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. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
49 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
51 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
52 selectors, in both main and reject logs.
54 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
57 JH/07 Add malware type "sock" for talking to simple daemon.
59 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
61 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
62 routers/transports under cutthrough routing.
64 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
65 numbers. Touch up "bool" conditional to keep the same definition.
67 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
69 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
71 JH/12 Expand items in router/transport headers_add or headers_remove lists
72 individually rather than the list as a whole. Bug 1452.
74 Required for reasonable handling of multiple headers_ options when
75 they may be empty; requires that headers_remove items with embedded
76 colons must have them doubled (or the list-separator changed).
78 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
79 view the policy declared in the DMARC record. Currently, $dmarc_status
80 is a combined value of both the record presence and the result of the
86 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
88 PP/02 Make -n do something, by making it not do something.
89 When combined with -bP, the name of an option is not output.
91 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
94 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
95 $sender_host_name and config options to manage this, and basic check
98 PP/05 DSCP support for outbound connections and control modifier for inbound.
100 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
101 (Only plugin which currently uses this is kerberos4, which nobody should
102 be using, but we should make it available and other future plugins might
103 conceivably use it, even though it would break NAT; stuff *should* be
104 using channel bindings instead).
106 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
107 name; added for Sendmail compatibility; requires admin caller.
108 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
109 just ignore it); requires trusted caller.
110 Also parse but ignore: -Ac -Am -X<logfile>
113 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
115 TL/02 Add +smtp_confirmation as a default logging option.
117 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
118 Patch by Magnus Holmgren from 2007-02-20.
120 TL/04 Bugzilla 1281 - Spec typo.
121 Bugzilla 1283 - Spec typo.
122 Bugzilla 1290 - Spec grammar fixes.
124 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
126 TL/06 Add Experimental DMARC support using libopendmarc libraries.
128 TL/07 Fix an out of order global option causing a segfault. Reported to dev
129 mailing list by by Dmitry Isaikin.
131 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
133 JH/02 Support "G" suffix to numbers in ${if comparisons.
135 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
137 NM/01 Bugzilla 1197 - Spec typo
138 Bugzilla 1196 - Spec examples corrections
140 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
142 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
143 gnutls_enable_pkcs11, but renamed to more accurately indicate its
146 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
147 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
149 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
150 "acl {{name}{arg}...}", and optional args on acl condition
153 JH/05 Permit multiple router/transport headers_add/remove lines.
155 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
157 JH/07 Avoid using a waiting database for a single-message-only transport.
158 Performance patch from Paul Fisher. Bugzilla 1262.
160 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
163 JH/09 Add $headers_added variable, with content from use of ACL modifier
164 add_header (but not yet added to the message). Bugzilla 199.
166 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
167 Pulled from Bugzilla 817 by Wolfgang Breyha.
169 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
171 (nb: this is the same fix as in Exim 4.80.1)
173 JH/11 Add A= logging on delivery lines, and a client_set_id option on
176 JH/12 Add optional authenticated_sender logging to A= and a log_selector
179 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
181 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
182 advertise SMTP AUTH mechanism to us, instead of a generic
183 protocol violation error. Also, make Exim more robust to bad
184 data from the Dovecot auth socket.
186 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
188 When a queue runner is handling a message, Exim first routes the
189 recipient addresses, during which it prunes them based on the retry
190 hints database. After that it attempts to deliver the message to
191 any remaining recipients. It then updates the hints database using
194 So if a recipient address works intermittently, it can get repeatedly
195 deferred at routing time. The retry hints record remains fresh so the
196 address never reaches the final cutoff time.
198 This is a fairly common occurrence when a user is bumping up against
199 their storage quota. Exim had some logic in its local delivery code
200 to deal with this. However it did not apply to per-recipient defers
201 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
203 This change adds a proper retry rule check during routing so that the
204 final cutoff time is checked against the message's age. We only do
205 this check if there is an address retry record and there is not a
206 domain retry record; this implies that previous attempts to handle
207 the address had the retry_use_local_parts option turned on. We use
208 this as an approximation for the destination being like a local
209 delivery, as in LMTP.
211 I suspect this new check makes the old local delivery cutoff check
212 redundant, but I have not verified this so I left the code in place.
214 TF/02 Correct gecos expansion when From: is a prefix of the username.
216 Test 0254 submits a message to Exim with the header
220 When I ran the test suite under the user fanf2, Exim expanded
221 the header to contain my full name, whereas it should have added
222 a Resent-Sender: header. It erroneously treats any prefix of the
223 username as equal to the username.
225 This change corrects that bug.
227 GF/01 DCC debug and logging tidyup
228 Error conditions log to paniclog rather than rejectlog.
229 Debug lines prefixed by "DCC: " to remove any ambiguity.
231 TF/03 Avoid unnecessary rebuilds of lookup-related code.
233 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
234 Bug spotted by Jeremy Harris; was flawed since initial commit.
235 Would have resulted in OCSP responses post-SNI triggering an Exim
236 NULL dereference and crash.
238 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
240 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
241 Bug detection, analysis and fix by Samuel Thibault.
242 Bugzilla 1331, Debian bug #698092.
244 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
246 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
247 Server implementation by Todd Lyons, client by JH.
248 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
249 config variable "prdr_enable" controls whether the server
250 advertises the facility. If the client requests PRDR a new
251 acl_data_smtp_prdr ACL is called once for each recipient, after
252 the body content is received and before the acl_smtp_data ACL.
253 The client is controlled by bolth of: a hosts_try_prdr option
254 on the smtp transport, and the server advertisement.
255 Default client logging of deliveries and rejections involving
256 PRDR are flagged with the string "PRDR".
258 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
259 fclose(). Diagnosis by Todd Lyons.
261 PP/17 Update configure.default to handle IPv6 localhost better.
262 Patch by Alain Williams (plus minor tweaks).
265 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
266 This is now consistent with GnuTLS, and is now documented: the
267 previous undocumented portable approach to treating the option as
268 unset was to force an expansion failure. That still works, and
269 an empty string is now equivalent.
271 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
272 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
273 not performing validation itself.
275 PP/20 Added force_command boolean option to pipe transport.
276 Patch from Nick Koston, of cPanel Inc.
278 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
281 TF/04 Added udpsend ACL modifer and hexquote expansion operator
283 PP/21 Fix eximon continuous updating with timestamped log-files.
284 Broken in a format-string cleanup in 4.80, missed when I repaired the
285 other false fix of the same issue.
286 Report and fix from Heiko Schlichting.
289 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
290 Report from Prashanth Katuri.
292 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
293 It's SecureTransport, so affects any MacOS clients which use the
294 system-integrated TLS libraries, including email clients.
296 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
297 using a MIME ACL for non-SMTP local injection.
298 Report and assistance in diagnosis by Warren Baker.
300 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
302 JH/16 Fix comparisons for 64b. Bugzilla 1385.
304 TL/09 Add expansion variable $authenticated_fail_id to keep track of
305 last id that failed so it may be referenced in subsequent ACL's.
307 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
310 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
311 ldap library initialization, allowing self-signed CA's to be
312 used. Also properly sets require_cert option later in code by
313 using NULL (global ldap config) instead of ldap handle (per
314 session). Bug diagnosis and testing by alxgomz.
316 TL/12 Enhanced documentation in the ratelimit.pl script provided in
317 the src/util/ subdirectory.
319 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
320 renamed to Transport Post Delivery Action by Jeremy Harris, as
323 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
324 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
325 redis_servers = needs to be configured which will be used by the redis
326 lookup. Patch from Warren Baker, of The Packet Hub.
328 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
330 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
331 hostname or reverse DNS when processing a host list. Used suggestions
332 from multiple comments on this bug.
334 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
336 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
337 Missed a few lines, added it to make the runtest require no keyboard
340 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
341 contains upper case chars. Make router use caseful_local_part.
343 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
344 support when GnuTLS has been built with p11-kit.
350 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
352 This, or similar/improved, will also be change PP/11 of 4.82.
358 PP/01 Handle short writes when writing local log-files.
359 In practice, only affects FreeBSD (8 onwards).
360 Bugzilla 1053, with thanks to Dmitry Isaikin.
362 NM/01 Bugzilla 949 - Documentation tweak
364 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
367 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
369 PP/02 Implemented gsasl authenticator.
371 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
373 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
374 `pkg-config foo` for cflags/libs.
376 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
377 with rest of GSASL and with heimdal_gssapi.
379 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
380 `pkg-config foo` for cflags/libs for the TLS implementation.
382 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
383 properties get this fed in as external SSF. A number of robustness
384 and debugging improvements to the cyrus_sasl authenticator.
386 PP/08 cyrus_sasl server now expands the server_realm option.
388 PP/09 Bugzilla 1214 - Log authentication information in reject log.
389 Patch by Jeremy Harris.
391 PP/10 Added dbmjz lookup type.
393 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
395 PP/12 MAIL args handles TAB as well as SP, for better interop with
396 non-compliant senders.
397 Analysis and variant patch by Todd Lyons.
399 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
400 Bug report from Lars Müller <lars@samba.org> (via SUSE),
401 Patch from Dirk Mueller <dmueller@suse.com>
403 PP/13 tls_peerdn now print-escaped for spool files.
404 Observed some $tls_peerdn in wild which contained \n, which resulted
405 in spool file corruption.
407 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
408 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
409 or write after TLS renegotiation, which otherwise led to messages
412 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
413 as a tracking header (ie: a signed header comes before the signature).
414 Patch from Wolfgang Breyha.
416 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
417 comma-sep list; embedded commas doubled.
419 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
421 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
423 Report and patch from Dmitry Banschikov.
425 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
426 Removed SSL_clear() after SSL_new() which led to protocol negotiation
427 failures. We appear to now support TLS1.1+ with Exim.
429 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
430 lets Exim select keys and certificates based upon TLS SNI from client.
431 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
432 before an outbound SMTP session. New log_selector, +tls_sni.
434 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
435 NULL dereference. Report and patch from Alun Jones.
437 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
438 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
439 Not seeing resolver debug output on NetBSD, but suspect this is a
440 resolver implementation change.
442 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
443 Left warnings. Added "eximon gdb" invocation mode.
445 PP/21 Defaulting "accept_8bitmime" to true, not false.
447 PP/22 Added -bw for inetd wait mode support.
449 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
450 locate the relevant includes and libraries. Made this the default.
452 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
453 Bugzilla 1246, report and most of solution from Tomasz Kusy.
455 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
456 This may cause build issues on older platforms.
458 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
459 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
460 gnutls_require_mac & gnutls_require_protocols (no longer supported).
461 Added SNI support via GnuTLS too.
462 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
464 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
466 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
467 Applied second patch from Janne, implementing suggestion to default
468 multiple-strings-in-record handling to match SPF spec.
470 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
472 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
473 read-only, out of scope).
474 Patch from Wolfgang Breyha, report from Stuart Northfield.
476 PP/29 Fix three issues highlighted by clang analyser static analysis.
477 Only crash-plausible issue would require the Cambridge-specific
478 iplookup router and a misconfiguration.
479 Report from Marcin Mirosław.
481 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
483 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
484 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
485 As part of this, removing so much warning spew let me fix some minor
486 real issues in debug logging.
488 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
489 assignment on my part. Fixed.
491 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
492 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
493 Janne Snabb (who went above and beyond: thank you).
495 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
496 string otherwise requires a connection and a bunch more work and it's
497 relatively easy to get wrong. Should also expose TLS library linkage
500 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
501 64-bit ${eval} (JH/03).
503 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
504 GNU libc to support some of the 64-bit stuff, should not lead to
505 conflicts. Defined before os.h is pulled in, so if a given platform
506 needs to override this, it can.
508 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
509 protection layer was required, which is not implemented.
510 Bugzilla 1254, patch from Wolfgang Breyha.
512 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
513 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
514 tls_dhparam take prime identifiers. Also unbreak combination of
515 OpenSSL+DH_params+TLSSNI.
517 PP/39 Disable SSLv2 by default in OpenSSL support.
523 PP/01 Solaris build fix for Oracle's LDAP libraries.
524 Bugzilla 1109, patch from Stephen Usher.
526 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
528 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
531 TF/02 Fix a couple more cases where we did not log the error message
532 when unlink() failed. See also change 4.74-TF/03.
534 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
535 lock up or crash if it happened to be inside a call to libc when it
536 got a SIGUSR1 from exiwhat.
538 The SIGUSR1 handler appends the current process status to the process
539 log which is later printed by exiwhat. It used to use the general
540 purpose logging code to do this, but several functions it calls are
541 not safe for signals.
543 The new output code in the SIGUSR1 handler is specific to the process
544 log, and simple enough that it's easy to inspect for signal safety.
545 Removing some special cases also simplifies the general logging code.
546 Removing the spurious timestamps from the process log simplifies
549 TF/04 Improved ratelimit ACL condition.
551 The /noupdate option has been deprecated in favour of /readonly which
552 has clearer semantics. The /leaky, /strict, and /readonly update modes
553 are mutually exclusive. The update mode is no longer included in the
554 database key; it just determines when the database is updated. (This
555 means that when you upgrde Exim will forget old rate measurements.)
557 Exim now checks that the per_* options are used with an update mode that
558 makes sense for the current ACL. For example, when Exim is processing a
559 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
560 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
561 must specify per_mail/readonly. If you omit the update mode it defaults to
562 /leaky where that makes sense (as before) or /readonly where required.
564 The /noupdate option is now undocumented but still supported for
565 backwards compatibility. It is equivalent to /readonly except that in
566 ACLs where /readonly is required you may specify /leaky/noupdate or
567 /strict/noupdate which are treated the same as /readonly.
569 A useful new feature is the /count= option. This is a generalization
570 of the per_byte option, so that you can measure the throughput of other
571 aggregate values. For example, the per_byte option is now equivalent
572 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
574 The per_rcpt option has been generalized using the /count= mechanism
575 (though it's more complicated than the per_byte equivalence). When it is
576 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
577 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
578 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
579 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
580 Note that using per_rcpt with a non-readonly update mode in more than
581 one ACL will cause the recipients to be double-counted. (The per_mail
582 and per_byte options don't have this problem.)
584 The handling of very low rates has changed slightly. If the computed rate
585 is less than the event's count (usually one) then this event is the first
586 after a long gap. In this case the rate is set to the same as this event's
587 count, so that the first message of a spam run is counted properly.
589 The major new feature is a mechanism for counting the rate of unique
590 events. The new per_addr option counts the number of different
591 recipients that someone has sent messages to in the last time period. It
592 behaves like per_rcpt if all the recipient addresses are different, but
593 duplicate recipient addresses do not increase the measured rate. Like
594 the /count= option this is a general mechanism, so the per_addr option
595 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
596 example, measure the rate that a client uses different sender addresses
597 with the options per_mail/unique=$sender_address. There are further
598 details in the main documentation.
600 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
602 TF/06 Removed a few PCRE remnants.
604 TF/07 Automatically extract Exim's version number from tags in the git
605 repository when doing development or release builds.
607 PP/02 Raise smtp_cmd_buffer_size to 16kB.
608 Bugzilla 879. Patch from Paul Fisher.
610 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
611 Heavily based on revision 40f9a89a from Simon Arlott's tree.
614 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
616 PP/05 Variable $av_failed, true if the AV scanner deferred.
617 Bugzilla 1078. Patch from John Horne.
619 PP/06 Stop make process more reliably on build failure.
620 Bugzilla 1087. Patch from Heiko Schlittermann.
622 PP/07 Make maildir_use_size_file an _expandable_ boolean.
623 Bugzilla 1089. Patch from Heiko Schlittermann.
625 PP/08 Handle ${run} returning more data than OS pipe buffer size.
626 Bugzilla 1131. Patch from Holger Weiß.
628 PP/09 Handle IPv6 addresses with SPF.
629 Bugzilla 860. Patch from Wolfgang Breyha.
631 PP/10 GnuTLS: support TLS 1.2 & 1.1.
633 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
636 PP/11 match_* no longer expand right-hand-side by default.
637 New compile-time build option, EXPAND_LISTMATCH_RHS.
638 New expansion conditions, "inlist", "inlisti".
640 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
642 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
644 PP/14 fix log_write() format string regression from TF/03.
645 Bugzilla 1152. Patch from Dmitry Isaikin.
651 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
653 PP/02 Harmonised TLS library version reporting; only show if debugging.
654 Layout now matches that introduced for other libraries in 4.74 PP/03.
656 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
658 PP/04 New "dns_use_edns0" global option.
660 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
663 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
664 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
666 TK/01 Updated PolarSSL code to 0.14.2.
667 Bugzilla 1097. Patch from Andreas Metzler.
669 PP/07 Catch divide-by-zero in ${eval:...}.
672 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
675 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
676 format-string attack -- SECURITY: remote arbitrary code execution.
678 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
679 time unintentionally subject to list matching rules, letting the header
680 cause arbitrary Exim lookups (of items which can occur in lists, *not*
681 arbitrary string expansion). This allowed for information disclosure.
683 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
684 INT_MIN/-1 -- value coerced to INT_MAX.
690 NM/01 Workround for PCRE version dependency in version reporting
693 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
694 This fixes portability to compilers other than gcc, notably
695 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
697 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
698 makefiles for portability to HP-UX and POSIX correctness.
700 PP/01 Permit LOOKUP_foo enabling on the make command-line.
701 Also via indented variable definition in the Makefile.
702 (Debugging by Oliver Heesakkers).
704 PP/02 Restore caching of spamd results with expanded spamd_address.
705 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
707 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
708 Improves build reliability. Fix from: Frank Elsner
710 NM/02 Fix wide character breakage in the rfc2047 coding
711 Fixes bug 1064. Patch from Andrey N. Oktyabrski
713 NM/03 Allow underscore in dnslist lookups
714 Fixes bug 1026. Patch from Graeme Fowler
716 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
717 Code patches from Adam Ciarcinski of NetBSD.
719 NM/04 Fixed exiqgrep to cope with mailq missing size issue
722 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
723 is logged, to avoid truncation. Patch from John Horne.
725 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
726 Patch from Jakob Hirsch.
728 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
729 SQL string expansion failure details.
730 Patch from Andrey Oktyabrski.
732 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
733 Patch from Simon Arlott.
735 PP/09 New lookups functionality failed to compile on old gcc which rejects
736 extern declarations in function scope.
737 Patch from Oliver Fleischmann
739 PP/10 Use sig_atomic_t for flags set from signal handlers.
740 Check getgroups() return and improve debugging.
741 Fixed developed for diagnosis in bug 927 (which turned out to be
744 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
745 Patch from Mark Zealey.
747 PP/12 Bugzilla 1056: Improved spamd server selection.
748 Patch from Mark Zealey.
750 PP/13 Bugzilla 1086: Deal with maildir quota file races.
751 Based on patch from Heiko Schlittermann.
753 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
754 Patch from Uwe Doering, sign-off by Michael Haardt.
756 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
757 variable declaration deep within a block. Bug and patch from
760 PP/15 lookups-Makefile IRIX compatibilty coercion.
762 PP/16 Make DISABLE_DKIM build knob functional.
764 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
765 Patch by Simon Arlott
767 TF/03 Fix valgrind.h portability to C89 compilers that do not support
768 variable argument macros. Our copy now differs from upstream.
774 TF/01 Failure to get a lock on a hints database can have serious
775 consequences so log it to the panic log.
777 TF/02 Log LMTP confirmation messages in the same way as SMTP,
778 controlled using the smtp_confirmation log selector.
780 TF/03 Include the error message when we fail to unlink a spool file.
782 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
783 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
784 for maintaining out-of-tree patches for some time.
786 PP/01 Bugzilla 139: Documentation and portability issues.
787 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
788 Handle per-OS dynamic-module compilation flags.
790 PP/02 Let /dev/null have normal permissions.
791 The 4.73 fixes were a little too stringent and complained about the
792 permissions on /dev/null. Exempt it from some checks.
793 Reported by Andreas M. Kirchwitz.
795 PP/03 Report version information for many libraries, including
796 Exim version information for dynamically loaded libraries. Created
797 version.h, now support a version extension string for distributors
798 who patch heavily. Dynamic module ABI change.
800 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
801 privilege escalation vulnerability whereby the Exim run-time user
802 can cause root to append content of the attacker's choosing to
805 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
808 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
809 If dropping privileges for untrusted macros, we disabled normal logging
810 on the basis that it would fail; for the Exim run-time user, this is not
811 the case, and it resulted in successful deliveries going unlogged.
812 Fixed. Reported by Andreas Metzler.
818 PP/01 Date: & Message-Id: revert to normally being appended to a message,
819 only prepend for the Resent-* case. Fixes regression introduced in
820 Exim 4.70 by NM/22 for Bugzilla 607.
822 PP/02 Include check_rfc2047_length in configure.default because we're seeing
823 increasing numbers of administrators be bitten by this.
825 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
827 PP/03 Bugzilla 994: added openssl_options main configuration option.
829 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
831 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
833 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
835 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
836 without a peer certificate, leading to a segfault because of an
837 assumption that peers always have certificates. Be a little more
838 paranoid. Problem reported by Martin Tscholak.
840 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
841 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
842 NB: ClamAV planning to remove STREAM in "middle of 2010".
843 CL also introduces -bmalware, various -d+acl logging additions and
844 more caution in buffer sizes.
846 PP/09 Implemented reverse_ip expansion operator.
848 PP/10 Bugzilla 937: provide a "debug" ACL control.
850 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
852 PP/12 Bugzilla 973: Implement --version.
854 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
856 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
858 PP/15 Bugzilla 816: support multiple condition rules on Routers.
860 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
861 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
862 ignore trailing whitespace.
864 JJ/02 prevent non-panic DKIM error from being sent to paniclog
866 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
869 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
870 Notification from Dr Andrew Aitchison.
872 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
873 ExtendedDetectionInfo response format.
874 Notification from John Horne.
876 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
879 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
880 XSL and documented dependency on system catalogs, with examples of how
883 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
886 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
887 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
888 configuration file which is writeable by the Exim user or group.
890 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
891 of configuration files to cover files specified with the -C option if
892 they are going to be used with root privileges, not just the default
895 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
896 option (effectively making it always true).
898 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
899 files to be used while preserving root privileges.
901 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
902 that rogue child processes cannot use them.
904 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
905 run-time user, instead of root.
907 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
908 Exim run-time user without dropping privileges.
910 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
911 result string, instead of calling string_vformat() twice with the same
914 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
915 for other users. Others should always drop root privileges if they use
916 -C on the command line, even for a whitelisted configure file.
918 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
920 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
926 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
927 $data_path, and $header_path variables; fixed documentation bugs and
930 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
931 exipick to access non-standard spools, including the "frozen" queue
934 NM/01 Bugzilla 965: Support mysql stored procedures.
935 Patch from Alain Williams
937 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
939 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
940 Patch from Andreas Metzler
942 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
943 Patch from Kirill Miazine
945 NM/05 Bugzilla 671: Added umask to procmail example.
947 JJ/03 installed exipick 20100323.0, fixing doc bug
949 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
950 directory. Notification and patch from Dan Rosenberg.
952 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
954 TK/02 Improve log output when DKIM signing operation fails.
956 MH/01 Treat the transport option dkim_domain as a colon separated
957 list, not as a single string, and sign the message with each element,
958 omitting multiple occurences of the same signer.
960 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
961 Bugzilla 985, 986. Patch by Simon Arlott
963 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
964 Patch by Simon Arlott
966 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
967 MBX locking. Notification from Dan Rosenberg.
973 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
975 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
977 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
979 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
981 NM/04 Bugzilla 915: Flag broken perl installation during build.
987 TK/01 Added patch by Johannes Berg that expands the main option
988 "spamd_address" if it starts with a dollar sign.
990 TK/02 Write list of recipients to X-Envelope-Sender header when building
991 the mbox-format spool file for content scanning (suggested by Jakob
994 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
995 (http://www.dcc-servers.net/) support via dccifd. Activated by
996 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
998 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
999 by Mark Daniel Reidel <mr@df.eu>.
1001 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1002 When building exim an external PCRE library is now needed -
1003 PCRE is a system library on the majority of modern systems.
1004 See entry on PCRE_LIBS in EDITME file.
1006 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1007 conversation. Added nologin parameter to request.
1008 Patch contributed by Kirill Miazine.
1010 TF/01 Do not log submission mode rewrites if they do not change the address.
1012 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1014 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1015 log files in place. Contributed by Roberto Lima.
1017 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1019 TF/03 Bugzilla 615: When checking the local_parts router precondition
1020 after a local_part_suffix or local_part_prefix option, Exim now
1021 does not use the address's named list lookup cache, since this
1022 contains cached lookups for the whole local part.
1024 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1025 Robert Millan. Documentation is in experimental-spec.txt.
1027 TF/04 Bugzilla 668: Fix parallel build (make -j).
1029 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1031 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1032 Patch provided by Jan Srzednicki.
1034 TF/05 Leading white space used to be stripped from $spam_report which
1035 wrecked the formatting. Now it is preserved.
1037 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1038 that they are available at delivery time.
1040 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1042 TF/08 TLS error reporting now respects the incoming_interface and
1043 incoming_port log selectors.
1045 TF/09 Produce a more useful error message if an SMTP transport's hosts
1046 setting expands to an empty string.
1048 NM/06 Bugzilla 744: EXPN did not work under TLS.
1049 Patch provided by Phil Pennock.
1051 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1052 Patch provided by Richard Godbee.
1054 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1055 acl_smtp_notquit, added index entry.
1057 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1058 Patch provided by Eugene Bujak.
1060 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1061 accept(). Patch provided by Maxim Dounin.
1063 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1064 Patch provided by Phil Pennock.
1066 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1068 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1069 Patch provided by Brad "anomie" Jorsch.
1071 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1072 Patch provided by Dean Brooks.
1074 TK/05 Add native DKIM support (does not depend on external libraries).
1076 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1077 Patch provided by Graeme Fowler.
1079 NM/16 Bugzilla 851: Documentation example syntax fix.
1081 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1083 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1086 NM/19 Bugzilla 745: TLS version reporting.
1087 Patch provided by Phil Pennock.
1089 NM/20 Bugzilla 167: bool: condition support.
1090 Patch provided by Phil Pennock.
1092 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1093 clients. Patch provided by Phil Pennock.
1095 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1096 Patch provided by Brad "anomie" Jorsch.
1098 NM/23 Bugzilla 687: Fix misparses in eximstats.
1099 Patch provided by Heiko Schlittermann.
1101 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1102 Patch provided by Heiko Schlittermann.
1104 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1105 plus update to original patch.
1107 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1109 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1110 Patch provided by David Brownlee.
1112 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1114 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1116 NM/30 Bugzilla 888: TLS documentation bugfixes.
1118 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1120 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1121 Unlike the original bugzilla I have changed all shell scripts in src tree.
1123 NM/33 Bugzilla 898: Transport filter timeout fix.
1124 Patch by Todd Rinaldo.
1126 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1127 Patch by Serge Demonchaux.
1129 NM/35 Bugzilla 39: Base64 decode bug fixes.
1130 Patch by Jakob Hirsch.
1132 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1134 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1136 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1138 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1144 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1145 ALT-N's libdkim that I have put here:
1146 http://duncanthrax.net/exim-experimental/
1148 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1149 were called 'true' and it seems that C99 defines that as a reserved
1150 keyword to be used with 'bool' variable types. That means you could
1151 not include C99-style headers which use bools without triggering
1152 build errors in sieve.c.
1154 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1155 as mailq or other aliases. Changed the --help handling significantly
1156 to do whats expected. exim_usage() emits usage/help information.
1158 SC/01 Added the -bylocaldomain option to eximstats.
1160 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1162 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1164 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1170 PH/01 Another patch from the Sieve maintainer.
1172 PH/02 When an IPv6 address is converted to a string for single-key lookup
1173 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1174 dots are used instead of colons so that keys in lsearch files need not
1175 contain colons. This was done some time before quoting was made available
1176 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1177 (notated using the quote facility) so as to distinguish them from IPv4
1178 keys. This meant that lookups for IP addresses in host lists did not work
1179 for iplsearch lookups.
1181 This has been fixed by arranging for IPv6 addresses to be expressed with
1182 colons if the lookup type is iplsearch. This is not incompatible, because
1183 previously such lookups could never work.
1185 The situation is now rather anomolous, since one *can* have colons in
1186 ordinary lsearch keys. However, making the change in all cases is
1187 incompatible and would probably break a number of configurations.
1189 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1192 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1193 conversion specification without a maximum field width, thereby enabling
1194 a rogue spamd server to cause a buffer overflow. While nobody in their
1195 right mind would setup Exim to query an untrusted spamd server, an
1196 attacker that gains access to a server running spamd could potentially
1197 exploit this vulnerability to run arbitrary code as the Exim user.
1199 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1200 $primary_hostname instead of what libspf2 thinks the hosts name is.
1202 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1203 a directory entry by the name of the lookup key. Previously, if a
1204 symlink pointed to a non-existing file or a file in a directory that
1205 Exim lacked permissions to read, a lookup for a key matching that
1206 symlink would fail. Now it is enough that a matching directory entry
1207 exists, symlink or not. (Bugzilla 503.)
1209 PH/03 The body_linecount and body_zerocount variables are now exported in the
1212 PH/04 Added the $dnslist_matched variable.
1214 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1215 This means they are set thereafter only if the connection becomes
1218 PH/06 Added the client_condition to authenticators so that some can be skipped
1219 by clients under certain conditions.
1221 PH/07 The error message for a badly-placed control=no_multiline_responses left
1222 "_responses" off the end of the name.
1224 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1226 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1227 (without spaces) instead of just copying the configuration text.
1229 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1231 PH/11 Added $max_received_linelength.
1233 PH/12 Added +ignore_defer and +include_defer to host lists.
1235 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1236 way in which PCRE > 7.0 is built.
1238 PH/14 Implemented queue_only_load_latch.
1240 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1241 MAIL command. The effect was to mangle the value on 64-bit systems.
1243 PH/16 Another patch from the Sieve maintainer.
1245 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1247 PH/18 If a system quota error occurred while trying to create the file for
1248 a maildir delivery, the message "Mailbox is full" was not appended to the
1249 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1250 only to a quota excession during the actual writing of the file.
1252 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1253 characters?) which causes problems in log lines. The DN values are now
1254 passed through string_printing() before being added to log lines.
1256 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1257 and InterBase are left for another time.)
1259 PH/21 Added message_body_newlines option.
1261 PH/22 Guard against possible overflow in moan_check_errorcopy().
1263 PH/23 POSIX allows open() to be a macro; guard against that.
1265 PH/24 If the recipient of an error message contained an @ in the local part
1266 (suitably quoted, of course), incorrect values were put in $domain and
1267 $local_part during the evaluation of errors_copy.
1273 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1274 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1277 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1278 issue a MAIL command.
1280 PH/02 In an ACL statement such as
1282 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1284 if a client was not listed at all, or was listed with a value other than
1285 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1286 the condition was not true (as it should be), so access was not denied.
1287 The bug was that the ! inversion was incorrectly passed on to the second
1288 item. This has been fixed.
1290 PH/03 Added additional dnslists conditions == and =& which are different from
1291 = and & when the dns lookup returns more than one IP address.
1293 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1294 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1296 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1297 FSYNC, which compiles an option called disable_fsync that allows for
1298 bypassing fsync(). The documentation is heavily laced with warnings.
1300 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1302 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1303 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1304 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1305 including adding "make clean"; (3) Added -fPIC when compiling the test
1306 dynamically loaded module, to get rid of a warning.
1308 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1309 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1310 = 0s. The bug is otherwise harmless.
1312 PH/07 There was a bug in the dovecot authenticator such that the value of
1313 $auth1 could be overwritten, and so not correctly preserved, after a
1314 successful authentication. This usually meant that the value preserved by
1315 the server_setid option was incorrect.
1317 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1319 PH/09 Installed PCRE release 7.0.
1321 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1322 run for batched SMTP input. It is now run at the start of every message
1323 in the batch. While fixing this I discovered that the process information
1324 (output by running exiwhat) was not always getting set for -bs and -bS
1325 input. This is fixed, and it now also says "batched" for BSMTP.
1327 PH/11 Added control=no_pipelining.
1329 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1330 patch, slightly modified), and move the expansion of helo_data till after
1331 the connection is made in the smtp transport (so it can use these
1334 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1336 PH/14 Added log_selector = +pid.
1338 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1340 PH/16 Add ${if forany and ${if forall.
1342 PH/17 Added dsn_from option to vary the From: line in DSNs.
1344 PH/18 Flush SMTP output before performing a callout, unless control =
1345 no_callout_flush is set.
1347 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1348 was true (the default) a successful delivery failed to delete the retry
1349 item, thus causing premature timeout of the address. The bug is now
1352 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1354 PH/21 Long custom messages for fakedefer and fakereject are now split up
1355 into multiline reponses in the same way that messages for "deny" and
1356 other ACL rejections are.
1358 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1359 with slight modification.
1361 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1362 draft, changing the syntax and factoring some duplicate code".
1364 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1365 for deliveries of the second and subsequent messages over the same SMTP
1368 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1369 ${reduce, with only minor "tidies".
1371 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1373 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1374 expansion side effects.
1376 PH/27 When a message times out after an over-quota error from an Exim-imposed
1377 quota, the bounce message says "mailbox is full". This message was not
1378 being given when it was a system quota that was exceeded. It now should
1381 MH/03 Made $recipients available in local_scan(). local_scan() already has
1382 better access to the recipient list through recipients_list[], but
1383 $recipients can be useful in postmaster-provided expansion strings.
1385 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1386 in the case of a MAIL command with additional options following the
1387 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1388 were accidentally chopped off.
1390 PH/29 SMTP synchronization checks are implemented when a command is read -
1391 there is a check that no more input is waiting when there shouldn't be
1392 any. However, for some commands, a delay in an ACL can mean that it is
1393 some time before the response is written. In this time, more input might
1394 arrive, invalidly. So now there are extra checks after an ACL has run for
1395 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1396 pipelining has not been advertised.
1398 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1400 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1401 returned in the SMTP response when smtp_return_error_details was false.
1402 This has been fixed.
1404 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1405 instead of the C I/O that was originally supplied, because problems were
1406 reported on Solaris.
1408 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1409 Exim which did not show up earlier: it was assuming that a call to
1410 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1411 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1412 was a macro that became an assignment, so it seemed to work. This has
1413 changed to a proper function call with a void return, hence the compile
1414 error. Exim's code has been fixed.
1416 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1419 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1421 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1423 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1424 because it turns out to be too big or there is a timeout) there is no
1425 "Completed" line in the log. When some messages of this type were
1426 selected by exigrep, they were listed as "not completed". Others were
1427 picked up by some special patterns. I have improved the selection
1428 criteria to be more general.
1430 PH/38 The host_find_failed option in the manualroute router can now be set
1431 to "ignore", to completely ignore a host whose IP address cannot be
1432 found. If all hosts are ignored, the behaviour is controlled by the new
1433 host_all_ignored option.
1435 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1436 homing or because of multiple MX records with /mx) generated more than
1437 one IP address, and the following item turned out to be the local host,
1438 all the secondary addresses of the first item were incorrectly removed
1439 from the list, along with the local host and any following hosts (which
1440 is what is supposed to happen).
1442 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1443 whoever called Exim into the -H file. In the case of the daemon it was
1444 behaving confusingly. When first started, it used values for whoever
1445 started the daemon, but after a SIGHUP it used the Exim user (because it
1446 calls itself on a restart). I have changed the code so that it now always
1449 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1450 message are rejected with the same error (e.g. no authentication or bad
1451 sender address), and a DATA command is nevertheless sent (as can happen
1452 with PIPELINING or a stupid MUA), the error message that was given to the
1453 RCPT commands is included in the rejection of the DATA command. This is
1454 intended to be helpful for MUAs that show only the final error to their
1457 PH/42 Another patch from the Sieve maintainer.
1459 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1460 Eximstats - Fixed some broken HTML links and added missing column headers
1462 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1463 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1465 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1467 PH/43 Yet another patch from the Sieve maintainer.
1469 PH/44 I found a way to check for a TCP/IP connection going away before sending
1470 the response to the final '.' that terminates a message, but only in the
1471 case where the client has not sent further data following the '.'
1472 (unfortunately, this is allowed). However, in many cases there won't be
1473 any further data because there won't be any more messages to send. A call
1474 to select() can be used: if it shows that the input is "ready", there is
1475 either input waiting, or the socket has been closed. An attempt to read
1476 the next input character can distinguish the two cases. Previously, Exim
1477 would have sent an OK response which the client would never have see.
1478 This could lead to message repetition. This fix should cure that, at
1479 least in a lot of common cases.
1481 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1482 advertised in response to EHLO.
1488 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1489 fixed by 4.65/MH/01 (is this a record?) are fixed:
1491 (i) An empty string was always treated as zero by the numeric comparison
1492 operators. This behaviour has been restored.
1494 (ii) It is documented that the numeric comparison operators always treat
1495 their arguments as decimal numbers. This was broken in that numbers
1496 starting with 0 were being interpreted as octal.
1498 While fixing these problems I realized that there was another issue that
1499 hadn't been noticed. Values of message_size_limit (both the global option
1500 and the transport option) were treated as octal if they started with 0.
1501 The documentation was vague. These values are now always treated as
1502 decimal, and I will make that clear in the documentation.
1508 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1509 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1512 MH/01 Don't check that the operands of numeric comparison operators are
1513 integers when their expansion is in "skipping" mode (fixes bug
1514 introduced by 4.64-PH/07).
1516 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1517 child addresses, Exim now panics and dies. Previously, because the count
1518 is held in a short int, deliveries were likely to be lost. As such a
1519 large number of recipients for a single message is ridiculous
1520 (performance will be very, very poor), I have chosen to impose a limit
1521 rather than extend the field.
1527 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1528 leftover -K file (the existence of which was triggered by #402).
1529 While we were at it, introduced process PID as part of the -K
1530 filename. This should rule out race conditions when creating
1533 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1534 processing considerably. Previous code took too long for large mails,
1535 triggering a timeout which in turn triggers #401.
1537 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1538 in the DK code in transports.c. sendfile() is not really portable,
1539 hence the _LINUX specificness.
1541 TF/01 In the add_headers option to the mail command in an Exim filter,
1542 there was a bug that Exim would claim a syntax error in any
1543 header after the first one which had an odd number of characters
1546 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1547 callout verification, Exim cached a "reject" for the entire domain. This
1548 is correct for most verifications, but it is not correct for a recipient
1549 verification with use_sender or use_postmaster set, because in that case
1550 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1551 case of MAIL FROM:<> rejection from other early rejections (e.g.
1552 rejection of HELO). When verifying a recipient using a non-null MAIL
1553 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1554 Whatever the result of the callout, the value of the domain cache is
1555 left unchanged (for any other kind of callout, getting as far as trying
1556 RCPT means that the domain itself is ok).
1558 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1561 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1562 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1563 session, the daemon ignores ECONNECT errors and logs others; it now
1564 ignores EPIPE as well.
1566 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1567 (quoted-printable decoding).
1569 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1570 later the small subsequent patch to fix an introduced bug.
1572 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1574 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1576 PH/08 An error is now given if message_size_limit is specified negative.
1578 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1579 to be given (somewhat) arbitrary names.
1581 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1584 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1585 miscellaneous code fixes
1587 PH/10 Added the log_reject_target ACL modifier to specify where to log
1590 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1591 hostname. This is wrong, because it relates to the incoming message (and
1592 probably the interface on which it is arriving) and not to the outgoing
1593 callout (which could be using a different interface). This has been
1594 changed to use the value of the helo_data option from the smtp transport
1595 instead - this is what is used when a message is actually being sent. If
1596 there is no remote transport (possible with a router that sets up host
1597 addresses), $smtp_active_hostname is used.
1599 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1600 tweaks were necessary in order to get it to work (see also 21 below):
1601 (a) The code assumed that strncpy() returns a negative number on buffer
1602 overflow, which isn't the case. Replaced with Exim's string_format()
1604 (b) There were several signed/unsigned issues. I just did the minimum
1605 hacking in of casts. There is scope for a larger refactoring.
1606 (c) The code used strcasecmp() which is not a standard C function.
1607 Replaced with Exim's strcmpic() function.
1608 (d) The code set only $1; it now sets $auth1 as well.
1609 (e) A simple test gave the error "authentication client didn't specify
1610 service in request". It would seem that Dovecot has changed its
1611 interface. Fortunately there's a specification; I followed it and
1612 changed what the client sends and it appears to be working now.
1614 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1617 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1618 address A is aliased to B and C, where B exists and C does not. Without
1619 -v the output is "A verified" because verification stops after a
1620 successful redirection if more than one address is generated. However,
1621 with -v the child addresses are also verified. Exim was outputting "A
1622 failed to verify" and then showing the successful verification for C,
1623 with its parentage. It now outputs "B failed to verify", showing B's
1624 parentage before showing the successful verification of C.
1626 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1627 look up a TXT record in a specific list after matching in a combined
1630 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1631 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1632 they consult the DNS. I had assumed they would set it the way they
1633 wanted; and indeed my experiments on Linux seem to show that in some
1634 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1635 To be on the safe side, however, I have now made the interface to
1636 host_find_byname() similar to host_find_bydns(), with an argument
1637 containing the DNS resolver options. The host_find_byname() function now
1638 sets these options at its start, just as host_find_bydns() does. The smtp
1639 transport options dns_qualify_single and dns_search_parents are passed to
1640 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1641 of host_find_byname() use the default settings of RES_DEFNAMES
1642 (qualify_single) but not RES_DNSRCH (search_parents).
1644 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1645 spool_read_header() do less string testing, by means of a preliminary
1646 switch on the second character of optional "-foo" lines. (This is
1647 overdue, caused by the large number of possibilities that now exist.
1648 Originally there were few.) While I was there, I also converted the
1649 str(n)cmp tests so they don't re-test the leading "-" and the first
1650 character, in the hope this might squeeze out yet more improvement.
1652 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1653 flag allowing group syntax was set by the header_syntax check but not
1654 turned off, possible causing trouble later; (2) The flag was not being
1655 set at all for the header_verify test, causing "group"-style headers to
1656 be rejected. I have now set it in this case, and also caused header_
1657 verify to ignore an empty address taken from a group. While doing this, I
1658 came across some other cases where the code for allowing group syntax
1659 while scanning a header line wasn't quite right (mostly, not resetting
1660 the flag correctly in the right place). These bugs could have caused
1661 trouble for malformed header lines. I hope it is now all correct.
1663 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1664 with the "reply" argument non-NULL. The code, however (which originally
1665 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1666 but it didn't always do it. This confused somebody who was copying the
1667 code for some other use. I have removed all the tests.
1669 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1670 feature that was used to support insecure browsers during the U.S. crypto
1671 embargo. It requires special client support, and Exim is probably the
1672 only MTA that supported it -- and would never use it because real RSA is
1673 always available. This code has been removed, because it had the bad
1674 effect of slowing Exim down by computing (never used) parameters for the
1675 RSA_EXPORT functionality.
1677 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1678 authenticator to fail if there's a tab character in the incoming data
1679 (there should never be unless someone is messing about, as it's supposed
1680 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1681 if the connection is using TLS or if the remote IP is the same as the
1682 local IP, and the "valid-client-cert option" if a client certificate has
1685 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1686 authenticators. This can be used for authorization after authentication
1687 succeeds. (In the case of plaintext, it servers for both authentication
1690 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1691 if any retry times were supplied.
1693 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1694 connection, where there is no client IP address to check. In this
1695 situation, the verify now always succeeds.
1697 PH/25 Applied John Jetmore's -Mset patch.
1699 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1701 PH/27 In a string expansion for a processed (not raw) header when multiple
1702 headers of the same name were present, leading whitespace was being
1703 removed from all of them, but trailing whitespace was being removed only
1704 from the last one. Now trailing whitespace is removed from each header
1705 before concatenation. Completely empty headers in a concatenation (as
1706 before) are ignored.
1708 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1709 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1711 PH/29 [Removed. This was a change that I later backed out, and forgot to
1712 correct the ChangeLog entry (that I had efficiently created) before
1713 committing the later change.]
1715 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1716 address errors (4xx response to RCPT) over the same connection as other
1717 messages routed to the same hosts. Such deliveries are always "forced",
1718 so retry times are not inspected. This resulted in far too many retries
1719 for the affected addresses. The effect occurred only when there were more
1720 hosts than the hosts_max_try setting in the smtp transport when it had
1721 the 4xx errors. Those hosts that it had tried were not added to the list
1722 of hosts for which the message was waiting, so if all were tried, there
1723 was no problem. Two fixes have been applied:
1725 (i) If there are any address or message errors in an SMTP delivery, none
1726 of the hosts (tried or untried) are now added to the list of hosts
1727 for which the message is waiting, so the message should not be a
1728 candidate for sending over the same connection that was used for a
1729 successful delivery of some other message. This seems entirely
1730 reasonable: after all the message is NOT "waiting for some host".
1731 This is so "obvious" that I'm not sure why it wasn't done
1732 previously. Hope I haven't missed anything, but it can't do any
1733 harm, as the worst effect is to miss an optimization.
1735 (ii) If, despite (i), such a delivery is accidentally attempted, the
1736 routing retry time is respected, so at least it doesn't keep
1737 hammering the server.
1739 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1740 in ${readsocket because some servers need this prod.
1742 PH/32 Added some extra debug output when updating a wait-xxx database.
1744 PH/33 The hint "could be header name not terminated by colon", which has been
1745 given for certain expansion errors for a long time, was not being given
1746 for the ${if def:h_colon_omitted{... case.
1748 PH/34 The spec says: "With one important exception, whenever a domain list is
1749 being scanned, $domain contains the subject domain." There was at least
1750 one case where this was not true.
1752 PH/35 The error "getsockname() failed: connection reset by peer" was being
1753 written to the panic log as well as the main log, but it isn't really
1754 panic-worthy as it just means the connection died rather early on. I have
1755 removed the panic log writing for the ECONNRESET error when getsockname()
1758 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1759 runs only) independently of the message's sender address. This meant
1760 that, if the 4xx error was in fact related to the sender, a different
1761 message to the same recipient with a different sender could confuse
1762 things. In particualar, this can happen when sending to a greylisting
1763 server, but other circumstances could also provoke similar problems.
1764 I have changed the default so that the retry time for these errors is now
1765 based a combination of the sender and recipient addresses. This change
1766 can be overridden by setting address_retry_include_sender=false in the
1769 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1770 remote server are returned as part of bounce messages. This was not
1771 happening for LMTP over a pipe (the lmtp transport), but now it is the
1772 same for both kinds of LMTP.
1774 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1775 in header lines that were in fact CNAMEs. This is no longer the case.
1777 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1778 and queue runs started by the daemon processed all messages. This has
1779 been fixed so that -R and -S can now usefully be given with -q<time>.
1781 PH/40 Import PCRE release 6.7 (fixes some bugs).
1783 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1785 PH/42 Give an error if -q is specified more than once.
1787 PH/43 Renamed the variables $interface_address and $interface_port as
1788 $received_ip_address and $received_port, to make it clear that these
1789 values apply to message reception, and not to the outgoing interface when
1790 a message is delivered. (The old names remain recognized, of course.)
1792 PH/44 There was no timeout on the connect() call when using a Unix domain
1793 socket in the ${readsocket expansion. There now is.
1795 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1796 be meaningful with "accept".
1798 SC/01 Eximstats V1.43
1799 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1801 SC/02 Eximstats V1.44
1802 Use a glob alias rather than an array ref in the generated
1803 parser. This improves both readability and performance.
1805 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1806 Collect SpamAssassin and rejection statistics.
1807 Don't display local sender or destination tables unless
1808 there is data to show.
1809 Added average volumes into the top table text output.
1811 SC/04 Eximstats V1.46
1812 Collect data on the number of addresses (recipients)
1813 as well as the number of messages.
1815 SC/05 Eximstats V1.47
1816 Added 'Message too big' to the list of mail rejection
1817 reasons (thanks to Marco Gaiarin).
1819 SC/06 Eximstats V1.48
1820 Mainlog lines which have GMT offsets and are too short to
1821 have a flag are now skipped.
1823 SC/07 Eximstats V1.49 (Alain Williams)
1824 Added the -emptyok flag.
1826 SC/08 Eximstats V1.50
1827 Fixes for obtaining the IP address from reject messages.
1829 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1830 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1831 whitesspace changes from 4.64-PH/27
1833 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1836 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1837 are found, allow negative numbers in numeric criteria)
1839 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1841 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1842 to match changes made in 4.64-PH/43
1844 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1846 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1847 "domain" is actually an IP address, and force a failure. This locks out
1848 those revolvers/nameservers that support "A-for-A" lookups, in
1849 contravention of the specifications.
1851 PH/48 When a host name was looked up from an IP address, and the subsequent
1852 forward lookup of the name timed out, the host name was left in
1853 $sender_host_name, contrary to the specification.
1855 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1856 restricted to single-key lookups, Exim was not diagnosing an error if
1857 * or *@ was used with a query-style lookup.
1859 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1861 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1862 long ago, but noone interested enough thought of it. Let's just say that
1863 the "1.1" means that there are some new functions that weren't there at
1864 some point in the past.
1866 PH/51 Error processing for expansion failure of helo_data from an smtp
1867 transport during callout processing was broken.
1869 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1870 tested/used via the -bh/-bhc/-bs options.
1872 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1873 bug, fixed in subsequent PCRE releases).
1875 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1876 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1878 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1884 SC/01 Use a glob alias rather than an array ref in eximstats generated
1885 parser. This improves both readability and performance.
1887 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1888 Don't display local sender or destination tables in eximstats unless
1889 there is data to show.
1890 Added average volumes into the eximstats top table text output.
1892 SC/03 Collect data on the number of addresses (recipients) as well
1893 as the number of messages in eximstats.
1895 TF/01 Correct an error in the documentation for the redirect router. Exim
1896 does (usually) call initgroups() when daemonizing.
1898 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1899 with consistent privilege compared to when running as a daemon.
1901 TF/03 Note in the spec that $authenticated_id is not set for local
1902 submissions from trusted users.
1904 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1905 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1907 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1908 by adding some example configuration directives to the default
1909 configuration file. A little bit of work is required to uncomment the
1910 directives and define how usernames and passwords are checked, but
1911 there is now a framework to start from.
1913 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1914 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1915 without this. I don't know how relevant this is to other LDAP libraries.
1917 PH/02 Add the verb name to the "unknown ACL verb" error.
1919 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1921 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1923 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1924 directory not expanded when it should be if an expanded home directory
1925 was set for the address (which is overridden by the transport).
1927 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1930 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1931 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1932 because it is too late at that time, and has no effect.
1934 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1935 security issue with \' (bugzilla #107). I could not use the
1936 PQescapeStringConn() function, because it needs a PGconn value as one of
1939 PH/08 When testing addresses using -bt, indicate those final addresses that
1940 are duplicates that would not cause an additional delivery. At least one
1941 person was confused, thinking that -bt output corresponded to deliveries.
1942 (Suppressing duplicates isn't a good idea as you lose the information
1943 about possibly different redirections that led to the duplicates.)
1945 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1946 systems where poll() doesn't work, in particular OS X.
1948 PH/10 Added more information to debugging output for retry time not reached.
1950 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1951 operations in malware.c.
1953 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1956 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1957 syslog_duplication set false, log lines that would normally be written
1958 both the the main log and to the reject log were not written to syslog at
1961 PH/14 In the default configuration, change the use of "message" in ACL warn
1962 statements to "add_header".
1964 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1965 not followed by a command (e.g. "seen endif").
1967 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1968 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1971 PH/17 Added extra conditions to the default value of delay_warning_condition
1975 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1976 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1977 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1980 The Auto-Submitted: and various List- headers are standardised, whereas I
1981 don't think Precedence: ever was.
1983 PH/18 Refactored debugging code in route_finduser() to show more information,
1984 in particular, the error code if getpwnam() issues one.
1986 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1987 This is apparently needed in addition to the PH/07 change above to avoid
1988 any possible encoding problems.
1990 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1991 but not after initializing Perl.
1993 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1994 output them only if debugging. By default they are written stderr,
1995 apparently, which is not desirable.
1997 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2000 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2003 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2005 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2006 authenticated or an ident call has been made. Suppress the default
2007 values for $authenticated_id and $authenticated_sender (but permit -oMai
2008 and -oMas) when testing with -bh.
2010 PH/24 Re-jigged the order of the tests in the default configuration so that the
2011 tests for valid domains and recipients precede the DNS black list and CSA
2012 tests, on the grounds that those ones are more expensive.
2014 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2015 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2016 This bug exists in every version of Exim that I still have, right back to
2019 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2020 However, an attempt to turn on case-sensitivity in a regex key by
2021 including (?-i) didn't work because the subject string was already
2022 lowercased, and the effects were non-intuitive. It turns out that a
2023 one-line patch can be used to allow (?-i) to work as expected.
2029 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2030 other effects) broke the use of negated acl sub-conditions.
2032 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2035 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2036 "Deny" causes Exim to reject the incoming connection with a 554 error.
2037 Unfortunately, if there is a major crisis, such as a disk failure,
2038 tcp-wrappers gives "deny", whereas what one would like would be some
2039 kind of temporary error. A kludge has been added to help with this.
2040 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2041 554 error is used if errno is still zero or contains ENOENT (which occurs
2042 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2045 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2047 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2048 errors. Otherwise a message that provokes a temporary error (when other
2049 messages do not) can cause a whole host to time out.
2051 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2052 addresses were routed directly to files or pipes from a redirect router.
2053 File deliveries just didn't batch; pipe deliveries might have suffered
2056 PH/06 A failure to get a lock for a hints database would erroneously always say
2057 "Failed to get write lock", even when it was really a read lock.
2059 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2060 of 0600. This has been changed to use the value of the lockfile_mode
2061 option (which defaults to 0600).
2063 PH/08 Applied small patch from the Sieve maintainer.
2065 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2066 folder from quota calculations, a direct delivery into this folder messed
2067 up the contents of the maildirsize file. This was because the regex was
2068 used only to exclude .Trash (or whatever) when the size of the mailbox
2069 was calculated. There was no check that a delivery was happening into an
2070 excluded directory. This bug has been fixed by ignoring all quota
2071 processing for deliveries into excluded directories.
2073 PH/10 Added the maildirfolder_create_regex option to appendfile.
2079 PH/01 The code for finding all the local interface addresses on a FreeBSD
2080 system running IPv6 was broken. This may well have applied to all BSD
2081 systems, as well as to others that have similar system calls. The broken
2082 code found IPv4 interfaces correctly, but gave incorrect values for the
2083 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2084 that it would not match correctly against @[] and not recognize the IPv6
2087 PH/02 The ipliteral router was not recognizing addresses of the form user@
2088 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2090 PH/03 Added disable_ipv6.
2092 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2093 decoded form, because it is most often used to construct To: headers
2094 lines in autoreplies, and the decoded form may well be syntactically
2095 invalid. However, $reply_address has leading white space removed, and all
2096 newlines turned into spaces so that the autoreply transport does not
2099 PH/05 If group was specified without a user on a router, and no group or user
2100 was specified on a transport, the group from the router was ignored.
2102 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2103 for visible compile-time settings that can be used to change these
2104 numbers, for those that want even more. Backwards compatibility with old
2105 spool files has been maintained. However, going back to a previous Exim
2106 release will lost any variables that are in spool files.
2108 PH/07 Two small changes when running in the test harness: increase delay when
2109 passing a TCP/IP connection to a new process, in case the original
2110 process has to generate a bounce, and remove special handling of
2111 127.0.0.2 (sic), which is no longer necessary.
2113 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2114 be the same on different OS.
2116 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2119 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2120 whether --show-vars was specified or not
2122 JJ/02 exipick: Added support for new ACL variable spool format introduced
2125 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2126 syntactically invalid From: or Reply-to: line, and a filter used this to
2127 generate an autoreply, and therefore failed to obtain an address for the
2128 autoreply, Exim could try to deliver to a non-existent relative file
2129 name, causing unrelated and misleading errors. What now happens is that
2130 it logs this as a hard delivery error, but does not attempt to create a
2133 PH/11 The exinext utility has a -C option for testing purposes, but although
2134 the given file was scanned by exinext itself; it wasn't being passed on
2135 when Exim was called.
2137 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2138 an end-of-file indication when reading a command response.
2140 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2141 compiled. In many other places in Exim, IPv6 addresses are always
2142 recognized, so I have changed this. It also means that IPv4 domain
2143 literals of the form [IPV4:n.n.n.n] are now always recognized.
2145 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2146 used if the router is not running as root, for example, when verifying at
2147 ACL time, or when using -bh. The debugging output from this situation was
2148 non-existent - all you got was a failure to exec. I have made two
2151 (a) Failures to set uid/gid, the current directory, or a process leader
2152 in a subprocess such as that created by queryprogram now generate
2153 suitable debugging ouput when -d is set.
2155 (b) The queryprogram router detects when it is not running as root,
2156 outputs suitable debugging information if -d is set, and then runs
2157 the subprocess without attempting to change uid/gid.
2159 PH/15 Minor change to Makefile for building test_host (undocumented testing
2162 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2163 additional section of a DNS packet that returns MX or SRV records.
2164 Instead, it always explicitly searches for A/AAAA records. This avoids
2165 major problems that occur when a DNS server includes only records of one
2166 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2167 fixed another bug: if SRV records were looked up and the corresponding
2168 address records were *not* found in the additional section, the port
2169 values from the SRV records were lost.
2171 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2172 using the correct key (the original address) when searching the retry
2173 rules in order to find which one to use for generating the retry hint.
2175 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2176 adding the default one. Similarly, if it contains a Reply-To: header, the
2177 errors_reply_to option, if set, is not used.
2179 PH/19 When calculating a retry time, Exim used to measure the "time since
2180 failure" by looking at the "first failed" field in the retry record. Now
2181 it does not use this if it is later than than the arrival time of the
2182 message. Instead it uses the arrival time. This makes for better
2183 behaviour in cases where some deliveries succeed, thus re-setting the
2184 "first failed" field. An example is a quota failure for a huge message
2185 when small messages continue to be delivered. Without this change, the
2186 "time since failure" will always be short, possible causing more frequent
2187 delivery attempts for the huge message than are intended.
2188 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2190 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2191 $1, $2, $3) because the numerical variables can be reset during some
2192 expansion items (e.g. "match"), thereby losing the authentication data.
2194 PH/21 Make -bV show the size of off_t variables so that the test suite can
2195 decide whether to run tests for quotas > 2G.
2197 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2198 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2199 filecount value is greater than 2G or if a quota value is greater than 2G
2200 on a system where the size of off_t is not greater than 4, a panic error
2203 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2204 never match. The debug and -bh output now contains an explicit error
2205 message indicating a malformed IPv4 address or mask.
2207 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2208 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2209 PH/23 above applies.
2211 PH/25 Do not write to syslog when running in the test harness. The only
2212 occasion when this arises is a failure to open the main or panic logs
2213 (for which there is an explicit test).
2215 PH/26 Added the /no_tell option to "control=freeze".
2217 PH/27 If a host name lookup failed very early in a connection, for example, if
2218 the IP address matched host_lookup and the reverse lookup yielded a name
2219 that did not have a forward lookup, an error message of the form "no IP
2220 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2221 could be logged. Now it outputs the IP address instead of "NULL".
2223 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2224 allows the sender and the authenticated sender to be set when
2225 submitting a message from within Exim. Since child_open_exim() is
2226 documented for local_scan(), the new function should be too.
2228 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2229 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2230 results in an empty string is now treated as unset.
2232 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2234 PH/31 Added sender_verify_fail logging option.
2236 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2237 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2238 tidied the source and removed it altogether.
2240 PH/33 When a queue run was abandoned because the load average was too high, a
2241 log line was always written; now it is written only if the queue_run log
2242 selector is set. In addition, the log line for abandonment now contains
2243 information about the queue run such as the pid. This is always present
2244 in "start" and "stop" lines but was omitted from the "abandon" line.
2246 PH/34 Omit spaces between a header name and the colon in the error message that
2247 is given when verify = headers_syntax fails (if there are lots of them,
2248 the message gets confusing).
2250 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2251 names, as there are now some PTR records that contain slashes. This check
2252 is only to protect against broken name servers that fall over on strange
2253 characters, so the fact that it applies to all lookups doesn't matter.
2255 PH/36 Now that the new test suite is complete, we can remove some of the
2256 special code in Exim that was needed for the old test suite. For example,
2257 sorting DNS records because real resolvers return them in an arbitrary
2258 order. The new test suite's fake resolver always returns records in the
2261 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2262 bounces) except when queue_only is set, to avoid logging races between
2263 the different processes.
2265 PH/38 Panic-die if .include specifies a non-absolute path.
2267 PH/39 A tweak to the "H" retry rule from its user.
2269 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2270 a label. They prevented compilation on older perls.
2272 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2273 a warning to be raised on newish perls.
2275 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2276 on queue. Changes to match documented behaviour of showing count of
2277 messages matching specified criteria.
2279 PH/40 Changed the default ident timeout from 30s to 5s.
2281 PH/41 Added support for the use of login_cap features, on those BSD systems
2282 that have them, for controlling the resources used by pipe deliveries.
2284 PH/42 The content-scanning code uses fopen() to create files in which to put
2285 message data. Previously it was not paying any attention to the mode of
2286 the files. Exim runs with umask(0) because the rest of the code creates
2287 files with open(), and sets the required mode explicitly. Thus, these
2288 files were ending up world-writeable. This was not a big issue, because,
2289 being within the spool directory, they were not world-accessible. I have
2290 created a function called modefopen, which takes an additional mode
2291 argument. It sets umask(777), creates the file, chmods it to the required
2292 mode, then resets the umask. All the relevant calls to fopen() in the
2293 content scanning code have been changed to use this function.
2295 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2296 to 24 hours. This avoids potential overflow problems when processing G
2297 and H retry rules. I suspect nobody ever tinkers with this value.
2299 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2301 PH/45 When the plaintext authenticator is running as a client, the server's
2302 challenges are checked to ensure they are valid base64 strings. By
2303 default, the authentication attempt is cancelled if an invalid string is
2304 received. Setting client_ignore_invalid_base64 true ignores these errors.
2305 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2306 they are received. Thus, the responses can be made to depend on the
2307 challenges. If an invalid string is ignored, an empty string is placed in
2310 PH/46 Messages that are created by the autoreply transport now contains a
2311 References: header, in accordance with RFCs 2822 and 3834.
2313 PH/47 Added authenticated_sender_force to the smtp transport.
2315 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2317 PH/49 Installed latest patch from the Sieve maintainer.
2319 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2320 was also set, the appendfile transport was unnecessarily scanning a
2321 directory of message files (e.g. for maildir delivery) to find the count
2322 of files (along with the size), even though it did not need this
2323 information. It now does the scan only if it needs to find either the
2324 size of the count of files.
2326 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2328 PH/52 Two bugs concerned with error handling when the smtp transport is
2331 (i) Exim was not creating retry information for temporary errors given
2332 for individual recipients after the DATA command when the smtp transport
2333 was used in LMTP mode. This meant that they could be retried too
2334 frequently, and not timed out correctly.
2336 (ii) Exim was setting the flag that allows error details to be returned
2337 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2338 recipients that were returned after the DATA command.
2340 PH/53 This is related to PH/52, but is more general: for any failing address,
2341 when detailed error information was permitted to be returned to the
2342 sender, but the error was temporary, then after the final timeout, only
2343 "retry timeout exceeded" was returned. Now it returns the full error as
2344 well as "retry timeout exceeded".
2346 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2347 do this, and (what is worse) MTAs that accept it.
2349 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2350 will now be deprecated.
2352 PH/56 New os.c-cygwin from the Cygwin maintainer.
2354 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2355 formats (previously only available in exim formats via -bpr, -bpru,
2356 and -bpra. Now also available in native and exiqgrep formats)
2358 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2359 with very large, slow to parse queues
2361 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2363 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2365 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2366 responses to authentication challenges, though it was showing the
2367 challenges; (ii) I've removed the CR characters from the debug output for
2370 PH/58 Allow for the insertion of a newline as well as a space when a string
2371 is turned into more than one encoded-word during RFC 2047 encoding. The
2372 Sieve code now uses this.
2374 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2375 data_4xx, lost_connection, tls_required.
2377 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2378 message was being sent as an SMTP response.
2380 PH/61 Add -l and -k options to exicyclog.
2382 PH/62 When verifying, if an address was redirected to one new address, so that
2383 verification continued, and the new address failed or deferred after
2384 having set something in $address_data, the value of $address_data was not
2385 passed back to the ACL. This was different to the case when no
2386 redirection occurred. The value is now passed back in both cases.
2388 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2389 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2390 use login_cap.h, so on its own it isn't the distinguishing feature. The
2391 new name refers directly to the setclassresources() function.
2393 PH/65 Added configuration files for NetBSD3.
2395 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2397 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2398 is preferred over IPv4.
2400 PH/68 The bounce_return_message and bounce_return_body options were not being
2401 honoured for bounces generated during the reception of non-SMTP messages.
2402 In particular, this applied to messages rejected by the ACL. This bug has
2403 been fixed. However, if bounce_return_message is true and bounce_return_
2404 body is false, the headers that are returned for a non-SMTP message
2405 include only those that have been read before the error was detected.
2406 (In the case of an ACL rejection, they have all been read.)
2408 PH/69 The HTML version of the specification is now built in a directory called
2409 spec_html instead of spec.html, because the latter looks like a path with
2410 a MIME-type, and this confuses some software.
2412 PH/70 Catch two compiler warnings in sieve.c.
2414 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2415 function verify_get_ident() calls ip_connect() to connect a socket, but
2416 if the "connect()" function timed out, ip_connect() used to close the
2417 socket. However, verify_get_ident() also closes the socket later, and in
2418 between Exim writes to the log, which may get opened at this point. When
2419 the socket was closed in ip_connect(), the log could get the same file
2420 descriptor number as the socket. This naturally causes chaos. The fix is
2421 not to close the socket in ip_connect(); the socket should be closed by
2422 the function that creates it. There was only one place in the code where
2423 this was missing, in the iplookup router, which I don't think anybody now
2424 uses, but I've fixed it anyway.
2426 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2427 well as to direct DNS lookups. Otherwise the handling of names in host
2428 lists is inconsistent and therefore confusing.
2434 PH/01 Two changes to the default runtime configuration:
2436 (1) Move the checks for relay_from_hosts and authenticated clients from
2437 after to before the (commented out) DNS black list checks.
2439 (2) Add control=submission to the relay_from_hosts and authenticated
2440 clients checks, on the grounds that messages accepted by these
2441 statements are most likely to be submissions.
2443 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2445 (1) Generate an error if the third argument for the ${prvs expansion is
2448 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2451 (3) Reset the variables that are obtained from the first argument of
2452 ${prvscheck and used in the second argument before leaving the code,
2453 because their memory is reclaimed, so using them afterwards may do
2456 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2457 one (it's much easier than Tom thought :-).
2459 (5) Because of (4), we can now allow for the use of $prvscheck_result
2460 inside the third argument.
2462 PH/03 For some reason, the default setting of PATH when running a command from
2463 a pipe transport was just "/usr/bin". I have changed it to
2466 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2467 anything to be listed in the output from -bV.
2469 PH/05 When a filter generated an autoreply, the entire To: header line was
2470 quoted in the delivery log line, like this:
2472 => >A.N.Other <ano@some.domain> <original@ddress> ...
2474 This has been changed so that it extracts the operative address. There
2475 may be more than one such address. If so, they are comma-separated, like
2478 => >ano@some.domain,ona@other.domain <original@ddress> ...
2480 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2481 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2482 not being looked up in the rDNS to get a host name, Exim was showing the
2483 IP address twice in Received: lines, even though the IP addresses were
2484 identical. For example:
2486 Received: from [1.2.3.4] (helo=[1.2.3.4])
2488 However, if the real host name was known, it was omitting the HELO data
2489 if it matched the actual IP address. This has been tidied up so that it
2490 doesn't show the same IP address twice.
2492 PH/07 When both +timestamp and +memory debugging was on, the value given by
2493 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2494 called by the debug printing, thereby overwriting the timestamp buffer.
2495 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2497 PH/08 When the original message was included in an autoreply transport, it
2498 always said "this is a copy of the message, including all the headers",
2499 even if body_only or headers_only was set. It now gives an appropriate
2502 PH/09 Applied a patch from the Sieve maintainer which:
2504 o fixes some comments
2505 o adds the (disabled) notify extension core
2506 o adds some debug output for the result of if/elsif tests
2507 o points to the current vacation draft in the documentation
2508 and documents the missing references header update
2512 o fixes a bug in processing the envelope test (when testing
2513 multiple envelope elements, the last element determinted the
2516 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2517 Electronic Mail") by including:
2519 Auto-submitted: auto-generated
2521 in the messages that it generates (bounce messages and others, such as
2522 warnings). In the case of bounce messages for non-SMTP mesages, there was
2523 also a typo: it was using "Auto_submitted" (underscore instead of
2524 hyphen). Since every message generated by Exim is necessarily in response
2525 to another message, thes have all been changed to:
2527 Auto-Submitted: auto-replied
2529 in accordance with these statements in the RFC:
2531 The auto-replied keyword:
2533 - SHOULD be used on messages sent in direct response to another
2534 message by an automatic process,
2536 - MUST NOT be used on manually-generated messages,
2538 - MAY be used on Delivery Status Notifications (DSNs) and Message
2539 Disposition Notifications (MDNs),
2541 - MUST NOT be used on messages generated by automatic or periodic
2542 processes, except for messages which are automatic responses to
2545 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2546 to the default Received: header definition.
2548 PH/12 Added log selector acl_warn_skipped (default on).
2550 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2551 variables because (a) they are in the wrong storage pool and (b) even if
2552 they were copied, it wouldn't work properly because of the caching.
2554 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2555 checking when decoding. Apparently there are clients that generate
2556 overlong encoded strings. Why am I not surprised?
2558 PH/15 If the first argument of "${if match_address" was not empty, but did not
2559 contain an "@" character, Exim crashed. Now it writes a panic log message
2560 and treats the condition as false.
2562 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2564 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2565 extension "envelope-auth". The code is finished and in agreement with
2566 other implementations, but there is no documentation so far and in fact,
2567 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2568 not changing the active code.
2570 Print executed "if" and "elsif" statements when debugging is used. This
2571 helps a great deal to understand what a filter does.
2573 Document more things not specified clearly in RFC3028. I had all this
2574 sorted out, when out of a sudden new issues came to my mind. Oops."
2576 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2579 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2580 canonical form (as documented). However, after a host name lookup from
2581 the IP address, check_host() was doing a simple string comparison with
2582 addresses acquired from the DNS when checking that the found name did
2583 have the original IP as one of its addresses. Since any found IPv6
2584 addresses are likely to be in abbreviated form, the comparison could
2585 fail. Luckily, there already exists a function for doing the comparison
2586 by converting both addresses to binary, so now that is used instead of
2587 the text comparison.
2589 PH/20 There was another similar case to PH/19, when a complete host name was
2590 given in a host list; looking up its IP address could give an abbreviated
2591 form, whereas the current host's name might or might not be abbreviated.
2592 The same fix has been applied.
2598 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2599 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2602 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2603 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2605 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2607 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2608 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2609 Also, for "Auto-Submitted", treat anything other than "no" as
2610 non-personal, in accordance with RFC 3834. (Previously it treated
2611 anything starting "auto-" as non-personal.)
2613 TF/01 The control=submission/name=... option had a problem with syntax
2614 errors if the name included a slash character. The /name= option
2615 now slurps the rest of the string, so it can include any characters
2616 but it must come last in the list of options (after /sender_retain
2619 PH/05 Some modifications to the interface to the fake nameserver for the new
2627 TK/01 Added the "success_on_redirect" address verification option. See
2628 NewStuff for rationale and an example.
2630 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2632 PH/02 Patch to exigrep to allow it to work on syslog lines.
2634 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2635 fread() to skip over the body file's header line, because in Cygwin the
2636 header line is locked and is inaccessible.
2638 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2639 co-exist for some time) to make it clear that it is the Exim ID that is
2640 referenced, not the Message-ID: header line.
2642 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2643 string_format() function, because snprintf() does not exist on all
2646 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2647 ${stat: expansion item.
2649 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2650 protocol synchronization error", to keep the pedants happy.
2652 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2653 well as for IRIX systems, when gcc is being used. See the host.c source
2656 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2658 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2661 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2662 transport and to the smtp transport in LMTP mode.
2664 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2666 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2667 run a filter in a subprocess. This could lead to confusion in subsequent
2668 lookups in the parent process. There should also be a search_tidyup() at
2669 the end of the subprocess.
2671 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2672 only if the host matched helo_try_verify_hosts, which caused the
2673 verification to occur when the EHLO/HELO command was issued. The ACL just
2674 tested the remembered result. Now, if a previous verification attempt has
2675 not happened, "verify = helo" does it there and then.
2677 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2679 TK/03 Fix log output including CR from clamd.
2681 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2682 exist provoked a memory error which could cause a segfault.
2684 PH/15 Installed PCRE 6.2
2686 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2688 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2689 of the problem. Specifically, suggested +O2 rather than +O1 for the
2692 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2694 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2695 record, the port was not correctly logged when the outgoing_port log
2696 selector was set (it logged the transort's default port).
2698 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2699 fallback_hosts, and "hosts" in the smtp transport.
2701 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2702 host errors such as "Connection refused".
2704 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2705 authentication with radiusclient 0.4.9:
2707 - Error returned from rc_read_config was caught wrongly
2708 - Username/password not passed on to radius server due to wrong length.
2710 The presumption is that some radiusclient API changes for 4.51/PH/17
2711 were not taken care of correctly. The code is still untested by me (my
2712 Linux distribution still has 0.3.2 of radiusclient), but it was
2713 contributed by a Radius user.
2715 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2716 expanding the "port" option of the smtp transport.
2718 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2719 while reading a MIME header. Thanks to Tom Hughes for a patch.
2721 PH/24 Include config.h inside local_scan.h so that configuration settings are
2724 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2725 that in an ACL for RCPT (for example), you can examine exactly what was
2728 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2729 commands, but it was not correctly comparing the address with the actual
2730 client host address. Thus, it would show the EHLO address in Received:
2731 header lines when this was not necessary.
2733 PH/27 Added the % operator to ${eval:}.
2735 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2736 it should be ignoring failures (because with -C, for example, it has lost
2737 privilege). It wasn't ignoring creation failures other than "already
2740 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2743 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2744 because an input line was too long, either on its own, or by virtue of
2745 too many continuations, the temporary file was not being removed, and the
2746 return code was incorrect.
2748 PH/31 Missing "BOOL" in function definition in filtertest.c.
2750 PH/32 Applied Sieve patches from the maintainer.
2752 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2754 PH/33 Added "verify = not_blind".
2756 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2757 Local/Makefile (with some defaults set). These are used in built scripts
2758 such as exicyclog, but they have never been used in the exim_install
2759 script (though there are many overriding facilities there). I have
2760 arranged that the exim_install script now takes note of these two
2763 PH/35 Installed configuration files for Dragonfly.
2765 PH/36 When a locally submitted message by a trusted user did not contain a
2766 From: header, and the sender address was obtained from -f or from an SMTP
2767 MAIL command, and the trusted user did not use -F to supply a sender
2768 name, $originator_name was incorrectly used when constructing a From:
2769 header. Furthermore, $originator_name was used for submission mode
2770 messages from external hosts without From: headers in a similar way,
2771 which is clearly wrong.
2773 PH/37 Added control=suppress_local_fixups.
2775 PH/38 When log_selector = +received_sender was set, and the addition of the
2776 sender made the log line's construction buffer exactly full, or one byte
2777 less than full, an overflow happened when the terminating "\n" was
2780 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2781 when the result of a list match is failure because a DNS lookup failed.
2783 PH/40 RM_COMMAND is now used in the building process.
2785 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2786 the "build-* directories that it finds.
2788 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2789 address in a domain literal was a prefix of an interface address.
2791 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2792 when verifying a sender address, unless rewrite_headers is false.
2794 PH/44 Wrote a long comment about why errors_to addresses are verified as
2795 recipients, not senders.
2797 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2798 the ratelimit ACL was added.
2800 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2802 PH/46 Added extra information about PostgreSQL errors to the error string.
2804 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2805 avoiding the need to install special zones in a real server. This is
2806 backwards compatible; if it can't find the fake resolver, it drops back.
2807 Thus, both old and new test suites can be run.
2809 TF/02 Added util/ratelimit.pl
2811 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2812 clock is set back in time.
2814 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2815 Candler <B.Candler@pobox.com>.
2817 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2818 OK for addresses that are the result of redirections.
2820 PH/48 A number of further additions for the benefit of the new test suite,
2821 including a fake gethostbyname() that interfaces to the fake DNS resolver
2824 TF/06 The fix for widen_domains has also been applied to qualify_single and
2825 search_parents which are the other dnslookup options that can cause
2828 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2831 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2833 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2834 service" if it encountered an MX record with an empty target hostname.
2835 The message is now "an MX or SRV record indicated no SMTP service".
2837 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2838 if the DNS of the sending site is misconfigured. This is quite a
2839 common situation. This change restores the behaviour of treating a
2840 helo verification defer as a failure.
2842 PH/51 If self=fail was set on a router, the bounce message did not include the
2843 actual error message.
2849 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2851 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2852 command itself ended in error, the underlying message about the transport
2853 filter timeout was being overwritten with the pipe command error. Now the
2854 underlying error message should be appended to the second error message.
2856 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2858 PH/02 Reduce the amount of output that "make" produces by default. Full output
2859 can still be requested.
2861 PH/03 The warning log line about a condition test deferring for a "warn" verb
2862 was being output only once per connection, rather than after each
2863 occurrence (because it was using the same function as for successful
2864 "warn" verbs). This seems wrong, so I have changed it.
2866 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2867 it should not have, which might have caused a crash in the right
2868 circumstances, but probably never did.
2870 PH/04 Installed a modified version of Tony Finch's patch to make submission
2871 mode fix the return path as well as the Sender: header line, and to
2872 add a /name= option so that you can make the user's friendly name appear
2875 TF/03 Added the control = fakedefer ACL modifier.
2877 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2878 Mark Lowes for thorough testing.
2880 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2882 TK/03 Merged latest SRS patch from Miles Wilton.
2884 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2885 with the definition in sysexits.h (which is #included earlier).
2886 Fortunately, Exim does not actually use EX_OK. The code used to try to
2887 preserve the sysexits.h value, by assumimg that macro definitions were
2888 scanned for macro replacements. I have been disabused of this notion,
2889 so now the code just undefines EX_OK before #including unistd.h.
2891 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2892 in the smtp transport. When a block could not be written in a single
2893 write() function, the timeout was being re-applied to each part-write.
2894 This seems wrong - if the receiver was accepting one byte at a time it
2895 would take for ever. The timeout is now adjusted when this happens. It
2896 doesn't have to be particularly precise.
2898 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2899 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2901 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2902 without a domain if the check to <postmaster@domain> fails.
2904 SC/01 Eximstats: added -xls and the ability to specify output files
2905 (patch written by Frank Heydlauf).
2907 SC/02 Eximstats: use FileHandles for outputing results.
2909 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2911 SC/04 Eximstats: fixed display of large numbers with -nvr option
2913 SC/05 Eximstats: fixed merging of reports with empty tables.
2915 SC/06 Eximstats: added the -include_original_destination flag
2917 SC/07 Eximstats: removed tabs and trailing whitespace.
2919 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2921 TK/06 MBOX spool code: Add real "From " MBOX separator line
2922 so the .eml file is really in mbox format (even though
2923 most programs do not really care). Patch from Alex Miller.
2925 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2926 The latter is generated from $received_to and is only set if the
2927 message has one envelope recipient. SA can use these headers,
2928 obviously out-of-the-box. Patch from Alex Miller.
2930 PH/08 The ${def test on a variable was returning false if the variable's
2931 value was "0", contrary to what the specification has always said!
2932 The result should be true unless the variable is empty.
2934 PH/09 The syntax error of a character other than { following "${if
2935 def:variable_name" (after optional whitespace) was not being diagnosed.
2936 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2937 accidental colon was present, for example, could give incorrect results.
2939 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2940 result is used (not including appendfile, where other changes are about
2943 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2944 This involved changing a lot of size variables from int to off_t. It
2945 should work with maildirs and everything.
2947 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2948 spamd dying while we are connected to it.
2950 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2953 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2954 function for BDB 4.3.
2956 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2958 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2959 the sources for PCRE have been re-arranged and more files are now
2962 PH/15 The code I had for printing potentially long long variables in PH/11
2963 above was not the best (it lost precision). The length of off_t variables
2964 is now inspected at build time, and an appropriate printing format (%ld
2965 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2966 to be "long long int" or "long int". This is needed for the internal
2967 formatting function string_vformat().
2969 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2970 the configuration file to be ":syslog", then the script "guesses" where
2971 the logs files are, rather than using the compiled in default. In our
2972 case the guess is not the same as the compiled default, so the script
2973 suddenly stopped working when I started to use syslog. The patch checks
2974 to see if log_file_path is "". If so, it attempts to read it from exim
2975 with no configuration file to get the compiled in version, before it
2976 falls back to the previous guessing code."
2978 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2979 implementing BATV in an Exim configuration. See NewStuff for the gory
2982 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2983 Makefile that are specific to HP-UX.
2985 PH/18 If the "use_postmaster" option was set for a recipient callout together
2986 with the "random" option, the postmaster address was used as the MAIL
2987 FROM address for the random test, but not for the subsequent recipient
2988 test. It is now used for both.
2990 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2991 patch removes a few documentation additions to RFC 3028, because the
2992 latest draft now contains them. It adds the new en;ascii-case comparator
2993 and a new error check for 8bit text in MIME parts. Comparator and
2994 require names are now matched exactly. I enabled the subaddress
2995 extension, but it is not well tested yet (read: it works for me)."
2997 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2998 rework some of the code of TK/09 above to avoid the hardwired use of
2999 "%lld" and "long long". Replaced the call to snprintf() with a call to
3002 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3003 records point to non-existent hosts", "retry timeout exceeded", and
3004 "retry time not reached for any host after a long failure period".
3006 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3007 experimental DomainKeys support:
3009 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3010 (2) On an error such as an illegally used "control", the wrong name for
3011 the control was given.
3013 These problems did NOT occur unless DomainKeys support was compiled.
3015 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3017 PH/24 Added ${if match_ip condition.
3019 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3020 databases so that it will be absolutely obvious if a crash occurs in the
3021 DB library. This is a regular occurrence (often caused by mis-matched
3024 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3025 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3026 picked up on a user's system that detects such things. There doesn't seem
3027 to be a gcc warning option for this - only an attribute that has to be
3028 put on the function's prototype. It seems that in Fedora Core 4 they have
3029 set this on a number of new functions. No doubt there will be more in due
3032 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3033 specify a transport. However, if an address that was verified by such a
3034 router was the subject of a callout, Exim crashed because it tried to
3035 read the rcpt_include_affixes from the non-existent transport. Now it
3036 just assumes that the setting of that option is false. This bug was
3037 introduced by 4.51/PH/31.
3039 PH/28 Changed -d+all to exclude +memory, because that information is very
3040 rarely of interest, but it makes the output a lot bigger. People tend to
3041 do -d+all out of habit.
3043 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3044 code in os-type was giving problems when libc.so lives in lib64, like on
3047 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3048 aren't the modern standard, and it seems that some systems' include files
3049 don't always have them. Exim was already checking for some of the newer
3050 ones like T_AAAA, and defining it itself. I've added checks for all the
3051 record types that Exim uses.
3053 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3054 not automatically generating a new one, as it is supposed to. This
3055 prevented TLS from working. If the file did exist, but contained invalid
3056 data, a new version was generated, as expected. It was only the case of a
3057 non-existent file that was broken.
3059 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3060 with a change in libdomainkeys > 0.64.
3062 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3063 from DNS. If the selector record carries the flag, it now has
3064 precedence over the domain-wide flag.
3066 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3068 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3069 the use of a port name that isn't defined in /etc/services) occurred, the
3070 message was deferred as in a normal delivery, and thus remained on the
3071 spool, instead of being failed because of the mua_wrapper setting. This
3072 is now fixed, and I tidied up some of the mua_wrapper messages at the
3075 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3076 the messages in a hash of arrays rather than using individual hashes.
3077 This is a bit cleaner and results in dramatic memory savings, albeit
3078 at a slight CPU cost.
3080 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3081 as requested by Marc Sherman.
3083 SC/10 Eximstats: added histograms for user specified patterns as requested
3086 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3088 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3089 fopen() in the content-scanning modules that did not already have it.
3095 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3096 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3098 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3100 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3102 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3103 to test Sieve filters that use "vacation".
3105 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3106 that changes the way the GnuTLS parameters are stored in the cache file.
3107 The new format can be generated externally. For backward compatibility,
3108 if the data in the cache doesn't make sense, Exim assumes it has read an
3109 old-format file, and it generates new data and writes a new file. This
3110 means that you can't go back to an older release without removing the
3113 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3114 work if there are any delivery delays because "one_time" forces the
3115 parent to be marked "delivered", so its unseen clone is never tried
3116 again. For this reason, Exim now forbids the simultaneous setting of
3119 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3120 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3121 entry if you want to know the details. The fix, however, neglected to
3122 consider the case when local delivery batching is involved. The test for
3123 "previously delivered" was not happening when checking to see if an
3124 address could be batched with a previous (undelivered) one; under
3125 certain circumstances this could lead to multiple deliveries to the same
3128 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3129 in its include files, and this causes problems building Exim.
3131 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3132 header_syntax) but Exim was just ignoring anything given after a slash.
3133 In particular, this caused confusion with an attempt to use "verify =
3134 reverse_host_lookup/defer_ok". An error is now given when options are
3135 supplied for verify items that do not have them. (Maybe reverse_host_
3136 lookup should have a defer_ok option, but that's a different point.)
3138 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3139 defined by RFC 821) to 2048, because there were problems with some AUTH
3140 commands, and RFC 1869 says the size should be increased for extended
3141 SMTP commands that take arguments.
3143 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3146 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3147 "unknown" error; now it says that the functionality isn't in the binary.
3149 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3150 an address' error message when a string expansion fails (syntax or
3151 whatever). Otherwise the password may appear in the log. Following change
3152 PH/42 below, there is no longer a chance of it appearing in a bounce
3155 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3157 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3158 of its addresses was ever tried. (Bugzilla bug #2.)
3160 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3161 the result incorrectly in the debug output. (It correctly added a newline
3162 to what was transported.)
3164 TF/01 Added $received_time.
3166 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3167 commented out examples of how to interface to a virus scanner and to
3168 SpamAssassin. Also added commented examples of av_scanner and
3169 spamd_address settings.
3171 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3172 and controls are allowed in which ACLs. There were a couple of minor
3173 errors. Some of the entries in the conditions table (which is a table of
3174 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3175 negation of where the condition IS allowed.
3177 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3179 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3180 header file does not have a version number, so I've had to invent a new
3181 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3182 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3183 radiusclient), but it was contributed by a Radius user.
3185 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3186 files or directories, for OpenSSL.
3188 PH/19 When an Exim process that is running as root has to create an Exim log
3189 file, it does so in a subprocess that runs as exim:exim so as to get the
3190 ownership right at creation (otherwise, other Exim processes might see
3191 the file with the wrong ownership). There was no test for failure of this
3192 fork() call, which would lead to the process getting stuck as it waited
3193 for a non-existent subprocess. Forks do occasionally fail when resources
3194 run out. I reviewed all the other calls to fork(); they all seem to check
3197 PH/20 When checking for unexpected SMTP input at connect time (before writing
3198 the banner), Exim was not dealing correctly with a non-positive return
3199 from the read() function. If the client had disconnected by this time,
3200 the result was a log entry for a synchronization error with an empty
3201 string after "input=" when read() returned zero. If read() returned -1
3202 (an event I could not check), uninitialized data bytes were printed.
3203 There were reports of junk text (parts of files, etc) appearing after
3206 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3208 PH/22 Added support for macro redefinition, and (re)definition in between
3209 driver and ACL definitions.
3211 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3212 forgetting to use the resulting value; it was using the unexpanded value.
3214 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3215 hadn't been configured. The fix is from Juergen Kreileder, who
3216 understands it better than I do:
3218 "Here's what I see happening with three configured cyrus_sasl
3219 authenticators configured (plain, login, cram-md5):
3221 On startup auth_cyrus_sasl_init() gets called for each of these.
3222 This means three calls to sasl_listmech() without a specified mech_list.
3223 => SASL tests which mechs of all available mechs actually work
3224 => three warnings about OTP not working
3225 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3227 With the patch, sasl_listmech() also gets called three times. But now
3228 SASL's mech_list option is set to the server_mech specified in the the
3229 authenticator. Or in other words, the answer from sasl_listmech()
3230 gets limited to just the mech you're testing for (which is different
3232 => the return list contains just 'plain' or 'login', 'cram-md5' or
3233 nothing depending on the value of ob->server_mech.
3235 I've just tested the patch: Authentication still works fine,
3236 unavailable mechs specified in the exim configuration are still
3237 caught, and the auth.log warnings about OTP are gone."
3239 PH/25 When debugging is enabled, the contents of the command line are added
3240 to the debugging output, even when log_selector=+arguments is not
3243 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3244 answer is "GNU", and only if the return is "GNU/something" is the answer
3247 PH/27 $acl_verify_message is now set immediately after the failure of a
3248 verification in an ACL, and so is available in subsequent modifiers. In
3249 particular, the message can be preserved by coding like this:
3251 warn !verify = sender
3252 set acl_m0 = $acl_verify_message
3254 Previously, $acl_verify_message was set only while expanding "message"
3255 and "log_message" when a very denied access.
3257 PH/28 Modified OS/os.c-Linux with
3259 -#ifndef OS_LOAD_AVERAGE
3260 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3262 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3263 nomenclature these days.)
3265 PH/29 Installed patch from the Sieve maintainer that adds the options
3266 sieve_useraddress and sieve_subaddress to the redirect router.
3268 PH/30 In these circumstances:
3269 . Two addresses routed to the same list of hosts;
3270 . First host does not offer TLS;
3271 . First host accepts first address;
3272 . First host gives temporary error to second address;
3273 . Second host offers TLS and a TLS session is established;
3274 . Second host accepts second address.
3275 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3276 and peerdn, if requested) that were in fact used only for the second
3279 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3280 attention to any local part prefix or suffix that was matched by the
3281 router that accepted the address. It now behaves in the same way as it
3282 does for delivery: the affixes are removed from the local part unless
3283 rcpt_include_affixes is set on the transport.
3285 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3286 timeout during the DATA phase of an incoming message.
3288 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3289 applied a patch sanctioned by the Sieve maintainer.
3291 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3292 the uid or gid is negative. A case of a negative gid caused this to be
3293 noticed. The fix allows for either to be negative.
3295 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3296 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3299 PH/36 The change PH/12 above was broken. Fixed it.
3301 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3302 the grounds that routing the same address twice would always produce the
3303 same answer. This might have been true once, but it is certainly no
3304 longer true now. Routing a child address may depend on the previous
3305 routing that produced that child. Some complicated redirection strategies
3306 went wrong when messages had multiple recipients, and made Exim's
3307 behaviour dependent on the order in which the addresses were given.
3309 I have moved the duplicate checking until after the routing is complete.
3310 Exim scans the addresses that are assigned to local and remote
3311 transports, and removes any duplicates. This means that more work will be
3312 done, as duplicates will always all be routed, but duplicates are
3313 presumably rare, so I don't expect this is of any significance.
3315 For deliveries to pipes, files, and autoreplies, the duplicate checking
3316 still happens during the routing process, since they are not going to be
3319 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3320 It corrects a timeout issue with spamd. This is Ian's comment: "The
3321 background is that sometimes spamd either never reads data from a
3322 connection it has accepted, or it never writes response data. The exiscan
3323 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3324 blindly assumes that writes won't block so it may never time out."
3326 PH/39 Allow G after quota size as well as K and M.
3328 PH/40 The value set for $authenticated_id in an authenticator may not contain
3329 binary zeroes or newlines because the value is written to log lines and
3330 to spool files. There was no check on this. Now the value is run through
3331 the string_printing() function so that such characters are converted to
3332 printable escape sequences.
3334 PH/41 $message_linecount is a new variable that contains the total number of
3335 lines in the message. Compare $body_linecount, which is the count for the
3338 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3339 bounce and delay warning messages, except in certain special cases, which
3342 (a) An SMTP error message from a remote host;
3343 (b) A message specified in a :fail: redirection;
3344 (c) A message specified in a "fail" command in a system filter;
3345 (d) A message specified in a FAIL return from the queryprogram router;
3346 (e) A message specified by the cannot_route_message router option.
3348 In these cases only, Exim does include the error details in bounce and
3349 warning messages. There are also a few cases where bland messages such
3350 as "unrouteable address" or "local delivery error" are given.
3352 PH/43 $value is now also set for the "else" part of a ${run expansion.
3354 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3355 being worked on, but at least Exim now implements the latest version to
3358 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3359 process to complete was treated as a delivery failure, a timeout while
3360 writing the message to the pipe was logged, but erroneously treated as a
3361 successful delivery. Such timeouts include transport filter timeouts. For
3362 consistency with the overall process timeout, these timeouts are now
3363 treated as errors, giving rise to delivery failures by default. However,
3364 there is now a new Boolean option for the pipe transport called
3365 timeout_defer, which, if set TRUE, converts the failures into defers for
3366 both kinds of timeout. A transport filter timeout is now identified in
3369 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3370 systems where "make" and "gmake" are different, calling "gmake" at top
3371 level broke things. I've arranged for the value of $(MAKE) to be passed
3372 from the Makefile to this script so that it can call the same version of
3376 A note about Exim versions 4.44 and 4.50
3377 ----------------------------------------
3379 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3380 changes of various kinds. As a consequence, a big documentation update was
3381 needed. This delayed the release for rather longer than seemed good, especially
3382 in the light of a couple of (minor) security issues. Therefore, the changes
3383 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3384 release. So 4.44 and 4.50 are in effect two different branches that both start
3387 I have left the 4.50 change log unchanged; it contains all the changes since
3388 4.43. The change log for 4.44 is below; many of its items are identical to
3389 those for 4.50. This seems to be the most sensible way to preserve the
3390 historical information.
3396 1. Minor wording change to the doc/README.SIEVE file.
3398 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3399 computation of the current number of files was incorrect.
3401 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3402 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3405 4. Give more explanation in the error message when the command for a transport
3406 filter fails to execute.
3408 5. There are several places where Exim runs a non-Exim command in a
3409 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3410 was being done only for the command run by the queryprogram router. It is
3411 now done for all such subprocesses. The other cases are: ${run, transport
3412 filters, and the commands run by the lmtp and pipe transports.
3414 6. Added CONFIGURE_GROUP build-time option.
3416 7. Some older OS have a limit of 256 on the maximum number of file
3417 descriptors. Exim was using setrlimit() to set 1000 as a large value
3418 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3419 systems. I've change it so that if it can't get 1000, it tries for 256.
3421 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3422 was an oversight, and furthermore, ever since the addition of extra
3423 controls (e.g. 4.43/32), the checks on when to allow different forms of
3424 "control" were broken. There should now be diagnostics for all cases when a
3425 control that does not make sense is encountered.
3427 9. Added the /retain_sender option to "control=submission".
3429 10. $recipients is now available in the predata ACL (oversight).
3431 11. Tidy the search cache before the fork to do a delivery from a message
3432 received from the command line. Otherwise the child will trigger a lookup
3433 failure and thereby defer the delivery if it tries to use (for example) a
3434 cached ldap connection that the parent has called unbind on.
3436 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3437 of $address_data from the recipient verification was clobbered by the
3438 sender verification.
3440 13. The value of address_data from a sender verification is now available in
3441 $sender_address_data in subsequent conditions in the ACL statement.
3443 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3445 15. Added a new option "connect=<time>" to callout options, to set a different
3448 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3449 was its contents. (It was OK if the option was not defined at all.)
3451 17. A "Completed" log line is now written for messages that are removed from
3452 the spool by the -Mrm option.
3454 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3455 information about exactly what failed.
3457 19. Added -dd to debug only the daemon process.
3459 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3460 handles timeouts, both on the server side and network timeouts. Renamed the
3461 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3463 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3464 It is now set to "smtps".
3466 22. $host_address is now set to the target address during the checking of
3467 ignore_target_hosts.
3469 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3470 being passed; this would have caused $sender_host_name to have been used if
3471 matching the list had actually called for a host name (not very likely,
3472 since this list is usually IP addresses). A host name is now passed as
3475 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3476 code that specifies a non-restarting handler (typically sigaction() in
3477 modern systems) in an attempt to fix a rare and obscure crash bug.
3479 25. Narrowed the window for a race in the daemon that could cause it to ignore
3480 SIGCHLD signals. This is not a major problem, because they are used only to
3481 wake it up if nothing else does.
3483 26. A malformed maildirsize file could cause Exim to calculate negative values
3484 for the mailbox size or file count. Odd effects could occur as a result.
3485 The maildirsize information is now recalculated if the size or filecount
3488 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3489 support for a long time. Removed HAVE_SYS_VFS_H.
3491 28. Installed the latest version of exipick from John Jetmore.
3493 29. In an address list, if the pattern was not a regular expression, an empty
3494 subject address (from a bounce message) matched only if the pattern was an
3495 empty string. Non-empty patterns were not even tested. This was the wrong
3496 because it is perfectly reasonable to use an empty address as part of a
3497 database query. An empty address is now tested by patterns that are
3498 lookups. However, all the other forms of pattern expect the subject to
3499 contain a local part and a domain, and therefore, for them, an empty
3500 address still always fails if the pattern is not itself empty.
3502 30. Exim went into a mad DNS loop when attempting to do a callout where the
3503 host was specified on an smtp transport, and looking it up yielded more
3504 than one IP address.
3506 31. Re-factored the code for checking spool and log partition space into a
3507 function that finds that data and another that does the check. The former
3508 is then used to implement four new variables: $spool_space, $log_space,
3509 $spool_inodes, and $log_inodes.
3511 32. The RFC2047 encoding function was originally intended for short strings
3512 such as real names; it was not keeping to the 75-character limit for
3513 encoded words that the RFC imposes. It now respects the limit, and
3514 generates multiple encoded words if necessary. To be on the safe side, I
3515 have increased the buffer size for the ${rfc2047: expansion operator from
3518 33. It is now permitted to omit both strings after an "if" condition; if the
3519 condition is true, the result is "true". As before, when the second string
3520 is omitted, a false condition yields an empty string. This makes it less
3521 cumbersome to write custom ACL and router conditions.
3523 34. Failure to deliver a bounce message always caused it to be frozen, even if
3524 there was an errors_to setting on the router. The errors_to setting is now
3527 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3528 canonical form (fully expanded) before being placed in
3529 $sender_host_address.
3531 36. The table in the code that translates DNS record types into text (T_A to
3532 "A" for instance) was missing entries for NS and CNAME. It is just possible
3533 that this could have caused confusion if both these types were looked up
3534 for the same domain, because the text type is used as part of Exim's
3535 per-process caching. But the chance of anyone hitting this buglet seems
3538 37. The dnsdb lookup has been extended in a number of ways.
3540 (1) There is a new type, "zns", which walks up the domain tree until it
3541 finds some nameserver records. It should be used with care.
3543 (2) There is a new type, "mxh", which is like "mx" except that it returns
3544 just the host names, not the priorities.
3546 (3) It is now possible to give a list of domains (or IP addresses) to be
3547 looked up. The behaviour when one of the lookups defers can be
3548 controlled by a keyword.
3550 (4) It is now possible to specify the separator character for use when
3551 multiple records are returned.
3553 38. The dnslists ACL condition has been extended: it is now possible to supply
3554 a list of IP addresses and/or domains to be looked up in a particular DNS
3557 39. Added log_selector=+queue_time_overall.
3559 40. When running the queue in the test harness, wait just a tad after forking a
3560 delivery process, to get repeatability of debugging output.
3562 41. Include certificate and key file names in error message when GnuTLS fails
3563 to set them up, because the GnuTLS error message doesn't include the name
3564 of the failing file when there is a problem reading it.
3566 42. Allow both -bf and -bF in the same test run.
3568 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3570 44. The "Exiscan patch" is now merged into the mainline Exim source.
3572 45. Sometimes the final signoff response after QUIT could fail to get
3573 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3574 < 0 before doing a fflush(). This bug looks as though it goes back to the
3575 introduction of TLS in release 3.20, but "sometimes" must have been rare
3576 because the tests only now provoked it.
3578 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3579 (this can affect the format of dates).
3581 47. exim_tidydb, when checking for the continued existence of a message for
3582 which it has found a message-specific retry record, was not finding
3583 messages that were in split spool directories. Consequently, it was
3584 deleting retry records that should have stayed in existence.
3586 48. Steve fixed some bugs in eximstats.
3588 49. The SPA authentication driver was not abandoning authentication and moving
3589 on to the next authenticator when an expansion was forced to fail,
3590 contradicting the general specification for all authenticators. Instead it
3591 was generating a temporary error. It now behaves as specified.
3593 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3594 (the order specifies the preference for clients). The order is now AES256,
3595 AES128, 3DES, ARCFOUR128.
3597 51. Small patch to Sieve code - explicitly set From: when generating an
3600 52. Exim crashed if a remote delivery caused a very long error message to be
3601 recorded - for instance if somebody sent an entire SpamAssassin report back
3602 as a large number of 550 error lines. This bug was coincidentally fixed by
3603 increasing the size of one of Exim's internal buffers (big_buffer) that
3604 happened as part of the Exiscan merge. However, to be on the safe side, I
3605 have made the code more robust (and fixed the comments that describe what
3608 53. Now that there can be additional text after "Completed" in log lines (if
3609 the queue_time_overall log selector is set), a one-byte patch to exigrep
3610 was needed to allow it to recognize "Completed" as not the last thing in
3613 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3614 patch that reportedly fixes this has been added. I am not expert enough to
3615 create a test for it. This is what the patch creator wrote:
3617 "I found a little strange behaviour of ldap code when working with
3618 Windows 2003 AD Domain, where users was placed in more than one
3619 Organization Units. When I tried to give exim partial DN, the exit code
3620 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3621 But simultaneously result of request was absolutely normal ldap result,
3622 so I produce this patch..."
3624 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3625 REFERENCE, so I have modified the code to exclude the patch when that macro
3628 55. Some experimental protocols are using DNS PTR records for new purposes. The
3629 keys for these records are domain names, not reversed IP addresses. The
3630 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3631 leaves it alone. Component reversal etc. now happens only for IP addresses.
3634 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3636 57. Double the size of the debug message buffer (to 2048) so that more of very
3637 long debug lines gets shown.
3639 58. The exicyclog utility now does better if the number of log files to keep
3640 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3642 59. Two changes related to the smtp_active_hostname option:
3644 (1) $smtp_active_hostname is now available as a variable.
3645 (2) The default for smtp_banner uses $smtp_active_hostname instead
3646 of $primary_hostname.
3648 60. The host_aton() function is supposed to be passed a string that is known
3649 to be a valid IP address. However, in the case of IPv6 addresses, it was
3650 not checking this. This is a hostage to fortune. Exim now panics and dies
3651 if the condition is not met. A case was found where this could be provoked
3652 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3653 components; fortuitously, this particular loophole had already been fixed
3654 by change 4.50/55 above.
3656 If there are any other similar loopholes, the new check in host_aton()
3657 itself should stop them being exploited. The report I received stated that
3658 data on the command line could provoke the exploit when Exim was running as
3659 exim, but did not say which command line option was involved. All I could
3660 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3661 running as the user.
3664 61. There was a buffer overflow vulnerability in the SPA authentication code
3665 (which came originally from the Samba project). I have added a test to the
3666 spa_base64_to_bits() function which I hope fixes it.
3669 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3670 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3672 63. The daemon start-up calls getloadavg() while still root for those OS that
3673 need the first call to be done as root, but it missed one case: when
3674 deliver_queue_load_max is set with deliver_drop_privilege. This is
3675 necessary for the benefit of the queue runner, because there is no re-exec
3676 when deliver_drop_privilege is set.
3678 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3679 This has been fixed.
3681 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3682 was in use, was not putting the data itself into the right store pool;
3683 consequently, it could be overwritten for a subsequent message in the same
3684 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3687 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3689 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3690 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3691 the function were treating the return as a boolean value, which happened to
3692 work because 0=false and not-0=true, but is not correct code.
3694 68. The host_aton() function was not handling scoped IPv6 addresses (those
3695 with, for example, "%eth0" on the end) correctly.
3697 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3698 negated items (that is, ~something) in unsigned ints. Some compilers
3699 apparently mutter when there is no cast.
3701 70. If an address verification called from an ACL failed, and did not produce a
3702 user-specific message (i.e. there was only a "system" message), nothing was
3703 put in $acl_verify_message. In this situation, it now puts the system
3706 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3707 session; change 4.31/43 added the unwanted input to the log line - except
3708 that it did not do this in the start of session case. It now does.
3710 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3711 This is wrong and can cause the other end to generate a synchronization
3712 error if it is another Exim or anything else that does the synchronization
3713 check. A QUIT command is no longer sent after a timeout.
3715 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3716 during host lookups.
3718 74. The defer_ok option of callout verification was not working if it was used
3719 when verifying addresses in header lines, that is, for this case:
3721 verify = header_sender/callout=defer_ok
3723 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3724 those file descriptors could be used for SMTP connections. If anything
3725 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3726 could be sent to the SMTP client, causing chaos. The daemon now opens
3727 stdin, stdout, and stderr to /dev/null when it puts itself into the
3730 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3731 log by default. The user can override this with suitable Perl magic.
3733 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3734 the log message when discard triggers, was not working for the DATA ACL or
3735 for the non-SMTP ACL.
3737 78. Error message wording change in sieve.c.
3739 79. If smtp_accept_max_per_host was set, the number of connections could be
3740 restricted to fewer than expected, because the daemon was trying to set up
3741 a new connection before checking whether the processes handling previous
3742 connections had finished. The check for completed processes is now done
3743 earlier. On busy systems, this bug wouldn't be noticed because something
3744 else would have woken the daemon, and it would have reaped the completed
3747 80. If a message was submitted locally by a user whose login name contained one
3748 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3749 It caused a spool format error. I have fixed the spool reading code. A
3750 related problem was that the "from" clause in the Received: line became
3751 illegal because of the space(s). It is now covered by ${quote_local_part.
3753 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3756 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3758 83. Patch from Sieve maintainer to fix unterminated string problem in
3759 "vacation" handling.
3761 84. Some minor changes to the Linux configuration files to help with other
3762 OS variants using glibc.
3764 85. One more patch for Sieve to update vacation handling to latest spec.
3767 ----------------------------------------------------
3768 See the note above about the 4.44 and 4.50 releases.
3769 ----------------------------------------------------
3775 1. Change 4.43/35 introduced a bug that caused file counts to be
3776 incorrectly computed when quota_filecount was set in an appendfile
3779 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3780 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3783 3. Give more explanation in the error message when the command for a transport
3784 filter fails to execute.
3786 4. There are several places where Exim runs a non-Exim command in a
3787 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3788 was being done only for the command run by the queryprogram router. It is
3789 now done for all such subprocesses. The other cases are: ${run, transport
3790 filters, and the commands run by the lmtp and pipe transports.
3792 5. Some older OS have a limit of 256 on the maximum number of file
3793 descriptors. Exim was using setrlimit() to set 1000 as a large value
3794 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3795 systems. I've change it so that if it can't get 1000, it tries for 256.
3797 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3798 was an oversight, and furthermore, ever since the addition of extra
3799 controls (e.g. 4.43/32), the checks on when to allow different forms of
3800 "control" were broken. There should now be diagnostics for all cases when a
3801 control that does not make sense is encountered.
3803 7. $recipients is now available in the predata ACL (oversight).
3805 8. Tidy the search cache before the fork to do a delivery from a message
3806 received from the command line. Otherwise the child will trigger a lookup
3807 failure and thereby defer the delivery if it tries to use (for example) a
3808 cached ldap connection that the parent has called unbind on.
3810 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3811 of $address_data from the recipient verification was clobbered by the
3812 sender verification.
3814 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3815 was its contents. (It was OK if the option was not defined at all.)
3817 11. A "Completed" log line is now written for messages that are removed from
3818 the spool by the -Mrm option.
3820 12. $host_address is now set to the target address during the checking of
3821 ignore_target_hosts.
3823 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3824 being passed; this would have caused $sender_host_name to have been used if
3825 matching the list had actually called for a host name (not very likely,
3826 since this list is usually IP addresses). A host name is now passed as
3829 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3830 code that specifies a non-restarting handler (typically sigaction() in
3831 modern systems) in an attempt to fix a rare and obscure crash bug.
3833 15. Narrowed the window for a race in the daemon that could cause it to ignore
3834 SIGCHLD signals. This is not a major problem, because they are used only to
3835 wake it up if nothing else does.
3837 16. A malformed maildirsize file could cause Exim to calculate negative values
3838 for the mailbox size or file count. Odd effects could occur as a result.
3839 The maildirsize information is now recalculated if the size or filecount
3842 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3843 support for a long time. Removed HAVE_SYS_VFS_H.
3845 18. Updated exipick to current release from John Jetmore.
3847 19. Allow an empty sender to be matched against a lookup in an address list.
3848 Previously the only cases considered were a regular expression, or an
3851 20. Exim went into a mad DNS lookup loop when doing a callout where the
3852 host was specified on the transport, if the DNS lookup yielded more than
3855 21. The RFC2047 encoding function was originally intended for short strings
3856 such as real names; it was not keeping to the 75-character limit for
3857 encoded words that the RFC imposes. It now respects the limit, and
3858 generates multiple encoded words if necessary. To be on the safe side, I
3859 have increased the buffer size for the ${rfc2047: expansion operator from
3862 22. Failure to deliver a bounce message always caused it to be frozen, even if
3863 there was an errors_to setting on the router. The errors_to setting is now
3866 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3867 canonical form (fully expanded) before being placed in
3868 $sender_host_address.
3870 24. Updated eximstats to version 1.33
3872 25. Include certificate and key file names in error message when GnuTLS fails
3873 to set them up, because the GnuTLS error message doesn't include the name
3874 of the failing file when there is a problem reading it.
3876 26. Expand error message when OpenSSL has problems setting up cert/key files.
3879 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3880 (this can affect the format of dates).
3882 28. exim_tidydb, when checking for the continued existence of a message for
3883 which it has found a message-specific retry record, was not finding
3884 messages that were in split spool directories. Consequently, it was
3885 deleting retry records that should have stayed in existence.
3887 29. eximstats updated to version 1.35
3888 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3889 1.35 - bugfix such that pie charts by volume are generated correctly
3891 30. The SPA authentication driver was not abandoning authentication and moving
3892 on to the next authenticator when an expansion was forced to fail,
3893 contradicting the general specification for all authenticators. Instead it
3894 was generating a temporary error. It now behaves as specified.
3896 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3897 (the order specifies the preference for clients). The order is now AES256,
3898 AES128, 3DES, ARCFOUR128.
3900 31. Small patch to Sieve code - explicitly set From: when generating an
3903 32. Exim crashed if a remote delivery caused a very long error message to be
3904 recorded - for instance if somebody sent an entire SpamAssassin report back
3905 as a large number of 550 error lines. This bug was coincidentally fixed by
3906 increasing the size of one of Exim's internal buffers (big_buffer) that
3907 happened as part of the Exiscan merge. However, to be on the safe side, I
3908 have made the code more robust (and fixed the comments that describe what
3911 33. Some experimental protocols are using DNS PTR records for new purposes. The
3912 keys for these records are domain names, not reversed IP addresses. The
3913 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3914 leaves it alone. Component reversal etc. now happens only for IP addresses.
3917 34. The host_aton() function is supposed to be passed a string that is known
3918 to be a valid IP address. However, in the case of IPv6 addresses, it was
3919 not checking this. This is a hostage to fortune. Exim now panics and dies
3920 if the condition is not met. A case was found where this could be provoked
3921 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3922 components; fortuitously, this particular loophole had already been fixed
3923 by change 4.50/55 or 4.44/33 above.
3925 If there are any other similar loopholes, the new check in host_aton()
3926 itself should stop them being exploited. The report I received stated that
3927 data on the command line could provoke the exploit when Exim was running as
3928 exim, but did not say which command line option was involved. All I could
3929 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3930 running as the user.
3933 35. There was a buffer overflow vulnerability in the SPA authentication code
3934 (which came originally from the Samba project). I have added a test to the
3935 spa_base64_to_bits() function which I hope fixes it.
3938 36. The daemon start-up calls getloadavg() while still root for those OS that
3939 need the first call to be done as root, but it missed one case: when
3940 deliver_queue_load_max is set with deliver_drop_privilege. This is
3941 necessary for the benefit of the queue runner, because there is no re-exec
3942 when deliver_drop_privilege is set.
3944 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3945 was in use, was not putting the data itself into the right store pool;
3946 consequently, it could be overwritten for a subsequent message in the same
3947 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3950 38. Sometimes the final signoff response after QUIT could fail to get
3951 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3952 < 0 before doing a fflush(). This bug looks as though it goes back to the
3953 introduction of TLS in release 3.20, but "sometimes" must have been rare
3954 because the tests only now provoked it.
3960 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3961 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3962 Now it returns an integer. A number of calls were still expecting a T/F
3963 return. Fortuitously, in all cases, the tests worked in OK situations,
3964 which is the norm. However, things would have gone wrong on any write
3965 failures on the smtp file descriptor. This function is used when sending
3966 messages over SMTP and also when doing verify callouts.
3968 2. When Exim is called to do synchronous delivery of a locally submitted
3969 message (the -odf or -odi options), it no longer closes stderr before doing
3972 3. Implemented the mua_wrapper option.
3974 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3976 5. Implemented the functions header_remove(), header_testname(),
3977 header_add_at_position(), and receive_remove_recipient(), and exported them
3980 6. If an ACL "warn" statement specified the addition of headers, Exim already
3981 inserted X-ACL-Warn: at the start if there was no header name. However, it
3982 was not making this test for the second and subsequent header lines if
3983 there were newlines in the string. This meant that an invalid header could
3984 be inserted if Exim was badly configured.
3986 7. Allow an ACL "warn" statement to add header lines at the start or after all
3987 the Received: headers, as well as at the end.
3989 8. Added the rcpt_4xx retry error code.
3991 9. Added postmaster_mailfrom=xxx to callout verification option.
3993 10. Added mailfrom=xxxx to the callout verification option, for verify=
3996 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3997 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3999 12. Inserted some casts to stop certain compilers warning when using pointer
4000 differences as field lengths or precisions in printf-type calls (mostly
4001 affecting debugging statements).
4003 13. Added optional readline() support for -be (dynamically loaded).
4005 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4006 same clock tick as a message's arrival, so that its received time was the
4007 same as the "first fail" time on the retry record, and that message
4008 remained on the queue past the ultimate address timeout, every queue runner
4009 would try a delivery (because it was past the ultimate address timeout) but
4010 after another failure, the ultimate address timeout, which should have then
4011 bounced the address, did not kick in. This was a "< instead of <=" error;
4012 in most cases the first failure would have been in the next clock tick
4013 after the received time, and all would be well.
4015 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4016 being recognized when the domain list was tested by the match_domain
4017 condition in an expansion string.
4019 16. Added the ${str2b64: operator.
4021 17. Exim was always calling setrlimit() to set a large limit for the number of
4022 processes, without checking whether the existing limit was already
4023 adequate. (It did check for the limit on file descriptors.) Furthermore,
4024 errors from getrlimit() and setrlimit() were being ignored. Now they are
4025 logged to the main and panic logs, but Exim does carry on, to try to do its
4026 job under whatever limits there are.
4028 18. Imported PCRE 5.0.
4030 19. Trivial typo in log message " temporarily refused connection" (the leading
4033 20. If the log selector return_path_on_delivery was set and an address was
4034 redirected to /dev/null, the delivery process crashed because it assumed
4035 that a return path would always be set for a "successful" delivery. In this
4036 case, the whole delivery is bypassed as an optimization, and therefore no
4039 21. Internal re-arrangement: the function for sending a challenge and reading
4040 a response while authentication was assuming a zero-terminated challenge
4041 string. It's now changed to take a pointer and a length, to allow for
4042 binary data in such strings.
4044 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4046 23. Exim was not respecting finduser_retries when seeking the login of the
4047 uid under which it was called; it was always trying 10 times. (The default
4048 setting of finduser_retries is zero.) Also, it was sleeping after the final
4049 failure, which is pointless.
4051 24. Implemented tls_on_connect_ports.
4053 25. Implemented acl_smtp_predata.
4055 26. If the domain in control=submission is set empty, Exim assumes that the
4056 authenticated id is a complete email address when it generates From: or
4057 Sender: header lines.
4059 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4060 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4061 chown and chgrp in /bin and hostname in /usr/bin.
4063 28. Exim was keeping the "process log" file open after each use, just as it
4064 does for the main log. This opens the possibility of it remaining open for
4065 long periods when the USR1 signal hits a daemon. Occasional processlog
4066 errors were reported, that could have been caused by this. Anyway, it seems
4067 much more sensible not to leave this file open at all, so that is what now
4070 29. The long-running daemon process does not normally write to the log once it
4071 has entered its main loop, and it closes the log before doing so. This is
4072 so that log files can straightforwardly be renamed and moved. However,
4073 there are a couple of unusual error situations where the daemon does write
4074 log entries, and I had neglected to close the log afterwards.
4076 30. The text of an SMTP error response that was received during a remote
4077 delivery was being truncated at 512 bytes. This is too short for some of
4078 the long messages that one sometimes sees. I've increased the limit to
4081 31. It is now possible to make retry rules that apply only when a message has a
4082 specific sender, in particular, an empty sender.
4084 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4085 it possible to be selective about when SMTP synchronization is enforced.
4087 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4089 32. Implemented hosts_connection_nolog.
4091 33. Added an ACL for QUIT.
4093 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4096 35. Added mailbox_size and mailbox_filecount to appendfile.
4098 36. Added control = no_multiline_responses to ACLs.
4100 37. There was a bug in the logic of the code that waits for the clock to tick
4101 in the case where the clock went backwards by a substantial amount such
4102 that the microsecond fraction of "now" was more than the microsecond
4103 fraction of "then" (but the whole seconds number was less).
4105 38. Added support for the libradius Radius client library this is found on
4106 FreeBSD (previously only the radiusclient library was supported).
4112 1. When certain lookups returned multiple values in the form name=value, the
4113 quoting of the values was not always being done properly. Specifically:
4114 (a) If the value started with a double quote, but contained no whitespace,
4116 (b) If the value contained whitespace other than a space character (i.e.
4117 tabs or newlines or carriage returns) it was not quoted.
4118 This fix has been applied to the mysql and pgsql lookups by writing a
4119 separate quoting function and calling it from the lookup code. The fix
4120 should probably also be applied to nisplus, ibase and oracle lookups, but
4121 since I cannot test any of those, I have not disturbed their existing code.
4123 2. A hit in the callout cache for a specific address caused a log line with no
4124 reason for rejecting RCPT. Now it says "Previous (cached) callout
4125 verification failure".
4127 3. There was an off-by-one bug in the queryprogram router. An over-long
4128 return line was truncated at 256 instead of 255 characters, thereby
4129 overflowing its buffer with the terminating zero. As well as fixing this, I
4130 have increased the buffer size to 1024 (and made a note to document this).
4132 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4133 when Exim is waiting for an SMTP response from a remote server, Exim
4134 restarts its select() call on the socket, thereby resetting its timeout.
4135 This is not a problem when such interrupts are rare. Somebody set up a cron
4136 job to run exiwhat every 2 minutes, which is less than the normal select()
4137 timeout (5 or 10 minutes). This meant that the select() timeout never
4138 kicked in because it was always reset. I have fixed this by comparing the
4139 time when an interrupt arrives with the time at the start of the first call
4140 to select(). If more time than the timeout has elapsed, the interrupt is
4141 treated as a timeout.
4143 5. Some internal re-factoring in preparation for the addition of Sieve
4144 extensions (by MH). In particular, the "personal" test is moved to a
4145 separate function, and given an option for scanning Cc: and Bcc: (which is
4146 not set for Exim filters).
4148 6. When Exim created an email address using the login of the caller as the
4149 local part (e.g. when creating a From: or Sender: header line), it was not
4150 quoting the local part when it contained special characters such as @.
4152 7. Installed new OpenBSD configuration files.
4154 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4155 try to make them clearer.
4157 9. Callout options, other than the timeout value, were being ignored when
4158 verifying sender addresses in header lines. For example, when using
4160 verify = header_sender/callout=no_cache
4162 the cache was (incorrectly) being used.
4164 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4165 only the Cygwin environment.
4167 11. When return_path_on_delivery was set as a log selector, if different remote
4168 addresses in the same message used different return paths and parallel
4169 remote delivery occurred, the wrong values would sometimes be logged.
4170 (Whenever a remote delivery process finished, the return path value from
4171 the most recently started remote delivery process was logged.)
4173 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4174 lines when AUTH and/or TLS are in use. This is the "received protocol"
4175 field. Exim used to use "asmtp" for authenticated SMTP, without any
4176 indication (in the protocol name) for TLS use. Now it follows the RFC and
4177 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4178 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4179 names appear in log lines as well as in Received: header lines.
4181 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4182 extensions, and comparison tests, and to fix some bugs.
4184 14. Changes to the "personal" filter test:
4186 (1) The test was buggy in that it was just doing the equivalent of
4187 "contains" tests on header lines. For example, if a user's address was
4188 anne@some.where, the "personal" test would incorrectly be true for
4190 To: susanne@some.where
4192 This test is now done by extracting each address from the header in turn,
4193 and checking the entire address. Other tests that are part of "personal"
4194 are now done using regular expressions (for example, to check local parts
4195 of addresses in From: header lines).
4197 (2) The list of non-personal local parts in From: addresses has been
4198 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4199 taken from the Sieve specification recommendations.
4201 (3) If the message contains any header line starting with "List-" it is
4202 treated as non-personal.
4204 (4) The test for "circular" in the Subject: header line has been removed
4205 because it now seems ill-conceived.
4207 15. Minor typos in src/EDITME comments corrected.
4209 16. Installed latest exipick from John Jetmore.
4211 17. If headers_add on a router specified a text string that was too long for
4212 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4213 of string_sprintf() is now avoided.
4215 18. $message_body_size was not set (it was always zero) when running the DATA
4216 ACL and the local_scan() function.
4218 19. For the "mail" command in an Exim filter, no default was being set for
4219 the once_repeat time, causing a random time value to be used if "once" was
4220 specified. (If the value happened to be <= 0, no repeat happened.) The
4221 default is now 0s, meaning "never repeat". The "vacation" command was OK
4222 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4223 (I found it when inspecting the code).
4225 20. There is now an overall timeout for performing a callout verification. It
4226 defaults to 4 times the callout timeout, which applies to individual SMTP
4227 commands during the callout. The overall timeout applies when there is more
4228 than one host that can be tried. The timeout is checked before trying the
4229 next host. This prevents very long delays if there are a large number of
4230 hosts and all are timing out (e.g. when the network connections are timing
4231 out). The value of the overall timeout can be changed by specifying an
4232 additional sub-option for "callout", called "maxwait". For example:
4234 verify = sender/callout=5s,maxwait=20s
4236 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4237 to the end before writing, but this should make it even safer).
4239 22. Exim was forgetting that it had advertised PIPELINING for the second and
4240 subsequent messages on an SMTP connection. It was also not resetting its
4241 memory on STARTTLS and an internal HELO.
4243 23. When Exim logs an SMTP synchronization error within a session, it now
4244 records whether PIPELINING has been advertised or not.
4246 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4247 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4248 rather than long int.
4250 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4252 26. Added the never_mail option to autoreply.
4258 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4259 crash if the getsockname() call failed; this can happen if a connection is
4260 closed very soon after it is established. The problem was simply in the
4261 order in which certain operations were done, causing Exim to try to write
4262 to the SMTP stream before it had set up the file descriptor. The bug has
4263 been fixed by making things happen in the correct order.
4269 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4270 before the connection was closed, thus losing the rejection response.
4272 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4273 some early Solaris releases, but causes trouble in current releases where
4274 socklen_t is defined.
4276 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4279 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4282 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4283 that releasing the top of it at the end releases what was used for sub-
4284 expansions (unless the block got too big). However, discard this block if
4285 the first thing is a variable or header, so that we can use its block when
4286 it is dynamic (useful for very large $message_headers, for example).
4288 6. Lookups now cache *every* query, not just the most recent. A new, separate
4289 store pool is used for this. It can be recovered when all lookup caches are
4290 flushed. Lookups now release memory at the end of their result strings.
4291 This has involved some general refactoring of the lookup sources.
4293 7. Some code has been added to the store_xxx() functions to reduce the amount
4294 of flapping under certain conditions.
4296 8. log_incoming_interface used to affect only the <= reception log lines. Now
4297 it causes the local interface and port to be added to several more SMTP log
4298 lines, for example "SMTP connection from", and rejection lines.
4300 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4302 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4304 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4305 could be overwritten at the end of the current message (or the start of a
4306 new message if it was set in a HELO ACL). The value is now preserved for
4307 the duration of the SMTP connection.
4309 12. If a transport had a headers_rewrite setting, and a matching header line
4310 contained an unqualified address, that address was qualified, even if it
4311 did not match any rewriting rules. The underlying bug was that the values
4312 of the flags that permit the existence of unqualified sender and recipient
4313 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4314 non-local messages, and by -bnq for local messages) were not being
4315 preserved with the message after it was received.
4317 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4318 "next input=" as part of the text comprising the host identity instead of
4319 the correct text. The code was using the same buffer for two different
4320 strings. However, depending on which order the printing function evaluated
4321 its arguments, the bug did not always show up. Under Linux, for example, my
4322 test suite worked just fine.
4324 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4325 doesn't work with some older versions of Perl. It has been changed to "my",
4326 which in any case is probably the better facility to use.
4328 15. A really picky compiler found some instances of statements for creating
4329 error messages that either had too many or two few arguments for the format
4332 16. The size of the buffer for calls to the DNS resolver has been increased
4333 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4334 for addresses that have a lot of PTR records. This alleviates a problem; it
4335 does not fully solve it.
4337 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4338 buffer now truncates the list and logs the incident, which is the same
4339 action as happens when Exim is looking up a host name and its aliases.
4340 Previously in this situation something unpredictable would happen;
4341 sometimes it was "internal error: store_reset failed".
4343 18. If a server dropped the connection unexpectedly when an Exim client was
4344 using GnuTLS and trying to read a response, the client delivery process
4345 crashed while trying to generate an error log message.
4347 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4348 string, for example:
4350 warn message = H1: something\nH2: something
4352 the text was added as a single header line from Exim's point of view
4353 though it ended up OK in the delivered message. However, searching for the
4354 second and subsequent header lines using $h_h2: did not work. This has been
4355 fixed. Similarly, if a system filter added multiple headers in this way,
4356 the routers could not see them.
4358 20. Expanded the error message when iplsearch is called with an invalid key to
4359 suggest using net-iplsearch in a host list.
4361 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4362 ACLs are no longer actually imposed (and a message to that effect is
4365 22. If a "gecos" field in a passwd entry contained escaped characters, in
4366 particular, if it contained a \" sequence, Exim got it wrong when building
4367 a From: or a Sender: header from that name. A second bug also caused
4368 incorrect handling when an unquoted " was present following a character
4369 that needed quoting.
4371 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4372 was not being matched caselessly.
4374 24. Arranged for all hyphens in the exim.8 source to be escaped with
4377 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4378 properly. Recipient callout cache records were still being keyed to include
4379 the sender, even when use_sender was set false. This led to far more
4380 callouts that were necessary. The sender is no longer included in the key
4381 when use_sender is false.
4383 26. Added "control = submission" modifier to ACLs.
4385 27. Added the ${base62d: operator to decode base 62 numbers.
4387 28. dnsdb lookups can now access SRV records.
4389 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4390 the configuration file.
4392 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4393 (-v) mode. This makes the output for a verbose queue run more intelligible.
4395 31. Added a use_postmaster feature to recipient callouts.
4397 32. Added the $body_zerocount variable, containing the number of binary zero
4398 bytes in the message body.
4400 33. The time of last modification of the "new" subdirectory is now used as the
4401 "mailbox time last read" when there is a quota error for a maildir
4404 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4406 35. Added +ignore_unknown as a special item in host lists.
4408 36. Code for decoding IPv6 addresses in host lists is now included, even if
4409 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4410 address was recognized as an IP address, but was then not correctly decoded
4411 into binary, causing unexpected and incorrect effects when compared with
4418 1. Very minor rewording of debugging text in manualroute to say "list of
4419 hosts" instead of "hostlist".
4421 2. If verify=header_syntax was set, and a header line with an unqualified
4422 address (no domain) and a large number of spaces between the end of the
4423 name and the colon was received, the reception process suffered a buffer
4424 overflow, and (when I tested it) crashed. This was caused by some obsolete
4425 code that should have been removed. The fix is to remove it!
4427 3. When running in the test harness, delay a bit after writing a bounce
4428 message to get a bit more predictability in the log output.
4430 4. Added a call to search_tidyup() just before forking a reception process. In
4431 theory, someone could use a lookup in the expansion of smtp_accept_max_
4432 per_host which, without the tidyup, could leave open a database connection.
4434 5. Added the variables $recipient_data and $sender_data which get set from a
4435 lookup success in an ACL "recipients" or "senders" condition, or a router
4436 "senders" option, similar to $domain_data and $local_part_data.
4438 6. Moved the writing of debug_print from before to after the "senders" test
4441 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4442 problems for message scanning, either using a data ACL, or using
4443 local_scan() because the Received: header was not generated till after they
4444 were called (in order to set the time as the time of reception completion).
4445 I have revised the way this works. The header is now generated after the
4446 body is received, but before the ACL or local_scan() are called. After they
4447 are run, the timestamp in the header is updated.
4453 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4454 before starting a queue runner without re-exec. This happened only when
4455 deliver_drop_privilege was set or when the Exim user was set to root. The
4456 effect of the bug was that timeouts during subsequent deliveries caused
4457 crashes instead of being properly handled. The handler is now left at its
4458 default (and expected) setting.
4460 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4461 message, again when deliver_drop_privilege is set or Exim is run as root.
4462 The bug described in (1) was not present in this case, but the tidying up
4463 of the other signals was missing. I have made the two cases consistent.
4465 3. The ignore_target_hosts setting on a manualroute router was being ignored
4466 for hosts that were looked up using the /MX notation.
4468 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4471 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4472 operated on the sender address. After changing the $sender_address to <>
4473 for the sender address verify, Exim was re-instated it as the original
4474 (before rewriting) address, but remembering that it had rewritten it, so it
4475 wasn't rewriting it again. This bug also had the effect of breaking the
4476 sender address verification caching when the sender address was rewritten.
4478 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4479 This has been changed so that if the ip literal address matches
4480 ignore_target_hosts, the router declines.
4482 7. Added expansion conditions match_domain, match_address, and match_local_
4483 part (NOT match_host).
4485 8. The placeholder for the Received: header didn't have a length field set.
4487 9. Added code to Exim itself and to exim_lock to test for a specific race
4488 condition that could lead to file corruption when using MBX delivery. The
4489 issue is with the lockfile that is created in /tmp. If this file is removed
4490 after a process has opened it but before that process has acquired a lock,
4491 there is the potential for a second process to recreate the file and also
4492 acquire a lock. This could lead to two Exim processes writing to the file
4493 at the same time. The added code performs the same test as UW imapd; it
4494 checks after acquiring the lock that its file descriptor still refers to
4495 the same named file.
4497 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4498 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4501 11. Added the smtp_active_hostname option. If used, this will typically be made
4502 to depend on the incoming interface address. Because $interface_address is
4503 not set up until the daemon has forked a reception process, error responses
4504 that can happen earlier (such as "too many connections") no longer contain
4507 12. If an expansion in a condition on a "warn" statement fails because a lookup
4508 defers, the "warn" statement is abandoned, and the next ACL statement is
4509 processed. Previously this caused the whole ACL to be aborted.
4511 13. Added the iplsearch lookup type.
4513 14. Added ident_timeout as a log selector.
4515 15. Added tls_certificate_verified as a log selector.
4517 16. Added a global option tls_require_ciphers (compare the smtp transport
4518 option of the same name). This controls incoming TLS connections.
4520 17. I finally figured out how to make tls_require_ciphers do a similar thing
4521 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4522 before starting the TLS session.
4524 18. Tabs are now shown as \t in -bP output.
4526 19. If the log selector return_path_on_delivery was set, Exim crashed when
4527 bouncing a message because it had too many Received: header lines.
4529 20. If two routers both had headers_remove settings, and the first one included
4530 a superfluous trailing colon, the final name in the first list and the
4531 first name in the second list were incorrectly joined into one item (with a
4532 colon in the middle).
4538 1. Added -C and -D options to the exinext utility, mainly to make it easier
4539 to include in the automated testing, but these could be helpful when
4540 multiple configurations are in use.
4542 2. The exinext utility was not formatting the output nicely when there was
4543 an alternate port involved in the retry record key, nor when there was a
4544 message id as well (for retries that were specific to a specific message
4545 and a specific host). It was also confused by IPv6 addresses, because of
4546 the additional colons they contain. I have fixed the IPv4 problem, and
4547 patched it up to do a reasonable job for IPv6.
4549 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4550 delivery, the log line now contains "pipelined" if PIPELINING was used.
4552 4. An SMTP transport process used to panic and die if the bind() call to set
4553 an explicit outgoing interface failed. This has been changed; it is now
4554 treated in the same way as a connect() failure.
4556 5. A reference to $sender_host_name in the part of a conditional expansion
4557 that was being skipped was still causing a DNS lookup. This no longer
4560 6. The def: expansion condition was not recognizing references to header lines
4561 that used bh_ and bheader_.
4563 7. Added the _cache feature to named lists.
4565 8. The code for checking quota_filecount in the appendfile transport was
4566 allowing one more file than it should have been.
4568 9. For compatibility with Sendmail, the command line option
4576 and sets the incoming protocol and host name (for trusted callers). The
4577 host name and its colon can be omitted when only the protocol is to be set.
4578 Note the Exim already has two private options, -pd and -ps, that refer to
4579 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4580 "s", but I don't think that's a major issue.
4582 10. A number of refactoring changes to the code, none of which should affect
4585 (a) The number of logging options was getting close to filling up the
4586 32-bit word that was used as a bit map. I have split them into two classes:
4587 those that are passed in the argument to log_write(), and those that are
4588 only ever tested independently outside of that function. These are now in
4589 separate 32-bit words, so there is plenty of room for expansion again.
4590 There is no change in the user interface or the logging behaviour.
4592 (b) When building, for example, log lines, the code previously used a
4593 macro that called string_cat() twice, in order to add two strings. This is
4594 not really sufficiently general. Furthermore, there was one instance where
4595 it was actually wrong because one of the argument was used twice, and in
4596 one call a function was used. (As it happened, calling the function twice
4597 did not affect the overall behaviour.) The macro has been replaced by a
4598 function that can join an arbitrary number of extra strings onto a growing
4601 (c) The code for expansion conditions now uses a table and a binary chop
4602 instead of a serial search (which was left over from when there were very
4603 few conditions). Also, it now recognizes conditions like "pam" even when
4604 the relevant support is not compiled in: a suitably worded error message is
4605 given if an attempt is made to use such a condition.
4607 11. Added ${time_interval:xxxxx}.
4609 12. A bug was causing one of the ddress fields not to be passed back correctly
4610 from remote delivery subprocesses. The field in question was not being
4611 subsequently used, so this caused to problems in practice.
4613 13. Added new log selectors queue_time and deliver_time.
4615 14. Might have fixed a bug in maildirsizefile handling that threw up
4616 "unexpected character" debug warnings, and recalculated the data
4617 unnecessarily. In any case, I expanded the warning message to give more
4620 15. Added the message "Restricted characters in address" to the statements in
4621 the default ACL that block characters like @ and % in local parts.
4623 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4624 Three changes have been made:
4626 (a) There was a serious bug; a negative response to MAIL caused the whole
4627 recipient domain to be cached as invalid, thereby blocking all messages
4628 to all local parts at the same domain, from all senders. This bug has
4629 been fixed. The domain is no longer cached after a negative response to
4630 MAIL if the sender used is not empty.
4632 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4635 (c) A new callout option, "use_sender" has been added for people who want
4636 the modified behaviour.
4642 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4645 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4646 indeed breaks things for older releases.
4648 3. Added additional logging to the case where there is a problem reading data
4649 from a filter that is running in a subprocess using a pipe, in order to
4650 try to track down a specific problem.
4652 4. Testing facility fudge: when running in the test harness and attempting
4653 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4654 getting "No route to host". Convert this to a timeout.
4656 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4659 6. Some OS don't have socklen_t but use size_t instead. This affects the
4660 fifth argument of getsockopt() amongst other things. This is now
4661 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4662 can be set for individual OS. I have set it for SunOS5, OSF1, and
4663 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4664 some earlier ones do not.
4666 7. Change 4.30/15 was not doing the test caselessly.
4668 8. The standard form for an IPv6 address literal was being rejected by address
4669 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4670 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4671 this, as well as the form without the "IPv6" on the front (but only when
4672 address literals are enabled, of course).
4674 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4676 10. Exim crashed if a message with an empty sender address specified by -f
4677 encountered a router with an errors_to setting. This could be provoked only
4678 by a command such as
4682 where an empty string was supplied; "<>" did not hit this bug.
4684 11. Installed PCRE release 4.5.
4686 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4687 remained set. It is now erased.
4689 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4690 times from message ids (which are base 36 rather than the normal 62).
4692 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4693 were being counted as actual protocol errors, and logged if the log
4694 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4695 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4696 and DATA following a set of rejected RCPTs do not count as protocol errors.
4697 In other words, Exim assumes they were pipelined, though this may not
4698 actually be the case. Of course, in all cases the client gets an
4699 appropriate error code.
4701 15. If a lookup fails in an ACL condition, a message about the failure may
4702 be available; it is used if testing the ACL cannot continue, because most
4703 such messages specify what the cause of the deferral is. However, some
4704 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4705 that caused an old message to be retained and used if a later statement
4706 caused a defer, replacing the real cause of the deferral.
4708 16. If an IP address had so many PTR records that the DNS lookup buffer
4709 was not large enough to hold them, Exim could crash while trying to process
4710 the truncated data. It now detects and logs this case.
4712 17. Further to 4.21/58, another change has been made: if (and only if) the
4713 first line of a message (the first header line) ends with CRLF, a bare LF
4714 in a subsequent header line has a space inserted after it, so as not to
4715 terminate the header.
4717 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4718 unnecessarily, used atoi() instead of strtol(), and didn't check the
4719 termination when getting file sizes from file names by regex.
4721 19. Completely re-implemented the support for maildirsize files, in the light
4722 of a number of problems with the previous contributed implementation
4723 (4.30/29). In particular:
4725 . If the quota is zero, the maildirsize file is maintained, but no quota is
4728 . If the maildir directory does not exist, it is created before any attempt
4729 to write a maildirsize file.
4731 . The quota value in the file is just a cache; if the quota is changed in
4732 the transport, the new value overrides.
4734 . A regular expression is available for excluding directories from the
4737 20. The autoreply transport checks the characters in options that define the
4738 message's headers; it allows continued headers, but it was checking with
4739 isspace() after an embedded newline instead of explicitly looking for a
4742 21. If all the "regular" hosts to which an address was routed had passed their
4743 expiry times, and had not reached their retry times, the address was
4744 bounced, even if fallback hosts were defined. Now Exim should go on to try
4747 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4748 equivalent code in the SMTP transport. Some hosts send humungous responses
4749 to HELO/EHLO, more than 1024 it seems.
4751 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4752 gives compiler warnings in some environments. I've now done it "properly",
4755 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4756 (because of problems with the built-in one) was declared to return uschar *
4757 instead of char *, causing compiler failure.
4759 25. Fixed a file descriptor leak when processing alias/forward files.
4761 26. Fixed a minor format string issue in dbfn.c.
4763 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4765 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4766 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4769 29. When a sender address is verified, it is cached, to save repeating the test
4770 when there is more than one recipient in a message. However, when the
4771 verification involves a callout, it is possible for different callout
4772 options to be set for different recipients. It is too complicated to keep
4773 track of this in the cache, so now Exim always runs a verification when a
4774 callout is required, relying on the callout cache for the optimization.
4775 The overhead is duplication of the address routing, but this should not be
4778 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4779 to be verified with callout=postmaster, and the main callout worked but the
4780 postmaster check failed, the verification correctly failed. However, if a
4781 subsequent RCPT command asked for sender verification *without* the
4782 postmaster check, incorrect caching caused this verification also to fail,
4785 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4786 it was not caching the DNS options (qualify_single, search_parents) that
4787 were used when the lookup failed. A subsequent lookup with different
4788 options therefore always gave the same answer, though there were cases
4789 where it should not have. (Example: a "domains = !$mx_any" option on a
4790 dnslookup router: the "domains" option is always processed without any
4791 widening, but the router might have qualify_single set.) Now Exim uses the
4792 cached value only when the same options are set.
4794 32. Added John Jetmore's "exipick" utility to the distribution.
4796 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4797 than a timeout (e.g. a certificate is required, and is not provided), an
4798 Exim server now closes the connection immediately. Previously it waited for
4799 the client to close - but if the client is SSL, it seems that they each
4800 wait for each other, leading to a delay before one of them times out.
4802 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4803 maintained 0.8.x compatibility because I don't think many are using it, and
4804 it is clearly obsolete.
4806 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4809 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4810 client certificate was expired. A simple patch fixes this, though I don't
4811 understand the full logic of why the verify callback is called multiple
4814 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4815 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4816 which causes problems with some clients (such as the Certicom SSL Plus
4817 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4818 disables the coutermeasure allowing Eudora to connect."
4820 38. Exim was not checking that a write() to a log file succeeded. This could
4821 lead to Bad Things if a log got too big, in particular if it hit a file
4822 size limit. Exim now panics and dies if it cannot write to a log file, just
4823 as it does if it cannot open a log file.
4825 39. Modified OS/Makefile-Linux so that it now contains
4827 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4829 The two -D definitions ensure that Exim is compiled with large file
4830 support, which makes it possible to handle log files that are bigger than
4833 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4834 instance) a domain was checked against a named list that involved a lookup,
4835 causing $domain_data to be set, then another domain was checked against the
4836 same list, then the first domain was re-checked, the value of $domain_data
4837 after the final check could be wrong. In particular, if the second check
4838 failed, it could be set empty. This bug probably also applied to
4841 41. The strip_trailing_dot option was not being applied to the address given
4842 with the -f command-line option.
4844 42. The code for reading a message's header from the spool was incrementing
4845 $received_count, but never initializing it. This meant that the value was
4846 incorrect (doubled) while delivering a message in the same process in which
4847 it was received. In the most common configuration of Exim, this never
4848 happens - a fresh exec is done - but it can happen when
4849 deliver_drop_privilege is set.
4851 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4852 it now includes up to 150 characters of the unexpected data in the log
4855 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4856 and building data strings. The size of both of these buffers was 10 000
4857 bytes - far larger than anybody would *ever* want, thought I. Needless to
4858 say, somebody hit the limit. I have increased the maximum line length to
4859 20 000 and the maximum data length of concatenated lines to 100 000. I have
4860 also fixed two bugs, because there was no checking on these buffers. Tsk,
4861 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4862 buffer is too small.
4864 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4865 lsearch lookups. Now it does.
4867 46. When parsing a route_list item in a manualroute router, a fixed-length
4868 buffer was used for the list of hosts. I made this 1024 bytes long,
4869 thinking that nobody would ever have a list of hosts that long. Wrong.
4870 Somebody had a whole pile of complicated expansion conditions, and the
4871 string was silently truncated, leading to an expansion error. It turns out
4872 that it is easier to change to an unlimited length (owing to other changes
4873 that have happened since this code was originally written) than to build
4874 structure for giving a limitation error. The length of the item that
4875 expands into the list of hosts is now unlimited.
4877 47. The lsearch lookup could not handle data where the length of text line was
4878 more than 4095 characters. Such lines were truncated, leading to shortened
4879 data being returned. It should now handle lines of any length.
4881 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4882 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4885 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4886 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4887 no longer happens in comment lines. A list of replacements is now placed
4888 at the head of all of the source files, except those whose only change is
4889 to replace PERL_COMMAND in the very first #! line.
4891 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4892 messages on the queue, with a bottom-up merge sort, using code contributed
4893 by Michael Haardt. This should make operations like -bp somewhat faster on
4894 large queues. It won't affect queue runners, except when queue_run_in_order
4897 51. Installed eximstats 1.31 in the distribution.
4899 52. Added support for SRV lookups to the dnslookup router.
4901 53. If an ACL referred to $message_body or $message_body_end, the value was not
4902 reset for any messages that followed in the same SMTP session.
4904 54. The store-handling optimization for building very long strings was not
4905 differentiating between the different store pools. I don't think this
4906 actually made any difference in practice, but I've tidied it.
4908 55. While running the routers to verify a sender address, $sender_address
4909 was still set to the sender address. This is wrong, because when routing to
4910 send a bounce to the sender, it would be empty. Therefore, I have changed
4911 it so that, while verifying a sender address, $sender_address is set to <>.
4912 (There is no change to what happens when verifying a recipient address.)
4914 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4915 target A or AAAA records (if not already returned) without resetting the
4916 qualify_single or search_parents options of the DNS resolver. These are
4917 inappropriate in this case because the targets of MX and SRV records must
4918 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4919 target that, when qualified, matched something in the local domain. These
4920 two options are now turned off when doing these lookups.
4922 57. It seems that at least some releases of Reiserfs (which does not have the
4923 concept of a fixed number of inodes) returns zero and not -1 for the
4924 number of available inodes. This interacted badly with check_spool_inodes,
4925 which assumed that -1 was the "no such thing" setting. What I have done is
4926 to check that the total number of inodes is greater than zero before doing
4927 the test of how many are available.
4929 58. When a "warn" ACL statement has a log_message modifier, the message is
4930 remembered, and not repeated. This is to avoid a lot of repetition when a
4931 message has many recipients that cause the same warning to be written.
4932 Howewer, Exim was preserving the list of already written lines for an
4933 entire SMTP session, which doesn't seem right. The memory is now reset if a
4934 new message is started.
4936 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4937 debugging output unless log_rewrite was also set.
4939 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4940 of (int)(handle) when we know that handle contains (void *)(-1).
4942 61. The Exim daemon panic-logs an error return when it closes the incoming
4943 connection. However "connection reset by peer" seems to be common, and
4944 isn't really an error worthy of noting specially, so that particular error
4947 62. When Exim is trying to find all the local interfaces, it used to panic and
4948 die if the ioctl to get the interface flags failed. However, it seems that
4949 on at least one OS (Solaris 9) it is possible to have an interface that is
4950 included in the list of interfaces, but for which you get a failure error
4951 for this call. This happens when the interface is not "plumbed" into a
4952 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4953 failure of the "get flags" call assumes that the interface is down.
4955 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4956 makes life easier for people who are doing arithmetic on fields extracted
4957 from dates, where you often get leading zeros that should not be
4958 interpreted as octal.
4960 64. Added qualify_domain to the redirect router, to override the global
4963 65. If a pathologically long header line contained very many addresses (the
4964 report of this problem mentioned 10 000) and each of them was rewritten,
4965 Exim could use up a very large amount of memory. (It kept on making new
4966 copies of the header line as it rewrote, and never released the old ones.)
4967 At the expense of a bit more processing, the header rewriting function has
4968 been changed so that it no longer eats memory in this way.
4970 66. The generation of the Received: header has been moved from the time that a
4971 message starts to be received, to the time that it finishes. The timestamp
4972 in the Received: header should now be very close to that of the <= log
4973 line. There are two side-effects of this change:
4975 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4976 logged header lines no longer include the local Received: line, because
4977 it has not yet been created. The same applies to a copy of the message
4978 that is returned to a non-SMTP sender when a message is rejected.
4980 (b) When a filter file is tested using -bf, no additional Received: header
4981 is added to the test message. After some thought, I decided that this
4984 This change does not affect the value of $received_for. It is still set
4985 after address rewriting, but before local_scan() is called.
4987 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4989 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4990 gave an unhelpful panic error message, and a defer error. I have managed to
4991 change this behaviour so that it now rejects any supplied certificate,
4992 which seems right, as the list of acceptable certificates is empty.
4994 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4995 gave an unhelpful defer error. I have not managed to make this reject any
4996 supplied certificates, but the error message it gives is "no certificate
4997 supplied", which is not helpful.
4999 70. exigrep's output now also includes lines that are not associated with any
5000 message, but which match the given pattern. Implemented by a patch from
5001 Martin Sluka, which also tidied up the Perl a bit.
5003 71. Recipient callout verification, like sender verification, was using <> in
5004 the MAIL FROM command. This isn't really the right thing, since the actual
5005 sender may affect whether the remote host accepts the recipient or not. I
5006 have changed it to use the actual sender in the callout; this means that
5007 the cache record is now keyed on a recipient/sender pair, not just the
5008 recipient address. There doesn't seem to be a real danger of callout loops,
5009 since a callout by the remote host to check the sender would use <>.
5010 [SEE ABOVE: changed after hitting problems.]
5012 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5013 temporary errors. However, in the case of such a code being given after
5014 the end of a data transmission (i.e. after ".") Exim was failing to write
5015 a retry record for the message. (Yes, there was some broken host that was
5016 actually sending 8xx at this point.)
5018 73. An unknown lookup type in a host list could cause Exim to panic-die when
5019 the list was checked. (An example that provoked this was putting <; in the
5020 middle of a list instead of at the start.) If this happened during a DATA
5021 ACL check, a -D file could be left lying around. This kind of configuration
5022 error no longer causes Exim to die; instead it causes a defer errror. The
5023 incident is still logged to the main and panic logs.
5025 74. Buglet left over from Exim 3 conversion. The message "too many messages
5026 in one connection" was written to the rejectlog but not the mainlog, except
5027 when address rewriting (yes!) was being logged.
5029 75. Added write_rejectlog option.
5031 76. When a system filter was run not as root (that is, when system_filter_user
5032 was set), the values of the $n variables were not being returned to the
5033 main process; thus, they were not subsequently available in the $sn
5036 77. Added +return_path_on_delivery log selector.
5038 78. A connection timeout was being treated differently from recipients deferred
5039 when testing hosts_max_try with a message that was older than the host's
5040 retry timeout. (The host should not be counted, thus allowing all hosts to
5041 be tried at least once before bouncing.) This may have been the cause of an
5042 occasionally reported bug whereby a message would remain on the queue
5043 longer than the retry timeout, but would be bounced if a delivery was
5044 forced. I say "may" because I never totally pinned down the problem;
5045 setting up timeout/retry tests is difficult. See also the next item.
5047 79. The ultimate address timeout was not being applied to errors that involved
5048 a combination of host plus message (for example, a timeout on a MAIL
5049 command). When an address resolved to a number of possible hosts, and they
5050 were not all tried for each delivery (e.g. because of hosts_max_try), a
5051 message could remain on the queue longer than the retry timeout.
5053 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5056 81. Fixed an obscure SMTP outgoing bug which required at least the following
5057 conditions: (a) there was another message waiting for the same server;
5058 (b) the server returned 5xx to all RCPT commands in the first message so
5059 that the message was not completed; (c) the server dropped the connection
5060 or gave a negative response to the RSET that Exim sends to abort the
5061 transaction. The observed case was a dropped connection after DATA that had
5062 been sent in pipelining mode. That is, the server had advertised PIPELINING
5063 but was not implementing it correctly. The effect of the bug was incorrect
5064 behaviour, such as trying another host, and this could lead to a crash.
5070 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5071 and daemon.c were passed as pointers to ints; they should have been
5072 pointers to socklen_t variables (which are typically unsigned ints).
5074 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5077 3. Fixed a really odd bug that affected only the testing scheme; patching a
5078 certain fixed string in the binary changed the value of another string that
5079 happened to be identical to the end of the original first string.
5081 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5082 name", it returns that address as the IP address. On some operating
5083 systems (e.g. Solaris), it also passes back the IP address string as the
5084 "host name". However, on others (e.g. Linux), it passes back an empty
5085 string. Exim wasn't checking for this, and was changing the host name to an
5086 empty string, assuming it had been canonicized.
5088 5. Although rare, it is permitted to have more than one PTR record for a given
5089 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5090 all the names associated with an address, because they do in Solaris.
5091 However, it seems that they do not in Linux for data that comes from the
5092 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5093 I found this out when I moved to a new Linux workstation and tried to run
5094 the Exim test suite.
5096 To get round this problem I have changed the code so that it now does its
5097 own call to the DNS to look up PTR records when searching for a host name.
5098 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5099 addresses that are only in /etc/hosts are still found.
5101 This behaviour is, however, controlled by an option called host_lookup_
5102 order, which defaults to "bydns:byaddr". If people want to use the other
5103 order, or indeed, just use one or the other means of lookup, they can
5104 specify it in this variable.
5106 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5107 some operating systems, this comes back from gethostbyaddr() as an empty
5108 string, and this is what Exim used to test for. However, it seems that in
5109 other systems, "." is yielded. Exim now tests for this case too.
5111 7. The values of check_spool_space and check_log_space are now held internally
5112 as a number of kilobytes instead of an absolute number of bytes. If a
5113 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5114 kilobyte. This means that much larger values can be stored.
5116 8. Exim monitor: an attempt to get the action menu when not actually pointing
5117 at a message produces an empty menu entitled "No message selected". This
5118 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5119 no entries in it ("Shell widget menu has zero width and/or height"). So I
5120 have added a single, blank menu entry in this case.
5122 9. Added ${quote_local_part.
5124 10. MIME decoding is now applied to the contents of Subject: header lines when
5127 11. Now that a reference to $sender_host_address automatically causes a reverse
5128 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5129 host lookup before query-style lookups in lists that might use this
5130 variable. This has therefore been abolished, and the "net-" prefix is no
5131 longer necessary for query-style lookups.
5133 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5134 have been a typo for LFLAGS, so it has been changed.
5136 13. The install script calls Exim with "-C /dev/null" in order to find the
5137 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5138 to be output. Howeve, since Exim outputs its version number before the
5139 error, it didn't break the script. It just looked ugly. I fixed this by
5140 always allowing "-C /dev/null" if the caller is root.
5142 14. Ignore overlarge ACL variable number when reading spool file - insurance
5143 against a later release with more variables having written the file.
5145 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5146 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5147 this, as well as the form without the "IPv6" on the front.
5149 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5150 OS/Makefile-Darwin file.
5152 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5153 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5155 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5156 variable. However, some OS use ints for inodes. I've added cast to long int
5157 to get rid of the compiler warning.
5159 19. I had forgotten to lock out "/../" in configuration file names when
5160 ALT_CONFIG_PREFIX was set.
5162 20. Routers used for verification do not need to specify transports. However,
5163 if such a router generated a host list, and callout was configured, Exim
5164 crashed, because it could not find a port number from the (non-existent)
5165 transport. It now assumes port 25 in this circumstance.
5167 21. Added the -t option to exigrep.
5169 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5170 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5171 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5172 them is set, LOOKUP_LSEARCH is forced.
5174 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5176 24. Added sender and host information to the "rejected by local_scan()" log
5177 line; previously there was no indication of these.
5179 25. Added .include_if_exists.
5181 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5182 Linux. It turns out that not all file systems support this. Apparently some
5183 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5184 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5185 not supported on the file descriptor, is now ignored when Exim is trying to
5186 sync a directory. This applies only to Linux.
5188 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5190 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5191 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5194 29. Added support for maildirsize files from supplied patch (modified a bit).
5196 30. The use of :fail: followed by an empty string could lead Exim to respond to
5197 sender verification failures with (e.g.):
5199 550 Verification failed for <xxx>
5200 550 Sender verify failed
5202 where the first response line was missing the '-' that indicates it is not
5203 the final line of the response.
5205 31. The loop for finding the name of the user that called Exim had a hardwired
5206 limit of 10; it now uses the value of finduser_retries, which is used for
5207 all other user lookups.
5209 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5212 33. Exim was neglecting to zero errno before one call of strtol() when
5213 expanding a string and expecting an integer value. On some systems this
5214 resulted in spurious "integer overflow" errors. Also, it was casting the
5215 result into an int without checking.
5217 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5218 did not work. The code looks as if it has *never* worked, though it appears
5219 to have been documented since at least releast 1.62. I have made it work.
5221 35. The "timeout_DNS" error in retry rules, also documented since at least
5222 1.62, also never worked. As it isn't clear exactly what this means, and
5223 clearly it isn't a major issue, I have abolished the feature by treating it
5224 as "timeout", and writing a warning to the main and panic logs.
5226 36. The display of retry rules for -brt wasn't always showing the error code
5229 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5230 timeout_connect_A, timeout_connect_MX.
5232 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5233 to the empty sender.
5235 39. The daemon was not analysing the content of -oX till after it had closed
5236 stderr and disconnected from the controlling terminal. This meant that any
5237 syntax errors were only noted on the panic log, and the return code from
5238 the command was 0. By re-arranging the code a little, I've made the
5239 decoding happen first, so such errors now appear on stderr, and the return
5240 code is 1. However, the actual setting up of the sockets still happens in
5241 the disconnected process, so errors there are still only recorded on the
5244 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5245 connections (as happens on some IP stacks) was logged at start up time as
5246 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5247 it from "IPv6 and IPv4", which means that two separate sockets are being
5250 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5251 says whether AF_INET or AF_INET6 was passed as an argument.
5253 42. Exiwhat output was messed up when time zones were included in log
5256 43. Exiwhat now gives more information about the daemon's listening ports,
5257 and whether -tls-on-connect was used.
5259 44. The "port" option of the smtp transport is now expanded.
5261 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5262 silently ignored. Now an error message is written to the main and panic
5265 46. There's a new ACL modifier called "logwrite" which writes to a log file
5266 as soon as it is encountered.
5268 47. Added $local_user_uid and $local_user_gid at routing time.
5270 48. Exim crashed when trying to verify a sender address that was being
5273 49. Exim was recognizing only a space character after ".include". It now also
5274 recognizes a tab character.
5276 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5277 extracting the relevant information from the specification. The man page no
5278 longer contains scrambled data for the -d option, and I've added a section
5279 at the front about calling Exim under different names.
5281 51. Added "extra_headers" argument to the "mail" command in filter files.
5283 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5286 53. Installed eximstats 1.29.
5288 54. Added transport_filter_timeout as a generic transport option.
5290 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5291 Cc: header lines. This was required by RFC 822, but it not required by RFC
5294 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5295 incoming messages that did not have them. Now it does so only if the
5296 message originates locally, that is, if there is no associated remote host
5297 address. When Resent- header lines are present, this applies to the Resent-
5298 lines rather than the non-Resent- lines.
5300 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5301 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5303 58. Messages for configuration errors now include the name of the main
5304 configuration file - useful now that there may be more than one file in a
5305 list (.included file names were always shown).
5307 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5308 for those rare installations that do not start the daemon as root or run it
5309 setuid root. I've cut out the call to initgroups() if the daemon is not
5312 60. The Exim user and group can now be bound into the binary as text strings
5313 that are looked up at the start of Exim's processing.
5315 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5317 62. Added $mailstore_basename variable.
5319 63. Installed patch to sieve.c from Michael Haardt.
5321 64. When Exim failed to open the panic log after failing to open the main log,
5322 the original message it was trying to log was written to stderr and debug
5323 output, but if they were not available (the usual case in production), it
5324 was lost. Now it is written to syslog before the two lines that record the
5325 failures to open the logs.
5327 65. Users' Exim filters run in subprocesses under the user's uid. It is
5328 possible for a "deliver" command or an alias in a "personal" command to
5329 provoke an address rewrite. If logging of address rewriting is configured,
5330 this fails because the process is not running as root or exim. There may be
5331 a better way of dealing with this, but for the moment (because 4.30 needs
5332 to be released), I have disabled address rewrite logging when running a
5333 filter in a non-root, non-exim process.
5339 1. The buildconfig auxiliary program wasn't quoting the value set for
5340 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5341 not defined. This bug was masked in 4.22 by the effect that was fixed in
5344 2. Some messages that were rejected after a message id was allocated were
5345 shown as "incomplete" by exigrep. It no longer does this for messages that
5346 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5348 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5349 have allow_domain_literals set, the ID did not get logged in the <= line.
5350 Domain literals are now always recognized in Message-ID: header lines.
5352 4. The first argument for a ${extract expansion item is the key name or field
5353 number. Leading and trailing spaces in this item were not being ignored,
5354 causing some misleading effects.
5356 5. When deliver_drop_privilege was set, single queue runner processes started
5357 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5358 same command in the process it spins off) were not dropping privilege.
5360 6. When the daemon running as "exim" started a queue runner, it always
5361 re-executed Exim in the spun-off process. This is a waste of effort when
5362 deliver_drop_privilege is set. The new process now just calls the
5363 queue-runner function directly.
5369 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5372 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5373 ignored. Though the use of -oP was forcing the writing of a pid file, it
5374 was always written to the default place.
5376 3. If the message "no IP address found for host xxxx" is generated during
5377 incoming verification, it is now followed by identification of the incoming
5378 connection (so you can more easily find what provoked it).
5380 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5382 5. Added some features to "harden" Exim a bit more against certain attacks:
5384 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5385 be put in Local/Makefile. This is like the never_users runtime option,
5386 but it cannot be overridden. The default setting is "root".
5388 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5389 prefix string with which any file named in a -C command line option
5392 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5393 is retained for -C and -D only if the caller of Exim is root. Without
5394 it, the exim user may also use -C and -D and retain privilege.
5396 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5397 command line option is disabled.
5399 6. Macro names set by the -D option must start with an upper case letter, just
5400 like macro names defined in the configuration file.
5402 7. Added "dereference=" facility to LDAP.
5404 8. Two instances of the typo "uknown" in the source files are fixed.
5406 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5407 the Configure-Makefile script screwed up while processing it.
5409 10. Incorporated PCRE 4.4.
5411 11. The SMTP synchronization check was not operating right at the start of an
5412 SMTP session. For example, it could not catch a HELO sent before the client
5413 waited for the greeting. There is now a check for outstanding input at the
5414 point when the greeting is written. Because of the duplex, asynchronous
5415 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5416 way, but not yet received, when the check is performed.
5418 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5419 on TCP/IP sockets, because this apparently causes some broken clients to
5422 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5423 unchanged) from the Cygwin maintainer.
5425 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5426 was supported instead of testing for mbx. Effectively a typo.
5428 15. The spa authenticator server code was not checking that the input it
5429 received was valid base64.
5431 16. The debug output line for the "set" modifier in ACLs was not showing the
5432 name of the variable that was being set.
5434 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5436 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5437 lookup on its own. Something else was needed to trigger the lookup. For
5438 example, a match in host_lookup or the need for a host name in a host list.
5439 Now, if $sender_host_name is referenced and the host name has not yet been
5440 looked up, a lookup is performed. If the lookup fails, the variable remains
5441 empty, and $host_lookup_failed is set to "1".
5443 19. Added "eqi" as a case-independent comparison operator.
5445 20. The saslauthd authentication condition could segfault if neither service
5446 nor realm was specified.
5448 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5449 error message that was logged was misleading, and incoming SMTP
5450 connections were dropped. The message is now more accurate, and temporary
5451 errors are given to SMTP connections.
5453 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5454 (except RSET) with a 421 error, until QUIT is received. However, it was
5455 failing to send a response to QUIT.
5457 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5458 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5459 ACL is now run after the helo_try_verify_hosts code.
5461 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5462 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5463 case-independent, so other case variants are also recognized). Apparently
5464 some systems use these upper case variants.
5466 25. If more than two messages were waiting for the same host, and a transport
5467 filter was specified for the transport, Exim sent two messages over the
5468 same TCP/IP connection, and then failed with "socket operation on non-
5469 socket" when it tried to send the third.
5471 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5473 27. The extern definition of crypt16() in expand.c was not being excluded when
5474 the OS had its own crypt16() function.
5476 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5477 as a preferred synonym for return_size_limit, both as an option and as an
5480 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5482 30. Changed the default configuration ACL to relax the local part checking rule
5483 for addresses that are not in any local domains. For these addresses,
5484 slashes and pipe symbols are allowed within local parts, but the sequence
5485 /../ is explicitly forbidden.
5487 31. SPA server authentication was not clearing the challenge buffer before
5490 32. log_message in a "warn" ACL statement was writing to the reject log as
5491 well as to the main log, which contradicts the documentation and doesn't
5492 seem right (because no rejection is happening). So I have stopped it.
5494 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5495 However, I am unable to do any testing of this.
5497 34. Fixed an infelicity in the appendfile transport. When checking directories
5498 for a mailbox, to see if any needed to be created, it was accidentally
5499 using path names with one or more superfluous leading slashes; tracing
5500 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5502 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5503 contents are added to the log line that is written for every discarded
5504 recipient. (Previously a log_message setting was ignored.)
5506 36. The ${quote: operator now quotes the string if it is empty.
5508 37. The install script runs exim in order to find its version number. If for
5509 some reason other than non-existence or emptiness, which it checks, it
5510 could not run './exim', it was installing it with an empty version number,
5511 i.e. as "exim-". This error state is now caught, and the installation is
5514 38. An argument was missing from the function that creates an error message
5515 when Exim fails to connect to the socket for saslauthd authentication.
5516 This could cause Exim to crash, or give a corrupted message.
5518 39. Added isip, isip4, and isip6 to ${if conditions.
5520 40. The ACL variables $acl_xx are now saved with the message, and can be
5521 accessed later in routers, transports, and filters.
5523 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5524 strings in the file are not string-expanded.
5526 42. If a MAIL command specified a SIZE value that was too large to fit into an
5527 int variable, the check against message_size_limit failed. Such values are
5528 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5529 day this will have to be increased, but I don't think I want to be around
5530 when emails are that large.
5537 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5538 iconv() is not standard in FreeBSD.
5540 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5541 IPv6 enabled. The observed symptom was a segmentation fault on return from
5542 the function os_common_find_running_interfaces() in src/os.c.
5544 3. In the check_special_case() function in daemon.c I had used "errno" as an
5545 argument name, which causes warnings on some systems. This was basically a
5546 typo, since it was named "eno" in the comments!
5548 4. The code that waits for the clock to tick (at a resolution of some fraction
5549 of a second) so as to ensure message-id uniqueness was always waiting for
5550 at least one whole tick, when it could have waited for less. [This is
5551 almost certainly not relevant at current processor speeds, where it is
5552 unlikely to ever wait at all. But we try to future-proof.]
5554 5. The function that sleeps for a time interval that includes fractions of a
5555 second contained a race. It did not block SIGALRM between setting the
5556 timer, and suspending (a couple of lines later). If the interval was short
5557 and the sigsuspend() was delayed until after it had expired, the suspension
5558 never ended. On busy systems this could lead to processes getting stuck for
5561 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5562 process, before it forks any delivery processes. The open lookup caching
5563 mechanism meant that the open file or database connection was passed into
5564 the delivery process. The problem was that delivery processes always tidy
5565 up cached lookup data. This could cause a problem for the next delivery
5566 process started by the queue runner, because the external queue runner
5567 process does not know about the closure. So the next delivery process
5568 still has data in the lookup cache. In the case of a file lookup, there was
5569 no problem because closing a file descriptor in a subprocess doesn't affect
5570 the parent. However, if the lookup was caching a connection to a database,
5571 the connection was closed, and the second delivery process was likely to
5572 see errors such as "PGSQL: query failed: server closed the connection
5573 unexpectedly". The problem has been fixed by closing all cached lookups
5574 in a queue runner before running a delivery process.
5576 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5577 seem to have the "const" qualifier which it has on other OS. I've
5580 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5581 *of the same type* (client or server) with the same public name that an
5582 error should be diagnosed.
5584 9. When Exim looked up a host name for an IP address, but failed to find the
5585 original IP address when looking up the host name (a safety check), it
5586 output the message "<ip address> does not match any IP for NULL", which was
5587 confusing, to say the least. The bug was that the host name should have
5588 appeared instead of "NULL".
5590 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5591 user that is built into the binary, and the -C or -D options is used, root
5592 privilege is dropped before the configuration file is read. In addition,
5593 logging is switched to stderr instead of the normal log files. If the
5594 configuration then re-defines the Exim user, the unprivileged environment
5595 is probably not what is expected, so Exim logs a panic warning message (but
5598 However, if deliver_drop_privilege is set, the unprivileged state may well
5599 be exactly what is intended, so the warning has been cut out in that case,
5600 and Exim is allowed to try to write to its normal log files.
5606 1. smtp_return_error_details was not giving details for temporary sender
5607 or receiver verification errors.
5609 2. Diagnose a configuration error if two authenticators have the same public
5612 3. Exim used not to create the message log file for a message until the first
5613 delivery attempt. This could be confusing when incoming messages were held
5614 for policy or load reasons. The message log file is now created at the time
5615 the message is received, and an initial "Received" line is written to it.
5617 4. The automatically generated man page for command line options had a minor
5618 bug that caused no ill effects; however, a more serious problem was that
5619 the procedure for building the man page automatically didn't always
5620 operate. Consequently, release 4.20 contains an out-of-date version. This
5621 shouldn't happen again.
5623 5. When building Exim with embedded Perl support, the script that builds the
5624 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5625 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5627 6. The freeze_tell option was not being used for messages that were frozen on
5628 arrival, either by an ACL or by local_scan().
5630 7. Added the smtp_incomplete_transaction log selector.
5632 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5633 was accepting AUTH without a new EHLO.
5635 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5636 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5639 10. Logging of TCP/IP connections (when configured) now happens in the main
5640 daemon process instead of the child process, so that the TCP/IP connection
5641 count is more accurate (but it can never be perfect).
5643 11. The use of "drop" in a nested ACL was not being handled correctly in the
5644 outer ACL. Now, if condition failure induced by the nested "drop" causes
5645 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5646 or "require"), the connection is dropped.
5648 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5649 that yield "discard" can now be used with an "accept" or a "discard" verb,
5650 but an error is generated for any others (because I can't see a useful way
5651 to define what should happen).
5653 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5654 are now processed in the same way as lines in the Exim configuration file.
5655 In particular, continuation lines are supported.
5657 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5659 15. Added -ti meaning -t -i.
5661 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5662 domains, and warn by logging if others are found.
5664 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5665 returned from GIFCONF when Exim is trying to find the list of interfaces on
5666 a host. The code in os.c has been modified to copy each ifreq to an aligned
5667 structure in all cases.
5669 Also, in some cases, the returned ifreq's were being copied to a 'struct
5670 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5671 means the last couple of bytes of an IPv6 address could be chopped if the
5672 ifreq contained only a normal sockaddr (14 bytes storage).
5674 18. Named domain lists were not supported in the hosts_treat_as_local option.
5675 An entry such as +xxxx was not recognized, and was treated as a literal
5678 19. Ensure that header lines added by a DATA ACL are included in the reject log
5679 if the ACL subsequently rejects the message.
5681 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5682 MD5 (which is deprecated).
5684 21. When testing a filter file using -bf, Exim was writing a message when it
5685 took the sender from a "From " line in the message, but it was not doing so
5686 when it took $return_path from a Return-Path: header line. It now does.
5688 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5689 with a valid header line field name (a series of printing characters
5690 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5692 23. Changed "disc" in the source to "disk" to conform to the documentation and
5693 the book and for uniformity.
5695 24. Ignore Sendmail's -Ooption=value command line item.
5697 25. When execve() failed while trying to run a command in a pipe transport,
5698 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5699 could be confused with a return value of 69 from the command itself. This
5700 has been changed to 127, the value the shell returns if it is asked to run
5701 a non-existent command. The wording for the related log line suggests a
5702 non-existent command as the problem.
5704 26. If received_header_text expands to an empty string, do not add a Received:
5705 header line to the message. (Well, it adds a token one on the spool, but
5706 marks it "old" so that it doesn't get used or transmitted.)
5708 27. Installed eximstats 1.28 (addition of -nt option).
5710 28. There was no check for failure on the call to getsockname() in the daemon
5711 code. This can fail if there is a shortage of resources on the system, with
5712 ENOMEM, for example. A temporary error is now given on failure.
5714 29. Contrary to the C standard, it seems that in some environments, the
5715 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5716 program. Exim now does this explicitly; it affects the formatting of
5717 timestamps using strftime().
5719 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5720 complaints. I've now initialized all the variables, to avoid this.
5722 32. Header lines added by a system filter were not being "seen" during
5723 transport-time rewrites.
5725 33. The info_callback() function passed to OpenSSL is set up with type void
5726 (*)(SSL *, int, int), as described somewhere. However, when calling the
5727 function (actually a macro) that sets it up, the type void(*)() is
5728 expected. I've put in a cast to prevent warnings from picky compilers.
5730 34. If a DNS black list lookup found a CNAME record, but there were no A
5731 records associated with the domain it pointed at, Exim crashed.
5733 35. If a DNS black list lookup returned more than one A record, Exim ignored
5734 all but the first. It now scans all returned addresses if a particular IP
5735 value is being sought. In this situation, the contents of the
5736 $dnslist_value variable are a list of all the addresses, separated by a
5739 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5740 to accept a host name and all its aliases if the forward lookup for any of
5741 them yielded the IP address of the incoming connection. Now it accepts only
5742 those names whose forward lookup yields the correct IP address. Any other
5743 names are discarded. This closes a loophole whereby a rogue DNS
5744 administrator could create reverse DNS records to break through a
5745 wildcarded host restriction in an ACL.
5747 37. If a user filter or a system filter that ran in a subprocess used any of
5748 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5749 the wrong values were passed to the pipe command ($thisaddress had the
5750 value of $0, $0 had the value of $1, etc). This bug was introduced by
5751 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5753 38. Improved the line breaking for long SMTP error messages from ACLs.
5754 Previously, if there was no break point between 40 and 75 characters, Exim
5755 left the rest of the message alone. Two changes have been made: (a) I've
5756 reduced the minimum length to 35 characters; (b) if it can't find a break
5757 point between 35 and 75 characters, it looks ahead and uses the first one
5758 that it finds. This may give the occasional overlong line, but at least the
5759 remaining text gets split now.
5761 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5762 file descriptors might be low, and that setting 1000 would always raise it.
5763 It turns out that in some environments, the limit is already over 1000 and
5764 that lowering it causes trouble. So now Exim takes care not to decrease it.
5766 40. When delivering a message, the value of $return_path is set to $sender_
5767 address at the start of routing (routers may change the value). By an
5768 oversight, this default was not being set up when an address was tested by
5769 -bt or -bv, which affected the outcome if any router or filter referred to
5772 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5773 log only when "message" or "log_message" are set. However, if one of the
5774 conditions was an address verification, or a call to a nested ACL, the
5775 messages generated by the underlying test were being passed through. This
5776 no longer happens. The underlying message is available in $acl_verify_
5777 message for both "message" and "log_message" expansions, so it can be
5778 passed through if needed.
5780 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5781 new expansion $bh_ to give the encoded byte string without charset
5782 translation. Translation happens only if iconv() is available; HAVE_ICONV
5783 indicates this at build time. HEADERS_CHARSET gives the charset to
5784 translate to; headers_charset can change it in the configuration, and
5785 "headers charset" can change it in an individual filter file.
5787 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5788 that creates RFC 2047 encoded "words" labels them as that charset instead
5789 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5790 expansion operator; (ii) when Exim creates a From: line for a local
5791 message; (iii) when a header line is rewritten to include a "phrase" part.
5793 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5794 buggy, causing it to skip the first lines of messages whose message ID
5795 ended in 'D'. This would not have bitten before Exim release 4.14, because
5796 message IDs were unlikely to end in 'D' before then. The effect was to have
5797 incorrect size information for certain domains.
5799 45. #include "config.h" was missing at the start of the crypt16.c module. This
5800 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5803 46. If there was a timeout during a "random" callout check, Exim treated it as
5804 a failure of the random address, and carried on sending RSET and the real
5805 address. If the delay was just some slowness somewhere, the response to the
5806 original RCPT would be taken as a response to RSET and so on, causing
5807 mayhem of various kinds.
5809 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5810 when I implemented it. It didn't allow for the fact that some option values
5811 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5812 the right test for positive values.
5814 48. Domain names in DNS records are case-independent. Exim always looks them up
5815 in lower case. Some resolvers return domain names in exactly the case they
5816 appear in the zone file, that is, they may contain uppercase letters. Not
5817 all resolvers do this - some return always lower case. Exim was treating a
5818 change of case by a resolver as a change of domain, similar to a widening
5819 of a domain abbreviation. This triggered its re-routing code and so it was
5820 trying to route what was effectively the same domain again. This normally
5821 caused routing to fail (because the router wouldn't handle the domain
5822 twice). Now Exim checks for this case specially, and just changes the
5823 casing of the domain that it ultimately uses when it transmits the message
5826 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5829 50. If a filter generated a file delivery with a non-absolute name (possible if
5830 no home directory exists for the router), the forbid_file option was not
5833 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5834 the existing equality matching.
5836 52. Exim was using ints instead of ino_t variables in some places where it was
5837 dealing with inode numbers.
5839 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5840 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5841 called TMPDIR, and if it finds it is different, it changes its value.
5843 54. The smtp_printf() function is now made available to local_scan() so
5844 additional output lines can be written before returning. There is also an
5845 smtp_fflush() function to enable the detection of a dropped connection.
5846 The variables smtp_input and smtp_batched_input are exported to
5849 55. Changed the default runtime configuration: the message "Unknown user"
5850 has been removed from the ACL, and instead placed on the localuser router,
5851 using the cannot_route_message feature. This means that any verification
5852 failures that generate their own messages won't get overridden. Similarly,
5853 the "Unrouteable address" message that was in the ACL for unverifiable
5854 relay addresses has also been removed.
5856 56. Added hosts_avoid_esmtp to the smtp transport.
5858 57. The exicyclog script was not checking for the esoteric option
5859 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5860 will work only if exicyclog is run under the appropriate euid.
5862 58. Following a discussion on the list, the rules by which Exim recognises line
5863 endings on incoming messages have been changed. The -dropcr and drop_cr
5864 options are now no-ops, retained only for backwards compatibility. The
5865 following line terminators are recognized: LF CRLF CR. However, special
5866 processing applies to CR:
5868 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5869 nor a local message in the state where . is a terminator.
5871 (ii) If a bare CR is encountered in a header line, an extra space is added
5872 after the line terminator so as not to end the header. The reasoning
5873 behind this is that bare CRs in header lines are most likely either
5874 to be mistakes, or people trying to play silly games.
5876 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5877 was being incorrectly given as 18 bytes larger than it should have been.
5878 This is a VOB (very old bug).
5880 60. This may never have affected anything current, but just in case it has:
5881 When the local host is found other than at the start of a list of hosts,
5882 the local host, those with the same MX, and any that follow, are discarded.
5883 When the list in question was part of a longer list of hosts, the following
5884 hosts (not currently being processed) were also being discarded. This no
5885 longer happens. I'm not sure if this situation could ever has previously
5888 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5891 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5892 header. This is something that is recommended in a new Internet Draft, and
5893 is something that is documented as being done by Sendmail. There are two
5894 possible values. For messages generated by the autoreply transport, Exim
5897 Auto-Submitted: auto-replied
5899 whereas for all other generated messages (e.g. bounces) it adds
5901 Auto-Submitted: auto-generated
5903 63. The "personal" condition in filters now includes a test for the
5904 Auto-Submitted: header. If it contains the string "auto-" the message it
5905 not considered personal.
5907 64. Added rcpt_include_affixes as a generic transport option.
5909 65. Added queue_only_override (default true).
5911 66. Added the syslog_duplication option.
5913 67. If what should have been the first header line of a message consisted of
5914 a space followed by a colon, Exim was mis-interpreting it as a header line.
5915 It isn't of course - it is syntactically invalid and should therefore be
5916 treated as the start of the message body. The misbehaviour could have
5917 caused a number of strange effects, including loss of data in subsequent
5918 header lines, and spool format errors.
5920 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5921 client host had authenticated. This control can now be exercised by an ACL
5922 for more flexibility.
5924 69. By default, callouts do not happen when testing with -bh. There is now a
5925 variant, -bhc, which does actually run the callout code, including
5926 consulting and updating the callout cache.
5928 70. Added support for saslauthd authentication, courtesy of Alexander
5931 71. If statvfs() failed on the spool or log directories while checking their
5932 size for availability, Exim confusingly gave the error "space shortage".
5933 Furthermore, in debugging mode it crashed with a floating point exception.
5934 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5935 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5936 serious problem, Exim now writes to the main and panic logs when this
5937 happens, with details of the failure. It then refuses to accept the
5938 incoming message, giving the message "spool directory problem" or "log
5939 directory problem" with a 421 code for SMTP messages.
5941 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5942 0, 1, and 2 exist, because some OS complain for execs without them (see
5943 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5944 descriptors. However, the code omitted to check that the open succeeded,
5945 causing mysterious errors if for some reason the permissions on /dev/null
5946 got screwed. Now Exim writes a message to the main and panic logs, and
5947 bombs out if it can't open /dev/null.
5949 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5950 interact so that it is all more flexible. It is supposed to remain
5951 backwards compatible. Also added extra_local_interfaces.
5953 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5954 to bomb out with an assertion failure - to the client this appears as a
5955 connection drop. This problem occurs in the part of the code that was taken
5956 from the Samba project. Fortunately, the assertion is in a very simple
5957 function, so I have fixed this by reproducing the function inline in the
5958 one place where it is called, and arranging for authentication to fail
5959 instead of killing the process with assert().
5961 75. The SPA client code was not working when the server requested OEM rather
5962 than Unicode encoding.
5964 76. Added code to make require_files with a specific uid setting more usable in
5965 the case where statting the file as root fails - usually a non-root-mounted
5966 NFS file system. When this happens and the failure is EACCES, Exim now
5967 forks a subprocess and does the per-uid checking as the relevant uid.
5969 77. Added process_log_path.
5971 78. If log_file_path was not explicitly set, a setting of check_log_space or
5972 check_log_inodes was ignored.
5974 79. If a space check for the spool or log partitions fails, the incident is now
5975 logged. Of course, in the latter case the data may get lost...
5977 80. Added the %p formatting code to string_format() so that it can be used to
5978 print addresses in debug_print(). Adjusted all the address printing in the
5979 debugging in store.c to use %p rather than %d.
5981 81. There was a concern that a line of code in smtp_in.c could overflow a
5982 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5983 initially expressed, the concern was not well-founded, because trailing
5984 spaces are removed early. However, if the trailing spaces were followed by
5985 a NULL, they did not get removed, so the overflow was possible. Two fixes
5988 (a) I re-wrote the offending code in a cleaner fashion.
5989 (b) If an incoming SMTP command contains a NULL character, it is rejected
5992 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5993 runs initgroups(), so that if the Exim user is in any additional groups,
5994 they will be used during message reception.
6000 The change log for 4.20 and earlier releases has been archived.