to the list of mail rejection
1870 reasons (thanks to Marco Gaiarin).
1872 SC/06 Eximstats V1.48
1873 Mainlog lines which have GMT offsets and are too short to
1874 have a flag are now skipped.
1876 SC/07 Eximstats V1.49 (Alain Williams)
1877 Added the -emptyok flag.
1879 SC/08 Eximstats V1.50
1880 Fixes for obtaining the IP address from reject messages.
1882 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1883 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1884 whitesspace changes from 4.64-PH/27
1886 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1889 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1890 are found, allow negative numbers in numeric criteria)
1892 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1894 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1895 to match changes made in 4.64-PH/43
1897 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1899 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1900 "domain" is actually an IP address, and force a failure. This locks out
1901 those revolvers/nameservers that support "A-for-A" lookups, in
1902 contravention of the specifications.
1904 PH/48 When a host name was looked up from an IP address, and the subsequent
1905 forward lookup of the name timed out, the host name was left in
1906 $sender_host_name, contrary to the specification.
1908 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1909 restricted to single-key lookups, Exim was not diagnosing an error if
1910 * or *@ was used with a query-style lookup.
1912 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1914 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1915 long ago, but noone interested enough thought of it. Let's just say that
1916 the "1.1" means that there are some new functions that weren't there at
1917 some point in the past.
1919 PH/51 Error processing for expansion failure of helo_data from an smtp
1920 transport during callout processing was broken.
1922 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1923 tested/used via the -bh/-bhc/-bs options.
1925 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1926 bug, fixed in subsequent PCRE releases).
1928 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1929 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1931 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1937 SC/01 Use a glob alias rather than an array ref in eximstats generated
1938 parser. This improves both readability and performance.
1940 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1941 Don't display local sender or destination tables in eximstats unless
1942 there is data to show.
1943 Added average volumes into the eximstats top table text output.
1945 SC/03 Collect data on the number of addresses (recipients) as well
1946 as the number of messages in eximstats.
1948 TF/01 Correct an error in the documentation for the redirect router. Exim
1949 does (usually) call initgroups() when daemonizing.
1951 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1952 with consistent privilege compared to when running as a daemon.
1954 TF/03 Note in the spec that $authenticated_id is not set for local
1955 submissions from trusted users.
1957 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1958 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1960 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1961 by adding some example configuration directives to the default
1962 configuration file. A little bit of work is required to uncomment the
1963 directives and define how usernames and passwords are checked, but
1964 there is now a framework to start from.
1966 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1967 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1968 without this. I don't know how relevant this is to other LDAP libraries.
1970 PH/02 Add the verb name to the "unknown ACL verb" error.
1972 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1974 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1976 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1977 directory not expanded when it should be if an expanded home directory
1978 was set for the address (which is overridden by the transport).
1980 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1983 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1984 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1985 because it is too late at that time, and has no effect.
1987 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1988 security issue with \' (bugzilla #107). I could not use the
1989 PQescapeStringConn() function, because it needs a PGconn value as one of
1992 PH/08 When testing addresses using -bt, indicate those final addresses that
1993 are duplicates that would not cause an additional delivery. At least one
1994 person was confused, thinking that -bt output corresponded to deliveries.
1995 (Suppressing duplicates isn't a good idea as you lose the information
1996 about possibly different redirections that led to the duplicates.)
1998 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1999 systems where poll() doesn't work, in particular OS X.
2001 PH/10 Added more information to debugging output for retry time not reached.
2003 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2004 operations in malware.c.
2006 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2009 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2010 syslog_duplication set false, log lines that would normally be written
2011 both the the main log and to the reject log were not written to syslog at
2014 PH/14 In the default configuration, change the use of "message" in ACL warn
2015 statements to "add_header".
2017 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2018 not followed by a command (e.g. "seen endif").
2020 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2021 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2024 PH/17 Added extra conditions to the default value of delay_warning_condition
2028 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2029 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2030 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2033 The Auto-Submitted: and various List- headers are standardised, whereas I
2034 don't think Precedence: ever was.
2036 PH/18 Refactored debugging code in route_finduser() to show more information,
2037 in particular, the error code if getpwnam() issues one.
2039 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2040 This is apparently needed in addition to the PH/07 change above to avoid
2041 any possible encoding problems.
2043 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2044 but not after initializing Perl.
2046 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2047 output them only if debugging. By default they are written stderr,
2048 apparently, which is not desirable.
2050 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2053 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2056 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2058 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2059 authenticated or an ident call has been made. Suppress the default
2060 values for $authenticated_id and $authenticated_sender (but permit -oMai
2061 and -oMas) when testing with -bh.
2063 PH/24 Re-jigged the order of the tests in the default configuration so that the
2064 tests for valid domains and recipients precede the DNS black list and CSA
2065 tests, on the grounds that those ones are more expensive.
2067 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2068 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2069 This bug exists in every version of Exim that I still have, right back to
2072 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2073 However, an attempt to turn on case-sensitivity in a regex key by
2074 including (?-i) didn't work because the subject string was already
2075 lowercased, and the effects were non-intuitive. It turns out that a
2076 one-line patch can be used to allow (?-i) to work as expected.
2082 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2083 other effects) broke the use of negated acl sub-conditions.
2085 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2088 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2089 "Deny" causes Exim to reject the incoming connection with a 554 error.
2090 Unfortunately, if there is a major crisis, such as a disk failure,
2091 tcp-wrappers gives "deny", whereas what one would like would be some
2092 kind of temporary error. A kludge has been added to help with this.
2093 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2094 554 error is used if errno is still zero or contains ENOENT (which occurs
2095 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2098 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2100 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2101 errors. Otherwise a message that provokes a temporary error (when other
2102 messages do not) can cause a whole host to time out.
2104 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2105 addresses were routed directly to files or pipes from a redirect router.
2106 File deliveries just didn't batch; pipe deliveries might have suffered
2109 PH/06 A failure to get a lock for a hints database would erroneously always say
2110 "Failed to get write lock", even when it was really a read lock.
2112 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2113 of 0600. This has been changed to use the value of the lockfile_mode
2114 option (which defaults to 0600).
2116 PH/08 Applied small patch from the Sieve maintainer.
2118 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2119 folder from quota calculations, a direct delivery into this folder messed
2120 up the contents of the maildirsize file. This was because the regex was
2121 used only to exclude .Trash (or whatever) when the size of the mailbox
2122 was calculated. There was no check that a delivery was happening into an
2123 excluded directory. This bug has been fixed by ignoring all quota
2124 processing for deliveries into excluded directories.
2126 PH/10 Added the maildirfolder_create_regex option to appendfile.
2132 PH/01 The code for finding all the local interface addresses on a FreeBSD
2133 system running IPv6 was broken. This may well have applied to all BSD
2134 systems, as well as to others that have similar system calls. The broken
2135 code found IPv4 interfaces correctly, but gave incorrect values for the
2136 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2137 that it would not match correctly against @[] and not recognize the IPv6
2140 PH/02 The ipliteral router was not recognizing addresses of the form user@
2141 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2143 PH/03 Added disable_ipv6.
2145 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2146 decoded form, because it is most often used to construct To: headers
2147 lines in autoreplies, and the decoded form may well be syntactically
2148 invalid. However, $reply_address has leading white space removed, and all
2149 newlines turned into spaces so that the autoreply transport does not
2152 PH/05 If group was specified without a user on a router, and no group or user
2153 was specified on a transport, the group from the router was ignored.
2155 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2156 for visible compile-time settings that can be used to change these
2157 numbers, for those that want even more. Backwards compatibility with old
2158 spool files has been maintained. However, going back to a previous Exim
2159 release will lost any variables that are in spool files.
2161 PH/07 Two small changes when running in the test harness: increase delay when
2162 passing a TCP/IP connection to a new process, in case the original
2163 process has to generate a bounce, and remove special handling of
2164 127.0.0.2 (sic), which is no longer necessary.
2166 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2167 be the same on different OS.
2169 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2172 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2173 whether --show-vars was specified or not
2175 JJ/02 exipick: Added support for new ACL variable spool format introduced
2178 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2179 syntactically invalid From: or Reply-to: line, and a filter used this to
2180 generate an autoreply, and therefore failed to obtain an address for the
2181 autoreply, Exim could try to deliver to a non-existent relative file
2182 name, causing unrelated and misleading errors. What now happens is that
2183 it logs this as a hard delivery error, but does not attempt to create a
2186 PH/11 The exinext utility has a -C option for testing purposes, but although
2187 the given file was scanned by exinext itself; it wasn't being passed on
2188 when Exim was called.
2190 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2191 an end-of-file indication when reading a command response.
2193 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2194 compiled. In many other places in Exim, IPv6 addresses are always
2195 recognized, so I have changed this. It also means that IPv4 domain
2196 literals of the form [IPV4:n.n.n.n] are now always recognized.
2198 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2199 used if the router is not running as root, for example, when verifying at
2200 ACL time, or when using -bh. The debugging output from this situation was
2201 non-existent - all you got was a failure to exec. I have made two
2204 (a) Failures to set uid/gid, the current directory, or a process leader
2205 in a subprocess such as that created by queryprogram now generate
2206 suitable debugging ouput when -d is set.
2208 (b) The queryprogram router detects when it is not running as root,
2209 outputs suitable debugging information if -d is set, and then runs
2210 the subprocess without attempting to change uid/gid.
2212 PH/15 Minor change to Makefile for building test_host (undocumented testing
2215 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2216 additional section of a DNS packet that returns MX or SRV records.
2217 Instead, it always explicitly searches for A/AAAA records. This avoids
2218 major problems that occur when a DNS server includes only records of one
2219 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2220 fixed another bug: if SRV records were looked up and the corresponding
2221 address records were *not* found in the additional section, the port
2222 values from the SRV records were lost.
2224 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2225 using the correct key (the original address) when searching the retry
2226 rules in order to find which one to use for generating the retry hint.
2228 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2229 adding the default one. Similarly, if it contains a Reply-To: header, the
2230 errors_reply_to option, if set, is not used.
2232 PH/19 When calculating a retry time, Exim used to measure the "time since
2233 failure" by looking at the "first failed" field in the retry record. Now
2234 it does not use this if it is later than than the arrival time of the
2235 message. Instead it uses the arrival time. This makes for better
2236 behaviour in cases where some deliveries succeed, thus re-setting the
2237 "first failed" field. An example is a quota failure for a huge message
2238 when small messages continue to be delivered. Without this change, the
2239 "time since failure" will always be short, possible causing more frequent
2240 delivery attempts for the huge message than are intended.
2241 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2243 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2244 $1, $2, $3) because the numerical variables can be reset during some
2245 expansion items (e.g. "match"), thereby losing the authentication data.
2247 PH/21 Make -bV show the size of off_t variables so that the test suite can
2248 decide whether to run tests for quotas > 2G.
2250 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2251 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2252 filecount value is greater than 2G or if a quota value is greater than 2G
2253 on a system where the size of off_t is not greater than 4, a panic error
2256 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2257 never match. The debug and -bh output now contains an explicit error
2258 message indicating a malformed IPv4 address or mask.
2260 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2261 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2262 PH/23 above applies.
2264 PH/25 Do not write to syslog when running in the test harness. The only
2265 occasion when this arises is a failure to open the main or panic logs
2266 (for which there is an explicit test).
2268 PH/26 Added the /no_tell option to "control=freeze".
2270 PH/27 If a host name lookup failed very early in a connection, for example, if
2271 the IP address matched host_lookup and the reverse lookup yielded a name
2272 that did not have a forward lookup, an error message of the form "no IP
2273 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2274 could be logged. Now it outputs the IP address instead of "NULL".
2276 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2277 allows the sender and the authenticated sender to be set when
2278 submitting a message from within Exim. Since child_open_exim() is
2279 documented for local_scan(), the new function should be too.
2281 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2282 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2283 results in an empty string is now treated as unset.
2285 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2287 PH/31 Added sender_verify_fail logging option.
2289 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2290 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2291 tidied the source and removed it altogether.
2293 PH/33 When a queue run was abandoned because the load average was too high, a
2294 log line was always written; now it is written only if the queue_run log
2295 selector is set. In addition, the log line for abandonment now contains
2296 information about the queue run such as the pid. This is always present
2297 in "start" and "stop" lines but was omitted from the "abandon" line.
2299 PH/34 Omit spaces between a header name and the colon in the error message that
2300 is given when verify = headers_syntax fails (if there are lots of them,
2301 the message gets confusing).
2303 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2304 names, as there are now some PTR records that contain slashes. This check
2305 is only to protect against broken name servers that fall over on strange
2306 characters, so the fact that it applies to all lookups doesn't matter.
2308 PH/36 Now that the new test suite is complete, we can remove some of the
2309 special code in Exim that was needed for the old test suite. For example,
2310 sorting DNS records because real resolvers return them in an arbitrary
2311 order. The new test suite's fake resolver always returns records in the
2314 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2315 bounces) except when queue_only is set, to avoid logging races between
2316 the different processes.
2318 PH/38 Panic-die if .include specifies a non-absolute path.
2320 PH/39 A tweak to the "H" retry rule from its user.
2322 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2323 a label. They prevented compilation on older perls.
2325 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2326 a warning to be raised on newish perls.
2328 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2329 on queue. Changes to match documented behaviour of showing count of
2330 messages matching specified criteria.
2332 PH/40 Changed the default ident timeout from 30s to 5s.
2334 PH/41 Added support for the use of login_cap features, on those BSD systems
2335 that have them, for controlling the resources used by pipe deliveries.
2337 PH/42 The content-scanning code uses fopen() to create files in which to put
2338 message data. Previously it was not paying any attention to the mode of
2339 the files. Exim runs with umask(0) because the rest of the code creates
2340 files with open(), and sets the required mode explicitly. Thus, these
2341 files were ending up world-writeable. This was not a big issue, because,
2342 being within the spool directory, they were not world-accessible. I have
2343 created a function called modefopen, which takes an additional mode
2344 argument. It sets umask(777), creates the file, chmods it to the required
2345 mode, then resets the umask. All the relevant calls to fopen() in the
2346 content scanning code have been changed to use this function.
2348 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2349 to 24 hours. This avoids potential overflow problems when processing G
2350 and H retry rules. I suspect nobody ever tinkers with this value.
2352 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2354 PH/45 When the plaintext authenticator is running as a client, the server's
2355 challenges are checked to ensure they are valid base64 strings. By
2356 default, the authentication attempt is cancelled if an invalid string is
2357 received. Setting client_ignore_invalid_base64 true ignores these errors.
2358 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2359 they are received. Thus, the responses can be made to depend on the
2360 challenges. If an invalid string is ignored, an empty string is placed in
2363 PH/46 Messages that are created by the autoreply transport now contains a
2364 References: header, in accordance with RFCs 2822 and 3834.
2366 PH/47 Added authenticated_sender_force to the smtp transport.
2368 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2370 PH/49 Installed latest patch from the Sieve maintainer.
2372 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2373 was also set, the appendfile transport was unnecessarily scanning a
2374 directory of message files (e.g. for maildir delivery) to find the count
2375 of files (along with the size), even though it did not need this
2376 information. It now does the scan only if it needs to find either the
2377 size of the count of files.
2379 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2381 PH/52 Two bugs concerned with error handling when the smtp transport is
2384 (i) Exim was not creating retry information for temporary errors given
2385 for individual recipients after the DATA command when the smtp transport
2386 was used in LMTP mode. This meant that they could be retried too
2387 frequently, and not timed out correctly.
2389 (ii) Exim was setting the flag that allows error details to be returned
2390 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2391 recipients that were returned after the DATA command.
2393 PH/53 This is related to PH/52, but is more general: for any failing address,
2394 when detailed error information was permitted to be returned to the
2395 sender, but the error was temporary, then after the final timeout, only
2396 "retry timeout exceeded" was returned. Now it returns the full error as
2397 well as "retry timeout exceeded".
2399 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2400 do this, and (what is worse) MTAs that accept it.
2402 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2403 will now be deprecated.
2405 PH/56 New os.c-cygwin from the Cygwin maintainer.
2407 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2408 formats (previously only available in exim formats via -bpr, -bpru,
2409 and -bpra. Now also available in native and exiqgrep formats)
2411 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2412 with very large, slow to parse queues
2414 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2416 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2418 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2419 responses to authentication challenges, though it was showing the
2420 challenges; (ii) I've removed the CR characters from the debug output for
2423 PH/58 Allow for the insertion of a newline as well as a space when a string
2424 is turned into more than one encoded-word during RFC 2047 encoding. The
2425 Sieve code now uses this.
2427 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2428 data_4xx, lost_connection, tls_required.
2430 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2431 message was being sent as an SMTP response.
2433 PH/61 Add -l and -k options to exicyclog.
2435 PH/62 When verifying, if an address was redirected to one new address, so that
2436 verification continued, and the new address failed or deferred after
2437 having set something in $address_data, the value of $address_data was not
2438 passed back to the ACL. This was different to the case when no
2439 redirection occurred. The value is now passed back in both cases.
2441 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2442 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2443 use login_cap.h, so on its own it isn't the distinguishing feature. The
2444 new name refers directly to the setclassresources() function.
2446 PH/65 Added configuration files for NetBSD3.
2448 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2450 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2451 is preferred over IPv4.
2453 PH/68 The bounce_return_message and bounce_return_body options were not being
2454 honoured for bounces generated during the reception of non-SMTP messages.
2455 In particular, this applied to messages rejected by the ACL. This bug has
2456 been fixed. However, if bounce_return_message is true and bounce_return_
2457 body is false, the headers that are returned for a non-SMTP message
2458 include only those that have been read before the error was detected.
2459 (In the case of an ACL rejection, they have all been read.)
2461 PH/69 The HTML version of the specification is now built in a directory called
2462 spec_html instead of spec.html, because the latter looks like a path with
2463 a MIME-type, and this confuses some software.
2465 PH/70 Catch two compiler warnings in sieve.c.
2467 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2468 function verify_get_ident() calls ip_connect() to connect a socket, but
2469 if the "connect()" function timed out, ip_connect() used to close the
2470 socket. However, verify_get_ident() also closes the socket later, and in
2471 between Exim writes to the log, which may get opened at this point. When
2472 the socket was closed in ip_connect(), the log could get the same file
2473 descriptor number as the socket. This naturally causes chaos. The fix is
2474 not to close the socket in ip_connect(); the socket should be closed by
2475 the function that creates it. There was only one place in the code where
2476 this was missing, in the iplookup router, which I don't think anybody now
2477 uses, but I've fixed it anyway.
2479 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2480 well as to direct DNS lookups. Otherwise the handling of names in host
2481 lists is inconsistent and therefore confusing.
2487 PH/01 Two changes to the default runtime configuration:
2489 (1) Move the checks for relay_from_hosts and authenticated clients from
2490 after to before the (commented out) DNS black list checks.
2492 (2) Add control=submission to the relay_from_hosts and authenticated
2493 clients checks, on the grounds that messages accepted by these
2494 statements are most likely to be submissions.
2496 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2498 (1) Generate an error if the third argument for the ${prvs expansion is
2501 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2504 (3) Reset the variables that are obtained from the first argument of
2505 ${prvscheck and used in the second argument before leaving the code,
2506 because their memory is reclaimed, so using them afterwards may do
2509 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2510 one (it's much easier than Tom thought :-).
2512 (5) Because of (4), we can now allow for the use of $prvscheck_result
2513 inside the third argument.
2515 PH/03 For some reason, the default setting of PATH when running a command from
2516 a pipe transport was just "/usr/bin". I have changed it to
2519 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2520 anything to be listed in the output from -bV.
2522 PH/05 When a filter generated an autoreply, the entire To: header line was
2523 quoted in the delivery log line, like this:
2525 => >A.N.Other <ano@some.domain> <original@ddress> ...
2527 This has been changed so that it extracts the operative address. There
2528 may be more than one such address. If so, they are comma-separated, like
2531 => >ano@some.domain,ona@other.domain <original@ddress> ...
2533 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2534 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2535 not being looked up in the rDNS to get a host name, Exim was showing the
2536 IP address twice in Received: lines, even though the IP addresses were
2537 identical. For example:
2539 Received: from [1.2.3.4] (helo=[1.2.3.4])
2541 However, if the real host name was known, it was omitting the HELO data
2542 if it matched the actual IP address. This has been tidied up so that it
2543 doesn't show the same IP address twice.
2545 PH/07 When both +timestamp and +memory debugging was on, the value given by
2546 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2547 called by the debug printing, thereby overwriting the timestamp buffer.
2548 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2550 PH/08 When the original message was included in an autoreply transport, it
2551 always said "this is a copy of the message, including all the headers",
2552 even if body_only or headers_only was set. It now gives an appropriate
2555 PH/09 Applied a patch from the Sieve maintainer which:
2557 o fixes some comments
2558 o adds the (disabled) notify extension core
2559 o adds some debug output for the result of if/elsif tests
2560 o points to the current vacation draft in the documentation
2561 and documents the missing references header update
2565 o fixes a bug in processing the envelope test (when testing
2566 multiple envelope elements, the last element determinted the
2569 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2570 Electronic Mail") by including:
2572 Auto-submitted: auto-generated
2574 in the messages that it generates (bounce messages and others, such as
2575 warnings). In the case of bounce messages for non-SMTP mesages, there was
2576 also a typo: it was using "Auto_submitted" (underscore instead of
2577 hyphen). Since every message generated by Exim is necessarily in response
2578 to another message, thes have all been changed to:
2580 Auto-Submitted: auto-replied
2582 in accordance with these statements in the RFC:
2584 The auto-replied keyword:
2586 - SHOULD be used on messages sent in direct response to another
2587 message by an automatic process,
2589 - MUST NOT be used on manually-generated messages,
2591 - MAY be used on Delivery Status Notifications (DSNs) and Message
2592 Disposition Notifications (MDNs),
2594 - MUST NOT be used on messages generated by automatic or periodic
2595 processes, except for messages which are automatic responses to
2598 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2599 to the default Received: header definition.
2601 PH/12 Added log selector acl_warn_skipped (default on).
2603 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2604 variables because (a) they are in the wrong storage pool and (b) even if
2605 they were copied, it wouldn't work properly because of the caching.
2607 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2608 checking when decoding. Apparently there are clients that generate
2609 overlong encoded strings. Why am I not surprised?
2611 PH/15 If the first argument of "${if match_address" was not empty, but did not
2612 contain an "@" character, Exim crashed. Now it writes a panic log message
2613 and treats the condition as false.
2615 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2617 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2618 extension "envelope-auth". The code is finished and in agreement with
2619 other implementations, but there is no documentation so far and in fact,
2620 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2621 not changing the active code.
2623 Print executed "if" and "elsif" statements when debugging is used. This
2624 helps a great deal to understand what a filter does.
2626 Document more things not specified clearly in RFC3028. I had all this
2627 sorted out, when out of a sudden new issues came to my mind. Oops."
2629 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2632 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2633 canonical form (as documented). However, after a host name lookup from
2634 the IP address, check_host() was doing a simple string comparison with
2635 addresses acquired from the DNS when checking that the found name did
2636 have the original IP as one of its addresses. Since any found IPv6
2637 addresses are likely to be in abbreviated form, the comparison could
2638 fail. Luckily, there already exists a function for doing the comparison
2639 by converting both addresses to binary, so now that is used instead of
2640 the text comparison.
2642 PH/20 There was another similar case to PH/19, when a complete host name was
2643 given in a host list; looking up its IP address could give an abbreviated
2644 form, whereas the current host's name might or might not be abbreviated.
2645 The same fix has been applied.
2651 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2652 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2655 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2656 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2658 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2660 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2661 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2662 Also, for "Auto-Submitted", treat anything other than "no" as
2663 non-personal, in accordance with RFC 3834. (Previously it treated
2664 anything starting "auto-" as non-personal.)
2666 TF/01 The control=submission/name=... option had a problem with syntax
2667 errors if the name included a slash character. The /name= option
2668 now slurps the rest of the string, so it can include any characters
2669 but it must come last in the list of options (after /sender_retain
2672 PH/05 Some modifications to the interface to the fake nameserver for the new
2680 TK/01 Added the "success_on_redirect" address verification option. See
2681 NewStuff for rationale and an example.
2683 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2685 PH/02 Patch to exigrep to allow it to work on syslog lines.
2687 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2688 fread() to skip over the body file's header line, because in Cygwin the
2689 header line is locked and is inaccessible.
2691 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2692 co-exist for some time) to make it clear that it is the Exim ID that is
2693 referenced, not the Message-ID: header line.
2695 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2696 string_format() function, because snprintf() does not exist on all
2699 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2700 ${stat: expansion item.
2702 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2703 protocol synchronization error", to keep the pedants happy.
2705 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2706 well as for IRIX systems, when gcc is being used. See the host.c source
2709 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2711 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2714 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2715 transport and to the smtp transport in LMTP mode.
2717 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2719 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2720 run a filter in a subprocess. This could lead to confusion in subsequent
2721 lookups in the parent process. There should also be a search_tidyup() at
2722 the end of the subprocess.
2724 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2725 only if the host matched helo_try_verify_hosts, which caused the
2726 verification to occur when the EHLO/HELO command was issued. The ACL just
2727 tested the remembered result. Now, if a previous verification attempt has
2728 not happened, "verify = helo" does it there and then.
2730 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2732 TK/03 Fix log output including CR from clamd.
2734 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2735 exist provoked a memory error which could cause a segfault.
2737 PH/15 Installed PCRE 6.2
2739 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2741 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2742 of the problem. Specifically, suggested +O2 rather than +O1 for the
2745 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2747 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2748 record, the port was not correctly logged when the outgoing_port log
2749 selector was set (it logged the transort's default port).
2751 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2752 fallback_hosts, and "hosts" in the smtp transport.
2754 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2755 host errors such as "Connection refused".
2757 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2758 authentication with radiusclient 0.4.9:
2760 - Error returned from rc_read_config was caught wrongly
2761 - Username/password not passed on to radius server due to wrong length.
2763 The presumption is that some radiusclient API changes for 4.51/PH/17
2764 were not taken care of correctly. The code is still untested by me (my
2765 Linux distribution still has 0.3.2 of radiusclient), but it was
2766 contributed by a Radius user.
2768 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2769 expanding the "port" option of the smtp transport.
2771 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2772 while reading a MIME header. Thanks to Tom Hughes for a patch.
2774 PH/24 Include config.h inside local_scan.h so that configuration settings are
2777 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2778 that in an ACL for RCPT (for example), you can examine exactly what was
2781 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2782 commands, but it was not correctly comparing the address with the actual
2783 client host address. Thus, it would show the EHLO address in Received:
2784 header lines when this was not necessary.
2786 PH/27 Added the % operator to ${eval:}.
2788 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2789 it should be ignoring failures (because with -C, for example, it has lost
2790 privilege). It wasn't ignoring creation failures other than "already
2793 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2796 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2797 because an input line was too long, either on its own, or by virtue of
2798 too many continuations, the temporary file was not being removed, and the
2799 return code was incorrect.
2801 PH/31 Missing "BOOL" in function definition in filtertest.c.
2803 PH/32 Applied Sieve patches from the maintainer.
2805 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2807 PH/33 Added "verify = not_blind".
2809 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2810 Local/Makefile (with some defaults set). These are used in built scripts
2811 such as exicyclog, but they have never been used in the exim_install
2812 script (though there are many overriding facilities there). I have
2813 arranged that the exim_install script now takes note of these two
2816 PH/35 Installed configuration files for Dragonfly.
2818 PH/36 When a locally submitted message by a trusted user did not contain a
2819 From: header, and the sender address was obtained from -f or from an SMTP
2820 MAIL command, and the trusted user did not use -F to supply a sender
2821 name, $originator_name was incorrectly used when constructing a From:
2822 header. Furthermore, $originator_name was used for submission mode
2823 messages from external hosts without From: headers in a similar way,
2824 which is clearly wrong.
2826 PH/37 Added control=suppress_local_fixups.
2828 PH/38 When log_selector = +received_sender was set, and the addition of the
2829 sender made the log line's construction buffer exactly full, or one byte
2830 less than full, an overflow happened when the terminating "\n" was
2833 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2834 when the result of a list match is failure because a DNS lookup failed.
2836 PH/40 RM_COMMAND is now used in the building process.
2838 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2839 the "build-* directories that it finds.
2841 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2842 address in a domain literal was a prefix of an interface address.
2844 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2845 when verifying a sender address, unless rewrite_headers is false.
2847 PH/44 Wrote a long comment about why errors_to addresses are verified as
2848 recipients, not senders.
2850 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2851 the ratelimit ACL was added.
2853 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2855 PH/46 Added extra information about PostgreSQL errors to the error string.
2857 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2858 avoiding the need to install special zones in a real server. This is
2859 backwards compatible; if it can't find the fake resolver, it drops back.
2860 Thus, both old and new test suites can be run.
2862 TF/02 Added util/ratelimit.pl
2864 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2865 clock is set back in time.
2867 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2868 Candler <B.Candler@pobox.com>.
2870 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2871 OK for addresses that are the result of redirections.
2873 PH/48 A number of further additions for the benefit of the new test suite,
2874 including a fake gethostbyname() that interfaces to the fake DNS resolver
2877 TF/06 The fix for widen_domains has also been applied to qualify_single and
2878 search_parents which are the other dnslookup options that can cause
2881 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2884 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2886 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2887 service" if it encountered an MX record with an empty target hostname.
2888 The message is now "an MX or SRV record indicated no SMTP service".
2890 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2891 if the DNS of the sending site is misconfigured. This is quite a
2892 common situation. This change restores the behaviour of treating a
2893 helo verification defer as a failure.
2895 PH/51 If self=fail was set on a router, the bounce message did not include the
2896 actual error message.
2902 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2904 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2905 command itself ended in error, the underlying message about the transport
2906 filter timeout was being overwritten with the pipe command error. Now the
2907 underlying error message should be appended to the second error message.
2909 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2911 PH/02 Reduce the amount of output that "make" produces by default. Full output
2912 can still be requested.
2914 PH/03 The warning log line about a condition test deferring for a "warn" verb
2915 was being output only once per connection, rather than after each
2916 occurrence (because it was using the same function as for successful
2917 "warn" verbs). This seems wrong, so I have changed it.
2919 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2920 it should not have, which might have caused a crash in the right
2921 circumstances, but probably never did.
2923 PH/04 Installed a modified version of Tony Finch's patch to make submission
2924 mode fix the return path as well as the Sender: header line, and to
2925 add a /name= option so that you can make the user's friendly name appear
2928 TF/03 Added the control = fakedefer ACL modifier.
2930 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2931 Mark Lowes for thorough testing.
2933 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2935 TK/03 Merged latest SRS patch from Miles Wilton.
2937 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2938 with the definition in sysexits.h (which is #included earlier).
2939 Fortunately, Exim does not actually use EX_OK. The code used to try to
2940 preserve the sysexits.h value, by assumimg that macro definitions were
2941 scanned for macro replacements. I have been disabused of this notion,
2942 so now the code just undefines EX_OK before #including unistd.h.
2944 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2945 in the smtp transport. When a block could not be written in a single
2946 write() function, the timeout was being re-applied to each part-write.
2947 This seems wrong - if the receiver was accepting one byte at a time it
2948 would take for ever. The timeout is now adjusted when this happens. It
2949 doesn't have to be particularly precise.
2951 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2952 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2954 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2955 without a domain if the check to <postmaster@domain> fails.
2957 SC/01 Eximstats: added -xls and the ability to specify output files
2958 (patch written by Frank Heydlauf).
2960 SC/02 Eximstats: use FileHandles for outputing results.
2962 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2964 SC/04 Eximstats: fixed display of large numbers with -nvr option
2966 SC/05 Eximstats: fixed merging of reports with empty tables.
2968 SC/06 Eximstats: added the -include_original_destination flag
2970 SC/07 Eximstats: removed tabs and trailing whitespace.
2972 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2974 TK/06 MBOX spool code: Add real "From " MBOX separator line
2975 so the .eml file is really in mbox format (even though
2976 most programs do not really care). Patch from Alex Miller.
2978 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2979 The latter is generated from $received_to and is only set if the
2980 message has one envelope recipient. SA can use these headers,
2981 obviously out-of-the-box. Patch from Alex Miller.
2983 PH/08 The ${def test on a variable was returning false if the variable's
2984 value was "0", contrary to what the specification has always said!
2985 The result should be true unless the variable is empty.
2987 PH/09 The syntax error of a character other than { following "${if
2988 def:variable_name" (after optional whitespace) was not being diagnosed.
2989 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2990 accidental colon was present, for example, could give incorrect results.
2992 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2993 result is used (not including appendfile, where other changes are about
2996 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2997 This involved changing a lot of size variables from int to off_t. It
2998 should work with maildirs and everything.
3000 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3001 spamd dying while we are connected to it.
3003 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3006 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3007 function for BDB 4.3.
3009 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3011 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3012 the sources for PCRE have been re-arranged and more files are now
3015 PH/15 The code I had for printing potentially long long variables in PH/11
3016 above was not the best (it lost precision). The length of off_t variables
3017 is now inspected at build time, and an appropriate printing format (%ld
3018 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3019 to be "long long int" or "long int". This is needed for the internal
3020 formatting function string_vformat().
3022 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3023 the configuration file to be ":syslog", then the script "guesses" where
3024 the logs files are, rather than using the compiled in default. In our
3025 case the guess is not the same as the compiled default, so the script
3026 suddenly stopped working when I started to use syslog. The patch checks
3027 to see if log_file_path is "". If so, it attempts to read it from exim
3028 with no configuration file to get the compiled in version, before it
3029 falls back to the previous guessing code."
3031 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3032 implementing BATV in an Exim configuration. See NewStuff for the gory
3035 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3036 Makefile that are specific to HP-UX.
3038 PH/18 If the "use_postmaster" option was set for a recipient callout together
3039 with the "random" option, the postmaster address was used as the MAIL
3040 FROM address for the random test, but not for the subsequent recipient
3041 test. It is now used for both.
3043 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3044 patch removes a few documentation additions to RFC 3028, because the
3045 latest draft now contains them. It adds the new en;ascii-case comparator
3046 and a new error check for 8bit text in MIME parts. Comparator and
3047 require names are now matched exactly. I enabled the subaddress
3048 extension, but it is not well tested yet (read: it works for me)."
3050 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3051 rework some of the code of TK/09 above to avoid the hardwired use of
3052 "%lld" and "long long". Replaced the call to snprintf() with a call to
3055 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3056 records point to non-existent hosts", "retry timeout exceeded", and
3057 "retry time not reached for any host after a long failure period".
3059 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3060 experimental DomainKeys support:
3062 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3063 (2) On an error such as an illegally used "control", the wrong name for
3064 the control was given.
3066 These problems did NOT occur unless DomainKeys support was compiled.
3068 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3070 PH/24 Added ${if match_ip condition.
3072 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3073 databases so that it will be absolutely obvious if a crash occurs in the
3074 DB library. This is a regular occurrence (often caused by mis-matched
3077 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3078 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3079 picked up on a user's system that detects such things. There doesn't seem
3080 to be a gcc warning option for this - only an attribute that has to be
3081 put on the function's prototype. It seems that in Fedora Core 4 they have
3082 set this on a number of new functions. No doubt there will be more in due
3085 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3086 specify a transport. However, if an address that was verified by such a
3087 router was the subject of a callout, Exim crashed because it tried to
3088 read the rcpt_include_affixes from the non-existent transport. Now it
3089 just assumes that the setting of that option is false. This bug was
3090 introduced by 4.51/PH/31.
3092 PH/28 Changed -d+all to exclude +memory, because that information is very
3093 rarely of interest, but it makes the output a lot bigger. People tend to
3094 do -d+all out of habit.
3096 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3097 code in os-type was giving problems when libc.so lives in lib64, like on
3100 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3101 aren't the modern standard, and it seems that some systems' include files
3102 don't always have them. Exim was already checking for some of the newer
3103 ones like T_AAAA, and defining it itself. I've added checks for all the
3104 record types that Exim uses.
3106 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3107 not automatically generating a new one, as it is supposed to. This
3108 prevented TLS from working. If the file did exist, but contained invalid
3109 data, a new version was generated, as expected. It was only the case of a
3110 non-existent file that was broken.
3112 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3113 with a change in libdomainkeys > 0.64.
3115 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3116 from DNS. If the selector record carries the flag, it now has
3117 precedence over the domain-wide flag.
3119 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3121 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3122 the use of a port name that isn't defined in /etc/services) occurred, the
3123 message was deferred as in a normal delivery, and thus remained on the
3124 spool, instead of being failed because of the mua_wrapper setting. This
3125 is now fixed, and I tidied up some of the mua_wrapper messages at the
3128 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3129 the messages in a hash of arrays rather than using individual hashes.
3130 This is a bit cleaner and results in dramatic memory savings, albeit
3131 at a slight CPU cost.
3133 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3134 as requested by Marc Sherman.
3136 SC/10 Eximstats: added histograms for user specified patterns as requested
3139 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3141 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3142 fopen() in the content-scanning modules that did not already have it.
3148 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3149 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3151 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3153 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3155 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3156 to test Sieve filters that use "vacation".
3158 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3159 that changes the way the GnuTLS parameters are stored in the cache file.
3160 The new format can be generated externally. For backward compatibility,
3161 if the data in the cache doesn't make sense, Exim assumes it has read an
3162 old-format file, and it generates new data and writes a new file. This
3163 means that you can't go back to an older release without removing the
3166 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3167 work if there are any delivery delays because "one_time" forces the
3168 parent to be marked "delivered", so its unseen clone is never tried
3169 again. For this reason, Exim now forbids the simultaneous setting of
3172 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3173 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3174 entry if you want to know the details. The fix, however, neglected to
3175 consider the case when local delivery batching is involved. The test for
3176 "previously delivered" was not happening when checking to see if an
3177 address could be batched with a previous (undelivered) one; under
3178 certain circumstances this could lead to multiple deliveries to the same
3181 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3182 in its include files, and this causes problems building Exim.
3184 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3185 header_syntax) but Exim was just ignoring anything given after a slash.
3186 In particular, this caused confusion with an attempt to use "verify =
3187 reverse_host_lookup/defer_ok". An error is now given when options are
3188 supplied for verify items that do not have them. (Maybe reverse_host_
3189 lookup should have a defer_ok option, but that's a different point.)
3191 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3192 defined by RFC 821) to 2048, because there were problems with some AUTH
3193 commands, and RFC 1869 says the size should be increased for extended
3194 SMTP commands that take arguments.
3196 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3199 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3200 "unknown" error; now it says that the functionality isn't in the binary.
3202 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3203 an address' error message when a string expansion fails (syntax or
3204 whatever). Otherwise the password may appear in the log. Following change
3205 PH/42 below, there is no longer a chance of it appearing in a bounce
3208 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3210 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3211 of its addresses was ever tried. (Bugzilla bug #2.)
3213 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3214 the result incorrectly in the debug output. (It correctly added a newline
3215 to what was transported.)
3217 TF/01 Added $received_time.
3219 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3220 commented out examples of how to interface to a virus scanner and to
3221 SpamAssassin. Also added commented examples of av_scanner and
3222 spamd_address settings.
3224 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3225 and controls are allowed in which ACLs. There were a couple of minor
3226 errors. Some of the entries in the conditions table (which is a table of
3227 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3228 negation of where the condition IS allowed.
3230 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3232 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3233 header file does not have a version number, so I've had to invent a new
3234 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3235 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3236 radiusclient), but it was contributed by a Radius user.
3238 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3239 files or directories, for OpenSSL.
3241 PH/19 When an Exim process that is running as root has to create an Exim log
3242 file, it does so in a subprocess that runs as exim:exim so as to get the
3243 ownership right at creation (otherwise, other Exim processes might see
3244 the file with the wrong ownership). There was no test for failure of this
3245 fork() call, which would lead to the process getting stuck as it waited
3246 for a non-existent subprocess. Forks do occasionally fail when resources
3247 run out. I reviewed all the other calls to fork(); they all seem to check
3250 PH/20 When checking for unexpected SMTP input at connect time (before writing
3251 the banner), Exim was not dealing correctly with a non-positive return
3252 from the read() function. If the client had disconnected by this time,
3253 the result was a log entry for a synchronization error with an empty
3254 string after "input=" when read() returned zero. If read() returned -1
3255 (an event I could not check), uninitialized data bytes were printed.
3256 There were reports of junk text (parts of files, etc) appearing after
3259 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3261 PH/22 Added support for macro redefinition, and (re)definition in between
3262 driver and ACL definitions.
3264 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3265 forgetting to use the resulting value; it was using the unexpanded value.
3267 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3268 hadn't been configured. The fix is from Juergen Kreileder, who
3269 understands it better than I do:
3271 "Here's what I see happening with three configured cyrus_sasl
3272 authenticators configured (plain, login, cram-md5):
3274 On startup auth_cyrus_sasl_init() gets called for each of these.
3275 This means three calls to sasl_listmech() without a specified mech_list.
3276 => SASL tests which mechs of all available mechs actually work
3277 => three warnings about OTP not working
3278 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3280 With the patch, sasl_listmech() also gets called three times. But now
3281 SASL's mech_list option is set to the server_mech specified in the the
3282 authenticator. Or in other words, the answer from sasl_listmech()
3283 gets limited to just the mech you're testing for (which is different
3285 => the return list contains just 'plain' or 'login', 'cram-md5' or
3286 nothing depending on the value of ob->server_mech.
3288 I've just tested the patch: Authentication still works fine,
3289 unavailable mechs specified in the exim configuration are still
3290 caught, and the auth.log warnings about OTP are gone."
3292 PH/25 When debugging is enabled, the contents of the command line are added
3293 to the debugging output, even when log_selector=+arguments is not
3296 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3297 answer is "GNU", and only if the return is "GNU/something" is the answer
3300 PH/27 $acl_verify_message is now set immediately after the failure of a
3301 verification in an ACL, and so is available in subsequent modifiers. In
3302 particular, the message can be preserved by coding like this:
3304 warn !verify = sender
3305 set acl_m0 = $acl_verify_message
3307 Previously, $acl_verify_message was set only while expanding "message"
3308 and "log_message" when a very denied access.
3310 PH/28 Modified OS/os.c-Linux with
3312 -#ifndef OS_LOAD_AVERAGE
3313 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3315 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3316 nomenclature these days.)
3318 PH/29 Installed patch from the Sieve maintainer that adds the options
3319 sieve_useraddress and sieve_subaddress to the redirect router.
3321 PH/30 In these circumstances:
3322 . Two addresses routed to the same list of hosts;
3323 . First host does not offer TLS;
3324 . First host accepts first address;
3325 . First host gives temporary error to second address;
3326 . Second host offers TLS and a TLS session is established;
3327 . Second host accepts second address.
3328 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3329 and peerdn, if requested) that were in fact used only for the second
3332 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3333 attention to any local part prefix or suffix that was matched by the
3334 router that accepted the address. It now behaves in the same way as it
3335 does for delivery: the affixes are removed from the local part unless
3336 rcpt_include_affixes is set on the transport.
3338 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3339 timeout during the DATA phase of an incoming message.
3341 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3342 applied a patch sanctioned by the Sieve maintainer.
3344 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3345 the uid or gid is negative. A case of a negative gid caused this to be
3346 noticed. The fix allows for either to be negative.
3348 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3349 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3352 PH/36 The change PH/12 above was broken. Fixed it.
3354 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3355 the grounds that routing the same address twice would always produce the
3356 same answer. This might have been true once, but it is certainly no
3357 longer true now. Routing a child address may depend on the previous
3358 routing that produced that child. Some complicated redirection strategies
3359 went wrong when messages had multiple recipients, and made Exim's
3360 behaviour dependent on the order in which the addresses were given.
3362 I have moved the duplicate checking until after the routing is complete.
3363 Exim scans the addresses that are assigned to local and remote
3364 transports, and removes any duplicates. This means that more work will be
3365 done, as duplicates will always all be routed, but duplicates are
3366 presumably rare, so I don't expect this is of any significance.
3368 For deliveries to pipes, files, and autoreplies, the duplicate checking
3369 still happens during the routing process, since they are not going to be
3372 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3373 It corrects a timeout issue with spamd. This is Ian's comment: "The
3374 background is that sometimes spamd either never reads data from a
3375 connection it has accepted, or it never writes response data. The exiscan
3376 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3377 blindly assumes that writes won't block so it may never time out."
3379 PH/39 Allow G after quota size as well as K and M.
3381 PH/40 The value set for $authenticated_id in an authenticator may not contain
3382 binary zeroes or newlines because the value is written to log lines and
3383 to spool files. There was no check on this. Now the value is run through
3384 the string_printing() function so that such characters are converted to
3385 printable escape sequences.
3387 PH/41 $message_linecount is a new variable that contains the total number of
3388 lines in the message. Compare $body_linecount, which is the count for the
3391 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3392 bounce and delay warning messages, except in certain special cases, which
3395 (a) An SMTP error message from a remote host;
3396 (b) A message specified in a :fail: redirection;
3397 (c) A message specified in a "fail" command in a system filter;
3398 (d) A message specified in a FAIL return from the queryprogram router;
3399 (e) A message specified by the cannot_route_message router option.
3401 In these cases only, Exim does include the error details in bounce and
3402 warning messages. There are also a few cases where bland messages such
3403 as "unrouteable address" or "local delivery error" are given.
3405 PH/43 $value is now also set for the "else" part of a ${run expansion.
3407 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3408 being worked on, but at least Exim now implements the latest version to
3411 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3412 process to complete was treated as a delivery failure, a timeout while
3413 writing the message to the pipe was logged, but erroneously treated as a
3414 successful delivery. Such timeouts include transport filter timeouts. For
3415 consistency with the overall process timeout, these timeouts are now
3416 treated as errors, giving rise to delivery failures by default. However,
3417 there is now a new Boolean option for the pipe transport called
3418 timeout_defer, which, if set TRUE, converts the failures into defers for
3419 both kinds of timeout. A transport filter timeout is now identified in
3422 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3423 systems where "make" and "gmake" are different, calling "gmake" at top
3424 level broke things. I've arranged for the value of $(MAKE) to be passed
3425 from the Makefile to this script so that it can call the same version of
3429 A note about Exim versions 4.44 and 4.50
3430 ----------------------------------------
3432 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3433 changes of various kinds. As a consequence, a big documentation update was
3434 needed. This delayed the release for rather longer than seemed good, especially
3435 in the light of a couple of (minor) security issues. Therefore, the changes
3436 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3437 release. So 4.44 and 4.50 are in effect two different branches that both start
3440 I have left the 4.50 change log unchanged; it contains all the changes since
3441 4.43. The change log for 4.44 is below; many of its items are identical to
3442 those for 4.50. This seems to be the most sensible way to preserve the
3443 historical information.
3449 1. Minor wording change to the doc/README.SIEVE file.
3451 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3452 computation of the current number of files was incorrect.
3454 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3455 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3458 4. Give more explanation in the error message when the command for a transport
3459 filter fails to execute.
3461 5. There are several places where Exim runs a non-Exim command in a
3462 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3463 was being done only for the command run by the queryprogram router. It is
3464 now done for all such subprocesses. The other cases are: ${run, transport
3465 filters, and the commands run by the lmtp and pipe transports.
3467 6. Added CONFIGURE_GROUP build-time option.
3469 7. Some older OS have a limit of 256 on the maximum number of file
3470 descriptors. Exim was using setrlimit() to set 1000 as a large value
3471 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3472 systems. I've change it so that if it can't get 1000, it tries for 256.
3474 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3475 was an oversight, and furthermore, ever since the addition of extra
3476 controls (e.g. 4.43/32), the checks on when to allow different forms of
3477 "control" were broken. There should now be diagnostics for all cases when a
3478 control that does not make sense is encountered.
3480 9. Added the /retain_sender option to "control=submission".
3482 10. $recipients is now available in the predata ACL (oversight).
3484 11. Tidy the search cache before the fork to do a delivery from a message
3485 received from the command line. Otherwise the child will trigger a lookup
3486 failure and thereby defer the delivery if it tries to use (for example) a
3487 cached ldap connection that the parent has called unbind on.
3489 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3490 of $address_data from the recipient verification was clobbered by the
3491 sender verification.
3493 13. The value of address_data from a sender verification is now available in
3494 $sender_address_data in subsequent conditions in the ACL statement.
3496 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3498 15. Added a new option "connect=<time>" to callout options, to set a different
3501 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3502 was its contents. (It was OK if the option was not defined at all.)
3504 17. A "Completed" log line is now written for messages that are removed from
3505 the spool by the -Mrm option.
3507 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3508 information about exactly what failed.
3510 19. Added -dd to debug only the daemon process.
3512 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3513 handles timeouts, both on the server side and network timeouts. Renamed the
3514 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3516 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3517 It is now set to "smtps".
3519 22. $host_address is now set to the target address during the checking of
3520 ignore_target_hosts.
3522 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3523 being passed; this would have caused $sender_host_name to have been used if
3524 matching the list had actually called for a host name (not very likely,
3525 since this list is usually IP addresses). A host name is now passed as
3528 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3529 code that specifies a non-restarting handler (typically sigaction() in
3530 modern systems) in an attempt to fix a rare and obscure crash bug.
3532 25. Narrowed the window for a race in the daemon that could cause it to ignore
3533 SIGCHLD signals. This is not a major problem, because they are used only to
3534 wake it up if nothing else does.
3536 26. A malformed maildirsize file could cause Exim to calculate negative values
3537 for the mailbox size or file count. Odd effects could occur as a result.
3538 The maildirsize information is now recalculated if the size or filecount
3541 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3542 support for a long time. Removed HAVE_SYS_VFS_H.
3544 28. Installed the latest version of exipick from John Jetmore.
3546 29. In an address list, if the pattern was not a regular expression, an empty
3547 subject address (from a bounce message) matched only if the pattern was an
3548 empty string. Non-empty patterns were not even tested. This was the wrong
3549 because it is perfectly reasonable to use an empty address as part of a
3550 database query. An empty address is now tested by patterns that are
3551 lookups. However, all the other forms of pattern expect the subject to
3552 contain a local part and a domain, and therefore, for them, an empty
3553 address still always fails if the pattern is not itself empty.
3555 30. Exim went into a mad DNS loop when attempting to do a callout where the
3556 host was specified on an smtp transport, and looking it up yielded more
3557 than one IP address.
3559 31. Re-factored the code for checking spool and log partition space into a
3560 function that finds that data and another that does the check. The former
3561 is then used to implement four new variables: $spool_space, $log_space,
3562 $spool_inodes, and $log_inodes.
3564 32. The RFC2047 encoding function was originally intended for short strings
3565 such as real names; it was not keeping to the 75-character limit for
3566 encoded words that the RFC imposes. It now respects the limit, and
3567 generates multiple encoded words if necessary. To be on the safe side, I
3568 have increased the buffer size for the ${rfc2047: expansion operator from
3571 33. It is now permitted to omit both strings after an "if" condition; if the
3572 condition is true, the result is "true". As before, when the second string
3573 is omitted, a false condition yields an empty string. This makes it less
3574 cumbersome to write custom ACL and router conditions.
3576 34. Failure to deliver a bounce message always caused it to be frozen, even if
3577 there was an errors_to setting on the router. The errors_to setting is now
3580 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3581 canonical form (fully expanded) before being placed in
3582 $sender_host_address.
3584 36. The table in the code that translates DNS record types into text (T_A to
3585 "A" for instance) was missing entries for NS and CNAME. It is just possible
3586 that this could have caused confusion if both these types were looked up
3587 for the same domain, because the text type is used as part of Exim's
3588 per-process caching. But the chance of anyone hitting this buglet seems
3591 37. The dnsdb lookup has been extended in a number of ways.
3593 (1) There is a new type, "zns", which walks up the domain tree until it
3594 finds some nameserver records. It should be used with care.
3596 (2) There is a new type, "mxh", which is like "mx" except that it returns
3597 just the host names, not the priorities.
3599 (3) It is now possible to give a list of domains (or IP addresses) to be
3600 looked up. The behaviour when one of the lookups defers can be
3601 controlled by a keyword.
3603 (4) It is now possible to specify the separator character for use when
3604 multiple records are returned.
3606 38. The dnslists ACL condition has been extended: it is now possible to supply
3607 a list of IP addresses and/or domains to be looked up in a particular DNS
3610 39. Added log_selector=+queue_time_overall.
3612 40. When running the queue in the test harness, wait just a tad after forking a
3613 delivery process, to get repeatability of debugging output.
3615 41. Include certificate and key file names in error message when GnuTLS fails
3616 to set them up, because the GnuTLS error message doesn't include the name
3617 of the failing file when there is a problem reading it.
3619 42. Allow both -bf and -bF in the same test run.
3621 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3623 44. The "Exiscan patch" is now merged into the mainline Exim source.
3625 45. Sometimes the final signoff response after QUIT could fail to get
3626 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3627 < 0 before doing a fflush(). This bug looks as though it goes back to the
3628 introduction of TLS in release 3.20, but "sometimes" must have been rare
3629 because the tests only now provoked it.
3631 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3632 (this can affect the format of dates).
3634 47. exim_tidydb, when checking for the continued existence of a message for
3635 which it has found a message-specific retry record, was not finding
3636 messages that were in split spool directories. Consequently, it was
3637 deleting retry records that should have stayed in existence.
3639 48. Steve fixed some bugs in eximstats.
3641 49. The SPA authentication driver was not abandoning authentication and moving
3642 on to the next authenticator when an expansion was forced to fail,
3643 contradicting the general specification for all authenticators. Instead it
3644 was generating a temporary error. It now behaves as specified.
3646 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3647 (the order specifies the preference for clients). The order is now AES256,
3648 AES128, 3DES, ARCFOUR128.
3650 51. Small patch to Sieve code - explicitly set From: when generating an
3653 52. Exim crashed if a remote delivery caused a very long error message to be
3654 recorded - for instance if somebody sent an entire SpamAssassin report back
3655 as a large number of 550 error lines. This bug was coincidentally fixed by
3656 increasing the size of one of Exim's internal buffers (big_buffer) that
3657 happened as part of the Exiscan merge. However, to be on the safe side, I
3658 have made the code more robust (and fixed the comments that describe what
3661 53. Now that there can be additional text after "Completed" in log lines (if
3662 the queue_time_overall log selector is set), a one-byte patch to exigrep
3663 was needed to allow it to recognize "Completed" as not the last thing in
3666 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3667 patch that reportedly fixes this has been added. I am not expert enough to
3668 create a test for it. This is what the patch creator wrote:
3670 "I found a little strange behaviour of ldap code when working with
3671 Windows 2003 AD Domain, where users was placed in more than one
3672 Organization Units. When I tried to give exim partial DN, the exit code
3673 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3674 But simultaneously result of request was absolutely normal ldap result,
3675 so I produce this patch..."
3677 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3678 REFERENCE, so I have modified the code to exclude the patch when that macro
3681 55. Some experimental protocols are using DNS PTR records for new purposes. The
3682 keys for these records are domain names, not reversed IP addresses. The
3683 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3684 leaves it alone. Component reversal etc. now happens only for IP addresses.
3687 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3689 57. Double the size of the debug message buffer (to 2048) so that more of very
3690 long debug lines gets shown.
3692 58. The exicyclog utility now does better if the number of log files to keep
3693 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3695 59. Two changes related to the smtp_active_hostname option:
3697 (1) $smtp_active_hostname is now available as a variable.
3698 (2) The default for smtp_banner uses $smtp_active_hostname instead
3699 of $primary_hostname.
3701 60. The host_aton() function is supposed to be passed a string that is known
3702 to be a valid IP address. However, in the case of IPv6 addresses, it was
3703 not checking this. This is a hostage to fortune. Exim now panics and dies
3704 if the condition is not met. A case was found where this could be provoked
3705 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3706 components; fortuitously, this particular loophole had already been fixed
3707 by change 4.50/55 above.
3709 If there are any other similar loopholes, the new check in host_aton()
3710 itself should stop them being exploited. The report I received stated that
3711 data on the command line could provoke the exploit when Exim was running as
3712 exim, but did not say which command line option was involved. All I could
3713 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3714 running as the user.
3717 61. There was a buffer overflow vulnerability in the SPA authentication code
3718 (which came originally from the Samba project). I have added a test to the
3719 spa_base64_to_bits() function which I hope fixes it.
3722 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3723 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3725 63. The daemon start-up calls getloadavg() while still root for those OS that
3726 need the first call to be done as root, but it missed one case: when
3727 deliver_queue_load_max is set with deliver_drop_privilege. This is
3728 necessary for the benefit of the queue runner, because there is no re-exec
3729 when deliver_drop_privilege is set.
3731 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3732 This has been fixed.
3734 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3735 was in use, was not putting the data itself into the right store pool;
3736 consequently, it could be overwritten for a subsequent message in the same
3737 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3740 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3742 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3743 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3744 the function were treating the return as a boolean value, which happened to
3745 work because 0=false and not-0=true, but is not correct code.
3747 68. The host_aton() function was not handling scoped IPv6 addresses (those
3748 with, for example, "%eth0" on the end) correctly.
3750 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3751 negated items (that is, ~something) in unsigned ints. Some compilers
3752 apparently mutter when there is no cast.
3754 70. If an address verification called from an ACL failed, and did not produce a
3755 user-specific message (i.e. there was only a "system" message), nothing was
3756 put in $acl_verify_message. In this situation, it now puts the system
3759 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3760 session; change 4.31/43 added the unwanted input to the log line - except
3761 that it did not do this in the start of session case. It now does.
3763 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3764 This is wrong and can cause the other end to generate a synchronization
3765 error if it is another Exim or anything else that does the synchronization
3766 check. A QUIT command is no longer sent after a timeout.
3768 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3769 during host lookups.
3771 74. The defer_ok option of callout verification was not working if it was used
3772 when verifying addresses in header lines, that is, for this case:
3774 verify = header_sender/callout=defer_ok
3776 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3777 those file descriptors could be used for SMTP connections. If anything
3778 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3779 could be sent to the SMTP client, causing chaos. The daemon now opens
3780 stdin, stdout, and stderr to /dev/null when it puts itself into the
3783 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3784 log by default. The user can override this with suitable Perl magic.
3786 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3787 the log message when discard triggers, was not working for the DATA ACL or
3788 for the non-SMTP ACL.
3790 78. Error message wording change in sieve.c.
3792 79. If smtp_accept_max_per_host was set, the number of connections could be
3793 restricted to fewer than expected, because the daemon was trying to set up
3794 a new connection before checking whether the processes handling previous
3795 connections had finished. The check for completed processes is now done
3796 earlier. On busy systems, this bug wouldn't be noticed because something
3797 else would have woken the daemon, and it would have reaped the completed
3800 80. If a message was submitted locally by a user whose login name contained one
3801 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3802 It caused a spool format error. I have fixed the spool reading code. A
3803 related problem was that the "from" clause in the Received: line became
3804 illegal because of the space(s). It is now covered by ${quote_local_part.
3806 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3809 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3811 83. Patch from Sieve maintainer to fix unterminated string problem in
3812 "vacation" handling.
3814 84. Some minor changes to the Linux configuration files to help with other
3815 OS variants using glibc.
3817 85. One more patch for Sieve to update vacation handling to latest spec.
3820 ----------------------------------------------------
3821 See the note above about the 4.44 and 4.50 releases.
3822 ----------------------------------------------------
3828 1. Change 4.43/35 introduced a bug that caused file counts to be
3829 incorrectly computed when quota_filecount was set in an appendfile
3832 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3833 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3836 3. Give more explanation in the error message when the command for a transport
3837 filter fails to execute.
3839 4. There are several places where Exim runs a non-Exim command in a
3840 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3841 was being done only for the command run by the queryprogram router. It is
3842 now done for all such subprocesses. The other cases are: ${run, transport
3843 filters, and the commands run by the lmtp and pipe transports.
3845 5. Some older OS have a limit of 256 on the maximum number of file
3846 descriptors. Exim was using setrlimit() to set 1000 as a large value
3847 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3848 systems. I've change it so that if it can't get 1000, it tries for 256.
3850 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3851 was an oversight, and furthermore, ever since the addition of extra
3852 controls (e.g. 4.43/32), the checks on when to allow different forms of
3853 "control" were broken. There should now be diagnostics for all cases when a
3854 control that does not make sense is encountered.
3856 7. $recipients is now available in the predata ACL (oversight).
3858 8. Tidy the search cache before the fork to do a delivery from a message
3859 received from the command line. Otherwise the child will trigger a lookup
3860 failure and thereby defer the delivery if it tries to use (for example) a
3861 cached ldap connection that the parent has called unbind on.
3863 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3864 of $address_data from the recipient verification was clobbered by the
3865 sender verification.
3867 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3868 was its contents. (It was OK if the option was not defined at all.)
3870 11. A "Completed" log line is now written for messages that are removed from
3871 the spool by the -Mrm option.
3873 12. $host_address is now set to the target address during the checking of
3874 ignore_target_hosts.
3876 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3877 being passed; this would have caused $sender_host_name to have been used if
3878 matching the list had actually called for a host name (not very likely,
3879 since this list is usually IP addresses). A host name is now passed as
3882 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3883 code that specifies a non-restarting handler (typically sigaction() in
3884 modern systems) in an attempt to fix a rare and obscure crash bug.
3886 15. Narrowed the window for a race in the daemon that could cause it to ignore
3887 SIGCHLD signals. This is not a major problem, because they are used only to
3888 wake it up if nothing else does.
3890 16. A malformed maildirsize file could cause Exim to calculate negative values
3891 for the mailbox size or file count. Odd effects could occur as a result.
3892 The maildirsize information is now recalculated if the size or filecount
3895 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3896 support for a long time. Removed HAVE_SYS_VFS_H.
3898 18. Updated exipick to current release from John Jetmore.
3900 19. Allow an empty sender to be matched against a lookup in an address list.
3901 Previously the only cases considered were a regular expression, or an
3904 20. Exim went into a mad DNS lookup loop when doing a callout where the
3905 host was specified on the transport, if the DNS lookup yielded more than
3908 21. The RFC2047 encoding function was originally intended for short strings
3909 such as real names; it was not keeping to the 75-character limit for
3910 encoded words that the RFC imposes. It now respects the limit, and
3911 generates multiple encoded words if necessary. To be on the safe side, I
3912 have increased the buffer size for the ${rfc2047: expansion operator from
3915 22. Failure to deliver a bounce message always caused it to be frozen, even if
3916 there was an errors_to setting on the router. The errors_to setting is now
3919 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3920 canonical form (fully expanded) before being placed in
3921 $sender_host_address.
3923 24. Updated eximstats to version 1.33
3925 25. Include certificate and key file names in error message when GnuTLS fails
3926 to set them up, because the GnuTLS error message doesn't include the name
3927 of the failing file when there is a problem reading it.
3929 26. Expand error message when OpenSSL has problems setting up cert/key files.
3932 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3933 (this can affect the format of dates).
3935 28. exim_tidydb, when checking for the continued existence of a message for
3936 which it has found a message-specific retry record, was not finding
3937 messages that were in split spool directories. Consequently, it was
3938 deleting retry records that should have stayed in existence.
3940 29. eximstats updated to version 1.35
3941 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3942 1.35 - bugfix such that pie charts by volume are generated correctly
3944 30. The SPA authentication driver was not abandoning authentication and moving
3945 on to the next authenticator when an expansion was forced to fail,
3946 contradicting the general specification for all authenticators. Instead it
3947 was generating a temporary error. It now behaves as specified.
3949 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3950 (the order specifies the preference for clients). The order is now AES256,
3951 AES128, 3DES, ARCFOUR128.
3953 31. Small patch to Sieve code - explicitly set From: when generating an
3956 32. Exim crashed if a remote delivery caused a very long error message to be
3957 recorded - for instance if somebody sent an entire SpamAssassin report back
3958 as a large number of 550 error lines. This bug was coincidentally fixed by
3959 increasing the size of one of Exim's internal buffers (big_buffer) that
3960 happened as part of the Exiscan merge. However, to be on the safe side, I
3961 have made the code more robust (and fixed the comments that describe what
3964 33. Some experimental protocols are using DNS PTR records for new purposes. The
3965 keys for these records are domain names, not reversed IP addresses. The
3966 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3967 leaves it alone. Component reversal etc. now happens only for IP addresses.
3970 34. The host_aton() function is supposed to be passed a string that is known
3971 to be a valid IP address. However, in the case of IPv6 addresses, it was
3972 not checking this. This is a hostage to fortune. Exim now panics and dies
3973 if the condition is not met. A case was found where this could be provoked
3974 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3975 components; fortuitously, this particular loophole had already been fixed
3976 by change 4.50/55 or 4.44/33 above.
3978 If there are any other similar loopholes, the new check in host_aton()
3979 itself should stop them being exploited. The report I received stated that
3980 data on the command line could provoke the exploit when Exim was running as
3981 exim, but did not say which command line option was involved. All I could
3982 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3983 running as the user.
3986 35. There was a buffer overflow vulnerability in the SPA authentication code
3987 (which came originally from the Samba project). I have added a test to the
3988 spa_base64_to_bits() function which I hope fixes it.
3991 36. The daemon start-up calls getloadavg() while still root for those OS that
3992 need the first call to be done as root, but it missed one case: when
3993 deliver_queue_load_max is set with deliver_drop_privilege. This is
3994 necessary for the benefit of the queue runner, because there is no re-exec
3995 when deliver_drop_privilege is set.
3997 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3998 was in use, was not putting the data itself into the right store pool;
3999 consequently, it could be overwritten for a subsequent message in the same
4000 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4003 38. Sometimes the final signoff response after QUIT could fail to get
4004 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4005 < 0 before doing a fflush(). This bug looks as though it goes back to the
4006 introduction of TLS in release 3.20, but "sometimes" must have been rare
4007 because the tests only now provoked it.
4013 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4014 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4015 Now it returns an integer. A number of calls were still expecting a T/F
4016 return. Fortuitously, in all cases, the tests worked in OK situations,
4017 which is the norm. However, things would have gone wrong on any write
4018 failures on the smtp file descriptor. This function is used when sending
4019 messages over SMTP and also when doing verify callouts.
4021 2. When Exim is called to do synchronous delivery of a locally submitted
4022 message (the -odf or -odi options), it no longer closes stderr before doing
4025 3. Implemented the mua_wrapper option.
4027 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4029 5. Implemented the functions header_remove(), header_testname(),
4030 header_add_at_position(), and receive_remove_recipient(), and exported them
4033 6. If an ACL "warn" statement specified the addition of headers, Exim already
4034 inserted X-ACL-Warn: at the start if there was no header name. However, it
4035 was not making this test for the second and subsequent header lines if
4036 there were newlines in the string. This meant that an invalid header could
4037 be inserted if Exim was badly configured.
4039 7. Allow an ACL "warn" statement to add header lines at the start or after all
4040 the Received: headers, as well as at the end.
4042 8. Added the rcpt_4xx retry error code.
4044 9. Added postmaster_mailfrom=xxx to callout verification option.
4046 10. Added mailfrom=xxxx to the callout verification option, for verify=
4049 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4050 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4052 12. Inserted some casts to stop certain compilers warning when using pointer
4053 differences as field lengths or precisions in printf-type calls (mostly
4054 affecting debugging statements).
4056 13. Added optional readline() support for -be (dynamically loaded).
4058 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4059 same clock tick as a message's arrival, so that its received time was the
4060 same as the "first fail" time on the retry record, and that message
4061 remained on the queue past the ultimate address timeout, every queue runner
4062 would try a delivery (because it was past the ultimate address timeout) but
4063 after another failure, the ultimate address timeout, which should have then
4064 bounced the address, did not kick in. This was a "< instead of <=" error;
4065 in most cases the first failure would have been in the next clock tick
4066 after the received time, and all would be well.
4068 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4069 being recognized when the domain list was tested by the match_domain
4070 condition in an expansion string.
4072 16. Added the ${str2b64: operator.
4074 17. Exim was always calling setrlimit() to set a large limit for the number of
4075 processes, without checking whether the existing limit was already
4076 adequate. (It did check for the limit on file descriptors.) Furthermore,
4077 errors from getrlimit() and setrlimit() were being ignored. Now they are
4078 logged to the main and panic logs, but Exim does carry on, to try to do its
4079 job under whatever limits there are.
4081 18. Imported PCRE 5.0.
4083 19. Trivial typo in log message " temporarily refused connection" (the leading
4086 20. If the log selector return_path_on_delivery was set and an address was
4087 redirected to /dev/null, the delivery process crashed because it assumed
4088 that a return path would always be set for a "successful" delivery. In this
4089 case, the whole delivery is bypassed as an optimization, and therefore no
4092 21. Internal re-arrangement: the function for sending a challenge and reading
4093 a response while authentication was assuming a zero-terminated challenge
4094 string. It's now changed to take a pointer and a length, to allow for
4095 binary data in such strings.
4097 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4099 23. Exim was not respecting finduser_retries when seeking the login of the
4100 uid under which it was called; it was always trying 10 times. (The default
4101 setting of finduser_retries is zero.) Also, it was sleeping after the final
4102 failure, which is pointless.
4104 24. Implemented tls_on_connect_ports.
4106 25. Implemented acl_smtp_predata.
4108 26. If the domain in control=submission is set empty, Exim assumes that the
4109 authenticated id is a complete email address when it generates From: or
4110 Sender: header lines.
4112 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4113 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4114 chown and chgrp in /bin and hostname in /usr/bin.
4116 28. Exim was keeping the "process log" file open after each use, just as it
4117 does for the main log. This opens the possibility of it remaining open for
4118 long periods when the USR1 signal hits a daemon. Occasional processlog
4119 errors were reported, that could have been caused by this. Anyway, it seems
4120 much more sensible not to leave this file open at all, so that is what now
4123 29. The long-running daemon process does not normally write to the log once it
4124 has entered its main loop, and it closes the log before doing so. This is
4125 so that log files can straightforwardly be renamed and moved. However,
4126 there are a couple of unusual error situations where the daemon does write
4127 log entries, and I had neglected to close the log afterwards.
4129 30. The text of an SMTP error response that was received during a remote
4130 delivery was being truncated at 512 bytes. This is too short for some of
4131 the long messages that one sometimes sees. I've increased the limit to
4134 31. It is now possible to make retry rules that apply only when a message has a
4135 specific sender, in particular, an empty sender.
4137 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4138 it possible to be selective about when SMTP synchronization is enforced.
4140 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4142 32. Implemented hosts_connection_nolog.
4144 33. Added an ACL for QUIT.
4146 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4149 35. Added mailbox_size and mailbox_filecount to appendfile.
4151 36. Added control = no_multiline_responses to ACLs.
4153 37. There was a bug in the logic of the code that waits for the clock to tick
4154 in the case where the clock went backwards by a substantial amount such
4155 that the microsecond fraction of "now" was more than the microsecond
4156 fraction of "then" (but the whole seconds number was less).
4158 38. Added support for the libradius Radius client library this is found on
4159 FreeBSD (previously only the radiusclient library was supported).
4165 1. When certain lookups returned multiple values in the form name=value, the
4166 quoting of the values was not always being done properly. Specifically:
4167 (a) If the value started with a double quote, but contained no whitespace,
4169 (b) If the value contained whitespace other than a space character (i.e.
4170 tabs or newlines or carriage returns) it was not quoted.
4171 This fix has been applied to the mysql and pgsql lookups by writing a
4172 separate quoting function and calling it from the lookup code. The fix
4173 should probably also be applied to nisplus, ibase and oracle lookups, but
4174 since I cannot test any of those, I have not disturbed their existing code.
4176 2. A hit in the callout cache for a specific address caused a log line with no
4177 reason for rejecting RCPT. Now it says "Previous (cached) callout
4178 verification failure".
4180 3. There was an off-by-one bug in the queryprogram router. An over-long
4181 return line was truncated at 256 instead of 255 characters, thereby
4182 overflowing its buffer with the terminating zero. As well as fixing this, I
4183 have increased the buffer size to 1024 (and made a note to document this).
4185 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4186 when Exim is waiting for an SMTP response from a remote server, Exim
4187 restarts its select() call on the socket, thereby resetting its timeout.
4188 This is not a problem when such interrupts are rare. Somebody set up a cron
4189 job to run exiwhat every 2 minutes, which is less than the normal select()
4190 timeout (5 or 10 minutes). This meant that the select() timeout never
4191 kicked in because it was always reset. I have fixed this by comparing the
4192 time when an interrupt arrives with the time at the start of the first call
4193 to select(). If more time than the timeout has elapsed, the interrupt is
4194 treated as a timeout.
4196 5. Some internal re-factoring in preparation for the addition of Sieve
4197 extensions (by MH). In particular, the "personal" test is moved to a
4198 separate function, and given an option for scanning Cc: and Bcc: (which is
4199 not set for Exim filters).
4201 6. When Exim created an email address using the login of the caller as the
4202 local part (e.g. when creating a From: or Sender: header line), it was not
4203 quoting the local part when it contained special characters such as @.
4205 7. Installed new OpenBSD configuration files.
4207 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4208 try to make them clearer.
4210 9. Callout options, other than the timeout value, were being ignored when
4211 verifying sender addresses in header lines. For example, when using
4213 verify = header_sender/callout=no_cache
4215 the cache was (incorrectly) being used.
4217 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4218 only the Cygwin environment.
4220 11. When return_path_on_delivery was set as a log selector, if different remote
4221 addresses in the same message used different return paths and parallel
4222 remote delivery occurred, the wrong values would sometimes be logged.
4223 (Whenever a remote delivery process finished, the return path value from
4224 the most recently started remote delivery process was logged.)
4226 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4227 lines when AUTH and/or TLS are in use. This is the "received protocol"
4228 field. Exim used to use "asmtp" for authenticated SMTP, without any
4229 indication (in the protocol name) for TLS use. Now it follows the RFC and
4230 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4231 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4232 names appear in log lines as well as in Received: header lines.
4234 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4235 extensions, and comparison tests, and to fix some bugs.
4237 14. Changes to the "personal" filter test:
4239 (1) The test was buggy in that it was just doing the equivalent of
4240 "contains" tests on header lines. For example, if a user's address was
4241 anne@some.where, the "personal" test would incorrectly be true for
4243 To: susanne@some.where
4245 This test is now done by extracting each address from the header in turn,
4246 and checking the entire address. Other tests that are part of "personal"
4247 are now done using regular expressions (for example, to check local parts
4248 of addresses in From: header lines).
4250 (2) The list of non-personal local parts in From: addresses has been
4251 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4252 taken from the Sieve specification recommendations.
4254 (3) If the message contains any header line starting with "List-" it is
4255 treated as non-personal.
4257 (4) The test for "circular" in the Subject: header line has been removed
4258 because it now seems ill-conceived.
4260 15. Minor typos in src/EDITME comments corrected.
4262 16. Installed latest exipick from John Jetmore.
4264 17. If headers_add on a router specified a text string that was too long for
4265 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4266 of string_sprintf() is now avoided.
4268 18. $message_body_size was not set (it was always zero) when running the DATA
4269 ACL and the local_scan() function.
4271 19. For the "mail" command in an Exim filter, no default was being set for
4272 the once_repeat time, causing a random time value to be used if "once" was
4273 specified. (If the value happened to be <= 0, no repeat happened.) The
4274 default is now 0s, meaning "never repeat". The "vacation" command was OK
4275 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4276 (I found it when inspecting the code).
4278 20. There is now an overall timeout for performing a callout verification. It
4279 defaults to 4 times the callout timeout, which applies to individual SMTP
4280 commands during the callout. The overall timeout applies when there is more
4281 than one host that can be tried. The timeout is checked before trying the
4282 next host. This prevents very long delays if there are a large number of
4283 hosts and all are timing out (e.g. when the network connections are timing
4284 out). The value of the overall timeout can be changed by specifying an
4285 additional sub-option for "callout", called "maxwait". For example:
4287 verify = sender/callout=5s,maxwait=20s
4289 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4290 to the end before writing, but this should make it even safer).
4292 22. Exim was forgetting that it had advertised PIPELINING for the second and
4293 subsequent messages on an SMTP connection. It was also not resetting its
4294 memory on STARTTLS and an internal HELO.
4296 23. When Exim logs an SMTP synchronization error within a session, it now
4297 records whether PIPELINING has been advertised or not.
4299 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4300 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4301 rather than long int.
4303 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4305 26. Added the never_mail option to autoreply.
4311 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4312 crash if the getsockname() call failed; this can happen if a connection is
4313 closed very soon after it is established. The problem was simply in the
4314 order in which certain operations were done, causing Exim to try to write
4315 to the SMTP stream before it had set up the file descriptor. The bug has
4316 been fixed by making things happen in the correct order.
4322 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4323 before the connection was closed, thus losing the rejection response.
4325 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4326 some early Solaris releases, but causes trouble in current releases where
4327 socklen_t is defined.
4329 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4332 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4335 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4336 that releasing the top of it at the end releases what was used for sub-
4337 expansions (unless the block got too big). However, discard this block if
4338 the first thing is a variable or header, so that we can use its block when
4339 it is dynamic (useful for very large $message_headers, for example).
4341 6. Lookups now cache *every* query, not just the most recent. A new, separate
4342 store pool is used for this. It can be recovered when all lookup caches are
4343 flushed. Lookups now release memory at the end of their result strings.
4344 This has involved some general refactoring of the lookup sources.
4346 7. Some code has been added to the store_xxx() functions to reduce the amount
4347 of flapping under certain conditions.
4349 8. log_incoming_interface used to affect only the <= reception log lines. Now
4350 it causes the local interface and port to be added to several more SMTP log
4351 lines, for example "SMTP connection from", and rejection lines.
4353 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4355 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4357 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4358 could be overwritten at the end of the current message (or the start of a
4359 new message if it was set in a HELO ACL). The value is now preserved for
4360 the duration of the SMTP connection.
4362 12. If a transport had a headers_rewrite setting, and a matching header line
4363 contained an unqualified address, that address was qualified, even if it
4364 did not match any rewriting rules. The underlying bug was that the values
4365 of the flags that permit the existence of unqualified sender and recipient
4366 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4367 non-local messages, and by -bnq for local messages) were not being
4368 preserved with the message after it was received.
4370 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4371 "next input=" as part of the text comprising the host identity instead of
4372 the correct text. The code was using the same buffer for two different
4373 strings. However, depending on which order the printing function evaluated
4374 its arguments, the bug did not always show up. Under Linux, for example, my
4375 test suite worked just fine.
4377 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4378 doesn't work with some older versions of Perl. It has been changed to "my",
4379 which in any case is probably the better facility to use.
4381 15. A really picky compiler found some instances of statements for creating
4382 error messages that either had too many or two few arguments for the format
4385 16. The size of the buffer for calls to the DNS resolver has been increased
4386 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4387 for addresses that have a lot of PTR records. This alleviates a problem; it
4388 does not fully solve it.
4390 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4391 buffer now truncates the list and logs the incident, which is the same
4392 action as happens when Exim is looking up a host name and its aliases.
4393 Previously in this situation something unpredictable would happen;
4394 sometimes it was "internal error: store_reset failed".
4396 18. If a server dropped the connection unexpectedly when an Exim client was
4397 using GnuTLS and trying to read a response, the client delivery process
4398 crashed while trying to generate an error log message.
4400 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4401 string, for example:
4403 warn message = H1: something\nH2: something
4405 the text was added as a single header line from Exim's point of view
4406 though it ended up OK in the delivered message. However, searching for the
4407 second and subsequent header lines using $h_h2: did not work. This has been
4408 fixed. Similarly, if a system filter added multiple headers in this way,
4409 the routers could not see them.
4411 20. Expanded the error message when iplsearch is called with an invalid key to
4412 suggest using net-iplsearch in a host list.
4414 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4415 ACLs are no longer actually imposed (and a message to that effect is
4418 22. If a "gecos" field in a passwd entry contained escaped characters, in
4419 particular, if it contained a \" sequence, Exim got it wrong when building
4420 a From: or a Sender: header from that name. A second bug also caused
4421 incorrect handling when an unquoted " was present following a character
4422 that needed quoting.
4424 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4425 was not being matched caselessly.
4427 24. Arranged for all hyphens in the exim.8 source to be escaped with
4430 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4431 properly. Recipient callout cache records were still being keyed to include
4432 the sender, even when use_sender was set false. This led to far more
4433 callouts that were necessary. The sender is no longer included in the key
4434 when use_sender is false.
4436 26. Added "control = submission" modifier to ACLs.
4438 27. Added the ${base62d: operator to decode base 62 numbers.
4440 28. dnsdb lookups can now access SRV records.
4442 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4443 the configuration file.
4445 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4446 (-v) mode. This makes the output for a verbose queue run more intelligible.
4448 31. Added a use_postmaster feature to recipient callouts.
4450 32. Added the $body_zerocount variable, containing the number of binary zero
4451 bytes in the message body.
4453 33. The time of last modification of the "new" subdirectory is now used as the
4454 "mailbox time last read" when there is a quota error for a maildir
4457 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4459 35. Added +ignore_unknown as a special item in host lists.
4461 36. Code for decoding IPv6 addresses in host lists is now included, even if
4462 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4463 address was recognized as an IP address, but was then not correctly decoded
4464 into binary, causing unexpected and incorrect effects when compared with
4471 1. Very minor rewording of debugging text in manualroute to say "list of
4472 hosts" instead of "hostlist".
4474 2. If verify=header_syntax was set, and a header line with an unqualified
4475 address (no domain) and a large number of spaces between the end of the
4476 name and the colon was received, the reception process suffered a buffer
4477 overflow, and (when I tested it) crashed. This was caused by some obsolete
4478 code that should have been removed. The fix is to remove it!
4480 3. When running in the test harness, delay a bit after writing a bounce
4481 message to get a bit more predictability in the log output.
4483 4. Added a call to search_tidyup() just before forking a reception process. In
4484 theory, someone could use a lookup in the expansion of smtp_accept_max_
4485 per_host which, without the tidyup, could leave open a database connection.
4487 5. Added the variables $recipient_data and $sender_data which get set from a
4488 lookup success in an ACL "recipients" or "senders" condition, or a router
4489 "senders" option, similar to $domain_data and $local_part_data.
4491 6. Moved the writing of debug_print from before to after the "senders" test
4494 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4495 problems for message scanning, either using a data ACL, or using
4496 local_scan() because the Received: header was not generated till after they
4497 were called (in order to set the time as the time of reception completion).
4498 I have revised the way this works. The header is now generated after the
4499 body is received, but before the ACL or local_scan() are called. After they
4500 are run, the timestamp in the header is updated.
4506 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4507 before starting a queue runner without re-exec. This happened only when
4508 deliver_drop_privilege was set or when the Exim user was set to root. The
4509 effect of the bug was that timeouts during subsequent deliveries caused
4510 crashes instead of being properly handled. The handler is now left at its
4511 default (and expected) setting.
4513 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4514 message, again when deliver_drop_privilege is set or Exim is run as root.
4515 The bug described in (1) was not present in this case, but the tidying up
4516 of the other signals was missing. I have made the two cases consistent.
4518 3. The ignore_target_hosts setting on a manualroute router was being ignored
4519 for hosts that were looked up using the /MX notation.
4521 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4524 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4525 operated on the sender address. After changing the $sender_address to <>
4526 for the sender address verify, Exim was re-instated it as the original
4527 (before rewriting) address, but remembering that it had rewritten it, so it
4528 wasn't rewriting it again. This bug also had the effect of breaking the
4529 sender address verification caching when the sender address was rewritten.
4531 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4532 This has been changed so that if the ip literal address matches
4533 ignore_target_hosts, the router declines.
4535 7. Added expansion conditions match_domain, match_address, and match_local_
4536 part (NOT match_host).
4538 8. The placeholder for the Received: header didn't have a length field set.
4540 9. Added code to Exim itself and to exim_lock to test for a specific race
4541 condition that could lead to file corruption when using MBX delivery. The
4542 issue is with the lockfile that is created in /tmp. If this file is removed
4543 after a process has opened it but before that process has acquired a lock,
4544 there is the potential for a second process to recreate the file and also
4545 acquire a lock. This could lead to two Exim processes writing to the file
4546 at the same time. The added code performs the same test as UW imapd; it
4547 checks after acquiring the lock that its file descriptor still refers to
4548 the same named file.
4550 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4551 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4554 11. Added the smtp_active_hostname option. If used, this will typically be made
4555 to depend on the incoming interface address. Because $interface_address is
4556 not set up until the daemon has forked a reception process, error responses
4557 that can happen earlier (such as "too many connections") no longer contain
4560 12. If an expansion in a condition on a "warn" statement fails because a lookup
4561 defers, the "warn" statement is abandoned, and the next ACL statement is
4562 processed. Previously this caused the whole ACL to be aborted.
4564 13. Added the iplsearch lookup type.
4566 14. Added ident_timeout as a log selector.
4568 15. Added tls_certificate_verified as a log selector.
4570 16. Added a global option tls_require_ciphers (compare the smtp transport
4571 option of the same name). This controls incoming TLS connections.
4573 17. I finally figured out how to make tls_require_ciphers do a similar thing
4574 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4575 before starting the TLS session.
4577 18. Tabs are now shown as \t in -bP output.
4579 19. If the log selector return_path_on_delivery was set, Exim crashed when
4580 bouncing a message because it had too many Received: header lines.
4582 20. If two routers both had headers_remove settings, and the first one included
4583 a superfluous trailing colon, the final name in the first list and the
4584 first name in the second list were incorrectly joined into one item (with a
4585 colon in the middle).
4591 1. Added -C and -D options to the exinext utility, mainly to make it easier
4592 to include in the automated testing, but these could be helpful when
4593 multiple configurations are in use.
4595 2. The exinext utility was not formatting the output nicely when there was
4596 an alternate port involved in the retry record key, nor when there was a
4597 message id as well (for retries that were specific to a specific message
4598 and a specific host). It was also confused by IPv6 addresses, because of
4599 the additional colons they contain. I have fixed the IPv4 problem, and
4600 patched it up to do a reasonable job for IPv6.
4602 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4603 delivery, the log line now contains "pipelined" if PIPELINING was used.
4605 4. An SMTP transport process used to panic and die if the bind() call to set
4606 an explicit outgoing interface failed. This has been changed; it is now
4607 treated in the same way as a connect() failure.
4609 5. A reference to $sender_host_name in the part of a conditional expansion
4610 that was being skipped was still causing a DNS lookup. This no longer
4613 6. The def: expansion condition was not recognizing references to header lines
4614 that used bh_ and bheader_.
4616 7. Added the _cache feature to named lists.
4618 8. The code for checking quota_filecount in the appendfile transport was
4619 allowing one more file than it should have been.
4621 9. For compatibility with Sendmail, the command line option
4629 and sets the incoming protocol and host name (for trusted callers). The
4630 host name and its colon can be omitted when only the protocol is to be set.
4631 Note the Exim already has two private options, -pd and -ps, that refer to
4632 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4633 "s", but I don't think that's a major issue.
4635 10. A number of refactoring changes to the code, none of which should affect
4638 (a) The number of logging options was getting close to filling up the
4639 32-bit word that was used as a bit map. I have split them into two classes:
4640 those that are passed in the argument to log_write(), and those that are
4641 only ever tested independently outside of that function. These are now in
4642 separate 32-bit words, so there is plenty of room for expansion again.
4643 There is no change in the user interface or the logging behaviour.
4645 (b) When building, for example, log lines, the code previously used a
4646 macro that called string_cat() twice, in order to add two strings. This is
4647 not really sufficiently general. Furthermore, there was one instance where
4648 it was actually wrong because one of the argument was used twice, and in
4649 one call a function was used. (As it happened, calling the function twice
4650 did not affect the overall behaviour.) The macro has been replaced by a
4651 function that can join an arbitrary number of extra strings onto a growing
4654 (c) The code for expansion conditions now uses a table and a binary chop
4655 instead of a serial search (which was left over from when there were very
4656 few conditions). Also, it now recognizes conditions like "pam" even when
4657 the relevant support is not compiled in: a suitably worded error message is
4658 given if an attempt is made to use such a condition.
4660 11. Added ${time_interval:xxxxx}.
4662 12. A bug was causing one of the ddress fields not to be passed back correctly
4663 from remote delivery subprocesses. The field in question was not being
4664 subsequently used, so this caused to problems in practice.
4666 13. Added new log selectors queue_time and deliver_time.
4668 14. Might have fixed a bug in maildirsizefile handling that threw up
4669 "unexpected character" debug warnings, and recalculated the data
4670 unnecessarily. In any case, I expanded the warning message to give more
4673 15. Added the message "Restricted characters in address" to the statements in
4674 the default ACL that block characters like @ and % in local parts.
4676 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4677 Three changes have been made:
4679 (a) There was a serious bug; a negative response to MAIL caused the whole
4680 recipient domain to be cached as invalid, thereby blocking all messages
4681 to all local parts at the same domain, from all senders. This bug has
4682 been fixed. The domain is no longer cached after a negative response to
4683 MAIL if the sender used is not empty.
4685 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4688 (c) A new callout option, "use_sender" has been added for people who want
4689 the modified behaviour.
4695 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4698 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4699 indeed breaks things for older releases.
4701 3. Added additional logging to the case where there is a problem reading data
4702 from a filter that is running in a subprocess using a pipe, in order to
4703 try to track down a specific problem.
4705 4. Testing facility fudge: when running in the test harness and attempting
4706 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4707 getting "No route to host". Convert this to a timeout.
4709 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4712 6. Some OS don't have socklen_t but use size_t instead. This affects the
4713 fifth argument of getsockopt() amongst other things. This is now
4714 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4715 can be set for individual OS. I have set it for SunOS5, OSF1, and
4716 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4717 some earlier ones do not.
4719 7. Change 4.30/15 was not doing the test caselessly.
4721 8. The standard form for an IPv6 address literal was being rejected by address
4722 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4723 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4724 this, as well as the form without the "IPv6" on the front (but only when
4725 address literals are enabled, of course).
4727 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4729 10. Exim crashed if a message with an empty sender address specified by -f
4730 encountered a router with an errors_to setting. This could be provoked only
4731 by a command such as
4735 where an empty string was supplied; "<>" did not hit this bug.
4737 11. Installed PCRE release 4.5.
4739 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4740 remained set. It is now erased.
4742 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4743 times from message ids (which are base 36 rather than the normal 62).
4745 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4746 were being counted as actual protocol errors, and logged if the log
4747 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4748 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4749 and DATA following a set of rejected RCPTs do not count as protocol errors.
4750 In other words, Exim assumes they were pipelined, though this may not
4751 actually be the case. Of course, in all cases the client gets an
4752 appropriate error code.
4754 15. If a lookup fails in an ACL condition, a message about the failure may
4755 be available; it is used if testing the ACL cannot continue, because most
4756 such messages specify what the cause of the deferral is. However, some
4757 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4758 that caused an old message to be retained and used if a later statement
4759 caused a defer, replacing the real cause of the deferral.
4761 16. If an IP address had so many PTR records that the DNS lookup buffer
4762 was not large enough to hold them, Exim could crash while trying to process
4763 the truncated data. It now detects and logs this case.
4765 17. Further to 4.21/58, another change has been made: if (and only if) the
4766 first line of a message (the first header line) ends with CRLF, a bare LF
4767 in a subsequent header line has a space inserted after it, so as not to
4768 terminate the header.
4770 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4771 unnecessarily, used atoi() instead of strtol(), and didn't check the
4772 termination when getting file sizes from file names by regex.
4774 19. Completely re-implemented the support for maildirsize files, in the light
4775 of a number of problems with the previous contributed implementation
4776 (4.30/29). In particular:
4778 . If the quota is zero, the maildirsize file is maintained, but no quota is
4781 . If the maildir directory does not exist, it is created before any attempt
4782 to write a maildirsize file.
4784 . The quota value in the file is just a cache; if the quota is changed in
4785 the transport, the new value overrides.
4787 . A regular expression is available for excluding directories from the
4790 20. The autoreply transport checks the characters in options that define the
4791 message's headers; it allows continued headers, but it was checking with
4792 isspace() after an embedded newline instead of explicitly looking for a
4795 21. If all the "regular" hosts to which an address was routed had passed their
4796 expiry times, and had not reached their retry times, the address was
4797 bounced, even if fallback hosts were defined. Now Exim should go on to try
4800 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4801 equivalent code in the SMTP transport. Some hosts send humungous responses
4802 to HELO/EHLO, more than 1024 it seems.
4804 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4805 gives compiler warnings in some environments. I've now done it "properly",
4808 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4809 (because of problems with the built-in one) was declared to return uschar *
4810 instead of char *, causing compiler failure.
4812 25. Fixed a file descriptor leak when processing alias/forward files.
4814 26. Fixed a minor format string issue in dbfn.c.
4816 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4818 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4819 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4822 29. When a sender address is verified, it is cached, to save repeating the test
4823 when there is more than one recipient in a message. However, when the
4824 verification involves a callout, it is possible for different callout
4825 options to be set for different recipients. It is too complicated to keep
4826 track of this in the cache, so now Exim always runs a verification when a
4827 callout is required, relying on the callout cache for the optimization.
4828 The overhead is duplication of the address routing, but this should not be
4831 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4832 to be verified with callout=postmaster, and the main callout worked but the
4833 postmaster check failed, the verification correctly failed. However, if a
4834 subsequent RCPT command asked for sender verification *without* the
4835 postmaster check, incorrect caching caused this verification also to fail,
4838 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4839 it was not caching the DNS options (qualify_single, search_parents) that
4840 were used when the lookup failed. A subsequent lookup with different
4841 options therefore always gave the same answer, though there were cases
4842 where it should not have. (Example: a "domains = !$mx_any" option on a
4843 dnslookup router: the "domains" option is always processed without any
4844 widening, but the router might have qualify_single set.) Now Exim uses the
4845 cached value only when the same options are set.
4847 32. Added John Jetmore's "exipick" utility to the distribution.
4849 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4850 than a timeout (e.g. a certificate is required, and is not provided), an
4851 Exim server now closes the connection immediately. Previously it waited for
4852 the client to close - but if the client is SSL, it seems that they each
4853 wait for each other, leading to a delay before one of them times out.
4855 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4856 maintained 0.8.x compatibility because I don't think many are using it, and
4857 it is clearly obsolete.
4859 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4862 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4863 client certificate was expired. A simple patch fixes this, though I don't
4864 understand the full logic of why the verify callback is called multiple
4867 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4868 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4869 which causes problems with some clients (such as the Certicom SSL Plus
4870 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4871 disables the coutermeasure allowing Eudora to connect."
4873 38. Exim was not checking that a write() to a log file succeeded. This could
4874 lead to Bad Things if a log got too big, in particular if it hit a file
4875 size limit. Exim now panics and dies if it cannot write to a log file, just
4876 as it does if it cannot open a log file.
4878 39. Modified OS/Makefile-Linux so that it now contains
4880 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4882 The two -D definitions ensure that Exim is compiled with large file
4883 support, which makes it possible to handle log files that are bigger than
4886 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4887 instance) a domain was checked against a named list that involved a lookup,
4888 causing $domain_data to be set, then another domain was checked against the
4889 same list, then the first domain was re-checked, the value of $domain_data
4890 after the final check could be wrong. In particular, if the second check
4891 failed, it could be set empty. This bug probably also applied to
4894 41. The strip_trailing_dot option was not being applied to the address given
4895 with the -f command-line option.
4897 42. The code for reading a message's header from the spool was incrementing
4898 $received_count, but never initializing it. This meant that the value was
4899 incorrect (doubled) while delivering a message in the same process in which
4900 it was received. In the most common configuration of Exim, this never
4901 happens - a fresh exec is done - but it can happen when
4902 deliver_drop_privilege is set.
4904 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4905 it now includes up to 150 characters of the unexpected data in the log
4908 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4909 and building data strings. The size of both of these buffers was 10 000
4910 bytes - far larger than anybody would *ever* want, thought I. Needless to
4911 say, somebody hit the limit. I have increased the maximum line length to
4912 20 000 and the maximum data length of concatenated lines to 100 000. I have
4913 also fixed two bugs, because there was no checking on these buffers. Tsk,
4914 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4915 buffer is too small.
4917 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4918 lsearch lookups. Now it does.
4920 46. When parsing a route_list item in a manualroute router, a fixed-length
4921 buffer was used for the list of hosts. I made this 1024 bytes long,
4922 thinking that nobody would ever have a list of hosts that long. Wrong.
4923 Somebody had a whole pile of complicated expansion conditions, and the
4924 string was silently truncated, leading to an expansion error. It turns out
4925 that it is easier to change to an unlimited length (owing to other changes
4926 that have happened since this code was originally written) than to build
4927 structure for giving a limitation error. The length of the item that
4928 expands into the list of hosts is now unlimited.
4930 47. The lsearch lookup could not handle data where the length of text line was
4931 more than 4095 characters. Such lines were truncated, leading to shortened
4932 data being returned. It should now handle lines of any length.
4934 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4935 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4938 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4939 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4940 no longer happens in comment lines. A list of replacements is now placed
4941 at the head of all of the source files, except those whose only change is
4942 to replace PERL_COMMAND in the very first #! line.
4944 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4945 messages on the queue, with a bottom-up merge sort, using code contributed
4946 by Michael Haardt. This should make operations like -bp somewhat faster on
4947 large queues. It won't affect queue runners, except when queue_run_in_order
4950 51. Installed eximstats 1.31 in the distribution.
4952 52. Added support for SRV lookups to the dnslookup router.
4954 53. If an ACL referred to $message_body or $message_body_end, the value was not
4955 reset for any messages that followed in the same SMTP session.
4957 54. The store-handling optimization for building very long strings was not
4958 differentiating between the different store pools. I don't think this
4959 actually made any difference in practice, but I've tidied it.
4961 55. While running the routers to verify a sender address, $sender_address
4962 was still set to the sender address. This is wrong, because when routing to
4963 send a bounce to the sender, it would be empty. Therefore, I have changed
4964 it so that, while verifying a sender address, $sender_address is set to <>.
4965 (There is no change to what happens when verifying a recipient address.)
4967 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4968 target A or AAAA records (if not already returned) without resetting the
4969 qualify_single or search_parents options of the DNS resolver. These are
4970 inappropriate in this case because the targets of MX and SRV records must
4971 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4972 target that, when qualified, matched something in the local domain. These
4973 two options are now turned off when doing these lookups.
4975 57. It seems that at least some releases of Reiserfs (which does not have the
4976 concept of a fixed number of inodes) returns zero and not -1 for the
4977 number of available inodes. This interacted badly with check_spool_inodes,
4978 which assumed that -1 was the "no such thing" setting. What I have done is
4979 to check that the total number of inodes is greater than zero before doing
4980 the test of how many are available.
4982 58. When a "warn" ACL statement has a log_message modifier, the message is
4983 remembered, and not repeated. This is to avoid a lot of repetition when a
4984 message has many recipients that cause the same warning to be written.
4985 Howewer, Exim was preserving the list of already written lines for an
4986 entire SMTP session, which doesn't seem right. The memory is now reset if a
4987 new message is started.
4989 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4990 debugging output unless log_rewrite was also set.
4992 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4993 of (int)(handle) when we know that handle contains (void *)(-1).
4995 61. The Exim daemon panic-logs an error return when it closes the incoming
4996 connection. However "connection reset by peer" seems to be common, and
4997 isn't really an error worthy of noting specially, so that particular error
5000 62. When Exim is trying to find all the local interfaces, it used to panic and
5001 die if the ioctl to get the interface flags failed. However, it seems that
5002 on at least one OS (Solaris 9) it is possible to have an interface that is
5003 included in the list of interfaces, but for which you get a failure error
5004 for this call. This happens when the interface is not "plumbed" into a
5005 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5006 failure of the "get flags" call assumes that the interface is down.
5008 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5009 makes life easier for people who are doing arithmetic on fields extracted
5010 from dates, where you often get leading zeros that should not be
5011 interpreted as octal.
5013 64. Added qualify_domain to the redirect router, to override the global
5016 65. If a pathologically long header line contained very many addresses (the
5017 report of this problem mentioned 10 000) and each of them was rewritten,
5018 Exim could use up a very large amount of memory. (It kept on making new
5019 copies of the header line as it rewrote, and never released the old ones.)
5020 At the expense of a bit more processing, the header rewriting function has
5021 been changed so that it no longer eats memory in this way.
5023 66. The generation of the Received: header has been moved from the time that a
5024 message starts to be received, to the time that it finishes. The timestamp
5025 in the Received: header should now be very close to that of the <= log
5026 line. There are two side-effects of this change:
5028 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5029 logged header lines no longer include the local Received: line, because
5030 it has not yet been created. The same applies to a copy of the message
5031 that is returned to a non-SMTP sender when a message is rejected.
5033 (b) When a filter file is tested using -bf, no additional Received: header
5034 is added to the test message. After some thought, I decided that this
5037 This change does not affect the value of $received_for. It is still set
5038 after address rewriting, but before local_scan() is called.
5040 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5042 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5043 gave an unhelpful panic error message, and a defer error. I have managed to
5044 change this behaviour so that it now rejects any supplied certificate,
5045 which seems right, as the list of acceptable certificates is empty.
5047 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5048 gave an unhelpful defer error. I have not managed to make this reject any
5049 supplied certificates, but the error message it gives is "no certificate
5050 supplied", which is not helpful.
5052 70. exigrep's output now also includes lines that are not associated with any
5053 message, but which match the given pattern. Implemented by a patch from
5054 Martin Sluka, which also tidied up the Perl a bit.
5056 71. Recipient callout verification, like sender verification, was using <> in
5057 the MAIL FROM command. This isn't really the right thing, since the actual
5058 sender may affect whether the remote host accepts the recipient or not. I
5059 have changed it to use the actual sender in the callout; this means that
5060 the cache record is now keyed on a recipient/sender pair, not just the
5061 recipient address. There doesn't seem to be a real danger of callout loops,
5062 since a callout by the remote host to check the sender would use <>.
5063 [SEE ABOVE: changed after hitting problems.]
5065 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5066 temporary errors. However, in the case of such a code being given after
5067 the end of a data transmission (i.e. after ".") Exim was failing to write
5068 a retry record for the message. (Yes, there was some broken host that was
5069 actually sending 8xx at this point.)
5071 73. An unknown lookup type in a host list could cause Exim to panic-die when
5072 the list was checked. (An example that provoked this was putting <; in the
5073 middle of a list instead of at the start.) If this happened during a DATA
5074 ACL check, a -D file could be left lying around. This kind of configuration
5075 error no longer causes Exim to die; instead it causes a defer errror. The
5076 incident is still logged to the main and panic logs.
5078 74. Buglet left over from Exim 3 conversion. The message "too many messages
5079 in one connection" was written to the rejectlog but not the mainlog, except
5080 when address rewriting (yes!) was being logged.
5082 75. Added write_rejectlog option.
5084 76. When a system filter was run not as root (that is, when system_filter_user
5085 was set), the values of the $n variables were not being returned to the
5086 main process; thus, they were not subsequently available in the $sn
5089 77. Added +return_path_on_delivery log selector.
5091 78. A connection timeout was being treated differently from recipients deferred
5092 when testing hosts_max_try with a message that was older than the host's
5093 retry timeout. (The host should not be counted, thus allowing all hosts to
5094 be tried at least once before bouncing.) This may have been the cause of an
5095 occasionally reported bug whereby a message would remain on the queue
5096 longer than the retry timeout, but would be bounced if a delivery was
5097 forced. I say "may" because I never totally pinned down the problem;
5098 setting up timeout/retry tests is difficult. See also the next item.
5100 79. The ultimate address timeout was not being applied to errors that involved
5101 a combination of host plus message (for example, a timeout on a MAIL
5102 command). When an address resolved to a number of possible hosts, and they
5103 were not all tried for each delivery (e.g. because of hosts_max_try), a
5104 message could remain on the queue longer than the retry timeout.
5106 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5109 81. Fixed an obscure SMTP outgoing bug which required at least the following
5110 conditions: (a) there was another message waiting for the same server;
5111 (b) the server returned 5xx to all RCPT commands in the first message so
5112 that the message was not completed; (c) the server dropped the connection
5113 or gave a negative response to the RSET that Exim sends to abort the
5114 transaction. The observed case was a dropped connection after DATA that had
5115 been sent in pipelining mode. That is, the server had advertised PIPELINING
5116 but was not implementing it correctly. The effect of the bug was incorrect
5117 behaviour, such as trying another host, and this could lead to a crash.
5123 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5124 and daemon.c were passed as pointers to ints; they should have been
5125 pointers to socklen_t variables (which are typically unsigned ints).
5127 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5130 3. Fixed a really odd bug that affected only the testing scheme; patching a
5131 certain fixed string in the binary changed the value of another string that
5132 happened to be identical to the end of the original first string.
5134 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5135 name", it returns that address as the IP address. On some operating
5136 systems (e.g. Solaris), it also passes back the IP address string as the
5137 "host name". However, on others (e.g. Linux), it passes back an empty
5138 string. Exim wasn't checking for this, and was changing the host name to an
5139 empty string, assuming it had been canonicized.
5141 5. Although rare, it is permitted to have more than one PTR record for a given
5142 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5143 all the names associated with an address, because they do in Solaris.
5144 However, it seems that they do not in Linux for data that comes from the
5145 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5146 I found this out when I moved to a new Linux workstation and tried to run
5147 the Exim test suite.
5149 To get round this problem I have changed the code so that it now does its
5150 own call to the DNS to look up PTR records when searching for a host name.
5151 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5152 addresses that are only in /etc/hosts are still found.
5154 This behaviour is, however, controlled by an option called host_lookup_
5155 order, which defaults to "bydns:byaddr". If people want to use the other
5156 order, or indeed, just use one or the other means of lookup, they can
5157 specify it in this variable.
5159 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5160 some operating systems, this comes back from gethostbyaddr() as an empty
5161 string, and this is what Exim used to test for. However, it seems that in
5162 other systems, "." is yielded. Exim now tests for this case too.
5164 7. The values of check_spool_space and check_log_space are now held internally
5165 as a number of kilobytes instead of an absolute number of bytes. If a
5166 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5167 kilobyte. This means that much larger values can be stored.
5169 8. Exim monitor: an attempt to get the action menu when not actually pointing
5170 at a message produces an empty menu entitled "No message selected". This
5171 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5172 no entries in it ("Shell widget menu has zero width and/or height"). So I
5173 have added a single, blank menu entry in this case.
5175 9. Added ${quote_local_part.
5177 10. MIME decoding is now applied to the contents of Subject: header lines when
5180 11. Now that a reference to $sender_host_address automatically causes a reverse
5181 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5182 host lookup before query-style lookups in lists that might use this
5183 variable. This has therefore been abolished, and the "net-" prefix is no
5184 longer necessary for query-style lookups.
5186 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5187 have been a typo for LFLAGS, so it has been changed.
5189 13. The install script calls Exim with "-C /dev/null" in order to find the
5190 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5191 to be output. Howeve, since Exim outputs its version number before the
5192 error, it didn't break the script. It just looked ugly. I fixed this by
5193 always allowing "-C /dev/null" if the caller is root.
5195 14. Ignore overlarge ACL variable number when reading spool file - insurance
5196 against a later release with more variables having written the file.
5198 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5199 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5200 this, as well as the form without the "IPv6" on the front.
5202 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5203 OS/Makefile-Darwin file.
5205 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5206 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5208 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5209 variable. However, some OS use ints for inodes. I've added cast to long int
5210 to get rid of the compiler warning.
5212 19. I had forgotten to lock out "/../" in configuration file names when
5213 ALT_CONFIG_PREFIX was set.
5215 20. Routers used for verification do not need to specify transports. However,
5216 if such a router generated a host list, and callout was configured, Exim
5217 crashed, because it could not find a port number from the (non-existent)
5218 transport. It now assumes port 25 in this circumstance.
5220 21. Added the -t option to exigrep.
5222 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5223 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5224 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5225 them is set, LOOKUP_LSEARCH is forced.
5227 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5229 24. Added sender and host information to the "rejected by local_scan()" log
5230 line; previously there was no indication of these.
5232 25. Added .include_if_exists.
5234 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5235 Linux. It turns out that not all file systems support this. Apparently some
5236 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5237 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5238 not supported on the file descriptor, is now ignored when Exim is trying to
5239 sync a directory. This applies only to Linux.
5241 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5243 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5244 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5247 29. Added support for maildirsize files from supplied patch (modified a bit).
5249 30. The use of :fail: followed by an empty string could lead Exim to respond to
5250 sender verification failures with (e.g.):
5252 550 Verification failed for <xxx>
5253 550 Sender verify failed
5255 where the first response line was missing the '-' that indicates it is not
5256 the final line of the response.
5258 31. The loop for finding the name of the user that called Exim had a hardwired
5259 limit of 10; it now uses the value of finduser_retries, which is used for
5260 all other user lookups.
5262 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5265 33. Exim was neglecting to zero errno before one call of strtol() when
5266 expanding a string and expecting an integer value. On some systems this
5267 resulted in spurious "integer overflow" errors. Also, it was casting the
5268 result into an int without checking.
5270 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5271 did not work. The code looks as if it has *never* worked, though it appears
5272 to have been documented since at least releast 1.62. I have made it work.
5274 35. The "timeout_DNS" error in retry rules, also documented since at least
5275 1.62, also never worked. As it isn't clear exactly what this means, and
5276 clearly it isn't a major issue, I have abolished the feature by treating it
5277 as "timeout", and writing a warning to the main and panic logs.
5279 36. The display of retry rules for -brt wasn't always showing the error code
5282 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5283 timeout_connect_A, timeout_connect_MX.
5285 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5286 to the empty sender.
5288 39. The daemon was not analysing the content of -oX till after it had closed
5289 stderr and disconnected from the controlling terminal. This meant that any
5290 syntax errors were only noted on the panic log, and the return code from
5291 the command was 0. By re-arranging the code a little, I've made the
5292 decoding happen first, so such errors now appear on stderr, and the return
5293 code is 1. However, the actual setting up of the sockets still happens in
5294 the disconnected process, so errors there are still only recorded on the
5297 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5298 connections (as happens on some IP stacks) was logged at start up time as
5299 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5300 it from "IPv6 and IPv4", which means that two separate sockets are being
5303 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5304 says whether AF_INET or AF_INET6 was passed as an argument.
5306 42. Exiwhat output was messed up when time zones were included in log
5309 43. Exiwhat now gives more information about the daemon's listening ports,
5310 and whether -tls-on-connect was used.
5312 44. The "port" option of the smtp transport is now expanded.
5314 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5315 silently ignored. Now an error message is written to the main and panic
5318 46. There's a new ACL modifier called "logwrite" which writes to a log file
5319 as soon as it is encountered.
5321 47. Added $local_user_uid and $local_user_gid at routing time.
5323 48. Exim crashed when trying to verify a sender address that was being
5326 49. Exim was recognizing only a space character after ".include". It now also
5327 recognizes a tab character.
5329 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5330 extracting the relevant information from the specification. The man page no
5331 longer contains scrambled data for the -d option, and I've added a section
5332 at the front about calling Exim under different names.
5334 51. Added "extra_headers" argument to the "mail" command in filter files.
5336 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5339 53. Installed eximstats 1.29.
5341 54. Added transport_filter_timeout as a generic transport option.
5343 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5344 Cc: header lines. This was required by RFC 822, but it not required by RFC
5347 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5348 incoming messages that did not have them. Now it does so only if the
5349 message originates locally, that is, if there is no associated remote host
5350 address. When Resent- header lines are present, this applies to the Resent-
5351 lines rather than the non-Resent- lines.
5353 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5354 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5356 58. Messages for configuration errors now include the name of the main
5357 configuration file - useful now that there may be more than one file in a
5358 list (.included file names were always shown).
5360 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5361 for those rare installations that do not start the daemon as root or run it
5362 setuid root. I've cut out the call to initgroups() if the daemon is not
5365 60. The Exim user and group can now be bound into the binary as text strings
5366 that are looked up at the start of Exim's processing.
5368 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5370 62. Added $mailstore_basename variable.
5372 63. Installed patch to sieve.c from Michael Haardt.
5374 64. When Exim failed to open the panic log after failing to open the main log,
5375 the original message it was trying to log was written to stderr and debug
5376 output, but if they were not available (the usual case in production), it
5377 was lost. Now it is written to syslog before the two lines that record the
5378 failures to open the logs.
5380 65. Users' Exim filters run in subprocesses under the user's uid. It is
5381 possible for a "deliver" command or an alias in a "personal" command to
5382 provoke an address rewrite. If logging of address rewriting is configured,
5383 this fails because the process is not running as root or exim. There may be
5384 a better way of dealing with this, but for the moment (because 4.30 needs
5385 to be released), I have disabled address rewrite logging when running a
5386 filter in a non-root, non-exim process.
5392 1. The buildconfig auxiliary program wasn't quoting the value set for
5393 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5394 not defined. This bug was masked in 4.22 by the effect that was fixed in
5397 2. Some messages that were rejected after a message id was allocated were
5398 shown as "incomplete" by exigrep. It no longer does this for messages that
5399 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5401 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5402 have allow_domain_literals set, the ID did not get logged in the <= line.
5403 Domain literals are now always recognized in Message-ID: header lines.
5405 4. The first argument for a ${extract expansion item is the key name or field
5406 number. Leading and trailing spaces in this item were not being ignored,
5407 causing some misleading effects.
5409 5. When deliver_drop_privilege was set, single queue runner processes started
5410 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5411 same command in the process it spins off) were not dropping privilege.
5413 6. When the daemon running as "exim" started a queue runner, it always
5414 re-executed Exim in the spun-off process. This is a waste of effort when
5415 deliver_drop_privilege is set. The new process now just calls the
5416 queue-runner function directly.
5422 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5425 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5426 ignored. Though the use of -oP was forcing the writing of a pid file, it
5427 was always written to the default place.
5429 3. If the message "no IP address found for host xxxx" is generated during
5430 incoming verification, it is now followed by identification of the incoming
5431 connection (so you can more easily find what provoked it).
5433 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5435 5. Added some features to "harden" Exim a bit more against certain attacks:
5437 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5438 be put in Local/Makefile. This is like the never_users runtime option,
5439 but it cannot be overridden. The default setting is "root".
5441 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5442 prefix string with which any file named in a -C command line option
5445 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5446 is retained for -C and -D only if the caller of Exim is root. Without
5447 it, the exim user may also use -C and -D and retain privilege.
5449 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5450 command line option is disabled.
5452 6. Macro names set by the -D option must start with an upper case letter, just
5453 like macro names defined in the configuration file.
5455 7. Added "dereference=" facility to LDAP.
5457 8. Two instances of the typo "uknown" in the source files are fixed.
5459 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5460 the Configure-Makefile script screwed up while processing it.
5462 10. Incorporated PCRE 4.4.
5464 11. The SMTP synchronization check was not operating right at the start of an
5465 SMTP session. For example, it could not catch a HELO sent before the client
5466 waited for the greeting. There is now a check for outstanding input at the
5467 point when the greeting is written. Because of the duplex, asynchronous
5468 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5469 way, but not yet received, when the check is performed.
5471 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5472 on TCP/IP sockets, because this apparently causes some broken clients to
5475 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5476 unchanged) from the Cygwin maintainer.
5478 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5479 was supported instead of testing for mbx. Effectively a typo.
5481 15. The spa authenticator server code was not checking that the input it
5482 received was valid base64.
5484 16. The debug output line for the "set" modifier in ACLs was not showing the
5485 name of the variable that was being set.
5487 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5489 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5490 lookup on its own. Something else was needed to trigger the lookup. For
5491 example, a match in host_lookup or the need for a host name in a host list.
5492 Now, if $sender_host_name is referenced and the host name has not yet been
5493 looked up, a lookup is performed. If the lookup fails, the variable remains
5494 empty, and $host_lookup_failed is set to "1".
5496 19. Added "eqi" as a case-independent comparison operator.
5498 20. The saslauthd authentication condition could segfault if neither service
5499 nor realm was specified.
5501 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5502 error message that was logged was misleading, and incoming SMTP
5503 connections were dropped. The message is now more accurate, and temporary
5504 errors are given to SMTP connections.
5506 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5507 (except RSET) with a 421 error, until QUIT is received. However, it was
5508 failing to send a response to QUIT.
5510 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5511 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5512 ACL is now run after the helo_try_verify_hosts code.
5514 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5515 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5516 case-independent, so other case variants are also recognized). Apparently
5517 some systems use these upper case variants.
5519 25. If more than two messages were waiting for the same host, and a transport
5520 filter was specified for the transport, Exim sent two messages over the
5521 same TCP/IP connection, and then failed with "socket operation on non-
5522 socket" when it tried to send the third.
5524 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5526 27. The extern definition of crypt16() in expand.c was not being excluded when
5527 the OS had its own crypt16() function.
5529 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5530 as a preferred synonym for return_size_limit, both as an option and as an
5533 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5535 30. Changed the default configuration ACL to relax the local part checking rule
5536 for addresses that are not in any local domains. For these addresses,
5537 slashes and pipe symbols are allowed within local parts, but the sequence
5538 /../ is explicitly forbidden.
5540 31. SPA server authentication was not clearing the challenge buffer before
5543 32. log_message in a "warn" ACL statement was writing to the reject log as
5544 well as to the main log, which contradicts the documentation and doesn't
5545 seem right (because no rejection is happening). So I have stopped it.
5547 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5548 However, I am unable to do any testing of this.
5550 34. Fixed an infelicity in the appendfile transport. When checking directories
5551 for a mailbox, to see if any needed to be created, it was accidentally
5552 using path names with one or more superfluous leading slashes; tracing
5553 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5555 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5556 contents are added to the log line that is written for every discarded
5557 recipient. (Previously a log_message setting was ignored.)
5559 36. The ${quote: operator now quotes the string if it is empty.
5561 37. The install script runs exim in order to find its version number. If for
5562 some reason other than non-existence or emptiness, which it checks, it
5563 could not run './exim', it was installing it with an empty version number,
5564 i.e. as "exim-". This error state is now caught, and the installation is
5567 38. An argument was missing from the function that creates an error message
5568 when Exim fails to connect to the socket for saslauthd authentication.
5569 This could cause Exim to crash, or give a corrupted message.
5571 39. Added isip, isip4, and isip6 to ${if conditions.
5573 40. The ACL variables $acl_xx are now saved with the message, and can be
5574 accessed later in routers, transports, and filters.
5576 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5577 strings in the file are not string-expanded.
5579 42. If a MAIL command specified a SIZE value that was too large to fit into an
5580 int variable, the check against message_size_limit failed. Such values are
5581 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5582 day this will have to be increased, but I don't think I want to be around
5583 when emails are that large.
5590 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5591 iconv() is not standard in FreeBSD.
5593 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5594 IPv6 enabled. The observed symptom was a segmentation fault on return from
5595 the function os_common_find_running_interfaces() in src/os.c.
5597 3. In the check_special_case() function in daemon.c I had used "errno" as an
5598 argument name, which causes warnings on some systems. This was basically a
5599 typo, since it was named "eno" in the comments!
5601 4. The code that waits for the clock to tick (at a resolution of some fraction
5602 of a second) so as to ensure message-id uniqueness was always waiting for
5603 at least one whole tick, when it could have waited for less. [This is
5604 almost certainly not relevant at current processor speeds, where it is
5605 unlikely to ever wait at all. But we try to future-proof.]
5607 5. The function that sleeps for a time interval that includes fractions of a
5608 second contained a race. It did not block SIGALRM between setting the
5609 timer, and suspending (a couple of lines later). If the interval was short
5610 and the sigsuspend() was delayed until after it had expired, the suspension
5611 never ended. On busy systems this could lead to processes getting stuck for
5614 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5615 process, before it forks any delivery processes. The open lookup caching
5616 mechanism meant that the open file or database connection was passed into
5617 the delivery process. The problem was that delivery processes always tidy
5618 up cached lookup data. This could cause a problem for the next delivery
5619 process started by the queue runner, because the external queue runner
5620 process does not know about the closure. So the next delivery process
5621 still has data in the lookup cache. In the case of a file lookup, there was
5622 no problem because closing a file descriptor in a subprocess doesn't affect
5623 the parent. However, if the lookup was caching a connection to a database,
5624 the connection was closed, and the second delivery process was likely to
5625 see errors such as "PGSQL: query failed: server closed the connection
5626 unexpectedly". The problem has been fixed by closing all cached lookups
5627 in a queue runner before running a delivery process.
5629 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5630 seem to have the "const" qualifier which it has on other OS. I've
5633 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5634 *of the same type* (client or server) with the same public name that an
5635 error should be diagnosed.
5637 9. When Exim looked up a host name for an IP address, but failed to find the
5638 original IP address when looking up the host name (a safety check), it
5639 output the message "<ip address> does not match any IP for NULL", which was
5640 confusing, to say the least. The bug was that the host name should have
5641 appeared instead of "NULL".
5643 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5644 user that is built into the binary, and the -C or -D options is used, root
5645 privilege is dropped before the configuration file is read. In addition,
5646 logging is switched to stderr instead of the normal log files. If the
5647 configuration then re-defines the Exim user, the unprivileged environment
5648 is probably not what is expected, so Exim logs a panic warning message (but
5651 However, if deliver_drop_privilege is set, the unprivileged state may well
5652 be exactly what is intended, so the warning has been cut out in that case,
5653 and Exim is allowed to try to write to its normal log files.
5659 1. smtp_return_error_details was not giving details for temporary sender
5660 or receiver verification errors.
5662 2. Diagnose a configuration error if two authenticators have the same public
5665 3. Exim used not to create the message log file for a message until the first
5666 delivery attempt. This could be confusing when incoming messages were held
5667 for policy or load reasons. The message log file is now created at the time
5668 the message is received, and an initial "Received" line is written to it.
5670 4. The automatically generated man page for command line options had a minor
5671 bug that caused no ill effects; however, a more serious problem was that
5672 the procedure for building the man page automatically didn't always
5673 operate. Consequently, release 4.20 contains an out-of-date version. This
5674 shouldn't happen again.
5676 5. When building Exim with embedded Perl support, the script that builds the
5677 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5678 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5680 6. The freeze_tell option was not being used for messages that were frozen on
5681 arrival, either by an ACL or by local_scan().
5683 7. Added the smtp_incomplete_transaction log selector.
5685 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5686 was accepting AUTH without a new EHLO.
5688 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5689 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5692 10. Logging of TCP/IP connections (when configured) now happens in the main
5693 daemon process instead of the child process, so that the TCP/IP connection
5694 count is more accurate (but it can never be perfect).
5696 11. The use of "drop" in a nested ACL was not being handled correctly in the
5697 outer ACL. Now, if condition failure induced by the nested "drop" causes
5698 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5699 or "require"), the connection is dropped.
5701 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5702 that yield "discard" can now be used with an "accept" or a "discard" verb,
5703 but an error is generated for any others (because I can't see a useful way
5704 to define what should happen).
5706 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5707 are now processed in the same way as lines in the Exim configuration file.
5708 In particular, continuation lines are supported.
5710 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5712 15. Added -ti meaning -t -i.
5714 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5715 domains, and warn by logging if others are found.
5717 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5718 returned from GIFCONF when Exim is trying to find the list of interfaces on
5719 a host. The code in os.c has been modified to copy each ifreq to an aligned
5720 structure in all cases.
5722 Also, in some cases, the returned ifreq's were being copied to a 'struct
5723 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5724 means the last couple of bytes of an IPv6 address could be chopped if the
5725 ifreq contained only a normal sockaddr (14 bytes storage).
5727 18. Named domain lists were not supported in the hosts_treat_as_local option.
5728 An entry such as +xxxx was not recognized, and was treated as a literal
5731 19. Ensure that header lines added by a DATA ACL are included in the reject log
5732 if the ACL subsequently rejects the message.
5734 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5735 MD5 (which is deprecated).
5737 21. When testing a filter file using -bf, Exim was writing a message when it
5738 took the sender from a "From " line in the message, but it was not doing so
5739 when it took $return_path from a Return-Path: header line. It now does.
5741 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5742 with a valid header line field name (a series of printing characters
5743 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5745 23. Changed "disc" in the source to "disk" to conform to the documentation and
5746 the book and for uniformity.
5748 24. Ignore Sendmail's -Ooption=value command line item.
5750 25. When execve() failed while trying to run a command in a pipe transport,
5751 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5752 could be confused with a return value of 69 from the command itself. This
5753 has been changed to 127, the value the shell returns if it is asked to run
5754 a non-existent command. The wording for the related log line suggests a
5755 non-existent command as the problem.
5757 26. If received_header_text expands to an empty string, do not add a Received:
5758 header line to the message. (Well, it adds a token one on the spool, but
5759 marks it "old" so that it doesn't get used or transmitted.)
5761 27. Installed eximstats 1.28 (addition of -nt option).
5763 28. There was no check for failure on the call to getsockname() in the daemon
5764 code. This can fail if there is a shortage of resources on the system, with
5765 ENOMEM, for example. A temporary error is now given on failure.
5767 29. Contrary to the C standard, it seems that in some environments, the
5768 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5769 program. Exim now does this explicitly; it affects the formatting of
5770 timestamps using strftime().
5772 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5773 complaints. I've now initialized all the variables, to avoid this.
5775 32. Header lines added by a system filter were not being "seen" during
5776 transport-time rewrites.
5778 33. The info_callback() function passed to OpenSSL is set up with type void
5779 (*)(SSL *, int, int), as described somewhere. However, when calling the
5780 function (actually a macro) that sets it up, the type void(*)() is
5781 expected. I've put in a cast to prevent warnings from picky compilers.
5783 34. If a DNS black list lookup found a CNAME record, but there were no A
5784 records associated with the domain it pointed at, Exim crashed.
5786 35. If a DNS black list lookup returned more than one A record, Exim ignored
5787 all but the first. It now scans all returned addresses if a particular IP
5788 value is being sought. In this situation, the contents of the
5789 $dnslist_value variable are a list of all the addresses, separated by a
5792 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5793 to accept a host name and all its aliases if the forward lookup for any of
5794 them yielded the IP address of the incoming connection. Now it accepts only
5795 those names whose forward lookup yields the correct IP address. Any other
5796 names are discarded. This closes a loophole whereby a rogue DNS
5797 administrator could create reverse DNS records to break through a
5798 wildcarded host restriction in an ACL.
5800 37. If a user filter or a system filter that ran in a subprocess used any of
5801 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5802 the wrong values were passed to the pipe command ($thisaddress had the
5803 value of $0, $0 had the value of $1, etc). This bug was introduced by
5804 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5806 38. Improved the line breaking for long SMTP error messages from ACLs.
5807 Previously, if there was no break point between 40 and 75 characters, Exim
5808 left the rest of the message alone. Two changes have been made: (a) I've
5809 reduced the minimum length to 35 characters; (b) if it can't find a break
5810 point between 35 and 75 characters, it looks ahead and uses the first one
5811 that it finds. This may give the occasional overlong line, but at least the
5812 remaining text gets split now.
5814 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5815 file descriptors might be low, and that setting 1000 would always raise it.
5816 It turns out that in some environments, the limit is already over 1000 and
5817 that lowering it causes trouble. So now Exim takes care not to decrease it.
5819 40. When delivering a message, the value of $return_path is set to $sender_
5820 address at the start of routing (routers may change the value). By an
5821 oversight, this default was not being set up when an address was tested by
5822 -bt or -bv, which affected the outcome if any router or filter referred to
5825 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5826 log only when "message" or "log_message" are set. However, if one of the
5827 conditions was an address verification, or a call to a nested ACL, the
5828 messages generated by the underlying test were being passed through. This
5829 no longer happens. The underlying message is available in $acl_verify_
5830 message for both "message" and "log_message" expansions, so it can be
5831 passed through if needed.
5833 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5834 new expansion $bh_ to give the encoded byte string without charset
5835 translation. Translation happens only if iconv() is available; HAVE_ICONV
5836 indicates this at build time. HEADERS_CHARSET gives the charset to
5837 translate to; headers_charset can change it in the configuration, and
5838 "headers charset" can change it in an individual filter file.
5840 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5841 that creates RFC 2047 encoded "words" labels them as that charset instead
5842 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5843 expansion operator; (ii) when Exim creates a From: line for a local
5844 message; (iii) when a header line is rewritten to include a "phrase" part.
5846 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5847 buggy, causing it to skip the first lines of messages whose message ID
5848 ended in 'D'. This would not have bitten before Exim release 4.14, because
5849 message IDs were unlikely to end in 'D' before then. The effect was to have
5850 incorrect size information for certain domains.
5852 45. #include "config.h" was missing at the start of the crypt16.c module. This
5853 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5856 46. If there was a timeout during a "random" callout check, Exim treated it as
5857 a failure of the random address, and carried on sending RSET and the real
5858 address. If the delay was just some slowness somewhere, the response to the
5859 original RCPT would be taken as a response to RSET and so on, causing
5860 mayhem of various kinds.
5862 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5863 when I implemented it. It didn't allow for the fact that some option values
5864 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5865 the right test for positive values.
5867 48. Domain names in DNS records are case-independent. Exim always looks them up
5868 in lower case. Some resolvers return domain names in exactly the case they
5869 appear in the zone file, that is, they may contain uppercase letters. Not
5870 all resolvers do this - some return always lower case. Exim was treating a
5871 change of case by a resolver as a change of domain, similar to a widening
5872 of a domain abbreviation. This triggered its re-routing code and so it was
5873 trying to route what was effectively the same domain again. This normally
5874 caused routing to fail (because the router wouldn't handle the domain
5875 twice). Now Exim checks for this case specially, and just changes the
5876 casing of the domain that it ultimately uses when it transmits the message
5879 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5882 50. If a filter generated a file delivery with a non-absolute name (possible if
5883 no home directory exists for the router), the forbid_file option was not
5886 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5887 the existing equality matching.
5889 52. Exim was using ints instead of ino_t variables in some places where it was
5890 dealing with inode numbers.
5892 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5893 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5894 called TMPDIR, and if it finds it is different, it changes its value.
5896 54. The smtp_printf() function is now made available to local_scan() so
5897 additional output lines can be written before returning. There is also an
5898 smtp_fflush() function to enable the detection of a dropped connection.
5899 The variables smtp_input and smtp_batched_input are exported to
5902 55. Changed the default runtime configuration: the message "Unknown user"
5903 has been removed from the ACL, and instead placed on the localuser router,
5904 using the cannot_route_message feature. This means that any verification
5905 failures that generate their own messages won't get overridden. Similarly,
5906 the "Unrouteable address" message that was in the ACL for unverifiable
5907 relay addresses has also been removed.
5909 56. Added hosts_avoid_esmtp to the smtp transport.
5911 57. The exicyclog script was not checking for the esoteric option
5912 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5913 will work only if exicyclog is run under the appropriate euid.
5915 58. Following a discussion on the list, the rules by which Exim recognises line
5916 endings on incoming messages have been changed. The -dropcr and drop_cr
5917 options are now no-ops, retained only for backwards compatibility. The
5918 following line terminators are recognized: LF CRLF CR. However, special
5919 processing applies to CR:
5921 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5922 nor a local message in the state where . is a terminator.
5924 (ii) If a bare CR is encountered in a header line, an extra space is added
5925 after the line terminator so as not to end the header. The reasoning
5926 behind this is that bare CRs in header lines are most likely either
5927 to be mistakes, or people trying to play silly games.
5929 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5930 was being incorrectly given as 18 bytes larger than it should have been.
5931 This is a VOB (very old bug).
5933 60. This may never have affected anything current, but just in case it has:
5934 When the local host is found other than at the start of a list of hosts,
5935 the local host, those with the same MX, and any that follow, are discarded.
5936 When the list in question was part of a longer list of hosts, the following
5937 hosts (not currently being processed) were also being discarded. This no
5938 longer happens. I'm not sure if this situation could ever has previously
5941 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5944 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5945 header. This is something that is recommended in a new Internet Draft, and
5946 is something that is documented as being done by Sendmail. There are two
5947 possible values. For messages generated by the autoreply transport, Exim
5950 Auto-Submitted: auto-replied
5952 whereas for all other generated messages (e.g. bounces) it adds
5954 Auto-Submitted: auto-generated
5956 63. The "personal" condition in filters now includes a test for the
5957 Auto-Submitted: header. If it contains the string "auto-" the message it
5958 not considered personal.
5960 64. Added rcpt_include_affixes as a generic transport option.
5962 65. Added queue_only_override (default true).
5964 66. Added the syslog_duplication option.
5966 67. If what should have been the first header line of a message consisted of
5967 a space followed by a colon, Exim was mis-interpreting it as a header line.
5968 It isn't of course - it is syntactically invalid and should therefore be
5969 treated as the start of the message body. The misbehaviour could have
5970 caused a number of strange effects, including loss of data in subsequent
5971 header lines, and spool format errors.
5973 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5974 client host had authenticated. This control can now be exercised by an ACL
5975 for more flexibility.
5977 69. By default, callouts do not happen when testing with -bh. There is now a
5978 variant, -bhc, which does actually run the callout code, including
5979 consulting and updating the callout cache.
5981 70. Added support for saslauthd authentication, courtesy of Alexander
5984 71. If statvfs() failed on the spool or log directories while checking their
5985 size for availability, Exim confusingly gave the error "space shortage".
5986 Furthermore, in debugging mode it crashed with a floating point exception.
5987 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5988 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5989 serious problem, Exim now writes to the main and panic logs when this
5990 happens, with details of the failure. It then refuses to accept the
5991 incoming message, giving the message "spool directory problem" or "log
5992 directory problem" with a 421 code for SMTP messages.
5994 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5995 0, 1, and 2 exist, because some OS complain for execs without them (see
5996 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5997 descriptors. However, the code omitted to check that the open succeeded,
5998 causing mysterious errors if for some reason the permissions on /dev/null
5999 got screwed. Now Exim writes a message to the main and panic logs, and
6000 bombs out if it can't open /dev/null.
6002 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6003 interact so that it is all more flexible. It is supposed to remain
6004 backwards compatible. Also added extra_local_interfaces.
6006 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6007 to bomb out with an assertion failure - to the client this appears as a
6008 connection drop. This problem occurs in the part of the code that was taken
6009 from the Samba project. Fortunately, the assertion is in a very simple
6010 function, so I have fixed this by reproducing the function inline in the
6011 one place where it is called, and arranging for authentication to fail
6012 instead of killing the process with assert().
6014 75. The SPA client code was not working when the server requested OEM rather
6015 than Unicode encoding.
6017 76. Added code to make require_files with a specific uid setting more usable in
6018 the case where statting the file as root fails - usually a non-root-mounted
6019 NFS file system. When this happens and the failure is EACCES, Exim now
6020 forks a subprocess and does the per-uid checking as the relevant uid.
6022 77. Added process_log_path.
6024 78. If log_file_path was not explicitly set, a setting of check_log_space or
6025 check_log_inodes was ignored.
6027 79. If a space check for the spool or log partitions fails, the incident is now
6028 logged. Of course, in the latter case the data may get lost...
6030 80. Added the %p formatting code to string_format() so that it can be used to
6031 print addresses in debug_print(). Adjusted all the address printing in the
6032 debugging in store.c to use %p rather than %d.
6034 81. There was a concern that a line of code in smtp_in.c could overflow a
6035 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6036 initially expressed, the concern was not well-founded, because trailing
6037 spaces are removed early. However, if the trailing spaces were followed by
6038 a NULL, they did not get removed, so the overflow was possible. Two fixes
6041 (a) I re-wrote the offending code in a cleaner fashion.
6042 (b) If an incoming SMTP command contains a NULL character, it is rejected
6045 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6046 runs initgroups(), so that if the Exim user is in any additional groups,
6047 they will be used during message reception.
6053 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob