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.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router and the smtp transport (applying to the forward
91 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
92 of ldap servers used for a specific lookup. Patch provided by Heiko
95 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 New variable $lookup_dnssec_authenticated for observability.
98 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
99 Patch submitted by Lars Timman.
101 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
103 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
104 Requires trusted mode and valid format message id, aborts otherwise.
105 Patch contributed by Heiko Schlichting.
111 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
113 PP/02 Make -n do something, by making it not do something.
114 When combined with -bP, the name of an option is not output.
116 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
119 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
120 $sender_host_name and config options to manage this, and basic check
123 PP/05 DSCP support for outbound connections and control modifier for inbound.
125 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
126 (Only plugin which currently uses this is kerberos4, which nobody should
127 be using, but we should make it available and other future plugins might
128 conceivably use it, even though it would break NAT; stuff *should* be
129 using channel bindings instead).
131 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
132 name; added for Sendmail compatibility; requires admin caller.
133 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
134 just ignore it); requires trusted caller.
135 Also parse but ignore: -Ac -Am -X<logfile>
138 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
140 TL/02 Add +smtp_confirmation as a default logging option.
142 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
143 Patch by Magnus Holmgren from 2007-02-20.
145 TL/04 Bugzilla 1281 - Spec typo.
146 Bugzilla 1283 - Spec typo.
147 Bugzilla 1290 - Spec grammar fixes.
149 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
151 TL/06 Add Experimental DMARC support using libopendmarc libraries.
153 TL/07 Fix an out of order global option causing a segfault. Reported to dev
154 mailing list by by Dmitry Isaikin.
156 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
158 JH/02 Support "G" suffix to numbers in ${if comparisons.
160 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
162 NM/01 Bugzilla 1197 - Spec typo
163 Bugzilla 1196 - Spec examples corrections
165 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
167 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
168 gnutls_enable_pkcs11, but renamed to more accurately indicate its
171 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
172 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
174 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
175 "acl {{name}{arg}...}", and optional args on acl condition
178 JH/05 Permit multiple router/transport headers_add/remove lines.
180 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
182 JH/07 Avoid using a waiting database for a single-message-only transport.
183 Performance patch from Paul Fisher. Bugzilla 1262.
185 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
188 JH/09 Add $headers_added variable, with content from use of ACL modifier
189 add_header (but not yet added to the message). Bugzilla 199.
191 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
192 Pulled from Bugzilla 817 by Wolfgang Breyha.
194 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
196 (nb: this is the same fix as in Exim 4.80.1)
198 JH/11 Add A= logging on delivery lines, and a client_set_id option on
201 JH/12 Add optional authenticated_sender logging to A= and a log_selector
204 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
206 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
207 advertise SMTP AUTH mechanism to us, instead of a generic
208 protocol violation error. Also, make Exim more robust to bad
209 data from the Dovecot auth socket.
211 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
213 When a queue runner is handling a message, Exim first routes the
214 recipient addresses, during which it prunes them based on the retry
215 hints database. After that it attempts to deliver the message to
216 any remaining recipients. It then updates the hints database using
219 So if a recipient address works intermittently, it can get repeatedly
220 deferred at routing time. The retry hints record remains fresh so the
221 address never reaches the final cutoff time.
223 This is a fairly common occurrence when a user is bumping up against
224 their storage quota. Exim had some logic in its local delivery code
225 to deal with this. However it did not apply to per-recipient defers
226 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
228 This change adds a proper retry rule check during routing so that the
229 final cutoff time is checked against the message's age. We only do
230 this check if there is an address retry record and there is not a
231 domain retry record; this implies that previous attempts to handle
232 the address had the retry_use_local_parts option turned on. We use
233 this as an approximation for the destination being like a local
234 delivery, as in LMTP.
236 I suspect this new check makes the old local delivery cutoff check
237 redundant, but I have not verified this so I left the code in place.
239 TF/02 Correct gecos expansion when From: is a prefix of the username.
241 Test 0254 submits a message to Exim with the header
245 When I ran the test suite under the user fanf2, Exim expanded
246 the header to contain my full name, whereas it should have added
247 a Resent-Sender: header. It erroneously treats any prefix of the
248 username as equal to the username.
250 This change corrects that bug.
252 GF/01 DCC debug and logging tidyup
253 Error conditions log to paniclog rather than rejectlog.
254 Debug lines prefixed by "DCC: " to remove any ambiguity.
256 TF/03 Avoid unnecessary rebuilds of lookup-related code.
258 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
259 Bug spotted by Jeremy Harris; was flawed since initial commit.
260 Would have resulted in OCSP responses post-SNI triggering an Exim
261 NULL dereference and crash.
263 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
265 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
266 Bug detection, analysis and fix by Samuel Thibault.
267 Bugzilla 1331, Debian bug #698092.
269 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
271 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
272 Server implementation by Todd Lyons, client by JH.
273 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
274 config variable "prdr_enable" controls whether the server
275 advertises the facility. If the client requests PRDR a new
276 acl_data_smtp_prdr ACL is called once for each recipient, after
277 the body content is received and before the acl_smtp_data ACL.
278 The client is controlled by bolth of: a hosts_try_prdr option
279 on the smtp transport, and the server advertisement.
280 Default client logging of deliveries and rejections involving
281 PRDR are flagged with the string "PRDR".
283 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
284 fclose(). Diagnosis by Todd Lyons.
286 PP/17 Update configure.default to handle IPv6 localhost better.
287 Patch by Alain Williams (plus minor tweaks).
290 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
291 This is now consistent with GnuTLS, and is now documented: the
292 previous undocumented portable approach to treating the option as
293 unset was to force an expansion failure. That still works, and
294 an empty string is now equivalent.
296 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
297 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
298 not performing validation itself.
300 PP/20 Added force_command boolean option to pipe transport.
301 Patch from Nick Koston, of cPanel Inc.
303 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
306 TF/04 Added udpsend ACL modifer and hexquote expansion operator
308 PP/21 Fix eximon continuous updating with timestamped log-files.
309 Broken in a format-string cleanup in 4.80, missed when I repaired the
310 other false fix of the same issue.
311 Report and fix from Heiko Schlichting.
314 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
315 Report from Prashanth Katuri.
317 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
318 It's SecureTransport, so affects any MacOS clients which use the
319 system-integrated TLS libraries, including email clients.
321 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
322 using a MIME ACL for non-SMTP local injection.
323 Report and assistance in diagnosis by Warren Baker.
325 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
327 JH/16 Fix comparisons for 64b. Bugzilla 1385.
329 TL/09 Add expansion variable $authenticated_fail_id to keep track of
330 last id that failed so it may be referenced in subsequent ACL's.
332 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
335 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
336 ldap library initialization, allowing self-signed CA's to be
337 used. Also properly sets require_cert option later in code by
338 using NULL (global ldap config) instead of ldap handle (per
339 session). Bug diagnosis and testing by alxgomz.
341 TL/12 Enhanced documentation in the ratelimit.pl script provided in
342 the src/util/ subdirectory.
344 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
345 renamed to Transport Post Delivery Action by Jeremy Harris, as
348 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
349 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
350 redis_servers = needs to be configured which will be used by the redis
351 lookup. Patch from Warren Baker, of The Packet Hub.
353 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
355 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
356 hostname or reverse DNS when processing a host list. Used suggestions
357 from multiple comments on this bug.
359 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
361 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
362 Missed a few lines, added it to make the runtest require no keyboard
365 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
366 contains upper case chars. Make router use caseful_local_part.
368 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
369 support when GnuTLS has been built with p11-kit.
375 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
377 This, or similar/improved, will also be change PP/11 of 4.82.
383 PP/01 Handle short writes when writing local log-files.
384 In practice, only affects FreeBSD (8 onwards).
385 Bugzilla 1053, with thanks to Dmitry Isaikin.
387 NM/01 Bugzilla 949 - Documentation tweak
389 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
392 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
394 PP/02 Implemented gsasl authenticator.
396 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
398 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
399 `pkg-config foo` for cflags/libs.
401 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
402 with rest of GSASL and with heimdal_gssapi.
404 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
405 `pkg-config foo` for cflags/libs for the TLS implementation.
407 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
408 properties get this fed in as external SSF. A number of robustness
409 and debugging improvements to the cyrus_sasl authenticator.
411 PP/08 cyrus_sasl server now expands the server_realm option.
413 PP/09 Bugzilla 1214 - Log authentication information in reject log.
414 Patch by Jeremy Harris.
416 PP/10 Added dbmjz lookup type.
418 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
420 PP/12 MAIL args handles TAB as well as SP, for better interop with
421 non-compliant senders.
422 Analysis and variant patch by Todd Lyons.
424 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
425 Bug report from Lars Müller <lars@samba.org> (via SUSE),
426 Patch from Dirk Mueller <dmueller@suse.com>
428 PP/13 tls_peerdn now print-escaped for spool files.
429 Observed some $tls_peerdn in wild which contained \n, which resulted
430 in spool file corruption.
432 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
433 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
434 or write after TLS renegotiation, which otherwise led to messages
437 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
438 as a tracking header (ie: a signed header comes before the signature).
439 Patch from Wolfgang Breyha.
441 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
442 comma-sep list; embedded commas doubled.
444 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
446 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
448 Report and patch from Dmitry Banschikov.
450 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
451 Removed SSL_clear() after SSL_new() which led to protocol negotiation
452 failures. We appear to now support TLS1.1+ with Exim.
454 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
455 lets Exim select keys and certificates based upon TLS SNI from client.
456 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
457 before an outbound SMTP session. New log_selector, +tls_sni.
459 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
460 NULL dereference. Report and patch from Alun Jones.
462 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
463 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
464 Not seeing resolver debug output on NetBSD, but suspect this is a
465 resolver implementation change.
467 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
468 Left warnings. Added "eximon gdb" invocation mode.
470 PP/21 Defaulting "accept_8bitmime" to true, not false.
472 PP/22 Added -bw for inetd wait mode support.
474 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
475 locate the relevant includes and libraries. Made this the default.
477 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
478 Bugzilla 1246, report and most of solution from Tomasz Kusy.
480 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
481 This may cause build issues on older platforms.
483 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
484 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
485 gnutls_require_mac & gnutls_require_protocols (no longer supported).
486 Added SNI support via GnuTLS too.
487 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
489 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
491 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
492 Applied second patch from Janne, implementing suggestion to default
493 multiple-strings-in-record handling to match SPF spec.
495 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
497 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
498 read-only, out of scope).
499 Patch from Wolfgang Breyha, report from Stuart Northfield.
501 PP/29 Fix three issues highlighted by clang analyser static analysis.
502 Only crash-plausible issue would require the Cambridge-specific
503 iplookup router and a misconfiguration.
504 Report from Marcin Mirosław.
506 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
508 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
509 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
510 As part of this, removing so much warning spew let me fix some minor
511 real issues in debug logging.
513 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
514 assignment on my part. Fixed.
516 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
517 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
518 Janne Snabb (who went above and beyond: thank you).
520 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
521 string otherwise requires a connection and a bunch more work and it's
522 relatively easy to get wrong. Should also expose TLS library linkage
525 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
526 64-bit ${eval} (JH/03).
528 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
529 GNU libc to support some of the 64-bit stuff, should not lead to
530 conflicts. Defined before os.h is pulled in, so if a given platform
531 needs to override this, it can.
533 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
534 protection layer was required, which is not implemented.
535 Bugzilla 1254, patch from Wolfgang Breyha.
537 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
538 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
539 tls_dhparam take prime identifiers. Also unbreak combination of
540 OpenSSL+DH_params+TLSSNI.
542 PP/39 Disable SSLv2 by default in OpenSSL support.
548 PP/01 Solaris build fix for Oracle's LDAP libraries.
549 Bugzilla 1109, patch from Stephen Usher.
551 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
553 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
556 TF/02 Fix a couple more cases where we did not log the error message
557 when unlink() failed. See also change 4.74-TF/03.
559 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
560 lock up or crash if it happened to be inside a call to libc when it
561 got a SIGUSR1 from exiwhat.
563 The SIGUSR1 handler appends the current process status to the process
564 log which is later printed by exiwhat. It used to use the general
565 purpose logging code to do this, but several functions it calls are
566 not safe for signals.
568 The new output code in the SIGUSR1 handler is specific to the process
569 log, and simple enough that it's easy to inspect for signal safety.
570 Removing some special cases also simplifies the general logging code.
571 Removing the spurious timestamps from the process log simplifies
574 TF/04 Improved ratelimit ACL condition.
576 The /noupdate option has been deprecated in favour of /readonly which
577 has clearer semantics. The /leaky, /strict, and /readonly update modes
578 are mutually exclusive. The update mode is no longer included in the
579 database key; it just determines when the database is updated. (This
580 means that when you upgrde Exim will forget old rate measurements.)
582 Exim now checks that the per_* options are used with an update mode that
583 makes sense for the current ACL. For example, when Exim is processing a
584 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
585 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
586 must specify per_mail/readonly. If you omit the update mode it defaults to
587 /leaky where that makes sense (as before) or /readonly where required.
589 The /noupdate option is now undocumented but still supported for
590 backwards compatibility. It is equivalent to /readonly except that in
591 ACLs where /readonly is required you may specify /leaky/noupdate or
592 /strict/noupdate which are treated the same as /readonly.
594 A useful new feature is the /count= option. This is a generalization
595 of the per_byte option, so that you can measure the throughput of other
596 aggregate values. For example, the per_byte option is now equivalent
597 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
599 The per_rcpt option has been generalized using the /count= mechanism
600 (though it's more complicated than the per_byte equivalence). When it is
601 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
602 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
603 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
604 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
605 Note that using per_rcpt with a non-readonly update mode in more than
606 one ACL will cause the recipients to be double-counted. (The per_mail
607 and per_byte options don't have this problem.)
609 The handling of very low rates has changed slightly. If the computed rate
610 is less than the event's count (usually one) then this event is the first
611 after a long gap. In this case the rate is set to the same as this event's
612 count, so that the first message of a spam run is counted properly.
614 The major new feature is a mechanism for counting the rate of unique
615 events. The new per_addr option counts the number of different
616 recipients that someone has sent messages to in the last time period. It
617 behaves like per_rcpt if all the recipient addresses are different, but
618 duplicate recipient addresses do not increase the measured rate. Like
619 the /count= option this is a general mechanism, so the per_addr option
620 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
621 example, measure the rate that a client uses different sender addresses
622 with the options per_mail/unique=$sender_address. There are further
623 details in the main documentation.
625 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
627 TF/06 Removed a few PCRE remnants.
629 TF/07 Automatically extract Exim's version number from tags in the git
630 repository when doing development or release builds.
632 PP/02 Raise smtp_cmd_buffer_size to 16kB.
633 Bugzilla 879. Patch from Paul Fisher.
635 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
636 Heavily based on revision 40f9a89a from Simon Arlott's tree.
639 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
641 PP/05 Variable $av_failed, true if the AV scanner deferred.
642 Bugzilla 1078. Patch from John Horne.
644 PP/06 Stop make process more reliably on build failure.
645 Bugzilla 1087. Patch from Heiko Schlittermann.
647 PP/07 Make maildir_use_size_file an _expandable_ boolean.
648 Bugzilla 1089. Patch from Heiko Schlittermann.
650 PP/08 Handle ${run} returning more data than OS pipe buffer size.
651 Bugzilla 1131. Patch from Holger Weiß.
653 PP/09 Handle IPv6 addresses with SPF.
654 Bugzilla 860. Patch from Wolfgang Breyha.
656 PP/10 GnuTLS: support TLS 1.2 & 1.1.
658 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
661 PP/11 match_* no longer expand right-hand-side by default.
662 New compile-time build option, EXPAND_LISTMATCH_RHS.
663 New expansion conditions, "inlist", "inlisti".
665 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
667 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
669 PP/14 fix log_write() format string regression from TF/03.
670 Bugzilla 1152. Patch from Dmitry Isaikin.
676 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
678 PP/02 Harmonised TLS library version reporting; only show if debugging.
679 Layout now matches that introduced for other libraries in 4.74 PP/03.
681 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
683 PP/04 New "dns_use_edns0" global option.
685 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
688 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
689 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
691 TK/01 Updated PolarSSL code to 0.14.2.
692 Bugzilla 1097. Patch from Andreas Metzler.
694 PP/07 Catch divide-by-zero in ${eval:...}.
697 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
700 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
701 format-string attack -- SECURITY: remote arbitrary code execution.
703 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
704 time unintentionally subject to list matching rules, letting the header
705 cause arbitrary Exim lookups (of items which can occur in lists, *not*
706 arbitrary string expansion). This allowed for information disclosure.
708 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
709 INT_MIN/-1 -- value coerced to INT_MAX.
715 NM/01 Workround for PCRE version dependency in version reporting
718 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
719 This fixes portability to compilers other than gcc, notably
720 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
722 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
723 makefiles for portability to HP-UX and POSIX correctness.
725 PP/01 Permit LOOKUP_foo enabling on the make command-line.
726 Also via indented variable definition in the Makefile.
727 (Debugging by Oliver Heesakkers).
729 PP/02 Restore caching of spamd results with expanded spamd_address.
730 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
732 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
733 Improves build reliability. Fix from: Frank Elsner
735 NM/02 Fix wide character breakage in the rfc2047 coding
736 Fixes bug 1064. Patch from Andrey N. Oktyabrski
738 NM/03 Allow underscore in dnslist lookups
739 Fixes bug 1026. Patch from Graeme Fowler
741 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
742 Code patches from Adam Ciarcinski of NetBSD.
744 NM/04 Fixed exiqgrep to cope with mailq missing size issue
747 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
748 is logged, to avoid truncation. Patch from John Horne.
750 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
751 Patch from Jakob Hirsch.
753 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
754 SQL string expansion failure details.
755 Patch from Andrey Oktyabrski.
757 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
758 Patch from Simon Arlott.
760 PP/09 New lookups functionality failed to compile on old gcc which rejects
761 extern declarations in function scope.
762 Patch from Oliver Fleischmann
764 PP/10 Use sig_atomic_t for flags set from signal handlers.
765 Check getgroups() return and improve debugging.
766 Fixed developed for diagnosis in bug 927 (which turned out to be
769 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
770 Patch from Mark Zealey.
772 PP/12 Bugzilla 1056: Improved spamd server selection.
773 Patch from Mark Zealey.
775 PP/13 Bugzilla 1086: Deal with maildir quota file races.
776 Based on patch from Heiko Schlittermann.
778 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
779 Patch from Uwe Doering, sign-off by Michael Haardt.
781 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
782 variable declaration deep within a block. Bug and patch from
785 PP/15 lookups-Makefile IRIX compatibilty coercion.
787 PP/16 Make DISABLE_DKIM build knob functional.
789 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
790 Patch by Simon Arlott
792 TF/03 Fix valgrind.h portability to C89 compilers that do not support
793 variable argument macros. Our copy now differs from upstream.
799 TF/01 Failure to get a lock on a hints database can have serious
800 consequences so log it to the panic log.
802 TF/02 Log LMTP confirmation messages in the same way as SMTP,
803 controlled using the smtp_confirmation log selector.
805 TF/03 Include the error message when we fail to unlink a spool file.
807 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
808 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
809 for maintaining out-of-tree patches for some time.
811 PP/01 Bugzilla 139: Documentation and portability issues.
812 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
813 Handle per-OS dynamic-module compilation flags.
815 PP/02 Let /dev/null have normal permissions.
816 The 4.73 fixes were a little too stringent and complained about the
817 permissions on /dev/null. Exempt it from some checks.
818 Reported by Andreas M. Kirchwitz.
820 PP/03 Report version information for many libraries, including
821 Exim version information for dynamically loaded libraries. Created
822 version.h, now support a version extension string for distributors
823 who patch heavily. Dynamic module ABI change.
825 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
826 privilege escalation vulnerability whereby the Exim run-time user
827 can cause root to append content of the attacker's choosing to
830 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
833 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
834 If dropping privileges for untrusted macros, we disabled normal logging
835 on the basis that it would fail; for the Exim run-time user, this is not
836 the case, and it resulted in successful deliveries going unlogged.
837 Fixed. Reported by Andreas Metzler.
843 PP/01 Date: & Message-Id: revert to normally being appended to a message,
844 only prepend for the Resent-* case. Fixes regression introduced in
845 Exim 4.70 by NM/22 for Bugzilla 607.
847 PP/02 Include check_rfc2047_length in configure.default because we're seeing
848 increasing numbers of administrators be bitten by this.
850 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
852 PP/03 Bugzilla 994: added openssl_options main configuration option.
854 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
856 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
858 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
860 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
861 without a peer certificate, leading to a segfault because of an
862 assumption that peers always have certificates. Be a little more
863 paranoid. Problem reported by Martin Tscholak.
865 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
866 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
867 NB: ClamAV planning to remove STREAM in "middle of 2010".
868 CL also introduces -bmalware, various -d+acl logging additions and
869 more caution in buffer sizes.
871 PP/09 Implemented reverse_ip expansion operator.
873 PP/10 Bugzilla 937: provide a "debug" ACL control.
875 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
877 PP/12 Bugzilla 973: Implement --version.
879 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
881 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
883 PP/15 Bugzilla 816: support multiple condition rules on Routers.
885 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
886 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
887 ignore trailing whitespace.
889 JJ/02 prevent non-panic DKIM error from being sent to paniclog
891 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
894 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
895 Notification from Dr Andrew Aitchison.
897 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
898 ExtendedDetectionInfo response format.
899 Notification from John Horne.
901 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
904 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
905 XSL and documented dependency on system catalogs, with examples of how
908 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
911 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
912 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
913 configuration file which is writeable by the Exim user or group.
915 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
916 of configuration files to cover files specified with the -C option if
917 they are going to be used with root privileges, not just the default
920 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
921 option (effectively making it always true).
923 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
924 files to be used while preserving root privileges.
926 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
927 that rogue child processes cannot use them.
929 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
930 run-time user, instead of root.
932 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
933 Exim run-time user without dropping privileges.
935 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
936 result string, instead of calling string_vformat() twice with the same
939 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
940 for other users. Others should always drop root privileges if they use
941 -C on the command line, even for a whitelisted configure file.
943 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
945 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
951 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
952 $data_path, and $header_path variables; fixed documentation bugs and
955 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
956 exipick to access non-standard spools, including the "frozen" queue
959 NM/01 Bugzilla 965: Support mysql stored procedures.
960 Patch from Alain Williams
962 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
964 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
965 Patch from Andreas Metzler
967 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
968 Patch from Kirill Miazine
970 NM/05 Bugzilla 671: Added umask to procmail example.
972 JJ/03 installed exipick 20100323.0, fixing doc bug
974 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
975 directory. Notification and patch from Dan Rosenberg.
977 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
979 TK/02 Improve log output when DKIM signing operation fails.
981 MH/01 Treat the transport option dkim_domain as a colon separated
982 list, not as a single string, and sign the message with each element,
983 omitting multiple occurences of the same signer.
985 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
986 Bugzilla 985, 986. Patch by Simon Arlott
988 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
989 Patch by Simon Arlott
991 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
992 MBX locking. Notification from Dan Rosenberg.
998 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1000 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1002 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1004 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1006 NM/04 Bugzilla 915: Flag broken perl installation during build.
1012 TK/01 Added patch by Johannes Berg that expands the main option
1013 "spamd_address" if it starts with a dollar sign.
1015 TK/02 Write list of recipients to X-Envelope-Sender header when building
1016 the mbox-format spool file for content scanning (suggested by Jakob
1019 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1020 (http://www.dcc-servers.net/) support via dccifd. Activated by
1021 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1023 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1024 by Mark Daniel Reidel <mr@df.eu>.
1026 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1027 When building exim an external PCRE library is now needed -
1028 PCRE is a system library on the majority of modern systems.
1029 See entry on PCRE_LIBS in EDITME file.
1031 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1032 conversation. Added nologin parameter to request.
1033 Patch contributed by Kirill Miazine.
1035 TF/01 Do not log submission mode rewrites if they do not change the address.
1037 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1039 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1040 log files in place. Contributed by Roberto Lima.
1042 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1044 TF/03 Bugzilla 615: When checking the local_parts router precondition
1045 after a local_part_suffix or local_part_prefix option, Exim now
1046 does not use the address's named list lookup cache, since this
1047 contains cached lookups for the whole local part.
1049 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1050 Robert Millan. Documentation is in experimental-spec.txt.
1052 TF/04 Bugzilla 668: Fix parallel build (make -j).
1054 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1056 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1057 Patch provided by Jan Srzednicki.
1059 TF/05 Leading white space used to be stripped from $spam_report which
1060 wrecked the formatting. Now it is preserved.
1062 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1063 that they are available at delivery time.
1065 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1067 TF/08 TLS error reporting now respects the incoming_interface and
1068 incoming_port log selectors.
1070 TF/09 Produce a more useful error message if an SMTP transport's hosts
1071 setting expands to an empty string.
1073 NM/06 Bugzilla 744: EXPN did not work under TLS.
1074 Patch provided by Phil Pennock.
1076 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1077 Patch provided by Richard Godbee.
1079 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1080 acl_smtp_notquit, added index entry.
1082 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1083 Patch provided by Eugene Bujak.
1085 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1086 accept(). Patch provided by Maxim Dounin.
1088 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1089 Patch provided by Phil Pennock.
1091 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1093 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1094 Patch provided by Brad "anomie" Jorsch.
1096 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1097 Patch provided by Dean Brooks.
1099 TK/05 Add native DKIM support (does not depend on external libraries).
1101 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1102 Patch provided by Graeme Fowler.
1104 NM/16 Bugzilla 851: Documentation example syntax fix.
1106 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1108 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1111 NM/19 Bugzilla 745: TLS version reporting.
1112 Patch provided by Phil Pennock.
1114 NM/20 Bugzilla 167: bool: condition support.
1115 Patch provided by Phil Pennock.
1117 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1118 clients. Patch provided by Phil Pennock.
1120 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1121 Patch provided by Brad "anomie" Jorsch.
1123 NM/23 Bugzilla 687: Fix misparses in eximstats.
1124 Patch provided by Heiko Schlittermann.
1126 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1127 Patch provided by Heiko Schlittermann.
1129 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1130 plus update to original patch.
1132 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1134 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1135 Patch provided by David Brownlee.
1137 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1139 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1141 NM/30 Bugzilla 888: TLS documentation bugfixes.
1143 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1145 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1146 Unlike the original bugzilla I have changed all shell scripts in src tree.
1148 NM/33 Bugzilla 898: Transport filter timeout fix.
1149 Patch by Todd Rinaldo.
1151 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1152 Patch by Serge Demonchaux.
1154 NM/35 Bugzilla 39: Base64 decode bug fixes.
1155 Patch by Jakob Hirsch.
1157 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1159 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1161 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1163 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1169 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1170 ALT-N's libdkim that I have put here:
1171 http://duncanthrax.net/exim-experimental/
1173 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1174 were called 'true' and it seems that C99 defines that as a reserved
1175 keyword to be used with 'bool' variable types. That means you could
1176 not include C99-style headers which use bools without triggering
1177 build errors in sieve.c.
1179 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1180 as mailq or other aliases. Changed the --help handling significantly
1181 to do whats expected. exim_usage() emits usage/help information.
1183 SC/01 Added the -bylocaldomain option to eximstats.
1185 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1187 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1189 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1195 PH/01 Another patch from the Sieve maintainer.
1197 PH/02 When an IPv6 address is converted to a string for single-key lookup
1198 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1199 dots are used instead of colons so that keys in lsearch files need not
1200 contain colons. This was done some time before quoting was made available
1201 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1202 (notated using the quote facility) so as to distinguish them from IPv4
1203 keys. This meant that lookups for IP addresses in host lists did not work
1204 for iplsearch lookups.
1206 This has been fixed by arranging for IPv6 addresses to be expressed with
1207 colons if the lookup type is iplsearch. This is not incompatible, because
1208 previously such lookups could never work.
1210 The situation is now rather anomolous, since one *can* have colons in
1211 ordinary lsearch keys. However, making the change in all cases is
1212 incompatible and would probably break a number of configurations.
1214 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1217 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1218 conversion specification without a maximum field width, thereby enabling
1219 a rogue spamd server to cause a buffer overflow. While nobody in their
1220 right mind would setup Exim to query an untrusted spamd server, an
1221 attacker that gains access to a server running spamd could potentially
1222 exploit this vulnerability to run arbitrary code as the Exim user.
1224 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1225 $primary_hostname instead of what libspf2 thinks the hosts name is.
1227 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1228 a directory entry by the name of the lookup key. Previously, if a
1229 symlink pointed to a non-existing file or a file in a directory that
1230 Exim lacked permissions to read, a lookup for a key matching that
1231 symlink would fail. Now it is enough that a matching directory entry
1232 exists, symlink or not. (Bugzilla 503.)
1234 PH/03 The body_linecount and body_zerocount variables are now exported in the
1237 PH/04 Added the $dnslist_matched variable.
1239 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1240 This means they are set thereafter only if the connection becomes
1243 PH/06 Added the client_condition to authenticators so that some can be skipped
1244 by clients under certain conditions.
1246 PH/07 The error message for a badly-placed control=no_multiline_responses left
1247 "_responses" off the end of the name.
1249 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1251 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1252 (without spaces) instead of just copying the configuration text.
1254 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1256 PH/11 Added $max_received_linelength.
1258 PH/12 Added +ignore_defer and +include_defer to host lists.
1260 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1261 way in which PCRE > 7.0 is built.
1263 PH/14 Implemented queue_only_load_latch.
1265 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1266 MAIL command. The effect was to mangle the value on 64-bit systems.
1268 PH/16 Another patch from the Sieve maintainer.
1270 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1272 PH/18 If a system quota error occurred while trying to create the file for
1273 a maildir delivery, the message "Mailbox is full" was not appended to the
1274 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1275 only to a quota excession during the actual writing of the file.
1277 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1278 characters?) which causes problems in log lines. The DN values are now
1279 passed through string_printing() before being added to log lines.
1281 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1282 and InterBase are left for another time.)
1284 PH/21 Added message_body_newlines option.
1286 PH/22 Guard against possible overflow in moan_check_errorcopy().
1288 PH/23 POSIX allows open() to be a macro; guard against that.
1290 PH/24 If the recipient of an error message contained an @ in the local part
1291 (suitably quoted, of course), incorrect values were put in $domain and
1292 $local_part during the evaluation of errors_copy.
1298 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1299 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1302 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1303 issue a MAIL command.
1305 PH/02 In an ACL statement such as
1307 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1309 if a client was not listed at all, or was listed with a value other than
1310 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1311 the condition was not true (as it should be), so access was not denied.
1312 The bug was that the ! inversion was incorrectly passed on to the second
1313 item. This has been fixed.
1315 PH/03 Added additional dnslists conditions == and =& which are different from
1316 = and & when the dns lookup returns more than one IP address.
1318 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1319 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1321 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1322 FSYNC, which compiles an option called disable_fsync that allows for
1323 bypassing fsync(). The documentation is heavily laced with warnings.
1325 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1327 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1328 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1329 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1330 including adding "make clean"; (3) Added -fPIC when compiling the test
1331 dynamically loaded module, to get rid of a warning.
1333 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1334 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1335 = 0s. The bug is otherwise harmless.
1337 PH/07 There was a bug in the dovecot authenticator such that the value of
1338 $auth1 could be overwritten, and so not correctly preserved, after a
1339 successful authentication. This usually meant that the value preserved by
1340 the server_setid option was incorrect.
1342 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1344 PH/09 Installed PCRE release 7.0.
1346 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1347 run for batched SMTP input. It is now run at the start of every message
1348 in the batch. While fixing this I discovered that the process information
1349 (output by running exiwhat) was not always getting set for -bs and -bS
1350 input. This is fixed, and it now also says "batched" for BSMTP.
1352 PH/11 Added control=no_pipelining.
1354 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1355 patch, slightly modified), and move the expansion of helo_data till after
1356 the connection is made in the smtp transport (so it can use these
1359 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1361 PH/14 Added log_selector = +pid.
1363 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1365 PH/16 Add ${if forany and ${if forall.
1367 PH/17 Added dsn_from option to vary the From: line in DSNs.
1369 PH/18 Flush SMTP output before performing a callout, unless control =
1370 no_callout_flush is set.
1372 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1373 was true (the default) a successful delivery failed to delete the retry
1374 item, thus causing premature timeout of the address. The bug is now
1377 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1379 PH/21 Long custom messages for fakedefer and fakereject are now split up
1380 into multiline reponses in the same way that messages for "deny" and
1381 other ACL rejections are.
1383 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1384 with slight modification.
1386 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1387 draft, changing the syntax and factoring some duplicate code".
1389 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1390 for deliveries of the second and subsequent messages over the same SMTP
1393 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1394 ${reduce, with only minor "tidies".
1396 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1398 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1399 expansion side effects.
1401 PH/27 When a message times out after an over-quota error from an Exim-imposed
1402 quota, the bounce message says "mailbox is full". This message was not
1403 being given when it was a system quota that was exceeded. It now should
1406 MH/03 Made $recipients available in local_scan(). local_scan() already has
1407 better access to the recipient list through recipients_list[], but
1408 $recipients can be useful in postmaster-provided expansion strings.
1410 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1411 in the case of a MAIL command with additional options following the
1412 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1413 were accidentally chopped off.
1415 PH/29 SMTP synchronization checks are implemented when a command is read -
1416 there is a check that no more input is waiting when there shouldn't be
1417 any. However, for some commands, a delay in an ACL can mean that it is
1418 some time before the response is written. In this time, more input might
1419 arrive, invalidly. So now there are extra checks after an ACL has run for
1420 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1421 pipelining has not been advertised.
1423 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1425 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1426 returned in the SMTP response when smtp_return_error_details was false.
1427 This has been fixed.
1429 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1430 instead of the C I/O that was originally supplied, because problems were
1431 reported on Solaris.
1433 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1434 Exim which did not show up earlier: it was assuming that a call to
1435 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1436 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1437 was a macro that became an assignment, so it seemed to work. This has
1438 changed to a proper function call with a void return, hence the compile
1439 error. Exim's code has been fixed.
1441 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1444 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1446 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1448 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1449 because it turns out to be too big or there is a timeout) there is no
1450 "Completed" line in the log. When some messages of this type were
1451 selected by exigrep, they were listed as "not completed". Others were
1452 picked up by some special patterns. I have improved the selection
1453 criteria to be more general.
1455 PH/38 The host_find_failed option in the manualroute router can now be set
1456 to "ignore", to completely ignore a host whose IP address cannot be
1457 found. If all hosts are ignored, the behaviour is controlled by the new
1458 host_all_ignored option.
1460 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1461 homing or because of multiple MX records with /mx) generated more than
1462 one IP address, and the following item turned out to be the local host,
1463 all the secondary addresses of the first item were incorrectly removed
1464 from the list, along with the local host and any following hosts (which
1465 is what is supposed to happen).
1467 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1468 whoever called Exim into the -H file. In the case of the daemon it was
1469 behaving confusingly. When first started, it used values for whoever
1470 started the daemon, but after a SIGHUP it used the Exim user (because it
1471 calls itself on a restart). I have changed the code so that it now always
1474 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1475 message are rejected with the same error (e.g. no authentication or bad
1476 sender address), and a DATA command is nevertheless sent (as can happen
1477 with PIPELINING or a stupid MUA), the error message that was given to the
1478 RCPT commands is included in the rejection of the DATA command. This is
1479 intended to be helpful for MUAs that show only the final error to their
1482 PH/42 Another patch from the Sieve maintainer.
1484 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1485 Eximstats - Fixed some broken HTML links and added missing column headers
1487 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1488 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1490 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1492 PH/43 Yet another patch from the Sieve maintainer.
1494 PH/44 I found a way to check for a TCP/IP connection going away before sending
1495 the response to the final '.' that terminates a message, but only in the
1496 case where the client has not sent further data following the '.'
1497 (unfortunately, this is allowed). However, in many cases there won't be
1498 any further data because there won't be any more messages to send. A call
1499 to select() can be used: if it shows that the input is "ready", there is
1500 either input waiting, or the socket has been closed. An attempt to read
1501 the next input character can distinguish the two cases. Previously, Exim
1502 would have sent an OK response which the client would never have see.
1503 This could lead to message repetition. This fix should cure that, at
1504 least in a lot of common cases.
1506 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1507 advertised in response to EHLO.
1513 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1514 fixed by 4.65/MH/01 (is this a record?) are fixed:
1516 (i) An empty string was always treated as zero by the numeric comparison
1517 operators. This behaviour has been restored.
1519 (ii) It is documented that the numeric comparison operators always treat
1520 their arguments as decimal numbers. This was broken in that numbers
1521 starting with 0 were being interpreted as octal.
1523 While fixing these problems I realized that there was another issue that
1524 hadn't been noticed. Values of message_size_limit (both the global option
1525 and the transport option) were treated as octal if they started with 0.
1526 The documentation was vague. These values are now always treated as
1527 decimal, and I will make that clear in the documentation.
1533 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1534 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1537 MH/01 Don't check that the operands of numeric comparison operators are
1538 integers when their expansion is in "skipping" mode (fixes bug
1539 introduced by 4.64-PH/07).
1541 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1542 child addresses, Exim now panics and dies. Previously, because the count
1543 is held in a short int, deliveries were likely to be lost. As such a
1544 large number of recipients for a single message is ridiculous
1545 (performance will be very, very poor), I have chosen to impose a limit
1546 rather than extend the field.
1552 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1553 leftover -K file (the existence of which was triggered by #402).
1554 While we were at it, introduced process PID as part of the -K
1555 filename. This should rule out race conditions when creating
1558 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1559 processing considerably. Previous code took too long for large mails,
1560 triggering a timeout which in turn triggers #401.
1562 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1563 in the DK code in transports.c. sendfile() is not really portable,
1564 hence the _LINUX specificness.
1566 TF/01 In the add_headers option to the mail command in an Exim filter,
1567 there was a bug that Exim would claim a syntax error in any
1568 header after the first one which had an odd number of characters
1571 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1572 callout verification, Exim cached a "reject" for the entire domain. This
1573 is correct for most verifications, but it is not correct for a recipient
1574 verification with use_sender or use_postmaster set, because in that case
1575 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1576 case of MAIL FROM:<> rejection from other early rejections (e.g.
1577 rejection of HELO). When verifying a recipient using a non-null MAIL
1578 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1579 Whatever the result of the callout, the value of the domain cache is
1580 left unchanged (for any other kind of callout, getting as far as trying
1581 RCPT means that the domain itself is ok).
1583 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1586 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1587 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1588 session, the daemon ignores ECONNECT errors and logs others; it now
1589 ignores EPIPE as well.
1591 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1592 (quoted-printable decoding).
1594 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1595 later the small subsequent patch to fix an introduced bug.
1597 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1599 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1601 PH/08 An error is now given if message_size_limit is specified negative.
1603 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1604 to be given (somewhat) arbitrary names.
1606 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1609 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1610 miscellaneous code fixes
1612 PH/10 Added the log_reject_target ACL modifier to specify where to log
1615 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1616 hostname. This is wrong, because it relates to the incoming message (and
1617 probably the interface on which it is arriving) and not to the outgoing
1618 callout (which could be using a different interface). This has been
1619 changed to use the value of the helo_data option from the smtp transport
1620 instead - this is what is used when a message is actually being sent. If
1621 there is no remote transport (possible with a router that sets up host
1622 addresses), $smtp_active_hostname is used.
1624 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1625 tweaks were necessary in order to get it to work (see also 21 below):
1626 (a) The code assumed that strncpy() returns a negative number on buffer
1627 overflow, which isn't the case. Replaced with Exim's string_format()
1629 (b) There were several signed/unsigned issues. I just did the minimum
1630 hacking in of casts. There is scope for a larger refactoring.
1631 (c) The code used strcasecmp() which is not a standard C function.
1632 Replaced with Exim's strcmpic() function.
1633 (d) The code set only $1; it now sets $auth1 as well.
1634 (e) A simple test gave the error "authentication client didn't specify
1635 service in request". It would seem that Dovecot has changed its
1636 interface. Fortunately there's a specification; I followed it and
1637 changed what the client sends and it appears to be working now.
1639 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1642 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1643 address A is aliased to B and C, where B exists and C does not. Without
1644 -v the output is "A verified" because verification stops after a
1645 successful redirection if more than one address is generated. However,
1646 with -v the child addresses are also verified. Exim was outputting "A
1647 failed to verify" and then showing the successful verification for C,
1648 with its parentage. It now outputs "B failed to verify", showing B's
1649 parentage before showing the successful verification of C.
1651 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1652 look up a TXT record in a specific list after matching in a combined
1655 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1656 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1657 they consult the DNS. I had assumed they would set it the way they
1658 wanted; and indeed my experiments on Linux seem to show that in some
1659 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1660 To be on the safe side, however, I have now made the interface to
1661 host_find_byname() similar to host_find_bydns(), with an argument
1662 containing the DNS resolver options. The host_find_byname() function now
1663 sets these options at its start, just as host_find_bydns() does. The smtp
1664 transport options dns_qualify_single and dns_search_parents are passed to
1665 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1666 of host_find_byname() use the default settings of RES_DEFNAMES
1667 (qualify_single) but not RES_DNSRCH (search_parents).
1669 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1670 spool_read_header() do less string testing, by means of a preliminary
1671 switch on the second character of optional "-foo" lines. (This is
1672 overdue, caused by the large number of possibilities that now exist.
1673 Originally there were few.) While I was there, I also converted the
1674 str(n)cmp tests so they don't re-test the leading "-" and the first
1675 character, in the hope this might squeeze out yet more improvement.
1677 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1678 flag allowing group syntax was set by the header_syntax check but not
1679 turned off, possible causing trouble later; (2) The flag was not being
1680 set at all for the header_verify test, causing "group"-style headers to
1681 be rejected. I have now set it in this case, and also caused header_
1682 verify to ignore an empty address taken from a group. While doing this, I
1683 came across some other cases where the code for allowing group syntax
1684 while scanning a header line wasn't quite right (mostly, not resetting
1685 the flag correctly in the right place). These bugs could have caused
1686 trouble for malformed header lines. I hope it is now all correct.
1688 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1689 with the "reply" argument non-NULL. The code, however (which originally
1690 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1691 but it didn't always do it. This confused somebody who was copying the
1692 code for some other use. I have removed all the tests.
1694 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1695 feature that was used to support insecure browsers during the U.S. crypto
1696 embargo. It requires special client support, and Exim is probably the
1697 only MTA that supported it -- and would never use it because real RSA is
1698 always available. This code has been removed, because it had the bad
1699 effect of slowing Exim down by computing (never used) parameters for the
1700 RSA_EXPORT functionality.
1702 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1703 authenticator to fail if there's a tab character in the incoming data
1704 (there should never be unless someone is messing about, as it's supposed
1705 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1706 if the connection is using TLS or if the remote IP is the same as the
1707 local IP, and the "valid-client-cert option" if a client certificate has
1710 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1711 authenticators. This can be used for authorization after authentication
1712 succeeds. (In the case of plaintext, it servers for both authentication
1715 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1716 if any retry times were supplied.
1718 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1719 connection, where there is no client IP address to check. In this
1720 situation, the verify now always succeeds.
1722 PH/25 Applied John Jetmore's -Mset patch.
1724 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1726 PH/27 In a string expansion for a processed (not raw) header when multiple
1727 headers of the same name were present, leading whitespace was being
1728 removed from all of them, but trailing whitespace was being removed only
1729 from the last one. Now trailing whitespace is removed from each header
1730 before concatenation. Completely empty headers in a concatenation (as
1731 before) are ignored.
1733 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1734 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1736 PH/29 [Removed. This was a change that I later backed out, and forgot to
1737 correct the ChangeLog entry (that I had efficiently created) before
1738 committing the later change.]
1740 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1741 address errors (4xx response to RCPT) over the same connection as other
1742 messages routed to the same hosts. Such deliveries are always "forced",
1743 so retry times are not inspected. This resulted in far too many retries
1744 for the affected addresses. The effect occurred only when there were more
1745 hosts than the hosts_max_try setting in the smtp transport when it had
1746 the 4xx errors. Those hosts that it had tried were not added to the list
1747 of hosts for which the message was waiting, so if all were tried, there
1748 was no problem. Two fixes have been applied:
1750 (i) If there are any address or message errors in an SMTP delivery, none
1751 of the hosts (tried or untried) are now added to the list of hosts
1752 for which the message is waiting, so the message should not be a
1753 candidate for sending over the same connection that was used for a
1754 successful delivery of some other message. This seems entirely
1755 reasonable: after all the message is NOT "waiting for some host".
1756 This is so "obvious" that I'm not sure why it wasn't done
1757 previously. Hope I haven't missed anything, but it can't do any
1758 harm, as the worst effect is to miss an optimization.
1760 (ii) If, despite (i), such a delivery is accidentally attempted, the
1761 routing retry time is respected, so at least it doesn't keep
1762 hammering the server.
1764 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1765 in ${readsocket because some servers need this prod.
1767 PH/32 Added some extra debug output when updating a wait-xxx database.
1769 PH/33 The hint "could be header name not terminated by colon", which has been
1770 given for certain expansion errors for a long time, was not being given
1771 for the ${if def:h_colon_omitted{... case.
1773 PH/34 The spec says: "With one important exception, whenever a domain list is
1774 being scanned, $domain contains the subject domain." There was at least
1775 one case where this was not true.
1777 PH/35 The error "getsockname() failed: connection reset by peer" was being
1778 written to the panic log as well as the main log, but it isn't really
1779 panic-worthy as it just means the connection died rather early on. I have
1780 removed the panic log writing for the ECONNRESET error when getsockname()
1783 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1784 runs only) independently of the message's sender address. This meant
1785 that, if the 4xx error was in fact related to the sender, a different
1786 message to the same recipient with a different sender could confuse
1787 things. In particualar, this can happen when sending to a greylisting
1788 server, but other circumstances could also provoke similar problems.
1789 I have changed the default so that the retry time for these errors is now
1790 based a combination of the sender and recipient addresses. This change
1791 can be overridden by setting address_retry_include_sender=false in the
1794 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1795 remote server are returned as part of bounce messages. This was not
1796 happening for LMTP over a pipe (the lmtp transport), but now it is the
1797 same for both kinds of LMTP.
1799 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1800 in header lines that were in fact CNAMEs. This is no longer the case.
1802 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1803 and queue runs started by the daemon processed all messages. This has
1804 been fixed so that -R and -S can now usefully be given with -q<time>.
1806 PH/40 Import PCRE release 6.7 (fixes some bugs).
1808 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1810 PH/42 Give an error if -q is specified more than once.
1812 PH/43 Renamed the variables $interface_address and $interface_port as
1813 $received_ip_address and $received_port, to make it clear that these
1814 values apply to message reception, and not to the outgoing interface when
1815 a message is delivered. (The old names remain recognized, of course.)
1817 PH/44 There was no timeout on the connect() call when using a Unix domain
1818 socket in the ${readsocket expansion. There now is.
1820 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1821 be meaningful with "accept".
1823 SC/01 Eximstats V1.43
1824 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1826 SC/02 Eximstats V1.44
1827 Use a glob alias rather than an array ref in the generated
1828 parser. This improves both readability and performance.
1830 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1831 Collect SpamAssassin and rejection statistics.
1832 Don't display local sender or destination tables unless
1833 there is data to show.
1834 Added average volumes into the top table text output.
1836 SC/04 Eximstats V1.46
1837 Collect data on the number of addresses (recipients)
1838 as well as the number of messages.
1840 SC/05 Eximstats V1.47
1841 Added 'Message too big' to the list of mail rejection
1842 reasons (thanks to Marco Gaiarin).
1844 SC/06 Eximstats V1.48
1845 Mainlog lines which have GMT offsets and are too short to
1846 have a flag are now skipped.
1848 SC/07 Eximstats V1.49 (Alain Williams)
1849 Added the -emptyok flag.
1851 SC/08 Eximstats V1.50
1852 Fixes for obtaining the IP address from reject messages.
1854 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1855 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1856 whitesspace changes from 4.64-PH/27
1858 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1861 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1862 are found, allow negative numbers in numeric criteria)
1864 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1866 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1867 to match changes made in 4.64-PH/43
1869 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1871 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1872 "domain" is actually an IP address, and force a failure. This locks out
1873 those revolvers/nameservers that support "A-for-A" lookups, in
1874 contravention of the specifications.
1876 PH/48 When a host name was looked up from an IP address, and the subsequent
1877 forward lookup of the name timed out, the host name was left in
1878 $sender_host_name, contrary to the specification.
1880 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1881 restricted to single-key lookups, Exim was not diagnosing an error if
1882 * or *@ was used with a query-style lookup.
1884 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1886 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1887 long ago, but noone interested enough thought of it. Let's just say that
1888 the "1.1" means that there are some new functions that weren't there at
1889 some point in the past.
1891 PH/51 Error processing for expansion failure of helo_data from an smtp
1892 transport during callout processing was broken.
1894 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1895 tested/used via the -bh/-bhc/-bs options.
1897 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1898 bug, fixed in subsequent PCRE releases).
1900 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1901 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1903 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1909 SC/01 Use a glob alias rather than an array ref in eximstats generated
1910 parser. This improves both readability and performance.
1912 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1913 Don't display local sender or destination tables in eximstats unless
1914 there is data to show.
1915 Added average volumes into the eximstats top table text output.
1917 SC/03 Collect data on the number of addresses (recipients) as well
1918 as the number of messages in eximstats.
1920 TF/01 Correct an error in the documentation for the redirect router. Exim
1921 does (usually) call initgroups() when daemonizing.
1923 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1924 with consistent privilege compared to when running as a daemon.
1926 TF/03 Note in the spec that $authenticated_id is not set for local
1927 submissions from trusted users.
1929 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1930 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1932 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1933 by adding some example configuration directives to the default
1934 configuration file. A little bit of work is required to uncomment the
1935 directives and define how usernames and passwords are checked, but
1936 there is now a framework to start from.
1938 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1939 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1940 without this. I don't know how relevant this is to other LDAP libraries.
1942 PH/02 Add the verb name to the "unknown ACL verb" error.
1944 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1946 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1948 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1949 directory not expanded when it should be if an expanded home directory
1950 was set for the address (which is overridden by the transport).
1952 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1955 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1956 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1957 because it is too late at that time, and has no effect.
1959 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1960 security issue with \' (bugzilla #107). I could not use the
1961 PQescapeStringConn() function, because it needs a PGconn value as one of
1964 PH/08 When testing addresses using -bt, indicate those final addresses that
1965 are duplicates that would not cause an additional delivery. At least one
1966 person was confused, thinking that -bt output corresponded to deliveries.
1967 (Suppressing duplicates isn't a good idea as you lose the information
1968 about possibly different redirections that led to the duplicates.)
1970 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1971 systems where poll() doesn't work, in particular OS X.
1973 PH/10 Added more information to debugging output for retry time not reached.
1975 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1976 operations in malware.c.
1978 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1981 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1982 syslog_duplication set false, log lines that would normally be written
1983 both the the main log and to the reject log were not written to syslog at
1986 PH/14 In the default configuration, change the use of "message" in ACL warn
1987 statements to "add_header".
1989 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1990 not followed by a command (e.g. "seen endif").
1992 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1993 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1996 PH/17 Added extra conditions to the default value of delay_warning_condition
2000 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2001 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2002 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2005 The Auto-Submitted: and various List- headers are standardised, whereas I
2006 don't think Precedence: ever was.
2008 PH/18 Refactored debugging code in route_finduser() to show more information,
2009 in particular, the error code if getpwnam() issues one.
2011 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2012 This is apparently needed in addition to the PH/07 change above to avoid
2013 any possible encoding problems.
2015 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2016 but not after initializing Perl.
2018 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2019 output them only if debugging. By default they are written stderr,
2020 apparently, which is not desirable.
2022 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2025 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2028 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2030 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2031 authenticated or an ident call has been made. Suppress the default
2032 values for $authenticated_id and $authenticated_sender (but permit -oMai
2033 and -oMas) when testing with -bh.
2035 PH/24 Re-jigged the order of the tests in the default configuration so that the
2036 tests for valid domains and recipients precede the DNS black list and CSA
2037 tests, on the grounds that those ones are more expensive.
2039 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2040 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2041 This bug exists in every version of Exim that I still have, right back to
2044 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2045 However, an attempt to turn on case-sensitivity in a regex key by
2046 including (?-i) didn't work because the subject string was already
2047 lowercased, and the effects were non-intuitive. It turns out that a
2048 one-line patch can be used to allow (?-i) to work as expected.
2054 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2055 other effects) broke the use of negated acl sub-conditions.
2057 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2060 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2061 "Deny" causes Exim to reject the incoming connection with a 554 error.
2062 Unfortunately, if there is a major crisis, such as a disk failure,
2063 tcp-wrappers gives "deny", whereas what one would like would be some
2064 kind of temporary error. A kludge has been added to help with this.
2065 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2066 554 error is used if errno is still zero or contains ENOENT (which occurs
2067 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2070 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2072 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2073 errors. Otherwise a message that provokes a temporary error (when other
2074 messages do not) can cause a whole host to time out.
2076 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2077 addresses were routed directly to files or pipes from a redirect router.
2078 File deliveries just didn't batch; pipe deliveries might have suffered
2081 PH/06 A failure to get a lock for a hints database would erroneously always say
2082 "Failed to get write lock", even when it was really a read lock.
2084 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2085 of 0600. This has been changed to use the value of the lockfile_mode
2086 option (which defaults to 0600).
2088 PH/08 Applied small patch from the Sieve maintainer.
2090 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2091 folder from quota calculations, a direct delivery into this folder messed
2092 up the contents of the maildirsize file. This was because the regex was
2093 used only to exclude .Trash (or whatever) when the size of the mailbox
2094 was calculated. There was no check that a delivery was happening into an
2095 excluded directory. This bug has been fixed by ignoring all quota
2096 processing for deliveries into excluded directories.
2098 PH/10 Added the maildirfolder_create_regex option to appendfile.
2104 PH/01 The code for finding all the local interface addresses on a FreeBSD
2105 system running IPv6 was broken. This may well have applied to all BSD
2106 systems, as well as to others that have similar system calls. The broken
2107 code found IPv4 interfaces correctly, but gave incorrect values for the
2108 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2109 that it would not match correctly against @[] and not recognize the IPv6
2112 PH/02 The ipliteral router was not recognizing addresses of the form user@
2113 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2115 PH/03 Added disable_ipv6.
2117 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2118 decoded form, because it is most often used to construct To: headers
2119 lines in autoreplies, and the decoded form may well be syntactically
2120 invalid. However, $reply_address has leading white space removed, and all
2121 newlines turned into spaces so that the autoreply transport does not
2124 PH/05 If group was specified without a user on a router, and no group or user
2125 was specified on a transport, the group from the router was ignored.
2127 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2128 for visible compile-time settings that can be used to change these
2129 numbers, for those that want even more. Backwards compatibility with old
2130 spool files has been maintained. However, going back to a previous Exim
2131 release will lost any variables that are in spool files.
2133 PH/07 Two small changes when running in the test harness: increase delay when
2134 passing a TCP/IP connection to a new process, in case the original
2135 process has to generate a bounce, and remove special handling of
2136 127.0.0.2 (sic), which is no longer necessary.
2138 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2139 be the same on different OS.
2141 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2144 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2145 whether --show-vars was specified or not
2147 JJ/02 exipick: Added support for new ACL variable spool format introduced
2150 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2151 syntactically invalid From: or Reply-to: line, and a filter used this to
2152 generate an autoreply, and therefore failed to obtain an address for the
2153 autoreply, Exim could try to deliver to a non-existent relative file
2154 name, causing unrelated and misleading errors. What now happens is that
2155 it logs this as a hard delivery error, but does not attempt to create a
2158 PH/11 The exinext utility has a -C option for testing purposes, but although
2159 the given file was scanned by exinext itself; it wasn't being passed on
2160 when Exim was called.
2162 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2163 an end-of-file indication when reading a command response.
2165 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2166 compiled. In many other places in Exim, IPv6 addresses are always
2167 recognized, so I have changed this. It also means that IPv4 domain
2168 literals of the form [IPV4:n.n.n.n] are now always recognized.
2170 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2171 used if the router is not running as root, for example, when verifying at
2172 ACL time, or when using -bh. The debugging output from this situation was
2173 non-existent - all you got was a failure to exec. I have made two
2176 (a) Failures to set uid/gid, the current directory, or a process leader
2177 in a subprocess such as that created by queryprogram now generate
2178 suitable debugging ouput when -d is set.
2180 (b) The queryprogram router detects when it is not running as root,
2181 outputs suitable debugging information if -d is set, and then runs
2182 the subprocess without attempting to change uid/gid.
2184 PH/15 Minor change to Makefile for building test_host (undocumented testing
2187 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2188 additional section of a DNS packet that returns MX or SRV records.
2189 Instead, it always explicitly searches for A/AAAA records. This avoids
2190 major problems that occur when a DNS server includes only records of one
2191 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2192 fixed another bug: if SRV records were looked up and the corresponding
2193 address records were *not* found in the additional section, the port
2194 values from the SRV records were lost.
2196 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2197 using the correct key (the original address) when searching the retry
2198 rules in order to find which one to use for generating the retry hint.
2200 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2201 adding the default one. Similarly, if it contains a Reply-To: header, the
2202 errors_reply_to option, if set, is not used.
2204 PH/19 When calculating a retry time, Exim used to measure the "time since
2205 failure" by looking at the "first failed" field in the retry record. Now
2206 it does not use this if it is later than than the arrival time of the
2207 message. Instead it uses the arrival time. This makes for better
2208 behaviour in cases where some deliveries succeed, thus re-setting the
2209 "first failed" field. An example is a quota failure for a huge message
2210 when small messages continue to be delivered. Without this change, the
2211 "time since failure" will always be short, possible causing more frequent
2212 delivery attempts for the huge message than are intended.
2213 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2215 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2216 $1, $2, $3) because the numerical variables can be reset during some
2217 expansion items (e.g. "match"), thereby losing the authentication data.
2219 PH/21 Make -bV show the size of off_t variables so that the test suite can
2220 decide whether to run tests for quotas > 2G.
2222 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2223 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2224 filecount value is greater than 2G or if a quota value is greater than 2G
2225 on a system where the size of off_t is not greater than 4, a panic error
2228 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2229 never match. The debug and -bh output now contains an explicit error
2230 message indicating a malformed IPv4 address or mask.
2232 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2233 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2234 PH/23 above applies.
2236 PH/25 Do not write to syslog when running in the test harness. The only
2237 occasion when this arises is a failure to open the main or panic logs
2238 (for which there is an explicit test).
2240 PH/26 Added the /no_tell option to "control=freeze".
2242 PH/27 If a host name lookup failed very early in a connection, for example, if
2243 the IP address matched host_lookup and the reverse lookup yielded a name
2244 that did not have a forward lookup, an error message of the form "no IP
2245 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2246 could be logged. Now it outputs the IP address instead of "NULL".
2248 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2249 allows the sender and the authenticated sender to be set when
2250 submitting a message from within Exim. Since child_open_exim() is
2251 documented for local_scan(), the new function should be too.
2253 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2254 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2255 results in an empty string is now treated as unset.
2257 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2259 PH/31 Added sender_verify_fail logging option.
2261 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2262 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2263 tidied the source and removed it altogether.
2265 PH/33 When a queue run was abandoned because the load average was too high, a
2266 log line was always written; now it is written only if the queue_run log
2267 selector is set. In addition, the log line for abandonment now contains
2268 information about the queue run such as the pid. This is always present
2269 in "start" and "stop" lines but was omitted from the "abandon" line.
2271 PH/34 Omit spaces between a header name and the colon in the error message that
2272 is given when verify = headers_syntax fails (if there are lots of them,
2273 the message gets confusing).
2275 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2276 names, as there are now some PTR records that contain slashes. This check
2277 is only to protect against broken name servers that fall over on strange
2278 characters, so the fact that it applies to all lookups doesn't matter.
2280 PH/36 Now that the new test suite is complete, we can remove some of the
2281 special code in Exim that was needed for the old test suite. For example,
2282 sorting DNS records because real resolvers return them in an arbitrary
2283 order. The new test suite's fake resolver always returns records in the
2286 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2287 bounces) except when queue_only is set, to avoid logging races between
2288 the different processes.
2290 PH/38 Panic-die if .include specifies a non-absolute path.
2292 PH/39 A tweak to the "H" retry rule from its user.
2294 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2295 a label. They prevented compilation on older perls.
2297 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2298 a warning to be raised on newish perls.
2300 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2301 on queue. Changes to match documented behaviour of showing count of
2302 messages matching specified criteria.
2304 PH/40 Changed the default ident timeout from 30s to 5s.
2306 PH/41 Added support for the use of login_cap features, on those BSD systems
2307 that have them, for controlling the resources used by pipe deliveries.
2309 PH/42 The content-scanning code uses fopen() to create files in which to put
2310 message data. Previously it was not paying any attention to the mode of
2311 the files. Exim runs with umask(0) because the rest of the code creates
2312 files with open(), and sets the required mode explicitly. Thus, these
2313 files were ending up world-writeable. This was not a big issue, because,
2314 being within the spool directory, they were not world-accessible. I have
2315 created a function called modefopen, which takes an additional mode
2316 argument. It sets umask(777), creates the file, chmods it to the required
2317 mode, then resets the umask. All the relevant calls to fopen() in the
2318 content scanning code have been changed to use this function.
2320 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2321 to 24 hours. This avoids potential overflow problems when processing G
2322 and H retry rules. I suspect nobody ever tinkers with this value.
2324 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2326 PH/45 When the plaintext authenticator is running as a client, the server's
2327 challenges are checked to ensure they are valid base64 strings. By
2328 default, the authentication attempt is cancelled if an invalid string is
2329 received. Setting client_ignore_invalid_base64 true ignores these errors.
2330 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2331 they are received. Thus, the responses can be made to depend on the
2332 challenges. If an invalid string is ignored, an empty string is placed in
2335 PH/46 Messages that are created by the autoreply transport now contains a
2336 References: header, in accordance with RFCs 2822 and 3834.
2338 PH/47 Added authenticated_sender_force to the smtp transport.
2340 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2342 PH/49 Installed latest patch from the Sieve maintainer.
2344 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2345 was also set, the appendfile transport was unnecessarily scanning a
2346 directory of message files (e.g. for maildir delivery) to find the count
2347 of files (along with the size), even though it did not need this
2348 information. It now does the scan only if it needs to find either the
2349 size of the count of files.
2351 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2353 PH/52 Two bugs concerned with error handling when the smtp transport is
2356 (i) Exim was not creating retry information for temporary errors given
2357 for individual recipients after the DATA command when the smtp transport
2358 was used in LMTP mode. This meant that they could be retried too
2359 frequently, and not timed out correctly.
2361 (ii) Exim was setting the flag that allows error details to be returned
2362 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2363 recipients that were returned after the DATA command.
2365 PH/53 This is related to PH/52, but is more general: for any failing address,
2366 when detailed error information was permitted to be returned to the
2367 sender, but the error was temporary, then after the final timeout, only
2368 "retry timeout exceeded" was returned. Now it returns the full error as
2369 well as "retry timeout exceeded".
2371 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2372 do this, and (what is worse) MTAs that accept it.
2374 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2375 will now be deprecated.
2377 PH/56 New os.c-cygwin from the Cygwin maintainer.
2379 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2380 formats (previously only available in exim formats via -bpr, -bpru,
2381 and -bpra. Now also available in native and exiqgrep formats)
2383 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2384 with very large, slow to parse queues
2386 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2388 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2390 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2391 responses to authentication challenges, though it was showing the
2392 challenges; (ii) I've removed the CR characters from the debug output for
2395 PH/58 Allow for the insertion of a newline as well as a space when a string
2396 is turned into more than one encoded-word during RFC 2047 encoding. The
2397 Sieve code now uses this.
2399 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2400 data_4xx, lost_connection, tls_required.
2402 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2403 message was being sent as an SMTP response.
2405 PH/61 Add -l and -k options to exicyclog.
2407 PH/62 When verifying, if an address was redirected to one new address, so that
2408 verification continued, and the new address failed or deferred after
2409 having set something in $address_data, the value of $address_data was not
2410 passed back to the ACL. This was different to the case when no
2411 redirection occurred. The value is now passed back in both cases.
2413 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2414 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2415 use login_cap.h, so on its own it isn't the distinguishing feature. The
2416 new name refers directly to the setclassresources() function.
2418 PH/65 Added configuration files for NetBSD3.
2420 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2422 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2423 is preferred over IPv4.
2425 PH/68 The bounce_return_message and bounce_return_body options were not being
2426 honoured for bounces generated during the reception of non-SMTP messages.
2427 In particular, this applied to messages rejected by the ACL. This bug has
2428 been fixed. However, if bounce_return_message is true and bounce_return_
2429 body is false, the headers that are returned for a non-SMTP message
2430 include only those that have been read before the error was detected.
2431 (In the case of an ACL rejection, they have all been read.)
2433 PH/69 The HTML version of the specification is now built in a directory called
2434 spec_html instead of spec.html, because the latter looks like a path with
2435 a MIME-type, and this confuses some software.
2437 PH/70 Catch two compiler warnings in sieve.c.
2439 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2440 function verify_get_ident() calls ip_connect() to connect a socket, but
2441 if the "connect()" function timed out, ip_connect() used to close the
2442 socket. However, verify_get_ident() also closes the socket later, and in
2443 between Exim writes to the log, which may get opened at this point. When
2444 the socket was closed in ip_connect(), the log could get the same file
2445 descriptor number as the socket. This naturally causes chaos. The fix is
2446 not to close the socket in ip_connect(); the socket should be closed by
2447 the function that creates it. There was only one place in the code where
2448 this was missing, in the iplookup router, which I don't think anybody now
2449 uses, but I've fixed it anyway.
2451 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2452 well as to direct DNS lookups. Otherwise the handling of names in host
2453 lists is inconsistent and therefore confusing.
2459 PH/01 Two changes to the default runtime configuration:
2461 (1) Move the checks for relay_from_hosts and authenticated clients from
2462 after to before the (commented out) DNS black list checks.
2464 (2) Add control=submission to the relay_from_hosts and authenticated
2465 clients checks, on the grounds that messages accepted by these
2466 statements are most likely to be submissions.
2468 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2470 (1) Generate an error if the third argument for the ${prvs expansion is
2473 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2476 (3) Reset the variables that are obtained from the first argument of
2477 ${prvscheck and used in the second argument before leaving the code,
2478 because their memory is reclaimed, so using them afterwards may do
2481 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2482 one (it's much easier than Tom thought :-).
2484 (5) Because of (4), we can now allow for the use of $prvscheck_result
2485 inside the third argument.
2487 PH/03 For some reason, the default setting of PATH when running a command from
2488 a pipe transport was just "/usr/bin". I have changed it to
2491 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2492 anything to be listed in the output from -bV.
2494 PH/05 When a filter generated an autoreply, the entire To: header line was
2495 quoted in the delivery log line, like this:
2497 => >A.N.Other <ano@some.domain> <original@ddress> ...
2499 This has been changed so that it extracts the operative address. There
2500 may be more than one such address. If so, they are comma-separated, like
2503 => >ano@some.domain,ona@other.domain <original@ddress> ...
2505 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2506 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2507 not being looked up in the rDNS to get a host name, Exim was showing the
2508 IP address twice in Received: lines, even though the IP addresses were
2509 identical. For example:
2511 Received: from [1.2.3.4] (helo=[1.2.3.4])
2513 However, if the real host name was known, it was omitting the HELO data
2514 if it matched the actual IP address. This has been tidied up so that it
2515 doesn't show the same IP address twice.
2517 PH/07 When both +timestamp and +memory debugging was on, the value given by
2518 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2519 called by the debug printing, thereby overwriting the timestamp buffer.
2520 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2522 PH/08 When the original message was included in an autoreply transport, it
2523 always said "this is a copy of the message, including all the headers",
2524 even if body_only or headers_only was set. It now gives an appropriate
2527 PH/09 Applied a patch from the Sieve maintainer which:
2529 o fixes some comments
2530 o adds the (disabled) notify extension core
2531 o adds some debug output for the result of if/elsif tests
2532 o points to the current vacation draft in the documentation
2533 and documents the missing references header update
2537 o fixes a bug in processing the envelope test (when testing
2538 multiple envelope elements, the last element determinted the
2541 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2542 Electronic Mail") by including:
2544 Auto-submitted: auto-generated
2546 in the messages that it generates (bounce messages and others, such as
2547 warnings). In the case of bounce messages for non-SMTP mesages, there was
2548 also a typo: it was using "Auto_submitted" (underscore instead of
2549 hyphen). Since every message generated by Exim is necessarily in response
2550 to another message, thes have all been changed to:
2552 Auto-Submitted: auto-replied
2554 in accordance with these statements in the RFC:
2556 The auto-replied keyword:
2558 - SHOULD be used on messages sent in direct response to another
2559 message by an automatic process,
2561 - MUST NOT be used on manually-generated messages,
2563 - MAY be used on Delivery Status Notifications (DSNs) and Message
2564 Disposition Notifications (MDNs),
2566 - MUST NOT be used on messages generated by automatic or periodic
2567 processes, except for messages which are automatic responses to
2570 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2571 to the default Received: header definition.
2573 PH/12 Added log selector acl_warn_skipped (default on).
2575 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2576 variables because (a) they are in the wrong storage pool and (b) even if
2577 they were copied, it wouldn't work properly because of the caching.
2579 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2580 checking when decoding. Apparently there are clients that generate
2581 overlong encoded strings. Why am I not surprised?
2583 PH/15 If the first argument of "${if match_address" was not empty, but did not
2584 contain an "@" character, Exim crashed. Now it writes a panic log message
2585 and treats the condition as false.
2587 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2589 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2590 extension "envelope-auth". The code is finished and in agreement with
2591 other implementations, but there is no documentation so far and in fact,
2592 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2593 not changing the active code.
2595 Print executed "if" and "elsif" statements when debugging is used. This
2596 helps a great deal to understand what a filter does.
2598 Document more things not specified clearly in RFC3028. I had all this
2599 sorted out, when out of a sudden new issues came to my mind. Oops."
2601 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2604 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2605 canonical form (as documented). However, after a host name lookup from
2606 the IP address, check_host() was doing a simple string comparison with
2607 addresses acquired from the DNS when checking that the found name did
2608 have the original IP as one of its addresses. Since any found IPv6
2609 addresses are likely to be in abbreviated form, the comparison could
2610 fail. Luckily, there already exists a function for doing the comparison
2611 by converting both addresses to binary, so now that is used instead of
2612 the text comparison.
2614 PH/20 There was another similar case to PH/19, when a complete host name was
2615 given in a host list; looking up its IP address could give an abbreviated
2616 form, whereas the current host's name might or might not be abbreviated.
2617 The same fix has been applied.
2623 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2624 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2627 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2628 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2630 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2632 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2633 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2634 Also, for "Auto-Submitted", treat anything other than "no" as
2635 non-personal, in accordance with RFC 3834. (Previously it treated
2636 anything starting "auto-" as non-personal.)
2638 TF/01 The control=submission/name=... option had a problem with syntax
2639 errors if the name included a slash character. The /name= option
2640 now slurps the rest of the string, so it can include any characters
2641 but it must come last in the list of options (after /sender_retain
2644 PH/05 Some modifications to the interface to the fake nameserver for the new
2652 TK/01 Added the "success_on_redirect" address verification option. See
2653 NewStuff for rationale and an example.
2655 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2657 PH/02 Patch to exigrep to allow it to work on syslog lines.
2659 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2660 fread() to skip over the body file's header line, because in Cygwin the
2661 header line is locked and is inaccessible.
2663 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2664 co-exist for some time) to make it clear that it is the Exim ID that is
2665 referenced, not the Message-ID: header line.
2667 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2668 string_format() function, because snprintf() does not exist on all
2671 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2672 ${stat: expansion item.
2674 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2675 protocol synchronization error", to keep the pedants happy.
2677 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2678 well as for IRIX systems, when gcc is being used. See the host.c source
2681 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2683 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2686 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2687 transport and to the smtp transport in LMTP mode.
2689 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2691 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2692 run a filter in a subprocess. This could lead to confusion in subsequent
2693 lookups in the parent process. There should also be a search_tidyup() at
2694 the end of the subprocess.
2696 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2697 only if the host matched helo_try_verify_hosts, which caused the
2698 verification to occur when the EHLO/HELO command was issued. The ACL just
2699 tested the remembered result. Now, if a previous verification attempt has
2700 not happened, "verify = helo" does it there and then.
2702 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2704 TK/03 Fix log output including CR from clamd.
2706 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2707 exist provoked a memory error which could cause a segfault.
2709 PH/15 Installed PCRE 6.2
2711 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2713 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2714 of the problem. Specifically, suggested +O2 rather than +O1 for the
2717 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2719 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2720 record, the port was not correctly logged when the outgoing_port log
2721 selector was set (it logged the transort's default port).
2723 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2724 fallback_hosts, and "hosts" in the smtp transport.
2726 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2727 host errors such as "Connection refused".
2729 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2730 authentication with radiusclient 0.4.9:
2732 - Error returned from rc_read_config was caught wrongly
2733 - Username/password not passed on to radius server due to wrong length.
2735 The presumption is that some radiusclient API changes for 4.51/PH/17
2736 were not taken care of correctly. The code is still untested by me (my
2737 Linux distribution still has 0.3.2 of radiusclient), but it was
2738 contributed by a Radius user.
2740 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2741 expanding the "port" option of the smtp transport.
2743 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2744 while reading a MIME header. Thanks to Tom Hughes for a patch.
2746 PH/24 Include config.h inside local_scan.h so that configuration settings are
2749 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2750 that in an ACL for RCPT (for example), you can examine exactly what was
2753 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2754 commands, but it was not correctly comparing the address with the actual
2755 client host address. Thus, it would show the EHLO address in Received:
2756 header lines when this was not necessary.
2758 PH/27 Added the % operator to ${eval:}.
2760 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2761 it should be ignoring failures (because with -C, for example, it has lost
2762 privilege). It wasn't ignoring creation failures other than "already
2765 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2768 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2769 because an input line was too long, either on its own, or by virtue of
2770 too many continuations, the temporary file was not being removed, and the
2771 return code was incorrect.
2773 PH/31 Missing "BOOL" in function definition in filtertest.c.
2775 PH/32 Applied Sieve patches from the maintainer.
2777 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2779 PH/33 Added "verify = not_blind".
2781 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2782 Local/Makefile (with some defaults set). These are used in built scripts
2783 such as exicyclog, but they have never been used in the exim_install
2784 script (though there are many overriding facilities there). I have
2785 arranged that the exim_install script now takes note of these two
2788 PH/35 Installed configuration files for Dragonfly.
2790 PH/36 When a locally submitted message by a trusted user did not contain a
2791 From: header, and the sender address was obtained from -f or from an SMTP
2792 MAIL command, and the trusted user did not use -F to supply a sender
2793 name, $originator_name was incorrectly used when constructing a From:
2794 header. Furthermore, $originator_name was used for submission mode
2795 messages from external hosts without From: headers in a similar way,
2796 which is clearly wrong.
2798 PH/37 Added control=suppress_local_fixups.
2800 PH/38 When log_selector = +received_sender was set, and the addition of the
2801 sender made the log line's construction buffer exactly full, or one byte
2802 less than full, an overflow happened when the terminating "\n" was
2805 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2806 when the result of a list match is failure because a DNS lookup failed.
2808 PH/40 RM_COMMAND is now used in the building process.
2810 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2811 the "build-* directories that it finds.
2813 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2814 address in a domain literal was a prefix of an interface address.
2816 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2817 when verifying a sender address, unless rewrite_headers is false.
2819 PH/44 Wrote a long comment about why errors_to addresses are verified as
2820 recipients, not senders.
2822 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2823 the ratelimit ACL was added.
2825 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2827 PH/46 Added extra information about PostgreSQL errors to the error string.
2829 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2830 avoiding the need to install special zones in a real server. This is
2831 backwards compatible; if it can't find the fake resolver, it drops back.
2832 Thus, both old and new test suites can be run.
2834 TF/02 Added util/ratelimit.pl
2836 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2837 clock is set back in time.
2839 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2840 Candler <B.Candler@pobox.com>.
2842 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2843 OK for addresses that are the result of redirections.
2845 PH/48 A number of further additions for the benefit of the new test suite,
2846 including a fake gethostbyname() that interfaces to the fake DNS resolver
2849 TF/06 The fix for widen_domains has also been applied to qualify_single and
2850 search_parents which are the other dnslookup options that can cause
2853 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2856 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2858 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2859 service" if it encountered an MX record with an empty target hostname.
2860 The message is now "an MX or SRV record indicated no SMTP service".
2862 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2863 if the DNS of the sending site is misconfigured. This is quite a
2864 common situation. This change restores the behaviour of treating a
2865 helo verification defer as a failure.
2867 PH/51 If self=fail was set on a router, the bounce message did not include the
2868 actual error message.
2874 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2876 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2877 command itself ended in error, the underlying message about the transport
2878 filter timeout was being overwritten with the pipe command error. Now the
2879 underlying error message should be appended to the second error message.
2881 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2883 PH/02 Reduce the amount of output that "make" produces by default. Full output
2884 can still be requested.
2886 PH/03 The warning log line about a condition test deferring for a "warn" verb
2887 was being output only once per connection, rather than after each
2888 occurrence (because it was using the same function as for successful
2889 "warn" verbs). This seems wrong, so I have changed it.
2891 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2892 it should not have, which might have caused a crash in the right
2893 circumstances, but probably never did.
2895 PH/04 Installed a modified version of Tony Finch's patch to make submission
2896 mode fix the return path as well as the Sender: header line, and to
2897 add a /name= option so that you can make the user's friendly name appear
2900 TF/03 Added the control = fakedefer ACL modifier.
2902 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2903 Mark Lowes for thorough testing.
2905 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2907 TK/03 Merged latest SRS patch from Miles Wilton.
2909 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2910 with the definition in sysexits.h (which is #included earlier).
2911 Fortunately, Exim does not actually use EX_OK. The code used to try to
2912 preserve the sysexits.h value, by assumimg that macro definitions were
2913 scanned for macro replacements. I have been disabused of this notion,
2914 so now the code just undefines EX_OK before #including unistd.h.
2916 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2917 in the smtp transport. When a block could not be written in a single
2918 write() function, the timeout was being re-applied to each part-write.
2919 This seems wrong - if the receiver was accepting one byte at a time it
2920 would take for ever. The timeout is now adjusted when this happens. It
2921 doesn't have to be particularly precise.
2923 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2924 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2926 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2927 without a domain if the check to <postmaster@domain> fails.
2929 SC/01 Eximstats: added -xls and the ability to specify output files
2930 (patch written by Frank Heydlauf).
2932 SC/02 Eximstats: use FileHandles for outputing results.
2934 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2936 SC/04 Eximstats: fixed display of large numbers with -nvr option
2938 SC/05 Eximstats: fixed merging of reports with empty tables.
2940 SC/06 Eximstats: added the -include_original_destination flag
2942 SC/07 Eximstats: removed tabs and trailing whitespace.
2944 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2946 TK/06 MBOX spool code: Add real "From " MBOX separator line
2947 so the .eml file is really in mbox format (even though
2948 most programs do not really care). Patch from Alex Miller.
2950 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2951 The latter is generated from $received_to and is only set if the
2952 message has one envelope recipient. SA can use these headers,
2953 obviously out-of-the-box. Patch from Alex Miller.
2955 PH/08 The ${def test on a variable was returning false if the variable's
2956 value was "0", contrary to what the specification has always said!
2957 The result should be true unless the variable is empty.
2959 PH/09 The syntax error of a character other than { following "${if
2960 def:variable_name" (after optional whitespace) was not being diagnosed.
2961 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2962 accidental colon was present, for example, could give incorrect results.
2964 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2965 result is used (not including appendfile, where other changes are about
2968 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2969 This involved changing a lot of size variables from int to off_t. It
2970 should work with maildirs and everything.
2972 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2973 spamd dying while we are connected to it.
2975 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2978 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2979 function for BDB 4.3.
2981 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2983 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2984 the sources for PCRE have been re-arranged and more files are now
2987 PH/15 The code I had for printing potentially long long variables in PH/11
2988 above was not the best (it lost precision). The length of off_t variables
2989 is now inspected at build time, and an appropriate printing format (%ld
2990 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2991 to be "long long int" or "long int". This is needed for the internal
2992 formatting function string_vformat().
2994 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2995 the configuration file to be ":syslog", then the script "guesses" where
2996 the logs files are, rather than using the compiled in default. In our
2997 case the guess is not the same as the compiled default, so the script
2998 suddenly stopped working when I started to use syslog. The patch checks
2999 to see if log_file_path is "". If so, it attempts to read it from exim
3000 with no configuration file to get the compiled in version, before it
3001 falls back to the previous guessing code."
3003 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3004 implementing BATV in an Exim configuration. See NewStuff for the gory
3007 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3008 Makefile that are specific to HP-UX.
3010 PH/18 If the "use_postmaster" option was set for a recipient callout together
3011 with the "random" option, the postmaster address was used as the MAIL
3012 FROM address for the random test, but not for the subsequent recipient
3013 test. It is now used for both.
3015 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3016 patch removes a few documentation additions to RFC 3028, because the
3017 latest draft now contains them. It adds the new en;ascii-case comparator
3018 and a new error check for 8bit text in MIME parts. Comparator and
3019 require names are now matched exactly. I enabled the subaddress
3020 extension, but it is not well tested yet (read: it works for me)."
3022 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3023 rework some of the code of TK/09 above to avoid the hardwired use of
3024 "%lld" and "long long". Replaced the call to snprintf() with a call to
3027 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3028 records point to non-existent hosts", "retry timeout exceeded", and
3029 "retry time not reached for any host after a long failure period".
3031 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3032 experimental DomainKeys support:
3034 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3035 (2) On an error such as an illegally used "control", the wrong name for
3036 the control was given.
3038 These problems did NOT occur unless DomainKeys support was compiled.
3040 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3042 PH/24 Added ${if match_ip condition.
3044 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3045 databases so that it will be absolutely obvious if a crash occurs in the
3046 DB library. This is a regular occurrence (often caused by mis-matched
3049 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3050 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3051 picked up on a user's system that detects such things. There doesn't seem
3052 to be a gcc warning option for this - only an attribute that has to be
3053 put on the function's prototype. It seems that in Fedora Core 4 they have
3054 set this on a number of new functions. No doubt there will be more in due
3057 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3058 specify a transport. However, if an address that was verified by such a
3059 router was the subject of a callout, Exim crashed because it tried to
3060 read the rcpt_include_affixes from the non-existent transport. Now it
3061 just assumes that the setting of that option is false. This bug was
3062 introduced by 4.51/PH/31.
3064 PH/28 Changed -d+all to exclude +memory, because that information is very
3065 rarely of interest, but it makes the output a lot bigger. People tend to
3066 do -d+all out of habit.
3068 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3069 code in os-type was giving problems when libc.so lives in lib64, like on
3072 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3073 aren't the modern standard, and it seems that some systems' include files
3074 don't always have them. Exim was already checking for some of the newer
3075 ones like T_AAAA, and defining it itself. I've added checks for all the
3076 record types that Exim uses.
3078 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3079 not automatically generating a new one, as it is supposed to. This
3080 prevented TLS from working. If the file did exist, but contained invalid
3081 data, a new version was generated, as expected. It was only the case of a
3082 non-existent file that was broken.
3084 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3085 with a change in libdomainkeys > 0.64.
3087 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3088 from DNS. If the selector record carries the flag, it now has
3089 precedence over the domain-wide flag.
3091 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3093 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3094 the use of a port name that isn't defined in /etc/services) occurred, the
3095 message was deferred as in a normal delivery, and thus remained on the
3096 spool, instead of being failed because of the mua_wrapper setting. This
3097 is now fixed, and I tidied up some of the mua_wrapper messages at the
3100 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3101 the messages in a hash of arrays rather than using individual hashes.
3102 This is a bit cleaner and results in dramatic memory savings, albeit
3103 at a slight CPU cost.
3105 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3106 as requested by Marc Sherman.
3108 SC/10 Eximstats: added histograms for user specified patterns as requested
3111 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3113 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3114 fopen() in the content-scanning modules that did not already have it.
3120 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3121 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3123 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3125 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3127 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3128 to test Sieve filters that use "vacation".
3130 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3131 that changes the way the GnuTLS parameters are stored in the cache file.
3132 The new format can be generated externally. For backward compatibility,
3133 if the data in the cache doesn't make sense, Exim assumes it has read an
3134 old-format file, and it generates new data and writes a new file. This
3135 means that you can't go back to an older release without removing the
3138 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3139 work if there are any delivery delays because "one_time" forces the
3140 parent to be marked "delivered", so its unseen clone is never tried
3141 again. For this reason, Exim now forbids the simultaneous setting of
3144 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3145 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3146 entry if you want to know the details. The fix, however, neglected to
3147 consider the case when local delivery batching is involved. The test for
3148 "previously delivered" was not happening when checking to see if an
3149 address could be batched with a previous (undelivered) one; under
3150 certain circumstances this could lead to multiple deliveries to the same
3153 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3154 in its include files, and this causes problems building Exim.
3156 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3157 header_syntax) but Exim was just ignoring anything given after a slash.
3158 In particular, this caused confusion with an attempt to use "verify =
3159 reverse_host_lookup/defer_ok". An error is now given when options are
3160 supplied for verify items that do not have them. (Maybe reverse_host_
3161 lookup should have a defer_ok option, but that's a different point.)
3163 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3164 defined by RFC 821) to 2048, because there were problems with some AUTH
3165 commands, and RFC 1869 says the size should be increased for extended
3166 SMTP commands that take arguments.
3168 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3171 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3172 "unknown" error; now it says that the functionality isn't in the binary.
3174 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3175 an address' error message when a string expansion fails (syntax or
3176 whatever). Otherwise the password may appear in the log. Following change
3177 PH/42 below, there is no longer a chance of it appearing in a bounce
3180 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3182 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3183 of its addresses was ever tried. (Bugzilla bug #2.)
3185 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3186 the result incorrectly in the debug output. (It correctly added a newline
3187 to what was transported.)
3189 TF/01 Added $received_time.
3191 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3192 commented out examples of how to interface to a virus scanner and to
3193 SpamAssassin. Also added commented examples of av_scanner and
3194 spamd_address settings.
3196 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3197 and controls are allowed in which ACLs. There were a couple of minor
3198 errors. Some of the entries in the conditions table (which is a table of
3199 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3200 negation of where the condition IS allowed.
3202 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3204 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3205 header file does not have a version number, so I've had to invent a new
3206 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3207 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3208 radiusclient), but it was contributed by a Radius user.
3210 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3211 files or directories, for OpenSSL.
3213 PH/19 When an Exim process that is running as root has to create an Exim log
3214 file, it does so in a subprocess that runs as exim:exim so as to get the
3215 ownership right at creation (otherwise, other Exim processes might see
3216 the file with the wrong ownership). There was no test for failure of this
3217 fork() call, which would lead to the process getting stuck as it waited
3218 for a non-existent subprocess. Forks do occasionally fail when resources
3219 run out. I reviewed all the other calls to fork(); they all seem to check
3222 PH/20 When checking for unexpected SMTP input at connect time (before writing
3223 the banner), Exim was not dealing correctly with a non-positive return
3224 from the read() function. If the client had disconnected by this time,
3225 the result was a log entry for a synchronization error with an empty
3226 string after "input=" when read() returned zero. If read() returned -1
3227 (an event I could not check), uninitialized data bytes were printed.
3228 There were reports of junk text (parts of files, etc) appearing after
3231 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3233 PH/22 Added support for macro redefinition, and (re)definition in between
3234 driver and ACL definitions.
3236 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3237 forgetting to use the resulting value; it was using the unexpanded value.
3239 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3240 hadn't been configured. The fix is from Juergen Kreileder, who
3241 understands it better than I do:
3243 "Here's what I see happening with three configured cyrus_sasl
3244 authenticators configured (plain, login, cram-md5):
3246 On startup auth_cyrus_sasl_init() gets called for each of these.
3247 This means three calls to sasl_listmech() without a specified mech_list.
3248 => SASL tests which mechs of all available mechs actually work
3249 => three warnings about OTP not working
3250 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3252 With the patch, sasl_listmech() also gets called three times. But now
3253 SASL's mech_list option is set to the server_mech specified in the the
3254 authenticator. Or in other words, the answer from sasl_listmech()
3255 gets limited to just the mech you're testing for (which is different
3257 => the return list contains just 'plain' or 'login', 'cram-md5' or
3258 nothing depending on the value of ob->server_mech.
3260 I've just tested the patch: Authentication still works fine,
3261 unavailable mechs specified in the exim configuration are still
3262 caught, and the auth.log warnings about OTP are gone."
3264 PH/25 When debugging is enabled, the contents of the command line are added
3265 to the debugging output, even when log_selector=+arguments is not
3268 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3269 answer is "GNU", and only if the return is "GNU/something" is the answer
3272 PH/27 $acl_verify_message is now set immediately after the failure of a
3273 verification in an ACL, and so is available in subsequent modifiers. In
3274 particular, the message can be preserved by coding like this:
3276 warn !verify = sender
3277 set acl_m0 = $acl_verify_message
3279 Previously, $acl_verify_message was set only while expanding "message"
3280 and "log_message" when a very denied access.
3282 PH/28 Modified OS/os.c-Linux with
3284 -#ifndef OS_LOAD_AVERAGE
3285 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3287 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3288 nomenclature these days.)
3290 PH/29 Installed patch from the Sieve maintainer that adds the options
3291 sieve_useraddress and sieve_subaddress to the redirect router.
3293 PH/30 In these circumstances:
3294 . Two addresses routed to the same list of hosts;
3295 . First host does not offer TLS;
3296 . First host accepts first address;
3297 . First host gives temporary error to second address;
3298 . Second host offers TLS and a TLS session is established;
3299 . Second host accepts second address.
3300 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3301 and peerdn, if requested) that were in fact used only for the second
3304 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3305 attention to any local part prefix or suffix that was matched by the
3306 router that accepted the address. It now behaves in the same way as it
3307 does for delivery: the affixes are removed from the local part unless
3308 rcpt_include_affixes is set on the transport.
3310 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3311 timeout during the DATA phase of an incoming message.
3313 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3314 applied a patch sanctioned by the Sieve maintainer.
3316 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3317 the uid or gid is negative. A case of a negative gid caused this to be
3318 noticed. The fix allows for either to be negative.
3320 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3321 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3324 PH/36 The change PH/12 above was broken. Fixed it.
3326 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3327 the grounds that routing the same address twice would always produce the
3328 same answer. This might have been true once, but it is certainly no
3329 longer true now. Routing a child address may depend on the previous
3330 routing that produced that child. Some complicated redirection strategies
3331 went wrong when messages had multiple recipients, and made Exim's
3332 behaviour dependent on the order in which the addresses were given.
3334 I have moved the duplicate checking until after the routing is complete.
3335 Exim scans the addresses that are assigned to local and remote
3336 transports, and removes any duplicates. This means that more work will be
3337 done, as duplicates will always all be routed, but duplicates are
3338 presumably rare, so I don't expect this is of any significance.
3340 For deliveries to pipes, files, and autoreplies, the duplicate checking
3341 still happens during the routing process, since they are not going to be
3344 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3345 It corrects a timeout issue with spamd. This is Ian's comment: "The
3346 background is that sometimes spamd either never reads data from a
3347 connection it has accepted, or it never writes response data. The exiscan
3348 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3349 blindly assumes that writes won't block so it may never time out."
3351 PH/39 Allow G after quota size as well as K and M.
3353 PH/40 The value set for $authenticated_id in an authenticator may not contain
3354 binary zeroes or newlines because the value is written to log lines and
3355 to spool files. There was no check on this. Now the value is run through
3356 the string_printing() function so that such characters are converted to
3357 printable escape sequences.
3359 PH/41 $message_linecount is a new variable that contains the total number of
3360 lines in the message. Compare $body_linecount, which is the count for the
3363 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3364 bounce and delay warning messages, except in certain special cases, which
3367 (a) An SMTP error message from a remote host;
3368 (b) A message specified in a :fail: redirection;
3369 (c) A message specified in a "fail" command in a system filter;
3370 (d) A message specified in a FAIL return from the queryprogram router;
3371 (e) A message specified by the cannot_route_message router option.
3373 In these cases only, Exim does include the error details in bounce and
3374 warning messages. There are also a few cases where bland messages such
3375 as "unrouteable address" or "local delivery error" are given.
3377 PH/43 $value is now also set for the "else" part of a ${run expansion.
3379 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3380 being worked on, but at least Exim now implements the latest version to
3383 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3384 process to complete was treated as a delivery failure, a timeout while
3385 writing the message to the pipe was logged, but erroneously treated as a
3386 successful delivery. Such timeouts include transport filter timeouts. For
3387 consistency with the overall process timeout, these timeouts are now
3388 treated as errors, giving rise to delivery failures by default. However,
3389 there is now a new Boolean option for the pipe transport called
3390 timeout_defer, which, if set TRUE, converts the failures into defers for
3391 both kinds of timeout. A transport filter timeout is now identified in
3394 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3395 systems where "make" and "gmake" are different, calling "gmake" at top
3396 level broke things. I've arranged for the value of $(MAKE) to be passed
3397 from the Makefile to this script so that it can call the same version of
3401 A note about Exim versions 4.44 and 4.50
3402 ----------------------------------------
3404 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3405 changes of various kinds. As a consequence, a big documentation update was
3406 needed. This delayed the release for rather longer than seemed good, especially
3407 in the light of a couple of (minor) security issues. Therefore, the changes
3408 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3409 release. So 4.44 and 4.50 are in effect two different branches that both start
3412 I have left the 4.50 change log unchanged; it contains all the changes since
3413 4.43. The change log for 4.44 is below; many of its items are identical to
3414 those for 4.50. This seems to be the most sensible way to preserve the
3415 historical information.
3421 1. Minor wording change to the doc/README.SIEVE file.
3423 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3424 computation of the current number of files was incorrect.
3426 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3427 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3430 4. Give more explanation in the error message when the command for a transport
3431 filter fails to execute.
3433 5. There are several places where Exim runs a non-Exim command in a
3434 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3435 was being done only for the command run by the queryprogram router. It is
3436 now done for all such subprocesses. The other cases are: ${run, transport
3437 filters, and the commands run by the lmtp and pipe transports.
3439 6. Added CONFIGURE_GROUP build-time option.
3441 7. Some older OS have a limit of 256 on the maximum number of file
3442 descriptors. Exim was using setrlimit() to set 1000 as a large value
3443 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3444 systems. I've change it so that if it can't get 1000, it tries for 256.
3446 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3447 was an oversight, and furthermore, ever since the addition of extra
3448 controls (e.g. 4.43/32), the checks on when to allow different forms of
3449 "control" were broken. There should now be diagnostics for all cases when a
3450 control that does not make sense is encountered.
3452 9. Added the /retain_sender option to "control=submission".
3454 10. $recipients is now available in the predata ACL (oversight).
3456 11. Tidy the search cache before the fork to do a delivery from a message
3457 received from the command line. Otherwise the child will trigger a lookup
3458 failure and thereby defer the delivery if it tries to use (for example) a
3459 cached ldap connection that the parent has called unbind on.
3461 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3462 of $address_data from the recipient verification was clobbered by the
3463 sender verification.
3465 13. The value of address_data from a sender verification is now available in
3466 $sender_address_data in subsequent conditions in the ACL statement.
3468 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3470 15. Added a new option "connect=<time>" to callout options, to set a different
3473 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3474 was its contents. (It was OK if the option was not defined at all.)
3476 17. A "Completed" log line is now written for messages that are removed from
3477 the spool by the -Mrm option.
3479 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3480 information about exactly what failed.
3482 19. Added -dd to debug only the daemon process.
3484 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3485 handles timeouts, both on the server side and network timeouts. Renamed the
3486 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3488 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3489 It is now set to "smtps".
3491 22. $host_address is now set to the target address during the checking of
3492 ignore_target_hosts.
3494 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3495 being passed; this would have caused $sender_host_name to have been used if
3496 matching the list had actually called for a host name (not very likely,
3497 since this list is usually IP addresses). A host name is now passed as
3500 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3501 code that specifies a non-restarting handler (typically sigaction() in
3502 modern systems) in an attempt to fix a rare and obscure crash bug.
3504 25. Narrowed the window for a race in the daemon that could cause it to ignore
3505 SIGCHLD signals. This is not a major problem, because they are used only to
3506 wake it up if nothing else does.
3508 26. A malformed maildirsize file could cause Exim to calculate negative values
3509 for the mailbox size or file count. Odd effects could occur as a result.
3510 The maildirsize information is now recalculated if the size or filecount
3513 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3514 support for a long time. Removed HAVE_SYS_VFS_H.
3516 28. Installed the latest version of exipick from John Jetmore.
3518 29. In an address list, if the pattern was not a regular expression, an empty
3519 subject address (from a bounce message) matched only if the pattern was an
3520 empty string. Non-empty patterns were not even tested. This was the wrong
3521 because it is perfectly reasonable to use an empty address as part of a
3522 database query. An empty address is now tested by patterns that are
3523 lookups. However, all the other forms of pattern expect the subject to
3524 contain a local part and a domain, and therefore, for them, an empty
3525 address still always fails if the pattern is not itself empty.
3527 30. Exim went into a mad DNS loop when attempting to do a callout where the
3528 host was specified on an smtp transport, and looking it up yielded more
3529 than one IP address.
3531 31. Re-factored the code for checking spool and log partition space into a
3532 function that finds that data and another that does the check. The former
3533 is then used to implement four new variables: $spool_space, $log_space,
3534 $spool_inodes, and $log_inodes.
3536 32. The RFC2047 encoding function was originally intended for short strings
3537 such as real names; it was not keeping to the 75-character limit for
3538 encoded words that the RFC imposes. It now respects the limit, and
3539 generates multiple encoded words if necessary. To be on the safe side, I
3540 have increased the buffer size for the ${rfc2047: expansion operator from
3543 33. It is now permitted to omit both strings after an "if" condition; if the
3544 condition is true, the result is "true". As before, when the second string
3545 is omitted, a false condition yields an empty string. This makes it less
3546 cumbersome to write custom ACL and router conditions.
3548 34. Failure to deliver a bounce message always caused it to be frozen, even if
3549 there was an errors_to setting on the router. The errors_to setting is now
3552 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3553 canonical form (fully expanded) before being placed in
3554 $sender_host_address.
3556 36. The table in the code that translates DNS record types into text (T_A to
3557 "A" for instance) was missing entries for NS and CNAME. It is just possible
3558 that this could have caused confusion if both these types were looked up
3559 for the same domain, because the text type is used as part of Exim's
3560 per-process caching. But the chance of anyone hitting this buglet seems
3563 37. The dnsdb lookup has been extended in a number of ways.
3565 (1) There is a new type, "zns", which walks up the domain tree until it
3566 finds some nameserver records. It should be used with care.
3568 (2) There is a new type, "mxh", which is like "mx" except that it returns
3569 just the host names, not the priorities.
3571 (3) It is now possible to give a list of domains (or IP addresses) to be
3572 looked up. The behaviour when one of the lookups defers can be
3573 controlled by a keyword.
3575 (4) It is now possible to specify the separator character for use when
3576 multiple records are returned.
3578 38. The dnslists ACL condition has been extended: it is now possible to supply
3579 a list of IP addresses and/or domains to be looked up in a particular DNS
3582 39. Added log_selector=+queue_time_overall.
3584 40. When running the queue in the test harness, wait just a tad after forking a
3585 delivery process, to get repeatability of debugging output.
3587 41. Include certificate and key file names in error message when GnuTLS fails
3588 to set them up, because the GnuTLS error message doesn't include the name
3589 of the failing file when there is a problem reading it.
3591 42. Allow both -bf and -bF in the same test run.
3593 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3595 44. The "Exiscan patch" is now merged into the mainline Exim source.
3597 45. Sometimes the final signoff response after QUIT could fail to get
3598 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3599 < 0 before doing a fflush(). This bug looks as though it goes back to the
3600 introduction of TLS in release 3.20, but "sometimes" must have been rare
3601 because the tests only now provoked it.
3603 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3604 (this can affect the format of dates).
3606 47. exim_tidydb, when checking for the continued existence of a message for
3607 which it has found a message-specific retry record, was not finding
3608 messages that were in split spool directories. Consequently, it was
3609 deleting retry records that should have stayed in existence.
3611 48. Steve fixed some bugs in eximstats.
3613 49. The SPA authentication driver was not abandoning authentication and moving
3614 on to the next authenticator when an expansion was forced to fail,
3615 contradicting the general specification for all authenticators. Instead it
3616 was generating a temporary error. It now behaves as specified.
3618 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3619 (the order specifies the preference for clients). The order is now AES256,
3620 AES128, 3DES, ARCFOUR128.
3622 51. Small patch to Sieve code - explicitly set From: when generating an
3625 52. Exim crashed if a remote delivery caused a very long error message to be
3626 recorded - for instance if somebody sent an entire SpamAssassin report back
3627 as a large number of 550 error lines. This bug was coincidentally fixed by
3628 increasing the size of one of Exim's internal buffers (big_buffer) that
3629 happened as part of the Exiscan merge. However, to be on the safe side, I
3630 have made the code more robust (and fixed the comments that describe what
3633 53. Now that there can be additional text after "Completed" in log lines (if
3634 the queue_time_overall log selector is set), a one-byte patch to exigrep
3635 was needed to allow it to recognize "Completed" as not the last thing in
3638 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3639 patch that reportedly fixes this has been added. I am not expert enough to
3640 create a test for it. This is what the patch creator wrote:
3642 "I found a little strange behaviour of ldap code when working with
3643 Windows 2003 AD Domain, where users was placed in more than one
3644 Organization Units. When I tried to give exim partial DN, the exit code
3645 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3646 But simultaneously result of request was absolutely normal ldap result,
3647 so I produce this patch..."
3649 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3650 REFERENCE, so I have modified the code to exclude the patch when that macro
3653 55. Some experimental protocols are using DNS PTR records for new purposes. The
3654 keys for these records are domain names, not reversed IP addresses. The
3655 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3656 leaves it alone. Component reversal etc. now happens only for IP addresses.
3659 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3661 57. Double the size of the debug message buffer (to 2048) so that more of very
3662 long debug lines gets shown.
3664 58. The exicyclog utility now does better if the number of log files to keep
3665 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3667 59. Two changes related to the smtp_active_hostname option:
3669 (1) $smtp_active_hostname is now available as a variable.
3670 (2) The default for smtp_banner uses $smtp_active_hostname instead
3671 of $primary_hostname.
3673 60. The host_aton() function is supposed to be passed a string that is known
3674 to be a valid IP address. However, in the case of IPv6 addresses, it was
3675 not checking this. This is a hostage to fortune. Exim now panics and dies
3676 if the condition is not met. A case was found where this could be provoked
3677 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3678 components; fortuitously, this particular loophole had already been fixed
3679 by change 4.50/55 above.
3681 If there are any other similar loopholes, the new check in host_aton()
3682 itself should stop them being exploited. The report I received stated that
3683 data on the command line could provoke the exploit when Exim was running as
3684 exim, but did not say which command line option was involved. All I could
3685 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3686 running as the user.
3689 61. There was a buffer overflow vulnerability in the SPA authentication code
3690 (which came originally from the Samba project). I have added a test to the
3691 spa_base64_to_bits() function which I hope fixes it.
3694 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3695 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3697 63. The daemon start-up calls getloadavg() while still root for those OS that
3698 need the first call to be done as root, but it missed one case: when
3699 deliver_queue_load_max is set with deliver_drop_privilege. This is
3700 necessary for the benefit of the queue runner, because there is no re-exec
3701 when deliver_drop_privilege is set.
3703 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3704 This has been fixed.
3706 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3707 was in use, was not putting the data itself into the right store pool;
3708 consequently, it could be overwritten for a subsequent message in the same
3709 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3712 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3714 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3715 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3716 the function were treating the return as a boolean value, which happened to
3717 work because 0=false and not-0=true, but is not correct code.
3719 68. The host_aton() function was not handling scoped IPv6 addresses (those
3720 with, for example, "%eth0" on the end) correctly.
3722 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3723 negated items (that is, ~something) in unsigned ints. Some compilers
3724 apparently mutter when there is no cast.
3726 70. If an address verification called from an ACL failed, and did not produce a
3727 user-specific message (i.e. there was only a "system" message), nothing was
3728 put in $acl_verify_message. In this situation, it now puts the system
3731 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3732 session; change 4.31/43 added the unwanted input to the log line - except
3733 that it did not do this in the start of session case. It now does.
3735 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3736 This is wrong and can cause the other end to generate a synchronization
3737 error if it is another Exim or anything else that does the synchronization
3738 check. A QUIT command is no longer sent after a timeout.
3740 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3741 during host lookups.
3743 74. The defer_ok option of callout verification was not working if it was used
3744 when verifying addresses in header lines, that is, for this case:
3746 verify = header_sender/callout=defer_ok
3748 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3749 those file descriptors could be used for SMTP connections. If anything
3750 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3751 could be sent to the SMTP client, causing chaos. The daemon now opens
3752 stdin, stdout, and stderr to /dev/null when it puts itself into the
3755 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3756 log by default. The user can override this with suitable Perl magic.
3758 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3759 the log message when discard triggers, was not working for the DATA ACL or
3760 for the non-SMTP ACL.
3762 78. Error message wording change in sieve.c.
3764 79. If smtp_accept_max_per_host was set, the number of connections could be
3765 restricted to fewer than expected, because the daemon was trying to set up
3766 a new connection before checking whether the processes handling previous
3767 connections had finished. The check for completed processes is now done
3768 earlier. On busy systems, this bug wouldn't be noticed because something
3769 else would have woken the daemon, and it would have reaped the completed
3772 80. If a message was submitted locally by a user whose login name contained one
3773 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3774 It caused a spool format error. I have fixed the spool reading code. A
3775 related problem was that the "from" clause in the Received: line became
3776 illegal because of the space(s). It is now covered by ${quote_local_part.
3778 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3781 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3783 83. Patch from Sieve maintainer to fix unterminated string problem in
3784 "vacation" handling.
3786 84. Some minor changes to the Linux configuration files to help with other
3787 OS variants using glibc.
3789 85. One more patch for Sieve to update vacation handling to latest spec.
3792 ----------------------------------------------------
3793 See the note above about the 4.44 and 4.50 releases.
3794 ----------------------------------------------------
3800 1. Change 4.43/35 introduced a bug that caused file counts to be
3801 incorrectly computed when quota_filecount was set in an appendfile
3804 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3805 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3808 3. Give more explanation in the error message when the command for a transport
3809 filter fails to execute.
3811 4. There are several places where Exim runs a non-Exim command in a
3812 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3813 was being done only for the command run by the queryprogram router. It is
3814 now done for all such subprocesses. The other cases are: ${run, transport
3815 filters, and the commands run by the lmtp and pipe transports.
3817 5. Some older OS have a limit of 256 on the maximum number of file
3818 descriptors. Exim was using setrlimit() to set 1000 as a large value
3819 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3820 systems. I've change it so that if it can't get 1000, it tries for 256.
3822 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3823 was an oversight, and furthermore, ever since the addition of extra
3824 controls (e.g. 4.43/32), the checks on when to allow different forms of
3825 "control" were broken. There should now be diagnostics for all cases when a
3826 control that does not make sense is encountered.
3828 7. $recipients is now available in the predata ACL (oversight).
3830 8. Tidy the search cache before the fork to do a delivery from a message
3831 received from the command line. Otherwise the child will trigger a lookup
3832 failure and thereby defer the delivery if it tries to use (for example) a
3833 cached ldap connection that the parent has called unbind on.
3835 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3836 of $address_data from the recipient verification was clobbered by the
3837 sender verification.
3839 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3840 was its contents. (It was OK if the option was not defined at all.)
3842 11. A "Completed" log line is now written for messages that are removed from
3843 the spool by the -Mrm option.
3845 12. $host_address is now set to the target address during the checking of
3846 ignore_target_hosts.
3848 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3849 being passed; this would have caused $sender_host_name to have been used if
3850 matching the list had actually called for a host name (not very likely,
3851 since this list is usually IP addresses). A host name is now passed as
3854 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3855 code that specifies a non-restarting handler (typically sigaction() in
3856 modern systems) in an attempt to fix a rare and obscure crash bug.
3858 15. Narrowed the window for a race in the daemon that could cause it to ignore
3859 SIGCHLD signals. This is not a major problem, because they are used only to
3860 wake it up if nothing else does.
3862 16. A malformed maildirsize file could cause Exim to calculate negative values
3863 for the mailbox size or file count. Odd effects could occur as a result.
3864 The maildirsize information is now recalculated if the size or filecount
3867 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3868 support for a long time. Removed HAVE_SYS_VFS_H.
3870 18. Updated exipick to current release from John Jetmore.
3872 19. Allow an empty sender to be matched against a lookup in an address list.
3873 Previously the only cases considered were a regular expression, or an
3876 20. Exim went into a mad DNS lookup loop when doing a callout where the
3877 host was specified on the transport, if the DNS lookup yielded more than
3880 21. The RFC2047 encoding function was originally intended for short strings
3881 such as real names; it was not keeping to the 75-character limit for
3882 encoded words that the RFC imposes. It now respects the limit, and
3883 generates multiple encoded words if necessary. To be on the safe side, I
3884 have increased the buffer size for the ${rfc2047: expansion operator from
3887 22. Failure to deliver a bounce message always caused it to be frozen, even if
3888 there was an errors_to setting on the router. The errors_to setting is now
3891 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3892 canonical form (fully expanded) before being placed in
3893 $sender_host_address.
3895 24. Updated eximstats to version 1.33
3897 25. Include certificate and key file names in error message when GnuTLS fails
3898 to set them up, because the GnuTLS error message doesn't include the name
3899 of the failing file when there is a problem reading it.
3901 26. Expand error message when OpenSSL has problems setting up cert/key files.
3904 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3905 (this can affect the format of dates).
3907 28. exim_tidydb, when checking for the continued existence of a message for
3908 which it has found a message-specific retry record, was not finding
3909 messages that were in split spool directories. Consequently, it was
3910 deleting retry records that should have stayed in existence.
3912 29. eximstats updated to version 1.35
3913 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3914 1.35 - bugfix such that pie charts by volume are generated correctly
3916 30. The SPA authentication driver was not abandoning authentication and moving
3917 on to the next authenticator when an expansion was forced to fail,
3918 contradicting the general specification for all authenticators. Instead it
3919 was generating a temporary error. It now behaves as specified.
3921 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3922 (the order specifies the preference for clients). The order is now AES256,
3923 AES128, 3DES, ARCFOUR128.
3925 31. Small patch to Sieve code - explicitly set From: when generating an
3928 32. Exim crashed if a remote delivery caused a very long error message to be
3929 recorded - for instance if somebody sent an entire SpamAssassin report back
3930 as a large number of 550 error lines. This bug was coincidentally fixed by
3931 increasing the size of one of Exim's internal buffers (big_buffer) that
3932 happened as part of the Exiscan merge. However, to be on the safe side, I
3933 have made the code more robust (and fixed the comments that describe what
3936 33. Some experimental protocols are using DNS PTR records for new purposes. The
3937 keys for these records are domain names, not reversed IP addresses. The
3938 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3939 leaves it alone. Component reversal etc. now happens only for IP addresses.
3942 34. The host_aton() function is supposed to be passed a string that is known
3943 to be a valid IP address. However, in the case of IPv6 addresses, it was
3944 not checking this. This is a hostage to fortune. Exim now panics and dies
3945 if the condition is not met. A case was found where this could be provoked
3946 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3947 components; fortuitously, this particular loophole had already been fixed
3948 by change 4.50/55 or 4.44/33 above.
3950 If there are any other similar loopholes, the new check in host_aton()
3951 itself should stop them being exploited. The report I received stated that
3952 data on the command line could provoke the exploit when Exim was running as
3953 exim, but did not say which command line option was involved. All I could
3954 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3955 running as the user.
3958 35. There was a buffer overflow vulnerability in the SPA authentication code
3959 (which came originally from the Samba project). I have added a test to the
3960 spa_base64_to_bits() function which I hope fixes it.
3963 36. The daemon start-up calls getloadavg() while still root for those OS that
3964 need the first call to be done as root, but it missed one case: when
3965 deliver_queue_load_max is set with deliver_drop_privilege. This is
3966 necessary for the benefit of the queue runner, because there is no re-exec
3967 when deliver_drop_privilege is set.
3969 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3970 was in use, was not putting the data itself into the right store pool;
3971 consequently, it could be overwritten for a subsequent message in the same
3972 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3975 38. Sometimes the final signoff response after QUIT could fail to get
3976 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3977 < 0 before doing a fflush(). This bug looks as though it goes back to the
3978 introduction of TLS in release 3.20, but "sometimes" must have been rare
3979 because the tests only now provoked it.
3985 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3986 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3987 Now it returns an integer. A number of calls were still expecting a T/F
3988 return. Fortuitously, in all cases, the tests worked in OK situations,
3989 which is the norm. However, things would have gone wrong on any write
3990 failures on the smtp file descriptor. This function is used when sending
3991 messages over SMTP and also when doing verify callouts.
3993 2. When Exim is called to do synchronous delivery of a locally submitted
3994 message (the -odf or -odi options), it no longer closes stderr before doing
3997 3. Implemented the mua_wrapper option.
3999 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4001 5. Implemented the functions header_remove(), header_testname(),
4002 header_add_at_position(), and receive_remove_recipient(), and exported them
4005 6. If an ACL "warn" statement specified the addition of headers, Exim already
4006 inserted X-ACL-Warn: at the start if there was no header name. However, it
4007 was not making this test for the second and subsequent header lines if
4008 there were newlines in the string. This meant that an invalid header could
4009 be inserted if Exim was badly configured.
4011 7. Allow an ACL "warn" statement to add header lines at the start or after all
4012 the Received: headers, as well as at the end.
4014 8. Added the rcpt_4xx retry error code.
4016 9. Added postmaster_mailfrom=xxx to callout verification option.
4018 10. Added mailfrom=xxxx to the callout verification option, for verify=
4021 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4022 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4024 12. Inserted some casts to stop certain compilers warning when using pointer
4025 differences as field lengths or precisions in printf-type calls (mostly
4026 affecting debugging statements).
4028 13. Added optional readline() support for -be (dynamically loaded).
4030 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4031 same clock tick as a message's arrival, so that its received time was the
4032 same as the "first fail" time on the retry record, and that message
4033 remained on the queue past the ultimate address timeout, every queue runner
4034 would try a delivery (because it was past the ultimate address timeout) but
4035 after another failure, the ultimate address timeout, which should have then
4036 bounced the address, did not kick in. This was a "< instead of <=" error;
4037 in most cases the first failure would have been in the next clock tick
4038 after the received time, and all would be well.
4040 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4041 being recognized when the domain list was tested by the match_domain
4042 condition in an expansion string.
4044 16. Added the ${str2b64: operator.
4046 17. Exim was always calling setrlimit() to set a large limit for the number of
4047 processes, without checking whether the existing limit was already
4048 adequate. (It did check for the limit on file descriptors.) Furthermore,
4049 errors from getrlimit() and setrlimit() were being ignored. Now they are
4050 logged to the main and panic logs, but Exim does carry on, to try to do its
4051 job under whatever limits there are.
4053 18. Imported PCRE 5.0.
4055 19. Trivial typo in log message " temporarily refused connection" (the leading
4058 20. If the log selector return_path_on_delivery was set and an address was
4059 redirected to /dev/null, the delivery process crashed because it assumed
4060 that a return path would always be set for a "successful" delivery. In this
4061 case, the whole delivery is bypassed as an optimization, and therefore no
4064 21. Internal re-arrangement: the function for sending a challenge and reading
4065 a response while authentication was assuming a zero-terminated challenge
4066 string. It's now changed to take a pointer and a length, to allow for
4067 binary data in such strings.
4069 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4071 23. Exim was not respecting finduser_retries when seeking the login of the
4072 uid under which it was called; it was always trying 10 times. (The default
4073 setting of finduser_retries is zero.) Also, it was sleeping after the final
4074 failure, which is pointless.
4076 24. Implemented tls_on_connect_ports.
4078 25. Implemented acl_smtp_predata.
4080 26. If the domain in control=submission is set empty, Exim assumes that the
4081 authenticated id is a complete email address when it generates From: or
4082 Sender: header lines.
4084 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4085 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4086 chown and chgrp in /bin and hostname in /usr/bin.
4088 28. Exim was keeping the "process log" file open after each use, just as it
4089 does for the main log. This opens the possibility of it remaining open for
4090 long periods when the USR1 signal hits a daemon. Occasional processlog
4091 errors were reported, that could have been caused by this. Anyway, it seems
4092 much more sensible not to leave this file open at all, so that is what now
4095 29. The long-running daemon process does not normally write to the log once it
4096 has entered its main loop, and it closes the log before doing so. This is
4097 so that log files can straightforwardly be renamed and moved. However,
4098 there are a couple of unusual error situations where the daemon does write
4099 log entries, and I had neglected to close the log afterwards.
4101 30. The text of an SMTP error response that was received during a remote
4102 delivery was being truncated at 512 bytes. This is too short for some of
4103 the long messages that one sometimes sees. I've increased the limit to
4106 31. It is now possible to make retry rules that apply only when a message has a
4107 specific sender, in particular, an empty sender.
4109 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4110 it possible to be selective about when SMTP synchronization is enforced.
4112 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4114 32. Implemented hosts_connection_nolog.
4116 33. Added an ACL for QUIT.
4118 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4121 35. Added mailbox_size and mailbox_filecount to appendfile.
4123 36. Added control = no_multiline_responses to ACLs.
4125 37. There was a bug in the logic of the code that waits for the clock to tick
4126 in the case where the clock went backwards by a substantial amount such
4127 that the microsecond fraction of "now" was more than the microsecond
4128 fraction of "then" (but the whole seconds number was less).
4130 38. Added support for the libradius Radius client library this is found on
4131 FreeBSD (previously only the radiusclient library was supported).
4137 1. When certain lookups returned multiple values in the form name=value, the
4138 quoting of the values was not always being done properly. Specifically:
4139 (a) If the value started with a double quote, but contained no whitespace,
4141 (b) If the value contained whitespace other than a space character (i.e.
4142 tabs or newlines or carriage returns) it was not quoted.
4143 This fix has been applied to the mysql and pgsql lookups by writing a
4144 separate quoting function and calling it from the lookup code. The fix
4145 should probably also be applied to nisplus, ibase and oracle lookups, but
4146 since I cannot test any of those, I have not disturbed their existing code.
4148 2. A hit in the callout cache for a specific address caused a log line with no
4149 reason for rejecting RCPT. Now it says "Previous (cached) callout
4150 verification failure".
4152 3. There was an off-by-one bug in the queryprogram router. An over-long
4153 return line was truncated at 256 instead of 255 characters, thereby
4154 overflowing its buffer with the terminating zero. As well as fixing this, I
4155 have increased the buffer size to 1024 (and made a note to document this).
4157 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4158 when Exim is waiting for an SMTP response from a remote server, Exim
4159 restarts its select() call on the socket, thereby resetting its timeout.
4160 This is not a problem when such interrupts are rare. Somebody set up a cron
4161 job to run exiwhat every 2 minutes, which is less than the normal select()
4162 timeout (5 or 10 minutes). This meant that the select() timeout never
4163 kicked in because it was always reset. I have fixed this by comparing the
4164 time when an interrupt arrives with the time at the start of the first call
4165 to select(). If more time than the timeout has elapsed, the interrupt is
4166 treated as a timeout.
4168 5. Some internal re-factoring in preparation for the addition of Sieve
4169 extensions (by MH). In particular, the "personal" test is moved to a
4170 separate function, and given an option for scanning Cc: and Bcc: (which is
4171 not set for Exim filters).
4173 6. When Exim created an email address using the login of the caller as the
4174 local part (e.g. when creating a From: or Sender: header line), it was not
4175 quoting the local part when it contained special characters such as @.
4177 7. Installed new OpenBSD configuration files.
4179 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4180 try to make them clearer.
4182 9. Callout options, other than the timeout value, were being ignored when
4183 verifying sender addresses in header lines. For example, when using
4185 verify = header_sender/callout=no_cache
4187 the cache was (incorrectly) being used.
4189 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4190 only the Cygwin environment.
4192 11. When return_path_on_delivery was set as a log selector, if different remote
4193 addresses in the same message used different return paths and parallel
4194 remote delivery occurred, the wrong values would sometimes be logged.
4195 (Whenever a remote delivery process finished, the return path value from
4196 the most recently started remote delivery process was logged.)
4198 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4199 lines when AUTH and/or TLS are in use. This is the "received protocol"
4200 field. Exim used to use "asmtp" for authenticated SMTP, without any
4201 indication (in the protocol name) for TLS use. Now it follows the RFC and
4202 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4203 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4204 names appear in log lines as well as in Received: header lines.
4206 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4207 extensions, and comparison tests, and to fix some bugs.
4209 14. Changes to the "personal" filter test:
4211 (1) The test was buggy in that it was just doing the equivalent of
4212 "contains" tests on header lines. For example, if a user's address was
4213 anne@some.where, the "personal" test would incorrectly be true for
4215 To: susanne@some.where
4217 This test is now done by extracting each address from the header in turn,
4218 and checking the entire address. Other tests that are part of "personal"
4219 are now done using regular expressions (for example, to check local parts
4220 of addresses in From: header lines).
4222 (2) The list of non-personal local parts in From: addresses has been
4223 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4224 taken from the Sieve specification recommendations.
4226 (3) If the message contains any header line starting with "List-" it is
4227 treated as non-personal.
4229 (4) The test for "circular" in the Subject: header line has been removed
4230 because it now seems ill-conceived.
4232 15. Minor typos in src/EDITME comments corrected.
4234 16. Installed latest exipick from John Jetmore.
4236 17. If headers_add on a router specified a text string that was too long for
4237 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4238 of string_sprintf() is now avoided.
4240 18. $message_body_size was not set (it was always zero) when running the DATA
4241 ACL and the local_scan() function.
4243 19. For the "mail" command in an Exim filter, no default was being set for
4244 the once_repeat time, causing a random time value to be used if "once" was
4245 specified. (If the value happened to be <= 0, no repeat happened.) The
4246 default is now 0s, meaning "never repeat". The "vacation" command was OK
4247 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4248 (I found it when inspecting the code).
4250 20. There is now an overall timeout for performing a callout verification. It
4251 defaults to 4 times the callout timeout, which applies to individual SMTP
4252 commands during the callout. The overall timeout applies when there is more
4253 than one host that can be tried. The timeout is checked before trying the
4254 next host. This prevents very long delays if there are a large number of
4255 hosts and all are timing out (e.g. when the network connections are timing
4256 out). The value of the overall timeout can be changed by specifying an
4257 additional sub-option for "callout", called "maxwait". For example:
4259 verify = sender/callout=5s,maxwait=20s
4261 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4262 to the end before writing, but this should make it even safer).
4264 22. Exim was forgetting that it had advertised PIPELINING for the second and
4265 subsequent messages on an SMTP connection. It was also not resetting its
4266 memory on STARTTLS and an internal HELO.
4268 23. When Exim logs an SMTP synchronization error within a session, it now
4269 records whether PIPELINING has been advertised or not.
4271 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4272 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4273 rather than long int.
4275 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4277 26. Added the never_mail option to autoreply.
4283 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4284 crash if the getsockname() call failed; this can happen if a connection is
4285 closed very soon after it is established. The problem was simply in the
4286 order in which certain operations were done, causing Exim to try to write
4287 to the SMTP stream before it had set up the file descriptor. The bug has
4288 been fixed by making things happen in the correct order.
4294 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4295 before the connection was closed, thus losing the rejection response.
4297 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4298 some early Solaris releases, but causes trouble in current releases where
4299 socklen_t is defined.
4301 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4304 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4307 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4308 that releasing the top of it at the end releases what was used for sub-
4309 expansions (unless the block got too big). However, discard this block if
4310 the first thing is a variable or header, so that we can use its block when
4311 it is dynamic (useful for very large $message_headers, for example).
4313 6. Lookups now cache *every* query, not just the most recent. A new, separate
4314 store pool is used for this. It can be recovered when all lookup caches are
4315 flushed. Lookups now release memory at the end of their result strings.
4316 This has involved some general refactoring of the lookup sources.
4318 7. Some code has been added to the store_xxx() functions to reduce the amount
4319 of flapping under certain conditions.
4321 8. log_incoming_interface used to affect only the <= reception log lines. Now
4322 it causes the local interface and port to be added to several more SMTP log
4323 lines, for example "SMTP connection from", and rejection lines.
4325 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4327 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4329 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4330 could be overwritten at the end of the current message (or the start of a
4331 new message if it was set in a HELO ACL). The value is now preserved for
4332 the duration of the SMTP connection.
4334 12. If a transport had a headers_rewrite setting, and a matching header line
4335 contained an unqualified address, that address was qualified, even if it
4336 did not match any rewriting rules. The underlying bug was that the values
4337 of the flags that permit the existence of unqualified sender and recipient
4338 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4339 non-local messages, and by -bnq for local messages) were not being
4340 preserved with the message after it was received.
4342 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4343 "next input=" as part of the text comprising the host identity instead of
4344 the correct text. The code was using the same buffer for two different
4345 strings. However, depending on which order the printing function evaluated
4346 its arguments, the bug did not always show up. Under Linux, for example, my
4347 test suite worked just fine.
4349 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4350 doesn't work with some older versions of Perl. It has been changed to "my",
4351 which in any case is probably the better facility to use.
4353 15. A really picky compiler found some instances of statements for creating
4354 error messages that either had too many or two few arguments for the format
4357 16. The size of the buffer for calls to the DNS resolver has been increased
4358 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4359 for addresses that have a lot of PTR records. This alleviates a problem; it
4360 does not fully solve it.
4362 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4363 buffer now truncates the list and logs the incident, which is the same
4364 action as happens when Exim is looking up a host name and its aliases.
4365 Previously in this situation something unpredictable would happen;
4366 sometimes it was "internal error: store_reset failed".
4368 18. If a server dropped the connection unexpectedly when an Exim client was
4369 using GnuTLS and trying to read a response, the client delivery process
4370 crashed while trying to generate an error log message.
4372 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4373 string, for example:
4375 warn message = H1: something\nH2: something
4377 the text was added as a single header line from Exim's point of view
4378 though it ended up OK in the delivered message. However, searching for the
4379 second and subsequent header lines using $h_h2: did not work. This has been
4380 fixed. Similarly, if a system filter added multiple headers in this way,
4381 the routers could not see them.
4383 20. Expanded the error message when iplsearch is called with an invalid key to
4384 suggest using net-iplsearch in a host list.
4386 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4387 ACLs are no longer actually imposed (and a message to that effect is
4390 22. If a "gecos" field in a passwd entry contained escaped characters, in
4391 particular, if it contained a \" sequence, Exim got it wrong when building
4392 a From: or a Sender: header from that name. A second bug also caused
4393 incorrect handling when an unquoted " was present following a character
4394 that needed quoting.
4396 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4397 was not being matched caselessly.
4399 24. Arranged for all hyphens in the exim.8 source to be escaped with
4402 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4403 properly. Recipient callout cache records were still being keyed to include
4404 the sender, even when use_sender was set false. This led to far more
4405 callouts that were necessary. The sender is no longer included in the key
4406 when use_sender is false.
4408 26. Added "control = submission" modifier to ACLs.
4410 27. Added the ${base62d: operator to decode base 62 numbers.
4412 28. dnsdb lookups can now access SRV records.
4414 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4415 the configuration file.
4417 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4418 (-v) mode. This makes the output for a verbose queue run more intelligible.
4420 31. Added a use_postmaster feature to recipient callouts.
4422 32. Added the $body_zerocount variable, containing the number of binary zero
4423 bytes in the message body.
4425 33. The time of last modification of the "new" subdirectory is now used as the
4426 "mailbox time last read" when there is a quota error for a maildir
4429 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4431 35. Added +ignore_unknown as a special item in host lists.
4433 36. Code for decoding IPv6 addresses in host lists is now included, even if
4434 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4435 address was recognized as an IP address, but was then not correctly decoded
4436 into binary, causing unexpected and incorrect effects when compared with
4443 1. Very minor rewording of debugging text in manualroute to say "list of
4444 hosts" instead of "hostlist".
4446 2. If verify=header_syntax was set, and a header line with an unqualified
4447 address (no domain) and a large number of spaces between the end of the
4448 name and the colon was received, the reception process suffered a buffer
4449 overflow, and (when I tested it) crashed. This was caused by some obsolete
4450 code that should have been removed. The fix is to remove it!
4452 3. When running in the test harness, delay a bit after writing a bounce
4453 message to get a bit more predictability in the log output.
4455 4. Added a call to search_tidyup() just before forking a reception process. In
4456 theory, someone could use a lookup in the expansion of smtp_accept_max_
4457 per_host which, without the tidyup, could leave open a database connection.
4459 5. Added the variables $recipient_data and $sender_data which get set from a
4460 lookup success in an ACL "recipients" or "senders" condition, or a router
4461 "senders" option, similar to $domain_data and $local_part_data.
4463 6. Moved the writing of debug_print from before to after the "senders" test
4466 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4467 problems for message scanning, either using a data ACL, or using
4468 local_scan() because the Received: header was not generated till after they
4469 were called (in order to set the time as the time of reception completion).
4470 I have revised the way this works. The header is now generated after the
4471 body is received, but before the ACL or local_scan() are called. After they
4472 are run, the timestamp in the header is updated.
4478 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4479 before starting a queue runner without re-exec. This happened only when
4480 deliver_drop_privilege was set or when the Exim user was set to root. The
4481 effect of the bug was that timeouts during subsequent deliveries caused
4482 crashes instead of being properly handled. The handler is now left at its
4483 default (and expected) setting.
4485 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4486 message, again when deliver_drop_privilege is set or Exim is run as root.
4487 The bug described in (1) was not present in this case, but the tidying up
4488 of the other signals was missing. I have made the two cases consistent.
4490 3. The ignore_target_hosts setting on a manualroute router was being ignored
4491 for hosts that were looked up using the /MX notation.
4493 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4496 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4497 operated on the sender address. After changing the $sender_address to <>
4498 for the sender address verify, Exim was re-instated it as the original
4499 (before rewriting) address, but remembering that it had rewritten it, so it
4500 wasn't rewriting it again. This bug also had the effect of breaking the
4501 sender address verification caching when the sender address was rewritten.
4503 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4504 This has been changed so that if the ip literal address matches
4505 ignore_target_hosts, the router declines.
4507 7. Added expansion conditions match_domain, match_address, and match_local_
4508 part (NOT match_host).
4510 8. The placeholder for the Received: header didn't have a length field set.
4512 9. Added code to Exim itself and to exim_lock to test for a specific race
4513 condition that could lead to file corruption when using MBX delivery. The
4514 issue is with the lockfile that is created in /tmp. If this file is removed
4515 after a process has opened it but before that process has acquired a lock,
4516 there is the potential for a second process to recreate the file and also
4517 acquire a lock. This could lead to two Exim processes writing to the file
4518 at the same time. The added code performs the same test as UW imapd; it
4519 checks after acquiring the lock that its file descriptor still refers to
4520 the same named file.
4522 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4523 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4526 11. Added the smtp_active_hostname option. If used, this will typically be made
4527 to depend on the incoming interface address. Because $interface_address is
4528 not set up until the daemon has forked a reception process, error responses
4529 that can happen earlier (such as "too many connections") no longer contain
4532 12. If an expansion in a condition on a "warn" statement fails because a lookup
4533 defers, the "warn" statement is abandoned, and the next ACL statement is
4534 processed. Previously this caused the whole ACL to be aborted.
4536 13. Added the iplsearch lookup type.
4538 14. Added ident_timeout as a log selector.
4540 15. Added tls_certificate_verified as a log selector.
4542 16. Added a global option tls_require_ciphers (compare the smtp transport
4543 option of the same name). This controls incoming TLS connections.
4545 17. I finally figured out how to make tls_require_ciphers do a similar thing
4546 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4547 before starting the TLS session.
4549 18. Tabs are now shown as \t in -bP output.
4551 19. If the log selector return_path_on_delivery was set, Exim crashed when
4552 bouncing a message because it had too many Received: header lines.
4554 20. If two routers both had headers_remove settings, and the first one included
4555 a superfluous trailing colon, the final name in the first list and the
4556 first name in the second list were incorrectly joined into one item (with a
4557 colon in the middle).
4563 1. Added -C and -D options to the exinext utility, mainly to make it easier
4564 to include in the automated testing, but these could be helpful when
4565 multiple configurations are in use.
4567 2. The exinext utility was not formatting the output nicely when there was
4568 an alternate port involved in the retry record key, nor when there was a
4569 message id as well (for retries that were specific to a specific message
4570 and a specific host). It was also confused by IPv6 addresses, because of
4571 the additional colons they contain. I have fixed the IPv4 problem, and
4572 patched it up to do a reasonable job for IPv6.
4574 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4575 delivery, the log line now contains "pipelined" if PIPELINING was used.
4577 4. An SMTP transport process used to panic and die if the bind() call to set
4578 an explicit outgoing interface failed. This has been changed; it is now
4579 treated in the same way as a connect() failure.
4581 5. A reference to $sender_host_name in the part of a conditional expansion
4582 that was being skipped was still causing a DNS lookup. This no longer
4585 6. The def: expansion condition was not recognizing references to header lines
4586 that used bh_ and bheader_.
4588 7. Added the _cache feature to named lists.
4590 8. The code for checking quota_filecount in the appendfile transport was
4591 allowing one more file than it should have been.
4593 9. For compatibility with Sendmail, the command line option
4601 and sets the incoming protocol and host name (for trusted callers). The
4602 host name and its colon can be omitted when only the protocol is to be set.
4603 Note the Exim already has two private options, -pd and -ps, that refer to
4604 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4605 "s", but I don't think that's a major issue.
4607 10. A number of refactoring changes to the code, none of which should affect
4610 (a) The number of logging options was getting close to filling up the
4611 32-bit word that was used as a bit map. I have split them into two classes:
4612 those that are passed in the argument to log_write(), and those that are
4613 only ever tested independently outside of that function. These are now in
4614 separate 32-bit words, so there is plenty of room for expansion again.
4615 There is no change in the user interface or the logging behaviour.
4617 (b) When building, for example, log lines, the code previously used a
4618 macro that called string_cat() twice, in order to add two strings. This is
4619 not really sufficiently general. Furthermore, there was one instance where
4620 it was actually wrong because one of the argument was used twice, and in
4621 one call a function was used. (As it happened, calling the function twice
4622 did not affect the overall behaviour.) The macro has been replaced by a
4623 function that can join an arbitrary number of extra strings onto a growing
4626 (c) The code for expansion conditions now uses a table and a binary chop
4627 instead of a serial search (which was left over from when there were very
4628 few conditions). Also, it now recognizes conditions like "pam" even when
4629 the relevant support is not compiled in: a suitably worded error message is
4630 given if an attempt is made to use such a condition.
4632 11. Added ${time_interval:xxxxx}.
4634 12. A bug was causing one of the ddress fields not to be passed back correctly
4635 from remote delivery subprocesses. The field in question was not being
4636 subsequently used, so this caused to problems in practice.
4638 13. Added new log selectors queue_time and deliver_time.
4640 14. Might have fixed a bug in maildirsizefile handling that threw up
4641 "unexpected character" debug warnings, and recalculated the data
4642 unnecessarily. In any case, I expanded the warning message to give more
4645 15. Added the message "Restricted characters in address" to the statements in
4646 the default ACL that block characters like @ and % in local parts.
4648 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4649 Three changes have been made:
4651 (a) There was a serious bug; a negative response to MAIL caused the whole
4652 recipient domain to be cached as invalid, thereby blocking all messages
4653 to all local parts at the same domain, from all senders. This bug has
4654 been fixed. The domain is no longer cached after a negative response to
4655 MAIL if the sender used is not empty.
4657 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4660 (c) A new callout option, "use_sender" has been added for people who want
4661 the modified behaviour.
4667 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4670 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4671 indeed breaks things for older releases.
4673 3. Added additional logging to the case where there is a problem reading data
4674 from a filter that is running in a subprocess using a pipe, in order to
4675 try to track down a specific problem.
4677 4. Testing facility fudge: when running in the test harness and attempting
4678 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4679 getting "No route to host". Convert this to a timeout.
4681 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4684 6. Some OS don't have socklen_t but use size_t instead. This affects the
4685 fifth argument of getsockopt() amongst other things. This is now
4686 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4687 can be set for individual OS. I have set it for SunOS5, OSF1, and
4688 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4689 some earlier ones do not.
4691 7. Change 4.30/15 was not doing the test caselessly.
4693 8. The standard form for an IPv6 address literal was being rejected by address
4694 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4695 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4696 this, as well as the form without the "IPv6" on the front (but only when
4697 address literals are enabled, of course).
4699 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4701 10. Exim crashed if a message with an empty sender address specified by -f
4702 encountered a router with an errors_to setting. This could be provoked only
4703 by a command such as
4707 where an empty string was supplied; "<>" did not hit this bug.
4709 11. Installed PCRE release 4.5.
4711 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4712 remained set. It is now erased.
4714 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4715 times from message ids (which are base 36 rather than the normal 62).
4717 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4718 were being counted as actual protocol errors, and logged if the log
4719 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4720 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4721 and DATA following a set of rejected RCPTs do not count as protocol errors.
4722 In other words, Exim assumes they were pipelined, though this may not
4723 actually be the case. Of course, in all cases the client gets an
4724 appropriate error code.
4726 15. If a lookup fails in an ACL condition, a message about the failure may
4727 be available; it is used if testing the ACL cannot continue, because most
4728 such messages specify what the cause of the deferral is. However, some
4729 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4730 that caused an old message to be retained and used if a later statement
4731 caused a defer, replacing the real cause of the deferral.
4733 16. If an IP address had so many PTR records that the DNS lookup buffer
4734 was not large enough to hold them, Exim could crash while trying to process
4735 the truncated data. It now detects and logs this case.
4737 17. Further to 4.21/58, another change has been made: if (and only if) the
4738 first line of a message (the first header line) ends with CRLF, a bare LF
4739 in a subsequent header line has a space inserted after it, so as not to
4740 terminate the header.
4742 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4743 unnecessarily, used atoi() instead of strtol(), and didn't check the
4744 termination when getting file sizes from file names by regex.
4746 19. Completely re-implemented the support for maildirsize files, in the light
4747 of a number of problems with the previous contributed implementation
4748 (4.30/29). In particular:
4750 . If the quota is zero, the maildirsize file is maintained, but no quota is
4753 . If the maildir directory does not exist, it is created before any attempt
4754 to write a maildirsize file.
4756 . The quota value in the file is just a cache; if the quota is changed in
4757 the transport, the new value overrides.
4759 . A regular expression is available for excluding directories from the
4762 20. The autoreply transport checks the characters in options that define the
4763 message's headers; it allows continued headers, but it was checking with
4764 isspace() after an embedded newline instead of explicitly looking for a
4767 21. If all the "regular" hosts to which an address was routed had passed their
4768 expiry times, and had not reached their retry times, the address was
4769 bounced, even if fallback hosts were defined. Now Exim should go on to try
4772 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4773 equivalent code in the SMTP transport. Some hosts send humungous responses
4774 to HELO/EHLO, more than 1024 it seems.
4776 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4777 gives compiler warnings in some environments. I've now done it "properly",
4780 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4781 (because of problems with the built-in one) was declared to return uschar *
4782 instead of char *, causing compiler failure.
4784 25. Fixed a file descriptor leak when processing alias/forward files.
4786 26. Fixed a minor format string issue in dbfn.c.
4788 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4790 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4791 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4794 29. When a sender address is verified, it is cached, to save repeating the test
4795 when there is more than one recipient in a message. However, when the
4796 verification involves a callout, it is possible for different callout
4797 options to be set for different recipients. It is too complicated to keep
4798 track of this in the cache, so now Exim always runs a verification when a
4799 callout is required, relying on the callout cache for the optimization.
4800 The overhead is duplication of the address routing, but this should not be
4803 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4804 to be verified with callout=postmaster, and the main callout worked but the
4805 postmaster check failed, the verification correctly failed. However, if a
4806 subsequent RCPT command asked for sender verification *without* the
4807 postmaster check, incorrect caching caused this verification also to fail,
4810 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4811 it was not caching the DNS options (qualify_single, search_parents) that
4812 were used when the lookup failed. A subsequent lookup with different
4813 options therefore always gave the same answer, though there were cases
4814 where it should not have. (Example: a "domains = !$mx_any" option on a
4815 dnslookup router: the "domains" option is always processed without any
4816 widening, but the router might have qualify_single set.) Now Exim uses the
4817 cached value only when the same options are set.
4819 32. Added John Jetmore's "exipick" utility to the distribution.
4821 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4822 than a timeout (e.g. a certificate is required, and is not provided), an
4823 Exim server now closes the connection immediately. Previously it waited for
4824 the client to close - but if the client is SSL, it seems that they each
4825 wait for each other, leading to a delay before one of them times out.
4827 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4828 maintained 0.8.x compatibility because I don't think many are using it, and
4829 it is clearly obsolete.
4831 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4834 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4835 client certificate was expired. A simple patch fixes this, though I don't
4836 understand the full logic of why the verify callback is called multiple
4839 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4840 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4841 which causes problems with some clients (such as the Certicom SSL Plus
4842 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4843 disables the coutermeasure allowing Eudora to connect."
4845 38. Exim was not checking that a write() to a log file succeeded. This could
4846 lead to Bad Things if a log got too big, in particular if it hit a file
4847 size limit. Exim now panics and dies if it cannot write to a log file, just
4848 as it does if it cannot open a log file.
4850 39. Modified OS/Makefile-Linux so that it now contains
4852 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4854 The two -D definitions ensure that Exim is compiled with large file
4855 support, which makes it possible to handle log files that are bigger than
4858 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4859 instance) a domain was checked against a named list that involved a lookup,
4860 causing $domain_data to be set, then another domain was checked against the
4861 same list, then the first domain was re-checked, the value of $domain_data
4862 after the final check could be wrong. In particular, if the second check
4863 failed, it could be set empty. This bug probably also applied to
4866 41. The strip_trailing_dot option was not being applied to the address given
4867 with the -f command-line option.
4869 42. The code for reading a message's header from the spool was incrementing
4870 $received_count, but never initializing it. This meant that the value was
4871 incorrect (doubled) while delivering a message in the same process in which
4872 it was received. In the most common configuration of Exim, this never
4873 happens - a fresh exec is done - but it can happen when
4874 deliver_drop_privilege is set.
4876 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4877 it now includes up to 150 characters of the unexpected data in the log
4880 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4881 and building data strings. The size of both of these buffers was 10 000
4882 bytes - far larger than anybody would *ever* want, thought I. Needless to
4883 say, somebody hit the limit. I have increased the maximum line length to
4884 20 000 and the maximum data length of concatenated lines to 100 000. I have
4885 also fixed two bugs, because there was no checking on these buffers. Tsk,
4886 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4887 buffer is too small.
4889 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4890 lsearch lookups. Now it does.
4892 46. When parsing a route_list item in a manualroute router, a fixed-length
4893 buffer was used for the list of hosts. I made this 1024 bytes long,
4894 thinking that nobody would ever have a list of hosts that long. Wrong.
4895 Somebody had a whole pile of complicated expansion conditions, and the
4896 string was silently truncated, leading to an expansion error. It turns out
4897 that it is easier to change to an unlimited length (owing to other changes
4898 that have happened since this code was originally written) than to build
4899 structure for giving a limitation error. The length of the item that
4900 expands into the list of hosts is now unlimited.
4902 47. The lsearch lookup could not handle data where the length of text line was
4903 more than 4095 characters. Such lines were truncated, leading to shortened
4904 data being returned. It should now handle lines of any length.
4906 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4907 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4910 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4911 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4912 no longer happens in comment lines. A list of replacements is now placed
4913 at the head of all of the source files, except those whose only change is
4914 to replace PERL_COMMAND in the very first #! line.
4916 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4917 messages on the queue, with a bottom-up merge sort, using code contributed
4918 by Michael Haardt. This should make operations like -bp somewhat faster on
4919 large queues. It won't affect queue runners, except when queue_run_in_order
4922 51. Installed eximstats 1.31 in the distribution.
4924 52. Added support for SRV lookups to the dnslookup router.
4926 53. If an ACL referred to $message_body or $message_body_end, the value was not
4927 reset for any messages that followed in the same SMTP session.
4929 54. The store-handling optimization for building very long strings was not
4930 differentiating between the different store pools. I don't think this
4931 actually made any difference in practice, but I've tidied it.
4933 55. While running the routers to verify a sender address, $sender_address
4934 was still set to the sender address. This is wrong, because when routing to
4935 send a bounce to the sender, it would be empty. Therefore, I have changed
4936 it so that, while verifying a sender address, $sender_address is set to <>.
4937 (There is no change to what happens when verifying a recipient address.)
4939 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4940 target A or AAAA records (if not already returned) without resetting the
4941 qualify_single or search_parents options of the DNS resolver. These are
4942 inappropriate in this case because the targets of MX and SRV records must
4943 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4944 target that, when qualified, matched something in the local domain. These
4945 two options are now turned off when doing these lookups.
4947 57. It seems that at least some releases of Reiserfs (which does not have the
4948 concept of a fixed number of inodes) returns zero and not -1 for the
4949 number of available inodes. This interacted badly with check_spool_inodes,
4950 which assumed that -1 was the "no such thing" setting. What I have done is
4951 to check that the total number of inodes is greater than zero before doing
4952 the test of how many are available.
4954 58. When a "warn" ACL statement has a log_message modifier, the message is
4955 remembered, and not repeated. This is to avoid a lot of repetition when a
4956 message has many recipients that cause the same warning to be written.
4957 Howewer, Exim was preserving the list of already written lines for an
4958 entire SMTP session, which doesn't seem right. The memory is now reset if a
4959 new message is started.
4961 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4962 debugging output unless log_rewrite was also set.
4964 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4965 of (int)(handle) when we know that handle contains (void *)(-1).
4967 61. The Exim daemon panic-logs an error return when it closes the incoming
4968 connection. However "connection reset by peer" seems to be common, and
4969 isn't really an error worthy of noting specially, so that particular error
4972 62. When Exim is trying to find all the local interfaces, it used to panic and
4973 die if the ioctl to get the interface flags failed. However, it seems that
4974 on at least one OS (Solaris 9) it is possible to have an interface that is
4975 included in the list of interfaces, but for which you get a failure error
4976 for this call. This happens when the interface is not "plumbed" into a
4977 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4978 failure of the "get flags" call assumes that the interface is down.
4980 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4981 makes life easier for people who are doing arithmetic on fields extracted
4982 from dates, where you often get leading zeros that should not be
4983 interpreted as octal.
4985 64. Added qualify_domain to the redirect router, to override the global
4988 65. If a pathologically long header line contained very many addresses (the
4989 report of this problem mentioned 10 000) and each of them was rewritten,
4990 Exim could use up a very large amount of memory. (It kept on making new
4991 copies of the header line as it rewrote, and never released the old ones.)
4992 At the expense of a bit more processing, the header rewriting function has
4993 been changed so that it no longer eats memory in this way.
4995 66. The generation of the Received: header has been moved from the time that a
4996 message starts to be received, to the time that it finishes. The timestamp
4997 in the Received: header should now be very close to that of the <= log
4998 line. There are two side-effects of this change:
5000 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5001 logged header lines no longer include the local Received: line, because
5002 it has not yet been created. The same applies to a copy of the message
5003 that is returned to a non-SMTP sender when a message is rejected.
5005 (b) When a filter file is tested using -bf, no additional Received: header
5006 is added to the test message. After some thought, I decided that this
5009 This change does not affect the value of $received_for. It is still set
5010 after address rewriting, but before local_scan() is called.
5012 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5014 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5015 gave an unhelpful panic error message, and a defer error. I have managed to
5016 change this behaviour so that it now rejects any supplied certificate,
5017 which seems right, as the list of acceptable certificates is empty.
5019 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5020 gave an unhelpful defer error. I have not managed to make this reject any
5021 supplied certificates, but the error message it gives is "no certificate
5022 supplied", which is not helpful.
5024 70. exigrep's output now also includes lines that are not associated with any
5025 message, but which match the given pattern. Implemented by a patch from
5026 Martin Sluka, which also tidied up the Perl a bit.
5028 71. Recipient callout verification, like sender verification, was using <> in
5029 the MAIL FROM command. This isn't really the right thing, since the actual
5030 sender may affect whether the remote host accepts the recipient or not. I
5031 have changed it to use the actual sender in the callout; this means that
5032 the cache record is now keyed on a recipient/sender pair, not just the
5033 recipient address. There doesn't seem to be a real danger of callout loops,
5034 since a callout by the remote host to check the sender would use <>.
5035 [SEE ABOVE: changed after hitting problems.]
5037 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5038 temporary errors. However, in the case of such a code being given after
5039 the end of a data transmission (i.e. after ".") Exim was failing to write
5040 a retry record for the message. (Yes, there was some broken host that was
5041 actually sending 8xx at this point.)
5043 73. An unknown lookup type in a host list could cause Exim to panic-die when
5044 the list was checked. (An example that provoked this was putting <; in the
5045 middle of a list instead of at the start.) If this happened during a DATA
5046 ACL check, a -D file could be left lying around. This kind of configuration
5047 error no longer causes Exim to die; instead it causes a defer errror. The
5048 incident is still logged to the main and panic logs.
5050 74. Buglet left over from Exim 3 conversion. The message "too many messages
5051 in one connection" was written to the rejectlog but not the mainlog, except
5052 when address rewriting (yes!) was being logged.
5054 75. Added write_rejectlog option.
5056 76. When a system filter was run not as root (that is, when system_filter_user
5057 was set), the values of the $n variables were not being returned to the
5058 main process; thus, they were not subsequently available in the $sn
5061 77. Added +return_path_on_delivery log selector.
5063 78. A connection timeout was being treated differently from recipients deferred
5064 when testing hosts_max_try with a message that was older than the host's
5065 retry timeout. (The host should not be counted, thus allowing all hosts to
5066 be tried at least once before bouncing.) This may have been the cause of an
5067 occasionally reported bug whereby a message would remain on the queue
5068 longer than the retry timeout, but would be bounced if a delivery was
5069 forced. I say "may" because I never totally pinned down the problem;
5070 setting up timeout/retry tests is difficult. See also the next item.
5072 79. The ultimate address timeout was not being applied to errors that involved
5073 a combination of host plus message (for example, a timeout on a MAIL
5074 command). When an address resolved to a number of possible hosts, and they
5075 were not all tried for each delivery (e.g. because of hosts_max_try), a
5076 message could remain on the queue longer than the retry timeout.
5078 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5081 81. Fixed an obscure SMTP outgoing bug which required at least the following
5082 conditions: (a) there was another message waiting for the same server;
5083 (b) the server returned 5xx to all RCPT commands in the first message so
5084 that the message was not completed; (c) the server dropped the connection
5085 or gave a negative response to the RSET that Exim sends to abort the
5086 transaction. The observed case was a dropped connection after DATA that had
5087 been sent in pipelining mode. That is, the server had advertised PIPELINING
5088 but was not implementing it correctly. The effect of the bug was incorrect
5089 behaviour, such as trying another host, and this could lead to a crash.
5095 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5096 and daemon.c were passed as pointers to ints; they should have been
5097 pointers to socklen_t variables (which are typically unsigned ints).
5099 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5102 3. Fixed a really odd bug that affected only the testing scheme; patching a
5103 certain fixed string in the binary changed the value of another string that
5104 happened to be identical to the end of the original first string.
5106 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5107 name", it returns that address as the IP address. On some operating
5108 systems (e.g. Solaris), it also passes back the IP address string as the
5109 "host name". However, on others (e.g. Linux), it passes back an empty
5110 string. Exim wasn't checking for this, and was changing the host name to an
5111 empty string, assuming it had been canonicized.
5113 5. Although rare, it is permitted to have more than one PTR record for a given
5114 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5115 all the names associated with an address, because they do in Solaris.
5116 However, it seems that they do not in Linux for data that comes from the
5117 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5118 I found this out when I moved to a new Linux workstation and tried to run
5119 the Exim test suite.
5121 To get round this problem I have changed the code so that it now does its
5122 own call to the DNS to look up PTR records when searching for a host name.
5123 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5124 addresses that are only in /etc/hosts are still found.
5126 This behaviour is, however, controlled by an option called host_lookup_
5127 order, which defaults to "bydns:byaddr". If people want to use the other
5128 order, or indeed, just use one or the other means of lookup, they can
5129 specify it in this variable.
5131 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5132 some operating systems, this comes back from gethostbyaddr() as an empty
5133 string, and this is what Exim used to test for. However, it seems that in
5134 other systems, "." is yielded. Exim now tests for this case too.
5136 7. The values of check_spool_space and check_log_space are now held internally
5137 as a number of kilobytes instead of an absolute number of bytes. If a
5138 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5139 kilobyte. This means that much larger values can be stored.
5141 8. Exim monitor: an attempt to get the action menu when not actually pointing
5142 at a message produces an empty menu entitled "No message selected". This
5143 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5144 no entries in it ("Shell widget menu has zero width and/or height"). So I
5145 have added a single, blank menu entry in this case.
5147 9. Added ${quote_local_part.
5149 10. MIME decoding is now applied to the contents of Subject: header lines when
5152 11. Now that a reference to $sender_host_address automatically causes a reverse
5153 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5154 host lookup before query-style lookups in lists that might use this
5155 variable. This has therefore been abolished, and the "net-" prefix is no
5156 longer necessary for query-style lookups.
5158 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5159 have been a typo for LFLAGS, so it has been changed.
5161 13. The install script calls Exim with "-C /dev/null" in order to find the
5162 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5163 to be output. Howeve, since Exim outputs its version number before the
5164 error, it didn't break the script. It just looked ugly. I fixed this by
5165 always allowing "-C /dev/null" if the caller is root.
5167 14. Ignore overlarge ACL variable number when reading spool file - insurance
5168 against a later release with more variables having written the file.
5170 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5171 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5172 this, as well as the form without the "IPv6" on the front.
5174 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5175 OS/Makefile-Darwin file.
5177 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5178 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5180 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5181 variable. However, some OS use ints for inodes. I've added cast to long int
5182 to get rid of the compiler warning.
5184 19. I had forgotten to lock out "/../" in configuration file names when
5185 ALT_CONFIG_PREFIX was set.
5187 20. Routers used for verification do not need to specify transports. However,
5188 if such a router generated a host list, and callout was configured, Exim
5189 crashed, because it could not find a port number from the (non-existent)
5190 transport. It now assumes port 25 in this circumstance.
5192 21. Added the -t option to exigrep.
5194 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5195 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5196 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5197 them is set, LOOKUP_LSEARCH is forced.
5199 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5201 24. Added sender and host information to the "rejected by local_scan()" log
5202 line; previously there was no indication of these.
5204 25. Added .include_if_exists.
5206 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5207 Linux. It turns out that not all file systems support this. Apparently some
5208 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5209 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5210 not supported on the file descriptor, is now ignored when Exim is trying to
5211 sync a directory. This applies only to Linux.
5213 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5215 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5216 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5219 29. Added support for maildirsize files from supplied patch (modified a bit).
5221 30. The use of :fail: followed by an empty string could lead Exim to respond to
5222 sender verification failures with (e.g.):
5224 550 Verification failed for <xxx>
5225 550 Sender verify failed
5227 where the first response line was missing the '-' that indicates it is not
5228 the final line of the response.
5230 31. The loop for finding the name of the user that called Exim had a hardwired
5231 limit of 10; it now uses the value of finduser_retries, which is used for
5232 all other user lookups.
5234 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5237 33. Exim was neglecting to zero errno before one call of strtol() when
5238 expanding a string and expecting an integer value. On some systems this
5239 resulted in spurious "integer overflow" errors. Also, it was casting the
5240 result into an int without checking.
5242 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5243 did not work. The code looks as if it has *never* worked, though it appears
5244 to have been documented since at least releast 1.62. I have made it work.
5246 35. The "timeout_DNS" error in retry rules, also documented since at least
5247 1.62, also never worked. As it isn't clear exactly what this means, and
5248 clearly it isn't a major issue, I have abolished the feature by treating it
5249 as "timeout", and writing a warning to the main and panic logs.
5251 36. The display of retry rules for -brt wasn't always showing the error code
5254 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5255 timeout_connect_A, timeout_connect_MX.
5257 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5258 to the empty sender.
5260 39. The daemon was not analysing the content of -oX till after it had closed
5261 stderr and disconnected from the controlling terminal. This meant that any
5262 syntax errors were only noted on the panic log, and the return code from
5263 the command was 0. By re-arranging the code a little, I've made the
5264 decoding happen first, so such errors now appear on stderr, and the return
5265 code is 1. However, the actual setting up of the sockets still happens in
5266 the disconnected process, so errors there are still only recorded on the
5269 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5270 connections (as happens on some IP stacks) was logged at start up time as
5271 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5272 it from "IPv6 and IPv4", which means that two separate sockets are being
5275 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5276 says whether AF_INET or AF_INET6 was passed as an argument.
5278 42. Exiwhat output was messed up when time zones were included in log
5281 43. Exiwhat now gives more information about the daemon's listening ports,
5282 and whether -tls-on-connect was used.
5284 44. The "port" option of the smtp transport is now expanded.
5286 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5287 silently ignored. Now an error message is written to the main and panic
5290 46. There's a new ACL modifier called "logwrite" which writes to a log file
5291 as soon as it is encountered.
5293 47. Added $local_user_uid and $local_user_gid at routing time.
5295 48. Exim crashed when trying to verify a sender address that was being
5298 49. Exim was recognizing only a space character after ".include". It now also
5299 recognizes a tab character.
5301 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5302 extracting the relevant information from the specification. The man page no
5303 longer contains scrambled data for the -d option, and I've added a section
5304 at the front about calling Exim under different names.
5306 51. Added "extra_headers" argument to the "mail" command in filter files.
5308 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5311 53. Installed eximstats 1.29.
5313 54. Added transport_filter_timeout as a generic transport option.
5315 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5316 Cc: header lines. This was required by RFC 822, but it not required by RFC
5319 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5320 incoming messages that did not have them. Now it does so only if the
5321 message originates locally, that is, if there is no associated remote host
5322 address. When Resent- header lines are present, this applies to the Resent-
5323 lines rather than the non-Resent- lines.
5325 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5326 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5328 58. Messages for configuration errors now include the name of the main
5329 configuration file - useful now that there may be more than one file in a
5330 list (.included file names were always shown).
5332 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5333 for those rare installations that do not start the daemon as root or run it
5334 setuid root. I've cut out the call to initgroups() if the daemon is not
5337 60. The Exim user and group can now be bound into the binary as text strings
5338 that are looked up at the start of Exim's processing.
5340 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5342 62. Added $mailstore_basename variable.
5344 63. Installed patch to sieve.c from Michael Haardt.
5346 64. When Exim failed to open the panic log after failing to open the main log,
5347 the original message it was trying to log was written to stderr and debug
5348 output, but if they were not available (the usual case in production), it
5349 was lost. Now it is written to syslog before the two lines that record the
5350 failures to open the logs.
5352 65. Users' Exim filters run in subprocesses under the user's uid. It is
5353 possible for a "deliver" command or an alias in a "personal" command to
5354 provoke an address rewrite. If logging of address rewriting is configured,
5355 this fails because the process is not running as root or exim. There may be
5356 a better way of dealing with this, but for the moment (because 4.30 needs
5357 to be released), I have disabled address rewrite logging when running a
5358 filter in a non-root, non-exim process.
5364 1. The buildconfig auxiliary program wasn't quoting the value set for
5365 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5366 not defined. This bug was masked in 4.22 by the effect that was fixed in
5369 2. Some messages that were rejected after a message id was allocated were
5370 shown as "incomplete" by exigrep. It no longer does this for messages that
5371 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5373 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5374 have allow_domain_literals set, the ID did not get logged in the <= line.
5375 Domain literals are now always recognized in Message-ID: header lines.
5377 4. The first argument for a ${extract expansion item is the key name or field
5378 number. Leading and trailing spaces in this item were not being ignored,
5379 causing some misleading effects.
5381 5. When deliver_drop_privilege was set, single queue runner processes started
5382 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5383 same command in the process it spins off) were not dropping privilege.
5385 6. When the daemon running as "exim" started a queue runner, it always
5386 re-executed Exim in the spun-off process. This is a waste of effort when
5387 deliver_drop_privilege is set. The new process now just calls the
5388 queue-runner function directly.
5394 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5397 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5398 ignored. Though the use of -oP was forcing the writing of a pid file, it
5399 was always written to the default place.
5401 3. If the message "no IP address found for host xxxx" is generated during
5402 incoming verification, it is now followed by identification of the incoming
5403 connection (so you can more easily find what provoked it).
5405 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5407 5. Added some features to "harden" Exim a bit more against certain attacks:
5409 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5410 be put in Local/Makefile. This is like the never_users runtime option,
5411 but it cannot be overridden. The default setting is "root".
5413 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5414 prefix string with which any file named in a -C command line option
5417 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5418 is retained for -C and -D only if the caller of Exim is root. Without
5419 it, the exim user may also use -C and -D and retain privilege.
5421 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5422 command line option is disabled.
5424 6. Macro names set by the -D option must start with an upper case letter, just
5425 like macro names defined in the configuration file.
5427 7. Added "dereference=" facility to LDAP.
5429 8. Two instances of the typo "uknown" in the source files are fixed.
5431 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5432 the Configure-Makefile script screwed up while processing it.
5434 10. Incorporated PCRE 4.4.
5436 11. The SMTP synchronization check was not operating right at the start of an
5437 SMTP session. For example, it could not catch a HELO sent before the client
5438 waited for the greeting. There is now a check for outstanding input at the
5439 point when the greeting is written. Because of the duplex, asynchronous
5440 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5441 way, but not yet received, when the check is performed.
5443 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5444 on TCP/IP sockets, because this apparently causes some broken clients to
5447 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5448 unchanged) from the Cygwin maintainer.
5450 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5451 was supported instead of testing for mbx. Effectively a typo.
5453 15. The spa authenticator server code was not checking that the input it
5454 received was valid base64.
5456 16. The debug output line for the "set" modifier in ACLs was not showing the
5457 name of the variable that was being set.
5459 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5461 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5462 lookup on its own. Something else was needed to trigger the lookup. For
5463 example, a match in host_lookup or the need for a host name in a host list.
5464 Now, if $sender_host_name is referenced and the host name has not yet been
5465 looked up, a lookup is performed. If the lookup fails, the variable remains
5466 empty, and $host_lookup_failed is set to "1".
5468 19. Added "eqi" as a case-independent comparison operator.
5470 20. The saslauthd authentication condition could segfault if neither service
5471 nor realm was specified.
5473 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5474 error message that was logged was misleading, and incoming SMTP
5475 connections were dropped. The message is now more accurate, and temporary
5476 errors are given to SMTP connections.
5478 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5479 (except RSET) with a 421 error, until QUIT is received. However, it was
5480 failing to send a response to QUIT.
5482 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5483 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5484 ACL is now run after the helo_try_verify_hosts code.
5486 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5487 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5488 case-independent, so other case variants are also recognized). Apparently
5489 some systems use these upper case variants.
5491 25. If more than two messages were waiting for the same host, and a transport
5492 filter was specified for the transport, Exim sent two messages over the
5493 same TCP/IP connection, and then failed with "socket operation on non-
5494 socket" when it tried to send the third.
5496 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5498 27. The extern definition of crypt16() in expand.c was not being excluded when
5499 the OS had its own crypt16() function.
5501 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5502 as a preferred synonym for return_size_limit, both as an option and as an
5505 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5507 30. Changed the default configuration ACL to relax the local part checking rule
5508 for addresses that are not in any local domains. For these addresses,
5509 slashes and pipe symbols are allowed within local parts, but the sequence
5510 /../ is explicitly forbidden.
5512 31. SPA server authentication was not clearing the challenge buffer before
5515 32. log_message in a "warn" ACL statement was writing to the reject log as
5516 well as to the main log, which contradicts the documentation and doesn't
5517 seem right (because no rejection is happening). So I have stopped it.
5519 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5520 However, I am unable to do any testing of this.
5522 34. Fixed an infelicity in the appendfile transport. When checking directories
5523 for a mailbox, to see if any needed to be created, it was accidentally
5524 using path names with one or more superfluous leading slashes; tracing
5525 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5527 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5528 contents are added to the log line that is written for every discarded
5529 recipient. (Previously a log_message setting was ignored.)
5531 36. The ${quote: operator now quotes the string if it is empty.
5533 37. The install script runs exim in order to find its version number. If for
5534 some reason other than non-existence or emptiness, which it checks, it
5535 could not run './exim', it was installing it with an empty version number,
5536 i.e. as "exim-". This error state is now caught, and the installation is
5539 38. An argument was missing from the function that creates an error message
5540 when Exim fails to connect to the socket for saslauthd authentication.
5541 This could cause Exim to crash, or give a corrupted message.
5543 39. Added isip, isip4, and isip6 to ${if conditions.
5545 40. The ACL variables $acl_xx are now saved with the message, and can be
5546 accessed later in routers, transports, and filters.
5548 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5549 strings in the file are not string-expanded.
5551 42. If a MAIL command specified a SIZE value that was too large to fit into an
5552 int variable, the check against message_size_limit failed. Such values are
5553 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5554 day this will have to be increased, but I don't think I want to be around
5555 when emails are that large.
5562 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5563 iconv() is not standard in FreeBSD.
5565 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5566 IPv6 enabled. The observed symptom was a segmentation fault on return from
5567 the function os_common_find_running_interfaces() in src/os.c.
5569 3. In the check_special_case() function in daemon.c I had used "errno" as an
5570 argument name, which causes warnings on some systems. This was basically a
5571 typo, since it was named "eno" in the comments!
5573 4. The code that waits for the clock to tick (at a resolution of some fraction
5574 of a second) so as to ensure message-id uniqueness was always waiting for
5575 at least one whole tick, when it could have waited for less. [This is
5576 almost certainly not relevant at current processor speeds, where it is
5577 unlikely to ever wait at all. But we try to future-proof.]
5579 5. The function that sleeps for a time interval that includes fractions of a
5580 second contained a race. It did not block SIGALRM between setting the
5581 timer, and suspending (a couple of lines later). If the interval was short
5582 and the sigsuspend() was delayed until after it had expired, the suspension
5583 never ended. On busy systems this could lead to processes getting stuck for
5586 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5587 process, before it forks any delivery processes. The open lookup caching
5588 mechanism meant that the open file or database connection was passed into
5589 the delivery process. The problem was that delivery processes always tidy
5590 up cached lookup data. This could cause a problem for the next delivery
5591 process started by the queue runner, because the external queue runner
5592 process does not know about the closure. So the next delivery process
5593 still has data in the lookup cache. In the case of a file lookup, there was
5594 no problem because closing a file descriptor in a subprocess doesn't affect
5595 the parent. However, if the lookup was caching a connection to a database,
5596 the connection was closed, and the second delivery process was likely to
5597 see errors such as "PGSQL: query failed: server closed the connection
5598 unexpectedly". The problem has been fixed by closing all cached lookups
5599 in a queue runner before running a delivery process.
5601 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5602 seem to have the "const" qualifier which it has on other OS. I've
5605 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5606 *of the same type* (client or server) with the same public name that an
5607 error should be diagnosed.
5609 9. When Exim looked up a host name for an IP address, but failed to find the
5610 original IP address when looking up the host name (a safety check), it
5611 output the message "<ip address> does not match any IP for NULL", which was
5612 confusing, to say the least. The bug was that the host name should have
5613 appeared instead of "NULL".
5615 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5616 user that is built into the binary, and the -C or -D options is used, root
5617 privilege is dropped before the configuration file is read. In addition,
5618 logging is switched to stderr instead of the normal log files. If the
5619 configuration then re-defines the Exim user, the unprivileged environment
5620 is probably not what is expected, so Exim logs a panic warning message (but
5623 However, if deliver_drop_privilege is set, the unprivileged state may well
5624 be exactly what is intended, so the warning has been cut out in that case,
5625 and Exim is allowed to try to write to its normal log files.
5631 1. smtp_return_error_details was not giving details for temporary sender
5632 or receiver verification errors.
5634 2. Diagnose a configuration error if two authenticators have the same public
5637 3. Exim used not to create the message log file for a message until the first
5638 delivery attempt. This could be confusing when incoming messages were held
5639 for policy or load reasons. The message log file is now created at the time
5640 the message is received, and an initial "Received" line is written to it.
5642 4. The automatically generated man page for command line options had a minor
5643 bug that caused no ill effects; however, a more serious problem was that
5644 the procedure for building the man page automatically didn't always
5645 operate. Consequently, release 4.20 contains an out-of-date version. This
5646 shouldn't happen again.
5648 5. When building Exim with embedded Perl support, the script that builds the
5649 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5650 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5652 6. The freeze_tell option was not being used for messages that were frozen on
5653 arrival, either by an ACL or by local_scan().
5655 7. Added the smtp_incomplete_transaction log selector.
5657 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5658 was accepting AUTH without a new EHLO.
5660 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5661 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5664 10. Logging of TCP/IP connections (when configured) now happens in the main
5665 daemon process instead of the child process, so that the TCP/IP connection
5666 count is more accurate (but it can never be perfect).
5668 11. The use of "drop" in a nested ACL was not being handled correctly in the
5669 outer ACL. Now, if condition failure induced by the nested "drop" causes
5670 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5671 or "require"), the connection is dropped.
5673 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5674 that yield "discard" can now be used with an "accept" or a "discard" verb,
5675 but an error is generated for any others (because I can't see a useful way
5676 to define what should happen).
5678 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5679 are now processed in the same way as lines in the Exim configuration file.
5680 In particular, continuation lines are supported.
5682 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5684 15. Added -ti meaning -t -i.
5686 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5687 domains, and warn by logging if others are found.
5689 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5690 returned from GIFCONF when Exim is trying to find the list of interfaces on
5691 a host. The code in os.c has been modified to copy each ifreq to an aligned
5692 structure in all cases.
5694 Also, in some cases, the returned ifreq's were being copied to a 'struct
5695 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5696 means the last couple of bytes of an IPv6 address could be chopped if the
5697 ifreq contained only a normal sockaddr (14 bytes storage).
5699 18. Named domain lists were not supported in the hosts_treat_as_local option.
5700 An entry such as +xxxx was not recognized, and was treated as a literal
5703 19. Ensure that header lines added by a DATA ACL are included in the reject log
5704 if the ACL subsequently rejects the message.
5706 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5707 MD5 (which is deprecated).
5709 21. When testing a filter file using -bf, Exim was writing a message when it
5710 took the sender from a "From " line in the message, but it was not doing so
5711 when it took $return_path from a Return-Path: header line. It now does.
5713 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5714 with a valid header line field name (a series of printing characters
5715 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5717 23. Changed "disc" in the source to "disk" to conform to the documentation and
5718 the book and for uniformity.
5720 24. Ignore Sendmail's -Ooption=value command line item.
5722 25. When execve() failed while trying to run a command in a pipe transport,
5723 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5724 could be confused with a return value of 69 from the command itself. This
5725 has been changed to 127, the value the shell returns if it is asked to run
5726 a non-existent command. The wording for the related log line suggests a
5727 non-existent command as the problem.
5729 26. If received_header_text expands to an empty string, do not add a Received:
5730 header line to the message. (Well, it adds a token one on the spool, but
5731 marks it "old" so that it doesn't get used or transmitted.)
5733 27. Installed eximstats 1.28 (addition of -nt option).
5735 28. There was no check for failure on the call to getsockname() in the daemon
5736 code. This can fail if there is a shortage of resources on the system, with
5737 ENOMEM, for example. A temporary error is now given on failure.
5739 29. Contrary to the C standard, it seems that in some environments, the
5740 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5741 program. Exim now does this explicitly; it affects the formatting of
5742 timestamps using strftime().
5744 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5745 complaints. I've now initialized all the variables, to avoid this.
5747 32. Header lines added by a system filter were not being "seen" during
5748 transport-time rewrites.
5750 33. The info_callback() function passed to OpenSSL is set up with type void
5751 (*)(SSL *, int, int), as described somewhere. However, when calling the
5752 function (actually a macro) that sets it up, the type void(*)() is
5753 expected. I've put in a cast to prevent warnings from picky compilers.
5755 34. If a DNS black list lookup found a CNAME record, but there were no A
5756 records associated with the domain it pointed at, Exim crashed.
5758 35. If a DNS black list lookup returned more than one A record, Exim ignored
5759 all but the first. It now scans all returned addresses if a particular IP
5760 value is being sought. In this situation, the contents of the
5761 $dnslist_value variable are a list of all the addresses, separated by a
5764 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5765 to accept a host name and all its aliases if the forward lookup for any of
5766 them yielded the IP address of the incoming connection. Now it accepts only
5767 those names whose forward lookup yields the correct IP address. Any other
5768 names are discarded. This closes a loophole whereby a rogue DNS
5769 administrator could create reverse DNS records to break through a
5770 wildcarded host restriction in an ACL.
5772 37. If a user filter or a system filter that ran in a subprocess used any of
5773 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5774 the wrong values were passed to the pipe command ($thisaddress had the
5775 value of $0, $0 had the value of $1, etc). This bug was introduced by
5776 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5778 38. Improved the line breaking for long SMTP error messages from ACLs.
5779 Previously, if there was no break point between 40 and 75 characters, Exim
5780 left the rest of the message alone. Two changes have been made: (a) I've
5781 reduced the minimum length to 35 characters; (b) if it can't find a break
5782 point between 35 and 75 characters, it looks ahead and uses the first one
5783 that it finds. This may give the occasional overlong line, but at least the
5784 remaining text gets split now.
5786 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5787 file descriptors might be low, and that setting 1000 would always raise it.
5788 It turns out that in some environments, the limit is already over 1000 and
5789 that lowering it causes trouble. So now Exim takes care not to decrease it.
5791 40. When delivering a message, the value of $return_path is set to $sender_
5792 address at the start of routing (routers may change the value). By an
5793 oversight, this default was not being set up when an address was tested by
5794 -bt or -bv, which affected the outcome if any router or filter referred to
5797 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5798 log only when "message" or "log_message" are set. However, if one of the
5799 conditions was an address verification, or a call to a nested ACL, the
5800 messages generated by the underlying test were being passed through. This
5801 no longer happens. The underlying message is available in $acl_verify_
5802 message for both "message" and "log_message" expansions, so it can be
5803 passed through if needed.
5805 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5806 new expansion $bh_ to give the encoded byte string without charset
5807 translation. Translation happens only if iconv() is available; HAVE_ICONV
5808 indicates this at build time. HEADERS_CHARSET gives the charset to
5809 translate to; headers_charset can change it in the configuration, and
5810 "headers charset" can change it in an individual filter file.
5812 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5813 that creates RFC 2047 encoded "words" labels them as that charset instead
5814 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5815 expansion operator; (ii) when Exim creates a From: line for a local
5816 message; (iii) when a header line is rewritten to include a "phrase" part.
5818 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5819 buggy, causing it to skip the first lines of messages whose message ID
5820 ended in 'D'. This would not have bitten before Exim release 4.14, because
5821 message IDs were unlikely to end in 'D' before then. The effect was to have
5822 incorrect size information for certain domains.
5824 45. #include "config.h" was missing at the start of the crypt16.c module. This
5825 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5828 46. If there was a timeout during a "random" callout check, Exim treated it as
5829 a failure of the random address, and carried on sending RSET and the real
5830 address. If the delay was just some slowness somewhere, the response to the
5831 original RCPT would be taken as a response to RSET and so on, causing
5832 mayhem of various kinds.
5834 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5835 when I implemented it. It didn't allow for the fact that some option values
5836 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5837 the right test for positive values.
5839 48. Domain names in DNS records are case-independent. Exim always looks them up
5840 in lower case. Some resolvers return domain names in exactly the case they
5841 appear in the zone file, that is, they may contain uppercase letters. Not
5842 all resolvers do this - some return always lower case. Exim was treating a
5843 change of case by a resolver as a change of domain, similar to a widening
5844 of a domain abbreviation. This triggered its re-routing code and so it was
5845 trying to route what was effectively the same domain again. This normally
5846 caused routing to fail (because the router wouldn't handle the domain
5847 twice). Now Exim checks for this case specially, and just changes the
5848 casing of the domain that it ultimately uses when it transmits the message
5851 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5854 50. If a filter generated a file delivery with a non-absolute name (possible if
5855 no home directory exists for the router), the forbid_file option was not
5858 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5859 the existing equality matching.
5861 52. Exim was using ints instead of ino_t variables in some places where it was
5862 dealing with inode numbers.
5864 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5865 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5866 called TMPDIR, and if it finds it is different, it changes its value.
5868 54. The smtp_printf() function is now made available to local_scan() so
5869 additional output lines can be written before returning. There is also an
5870 smtp_fflush() function to enable the detection of a dropped connection.
5871 The variables smtp_input and smtp_batched_input are exported to
5874 55. Changed the default runtime configuration: the message "Unknown user"
5875 has been removed from the ACL, and instead placed on the localuser router,
5876 using the cannot_route_message feature. This means that any verification
5877 failures that generate their own messages won't get overridden. Similarly,
5878 the "Unrouteable address" message that was in the ACL for unverifiable
5879 relay addresses has also been removed.
5881 56. Added hosts_avoid_esmtp to the smtp transport.
5883 57. The exicyclog script was not checking for the esoteric option
5884 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5885 will work only if exicyclog is run under the appropriate euid.
5887 58. Following a discussion on the list, the rules by which Exim recognises line
5888 endings on incoming messages have been changed. The -dropcr and drop_cr
5889 options are now no-ops, retained only for backwards compatibility. The
5890 following line terminators are recognized: LF CRLF CR. However, special
5891 processing applies to CR:
5893 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5894 nor a local message in the state where . is a terminator.
5896 (ii) If a bare CR is encountered in a header line, an extra space is added
5897 after the line terminator so as not to end the header. The reasoning
5898 behind this is that bare CRs in header lines are most likely either
5899 to be mistakes, or people trying to play silly games.
5901 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5902 was being incorrectly given as 18 bytes larger than it should have been.
5903 This is a VOB (very old bug).
5905 60. This may never have affected anything current, but just in case it has:
5906 When the local host is found other than at the start of a list of hosts,
5907 the local host, those with the same MX, and any that follow, are discarded.
5908 When the list in question was part of a longer list of hosts, the following
5909 hosts (not currently being processed) were also being discarded. This no
5910 longer happens. I'm not sure if this situation could ever has previously
5913 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5916 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5917 header. This is something that is recommended in a new Internet Draft, and
5918 is something that is documented as being done by Sendmail. There are two
5919 possible values. For messages generated by the autoreply transport, Exim
5922 Auto-Submitted: auto-replied
5924 whereas for all other generated messages (e.g. bounces) it adds
5926 Auto-Submitted: auto-generated
5928 63. The "personal" condition in filters now includes a test for the
5929 Auto-Submitted: header. If it contains the string "auto-" the message it
5930 not considered personal.
5932 64. Added rcpt_include_affixes as a generic transport option.
5934 65. Added queue_only_override (default true).
5936 66. Added the syslog_duplication option.
5938 67. If what should have been the first header line of a message consisted of
5939 a space followed by a colon, Exim was mis-interpreting it as a header line.
5940 It isn't of course - it is syntactically invalid and should therefore be
5941 treated as the start of the message body. The misbehaviour could have
5942 caused a number of strange effects, including loss of data in subsequent
5943 header lines, and spool format errors.
5945 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5946 client host had authenticated. This control can now be exercised by an ACL
5947 for more flexibility.
5949 69. By default, callouts do not happen when testing with -bh. There is now a
5950 variant, -bhc, which does actually run the callout code, including
5951 consulting and updating the callout cache.
5953 70. Added support for saslauthd authentication, courtesy of Alexander
5956 71. If statvfs() failed on the spool or log directories while checking their
5957 size for availability, Exim confusingly gave the error "space shortage".
5958 Furthermore, in debugging mode it crashed with a floating point exception.
5959 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5960 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5961 serious problem, Exim now writes to the main and panic logs when this
5962 happens, with details of the failure. It then refuses to accept the
5963 incoming message, giving the message "spool directory problem" or "log
5964 directory problem" with a 421 code for SMTP messages.
5966 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5967 0, 1, and 2 exist, because some OS complain for execs without them (see
5968 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5969 descriptors. However, the code omitted to check that the open succeeded,
5970 causing mysterious errors if for some reason the permissions on /dev/null
5971 got screwed. Now Exim writes a message to the main and panic logs, and
5972 bombs out if it can't open /dev/null.
5974 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5975 interact so that it is all more flexible. It is supposed to remain
5976 backwards compatible. Also added extra_local_interfaces.
5978 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5979 to bomb out with an assertion failure - to the client this appears as a
5980 connection drop. This problem occurs in the part of the code that was taken
5981 from the Samba project. Fortunately, the assertion is in a very simple
5982 function, so I have fixed this by reproducing the function inline in the
5983 one place where it is called, and arranging for authentication to fail
5984 instead of killing the process with assert().
5986 75. The SPA client code was not working when the server requested OEM rather
5987 than Unicode encoding.
5989 76. Added code to make require_files with a specific uid setting more usable in
5990 the case where statting the file as root fails - usually a non-root-mounted
5991 NFS file system. When this happens and the failure is EACCES, Exim now
5992 forks a subprocess and does the per-uid checking as the relevant uid.
5994 77. Added process_log_path.
5996 78. If log_file_path was not explicitly set, a setting of check_log_space or
5997 check_log_inodes was ignored.
5999 79. If a space check for the spool or log partitions fails, the incident is now
6000 logged. Of course, in the latter case the data may get lost...
6002 80. Added the %p formatting code to string_format() so that it can be used to
6003 print addresses in debug_print(). Adjusted all the address printing in the
6004 debugging in store.c to use %p rather than %d.
6006 81. There was a concern that a line of code in smtp_in.c could overflow a
6007 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6008 initially expressed, the concern was not well-founded, because trailing
6009 spaces are removed early. However, if the trailing spaces were followed by
6010 a NULL, they did not get removed, so the overflow was possible. Two fixes
6013 (a) I re-wrote the offending code in a cleaner fashion.
6014 (b) If an incoming SMTP command contains a NULL character, it is rejected
6017 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6018 runs initgroups(), so that if the Exim user is in any additional groups,
6019 they will be used during message reception.
6025 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob