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.
62 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
63 routers/transports under cutthrough routing.
65 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
66 numbers. Touch up "bool" conditional to keep the same definition.
72 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
74 PP/02 Make -n do something, by making it not do something.
75 When combined with -bP, the name of an option is not output.
77 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
80 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
81 $sender_host_name and config options to manage this, and basic check
84 PP/05 DSCP support for outbound connections and control modifier for inbound.
86 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
87 (Only plugin which currently uses this is kerberos4, which nobody should
88 be using, but we should make it available and other future plugins might
89 conceivably use it, even though it would break NAT; stuff *should* be
90 using channel bindings instead).
92 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
93 name; added for Sendmail compatibility; requires admin caller.
94 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
95 just ignore it); requires trusted caller.
96 Also parse but ignore: -Ac -Am -X<logfile>
99 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
101 TL/02 Add +smtp_confirmation as a default logging option.
103 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
104 Patch by Magnus Holmgren from 2007-02-20.
106 TL/04 Bugzilla 1281 - Spec typo.
107 Bugzilla 1283 - Spec typo.
108 Bugzilla 1290 - Spec grammar fixes.
110 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
112 TL/06 Add Experimental DMARC support using libopendmarc libraries.
114 TL/07 Fix an out of order global option causing a segfault. Reported to dev
115 mailing list by by Dmitry Isaikin.
117 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
119 JH/02 Support "G" suffix to numbers in ${if comparisons.
121 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
123 NM/01 Bugzilla 1197 - Spec typo
124 Bugzilla 1196 - Spec examples corrections
126 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
128 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
129 gnutls_enable_pkcs11, but renamed to more accurately indicate its
132 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
133 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
135 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
136 "acl {{name}{arg}...}", and optional args on acl condition
139 JH/05 Permit multiple router/transport headers_add/remove lines.
141 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
143 JH/07 Avoid using a waiting database for a single-message-only transport.
144 Performance patch from Paul Fisher. Bugzilla 1262.
146 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
149 JH/09 Add $headers_added variable, with content from use of ACL modifier
150 add_header (but not yet added to the message). Bugzilla 199.
152 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
153 Pulled from Bugzilla 817 by Wolfgang Breyha.
155 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
157 (nb: this is the same fix as in Exim 4.80.1)
159 JH/11 Add A= logging on delivery lines, and a client_set_id option on
162 JH/12 Add optional authenticated_sender logging to A= and a log_selector
165 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
167 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
168 advertise SMTP AUTH mechanism to us, instead of a generic
169 protocol violation error. Also, make Exim more robust to bad
170 data from the Dovecot auth socket.
172 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
174 When a queue runner is handling a message, Exim first routes the
175 recipient addresses, during which it prunes them based on the retry
176 hints database. After that it attempts to deliver the message to
177 any remaining recipients. It then updates the hints database using
180 So if a recipient address works intermittently, it can get repeatedly
181 deferred at routing time. The retry hints record remains fresh so the
182 address never reaches the final cutoff time.
184 This is a fairly common occurrence when a user is bumping up against
185 their storage quota. Exim had some logic in its local delivery code
186 to deal with this. However it did not apply to per-recipient defers
187 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
189 This change adds a proper retry rule check during routing so that the
190 final cutoff time is checked against the message's age. We only do
191 this check if there is an address retry record and there is not a
192 domain retry record; this implies that previous attempts to handle
193 the address had the retry_use_local_parts option turned on. We use
194 this as an approximation for the destination being like a local
195 delivery, as in LMTP.
197 I suspect this new check makes the old local delivery cutoff check
198 redundant, but I have not verified this so I left the code in place.
200 TF/02 Correct gecos expansion when From: is a prefix of the username.
202 Test 0254 submits a message to Exim with the header
206 When I ran the test suite under the user fanf2, Exim expanded
207 the header to contain my full name, whereas it should have added
208 a Resent-Sender: header. It erroneously treats any prefix of the
209 username as equal to the username.
211 This change corrects that bug.
213 GF/01 DCC debug and logging tidyup
214 Error conditions log to paniclog rather than rejectlog.
215 Debug lines prefixed by "DCC: " to remove any ambiguity.
217 TF/03 Avoid unnecessary rebuilds of lookup-related code.
219 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
220 Bug spotted by Jeremy Harris; was flawed since initial commit.
221 Would have resulted in OCSP responses post-SNI triggering an Exim
222 NULL dereference and crash.
224 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
226 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
227 Bug detection, analysis and fix by Samuel Thibault.
228 Bugzilla 1331, Debian bug #698092.
230 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
232 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
233 Server implementation by Todd Lyons, client by JH.
234 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
235 config variable "prdr_enable" controls whether the server
236 advertises the facility. If the client requests PRDR a new
237 acl_data_smtp_prdr ACL is called once for each recipient, after
238 the body content is received and before the acl_smtp_data ACL.
239 The client is controlled by bolth of: a hosts_try_prdr option
240 on the smtp transport, and the server advertisement.
241 Default client logging of deliveries and rejections involving
242 PRDR are flagged with the string "PRDR".
244 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
245 fclose(). Diagnosis by Todd Lyons.
247 PP/17 Update configure.default to handle IPv6 localhost better.
248 Patch by Alain Williams (plus minor tweaks).
251 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
252 This is now consistent with GnuTLS, and is now documented: the
253 previous undocumented portable approach to treating the option as
254 unset was to force an expansion failure. That still works, and
255 an empty string is now equivalent.
257 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
258 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
259 not performing validation itself.
261 PP/20 Added force_command boolean option to pipe transport.
262 Patch from Nick Koston, of cPanel Inc.
264 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
267 TF/04 Added udpsend ACL modifer and hexquote expansion operator
269 PP/21 Fix eximon continuous updating with timestamped log-files.
270 Broken in a format-string cleanup in 4.80, missed when I repaired the
271 other false fix of the same issue.
272 Report and fix from Heiko Schlichting.
275 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
276 Report from Prashanth Katuri.
278 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
279 It's SecureTransport, so affects any MacOS clients which use the
280 system-integrated TLS libraries, including email clients.
282 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
283 using a MIME ACL for non-SMTP local injection.
284 Report and assistance in diagnosis by Warren Baker.
286 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
288 JH/16 Fix comparisons for 64b. Bugzilla 1385.
290 TL/09 Add expansion variable $authenticated_fail_id to keep track of
291 last id that failed so it may be referenced in subsequent ACL's.
293 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
296 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
297 ldap library initialization, allowing self-signed CA's to be
298 used. Also properly sets require_cert option later in code by
299 using NULL (global ldap config) instead of ldap handle (per
300 session). Bug diagnosis and testing by alxgomz.
302 TL/12 Enhanced documentation in the ratelimit.pl script provided in
303 the src/util/ subdirectory.
305 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
306 renamed to Transport Post Delivery Action by Jeremy Harris, as
309 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
310 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
311 redis_servers = needs to be configured which will be used by the redis
312 lookup. Patch from Warren Baker, of The Packet Hub.
314 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
316 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
317 hostname or reverse DNS when processing a host list. Used suggestions
318 from multiple comments on this bug.
320 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
322 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
323 Missed a few lines, added it to make the runtest require no keyboard
326 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
327 contains upper case chars. Make router use caseful_local_part.
329 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
330 support when GnuTLS has been built with p11-kit.
336 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
338 This, or similar/improved, will also be change PP/11 of 4.82.
344 PP/01 Handle short writes when writing local log-files.
345 In practice, only affects FreeBSD (8 onwards).
346 Bugzilla 1053, with thanks to Dmitry Isaikin.
348 NM/01 Bugzilla 949 - Documentation tweak
350 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
353 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
355 PP/02 Implemented gsasl authenticator.
357 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
359 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
360 `pkg-config foo` for cflags/libs.
362 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
363 with rest of GSASL and with heimdal_gssapi.
365 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
366 `pkg-config foo` for cflags/libs for the TLS implementation.
368 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
369 properties get this fed in as external SSF. A number of robustness
370 and debugging improvements to the cyrus_sasl authenticator.
372 PP/08 cyrus_sasl server now expands the server_realm option.
374 PP/09 Bugzilla 1214 - Log authentication information in reject log.
375 Patch by Jeremy Harris.
377 PP/10 Added dbmjz lookup type.
379 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
381 PP/12 MAIL args handles TAB as well as SP, for better interop with
382 non-compliant senders.
383 Analysis and variant patch by Todd Lyons.
385 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
386 Bug report from Lars Müller <lars@samba.org> (via SUSE),
387 Patch from Dirk Mueller <dmueller@suse.com>
389 PP/13 tls_peerdn now print-escaped for spool files.
390 Observed some $tls_peerdn in wild which contained \n, which resulted
391 in spool file corruption.
393 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
394 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
395 or write after TLS renegotiation, which otherwise led to messages
398 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
399 as a tracking header (ie: a signed header comes before the signature).
400 Patch from Wolfgang Breyha.
402 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
403 comma-sep list; embedded commas doubled.
405 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
407 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
409 Report and patch from Dmitry Banschikov.
411 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
412 Removed SSL_clear() after SSL_new() which led to protocol negotiation
413 failures. We appear to now support TLS1.1+ with Exim.
415 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
416 lets Exim select keys and certificates based upon TLS SNI from client.
417 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
418 before an outbound SMTP session. New log_selector, +tls_sni.
420 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
421 NULL dereference. Report and patch from Alun Jones.
423 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
424 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
425 Not seeing resolver debug output on NetBSD, but suspect this is a
426 resolver implementation change.
428 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
429 Left warnings. Added "eximon gdb" invocation mode.
431 PP/21 Defaulting "accept_8bitmime" to true, not false.
433 PP/22 Added -bw for inetd wait mode support.
435 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
436 locate the relevant includes and libraries. Made this the default.
438 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
439 Bugzilla 1246, report and most of solution from Tomasz Kusy.
441 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
442 This may cause build issues on older platforms.
444 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
445 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
446 gnutls_require_mac & gnutls_require_protocols (no longer supported).
447 Added SNI support via GnuTLS too.
448 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
450 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
452 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
453 Applied second patch from Janne, implementing suggestion to default
454 multiple-strings-in-record handling to match SPF spec.
456 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
458 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
459 read-only, out of scope).
460 Patch from Wolfgang Breyha, report from Stuart Northfield.
462 PP/29 Fix three issues highlighted by clang analyser static analysis.
463 Only crash-plausible issue would require the Cambridge-specific
464 iplookup router and a misconfiguration.
465 Report from Marcin Mirosław.
467 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
469 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
470 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
471 As part of this, removing so much warning spew let me fix some minor
472 real issues in debug logging.
474 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
475 assignment on my part. Fixed.
477 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
478 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
479 Janne Snabb (who went above and beyond: thank you).
481 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
482 string otherwise requires a connection and a bunch more work and it's
483 relatively easy to get wrong. Should also expose TLS library linkage
486 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
487 64-bit ${eval} (JH/03).
489 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
490 GNU libc to support some of the 64-bit stuff, should not lead to
491 conflicts. Defined before os.h is pulled in, so if a given platform
492 needs to override this, it can.
494 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
495 protection layer was required, which is not implemented.
496 Bugzilla 1254, patch from Wolfgang Breyha.
498 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
499 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
500 tls_dhparam take prime identifiers. Also unbreak combination of
501 OpenSSL+DH_params+TLSSNI.
503 PP/39 Disable SSLv2 by default in OpenSSL support.
509 PP/01 Solaris build fix for Oracle's LDAP libraries.
510 Bugzilla 1109, patch from Stephen Usher.
512 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
514 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
517 TF/02 Fix a couple more cases where we did not log the error message
518 when unlink() failed. See also change 4.74-TF/03.
520 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
521 lock up or crash if it happened to be inside a call to libc when it
522 got a SIGUSR1 from exiwhat.
524 The SIGUSR1 handler appends the current process status to the process
525 log which is later printed by exiwhat. It used to use the general
526 purpose logging code to do this, but several functions it calls are
527 not safe for signals.
529 The new output code in the SIGUSR1 handler is specific to the process
530 log, and simple enough that it's easy to inspect for signal safety.
531 Removing some special cases also simplifies the general logging code.
532 Removing the spurious timestamps from the process log simplifies
535 TF/04 Improved ratelimit ACL condition.
537 The /noupdate option has been deprecated in favour of /readonly which
538 has clearer semantics. The /leaky, /strict, and /readonly update modes
539 are mutually exclusive. The update mode is no longer included in the
540 database key; it just determines when the database is updated. (This
541 means that when you upgrde Exim will forget old rate measurements.)
543 Exim now checks that the per_* options are used with an update mode that
544 makes sense for the current ACL. For example, when Exim is processing a
545 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
546 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
547 must specify per_mail/readonly. If you omit the update mode it defaults to
548 /leaky where that makes sense (as before) or /readonly where required.
550 The /noupdate option is now undocumented but still supported for
551 backwards compatibility. It is equivalent to /readonly except that in
552 ACLs where /readonly is required you may specify /leaky/noupdate or
553 /strict/noupdate which are treated the same as /readonly.
555 A useful new feature is the /count= option. This is a generalization
556 of the per_byte option, so that you can measure the throughput of other
557 aggregate values. For example, the per_byte option is now equivalent
558 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
560 The per_rcpt option has been generalized using the /count= mechanism
561 (though it's more complicated than the per_byte equivalence). When it is
562 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
563 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
564 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
565 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
566 Note that using per_rcpt with a non-readonly update mode in more than
567 one ACL will cause the recipients to be double-counted. (The per_mail
568 and per_byte options don't have this problem.)
570 The handling of very low rates has changed slightly. If the computed rate
571 is less than the event's count (usually one) then this event is the first
572 after a long gap. In this case the rate is set to the same as this event's
573 count, so that the first message of a spam run is counted properly.
575 The major new feature is a mechanism for counting the rate of unique
576 events. The new per_addr option counts the number of different
577 recipients that someone has sent messages to in the last time period. It
578 behaves like per_rcpt if all the recipient addresses are different, but
579 duplicate recipient addresses do not increase the measured rate. Like
580 the /count= option this is a general mechanism, so the per_addr option
581 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
582 example, measure the rate that a client uses different sender addresses
583 with the options per_mail/unique=$sender_address. There are further
584 details in the main documentation.
586 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
588 TF/06 Removed a few PCRE remnants.
590 TF/07 Automatically extract Exim's version number from tags in the git
591 repository when doing development or release builds.
593 PP/02 Raise smtp_cmd_buffer_size to 16kB.
594 Bugzilla 879. Patch from Paul Fisher.
596 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
597 Heavily based on revision 40f9a89a from Simon Arlott's tree.
600 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
602 PP/05 Variable $av_failed, true if the AV scanner deferred.
603 Bugzilla 1078. Patch from John Horne.
605 PP/06 Stop make process more reliably on build failure.
606 Bugzilla 1087. Patch from Heiko Schlittermann.
608 PP/07 Make maildir_use_size_file an _expandable_ boolean.
609 Bugzilla 1089. Patch from Heiko Schlittermann.
611 PP/08 Handle ${run} returning more data than OS pipe buffer size.
612 Bugzilla 1131. Patch from Holger Weiß.
614 PP/09 Handle IPv6 addresses with SPF.
615 Bugzilla 860. Patch from Wolfgang Breyha.
617 PP/10 GnuTLS: support TLS 1.2 & 1.1.
619 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
622 PP/11 match_* no longer expand right-hand-side by default.
623 New compile-time build option, EXPAND_LISTMATCH_RHS.
624 New expansion conditions, "inlist", "inlisti".
626 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
628 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
630 PP/14 fix log_write() format string regression from TF/03.
631 Bugzilla 1152. Patch from Dmitry Isaikin.
637 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
639 PP/02 Harmonised TLS library version reporting; only show if debugging.
640 Layout now matches that introduced for other libraries in 4.74 PP/03.
642 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
644 PP/04 New "dns_use_edns0" global option.
646 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
649 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
650 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
652 TK/01 Updated PolarSSL code to 0.14.2.
653 Bugzilla 1097. Patch from Andreas Metzler.
655 PP/07 Catch divide-by-zero in ${eval:...}.
658 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
661 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
662 format-string attack -- SECURITY: remote arbitrary code execution.
664 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
665 time unintentionally subject to list matching rules, letting the header
666 cause arbitrary Exim lookups (of items which can occur in lists, *not*
667 arbitrary string expansion). This allowed for information disclosure.
669 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
670 INT_MIN/-1 -- value coerced to INT_MAX.
676 NM/01 Workround for PCRE version dependency in version reporting
679 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
680 This fixes portability to compilers other than gcc, notably
681 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
683 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
684 makefiles for portability to HP-UX and POSIX correctness.
686 PP/01 Permit LOOKUP_foo enabling on the make command-line.
687 Also via indented variable definition in the Makefile.
688 (Debugging by Oliver Heesakkers).
690 PP/02 Restore caching of spamd results with expanded spamd_address.
691 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
693 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
694 Improves build reliability. Fix from: Frank Elsner
696 NM/02 Fix wide character breakage in the rfc2047 coding
697 Fixes bug 1064. Patch from Andrey N. Oktyabrski
699 NM/03 Allow underscore in dnslist lookups
700 Fixes bug 1026. Patch from Graeme Fowler
702 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
703 Code patches from Adam Ciarcinski of NetBSD.
705 NM/04 Fixed exiqgrep to cope with mailq missing size issue
708 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
709 is logged, to avoid truncation. Patch from John Horne.
711 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
712 Patch from Jakob Hirsch.
714 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
715 SQL string expansion failure details.
716 Patch from Andrey Oktyabrski.
718 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
719 Patch from Simon Arlott.
721 PP/09 New lookups functionality failed to compile on old gcc which rejects
722 extern declarations in function scope.
723 Patch from Oliver Fleischmann
725 PP/10 Use sig_atomic_t for flags set from signal handlers.
726 Check getgroups() return and improve debugging.
727 Fixed developed for diagnosis in bug 927 (which turned out to be
730 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
731 Patch from Mark Zealey.
733 PP/12 Bugzilla 1056: Improved spamd server selection.
734 Patch from Mark Zealey.
736 PP/13 Bugzilla 1086: Deal with maildir quota file races.
737 Based on patch from Heiko Schlittermann.
739 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
740 Patch from Uwe Doering, sign-off by Michael Haardt.
742 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
743 variable declaration deep within a block. Bug and patch from
746 PP/15 lookups-Makefile IRIX compatibilty coercion.
748 PP/16 Make DISABLE_DKIM build knob functional.
750 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
751 Patch by Simon Arlott
753 TF/03 Fix valgrind.h portability to C89 compilers that do not support
754 variable argument macros. Our copy now differs from upstream.
760 TF/01 Failure to get a lock on a hints database can have serious
761 consequences so log it to the panic log.
763 TF/02 Log LMTP confirmation messages in the same way as SMTP,
764 controlled using the smtp_confirmation log selector.
766 TF/03 Include the error message when we fail to unlink a spool file.
768 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
769 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
770 for maintaining out-of-tree patches for some time.
772 PP/01 Bugzilla 139: Documentation and portability issues.
773 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
774 Handle per-OS dynamic-module compilation flags.
776 PP/02 Let /dev/null have normal permissions.
777 The 4.73 fixes were a little too stringent and complained about the
778 permissions on /dev/null. Exempt it from some checks.
779 Reported by Andreas M. Kirchwitz.
781 PP/03 Report version information for many libraries, including
782 Exim version information for dynamically loaded libraries. Created
783 version.h, now support a version extension string for distributors
784 who patch heavily. Dynamic module ABI change.
786 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
787 privilege escalation vulnerability whereby the Exim run-time user
788 can cause root to append content of the attacker's choosing to
791 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
794 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
795 If dropping privileges for untrusted macros, we disabled normal logging
796 on the basis that it would fail; for the Exim run-time user, this is not
797 the case, and it resulted in successful deliveries going unlogged.
798 Fixed. Reported by Andreas Metzler.
804 PP/01 Date: & Message-Id: revert to normally being appended to a message,
805 only prepend for the Resent-* case. Fixes regression introduced in
806 Exim 4.70 by NM/22 for Bugzilla 607.
808 PP/02 Include check_rfc2047_length in configure.default because we're seeing
809 increasing numbers of administrators be bitten by this.
811 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
813 PP/03 Bugzilla 994: added openssl_options main configuration option.
815 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
817 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
819 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
821 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
822 without a peer certificate, leading to a segfault because of an
823 assumption that peers always have certificates. Be a little more
824 paranoid. Problem reported by Martin Tscholak.
826 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
827 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
828 NB: ClamAV planning to remove STREAM in "middle of 2010".
829 CL also introduces -bmalware, various -d+acl logging additions and
830 more caution in buffer sizes.
832 PP/09 Implemented reverse_ip expansion operator.
834 PP/10 Bugzilla 937: provide a "debug" ACL control.
836 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
838 PP/12 Bugzilla 973: Implement --version.
840 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
842 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
844 PP/15 Bugzilla 816: support multiple condition rules on Routers.
846 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
847 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
848 ignore trailing whitespace.
850 JJ/02 prevent non-panic DKIM error from being sent to paniclog
852 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
855 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
856 Notification from Dr Andrew Aitchison.
858 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
859 ExtendedDetectionInfo response format.
860 Notification from John Horne.
862 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
865 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
866 XSL and documented dependency on system catalogs, with examples of how
869 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
872 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
873 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
874 configuration file which is writeable by the Exim user or group.
876 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
877 of configuration files to cover files specified with the -C option if
878 they are going to be used with root privileges, not just the default
881 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
882 option (effectively making it always true).
884 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
885 files to be used while preserving root privileges.
887 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
888 that rogue child processes cannot use them.
890 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
891 run-time user, instead of root.
893 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
894 Exim run-time user without dropping privileges.
896 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
897 result string, instead of calling string_vformat() twice with the same
900 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
901 for other users. Others should always drop root privileges if they use
902 -C on the command line, even for a whitelisted configure file.
904 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
906 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
912 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
913 $data_path, and $header_path variables; fixed documentation bugs and
916 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
917 exipick to access non-standard spools, including the "frozen" queue
920 NM/01 Bugzilla 965: Support mysql stored procedures.
921 Patch from Alain Williams
923 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
925 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
926 Patch from Andreas Metzler
928 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
929 Patch from Kirill Miazine
931 NM/05 Bugzilla 671: Added umask to procmail example.
933 JJ/03 installed exipick 20100323.0, fixing doc bug
935 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
936 directory. Notification and patch from Dan Rosenberg.
938 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
940 TK/02 Improve log output when DKIM signing operation fails.
942 MH/01 Treat the transport option dkim_domain as a colon separated
943 list, not as a single string, and sign the message with each element,
944 omitting multiple occurences of the same signer.
946 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
947 Bugzilla 985, 986. Patch by Simon Arlott
949 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
950 Patch by Simon Arlott
952 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
953 MBX locking. Notification from Dan Rosenberg.
959 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
961 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
963 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
965 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
967 NM/04 Bugzilla 915: Flag broken perl installation during build.
973 TK/01 Added patch by Johannes Berg that expands the main option
974 "spamd_address" if it starts with a dollar sign.
976 TK/02 Write list of recipients to X-Envelope-Sender header when building
977 the mbox-format spool file for content scanning (suggested by Jakob
980 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
981 (http://www.dcc-servers.net/) support via dccifd. Activated by
982 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
984 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
985 by Mark Daniel Reidel <mr@df.eu>.
987 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
988 When building exim an external PCRE library is now needed -
989 PCRE is a system library on the majority of modern systems.
990 See entry on PCRE_LIBS in EDITME file.
992 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
993 conversation. Added nologin parameter to request.
994 Patch contributed by Kirill Miazine.
996 TF/01 Do not log submission mode rewrites if they do not change the address.
998 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1000 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1001 log files in place. Contributed by Roberto Lima.
1003 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1005 TF/03 Bugzilla 615: When checking the local_parts router precondition
1006 after a local_part_suffix or local_part_prefix option, Exim now
1007 does not use the address's named list lookup cache, since this
1008 contains cached lookups for the whole local part.
1010 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1011 Robert Millan. Documentation is in experimental-spec.txt.
1013 TF/04 Bugzilla 668: Fix parallel build (make -j).
1015 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1017 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1018 Patch provided by Jan Srzednicki.
1020 TF/05 Leading white space used to be stripped from $spam_report which
1021 wrecked the formatting. Now it is preserved.
1023 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1024 that they are available at delivery time.
1026 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1028 TF/08 TLS error reporting now respects the incoming_interface and
1029 incoming_port log selectors.
1031 TF/09 Produce a more useful error message if an SMTP transport's hosts
1032 setting expands to an empty string.
1034 NM/06 Bugzilla 744: EXPN did not work under TLS.
1035 Patch provided by Phil Pennock.
1037 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1038 Patch provided by Richard Godbee.
1040 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1041 acl_smtp_notquit, added index entry.
1043 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1044 Patch provided by Eugene Bujak.
1046 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1047 accept(). Patch provided by Maxim Dounin.
1049 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1050 Patch provided by Phil Pennock.
1052 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1054 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1055 Patch provided by Brad "anomie" Jorsch.
1057 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1058 Patch provided by Dean Brooks.
1060 TK/05 Add native DKIM support (does not depend on external libraries).
1062 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1063 Patch provided by Graeme Fowler.
1065 NM/16 Bugzilla 851: Documentation example syntax fix.
1067 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1069 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1072 NM/19 Bugzilla 745: TLS version reporting.
1073 Patch provided by Phil Pennock.
1075 NM/20 Bugzilla 167: bool: condition support.
1076 Patch provided by Phil Pennock.
1078 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1079 clients. Patch provided by Phil Pennock.
1081 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1082 Patch provided by Brad "anomie" Jorsch.
1084 NM/23 Bugzilla 687: Fix misparses in eximstats.
1085 Patch provided by Heiko Schlittermann.
1087 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1088 Patch provided by Heiko Schlittermann.
1090 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1091 plus update to original patch.
1093 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1095 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1096 Patch provided by David Brownlee.
1098 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1100 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1102 NM/30 Bugzilla 888: TLS documentation bugfixes.
1104 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1106 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1107 Unlike the original bugzilla I have changed all shell scripts in src tree.
1109 NM/33 Bugzilla 898: Transport filter timeout fix.
1110 Patch by Todd Rinaldo.
1112 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1113 Patch by Serge Demonchaux.
1115 NM/35 Bugzilla 39: Base64 decode bug fixes.
1116 Patch by Jakob Hirsch.
1118 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1120 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1122 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1124 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1130 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1131 ALT-N's libdkim that I have put here:
1132 http://duncanthrax.net/exim-experimental/
1134 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1135 were called 'true' and it seems that C99 defines that as a reserved
1136 keyword to be used with 'bool' variable types. That means you could
1137 not include C99-style headers which use bools without triggering
1138 build errors in sieve.c.
1140 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1141 as mailq or other aliases. Changed the --help handling significantly
1142 to do whats expected. exim_usage() emits usage/help information.
1144 SC/01 Added the -bylocaldomain option to eximstats.
1146 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1148 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1150 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1156 PH/01 Another patch from the Sieve maintainer.
1158 PH/02 When an IPv6 address is converted to a string for single-key lookup
1159 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1160 dots are used instead of colons so that keys in lsearch files need not
1161 contain colons. This was done some time before quoting was made available
1162 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1163 (notated using the quote facility) so as to distinguish them from IPv4
1164 keys. This meant that lookups for IP addresses in host lists did not work
1165 for iplsearch lookups.
1167 This has been fixed by arranging for IPv6 addresses to be expressed with
1168 colons if the lookup type is iplsearch. This is not incompatible, because
1169 previously such lookups could never work.
1171 The situation is now rather anomolous, since one *can* have colons in
1172 ordinary lsearch keys. However, making the change in all cases is
1173 incompatible and would probably break a number of configurations.
1175 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1178 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1179 conversion specification without a maximum field width, thereby enabling
1180 a rogue spamd server to cause a buffer overflow. While nobody in their
1181 right mind would setup Exim to query an untrusted spamd server, an
1182 attacker that gains access to a server running spamd could potentially
1183 exploit this vulnerability to run arbitrary code as the Exim user.
1185 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1186 $primary_hostname instead of what libspf2 thinks the hosts name is.
1188 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1189 a directory entry by the name of the lookup key. Previously, if a
1190 symlink pointed to a non-existing file or a file in a directory that
1191 Exim lacked permissions to read, a lookup for a key matching that
1192 symlink would fail. Now it is enough that a matching directory entry
1193 exists, symlink or not. (Bugzilla 503.)
1195 PH/03 The body_linecount and body_zerocount variables are now exported in the
1198 PH/04 Added the $dnslist_matched variable.
1200 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1201 This means they are set thereafter only if the connection becomes
1204 PH/06 Added the client_condition to authenticators so that some can be skipped
1205 by clients under certain conditions.
1207 PH/07 The error message for a badly-placed control=no_multiline_responses left
1208 "_responses" off the end of the name.
1210 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1212 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1213 (without spaces) instead of just copying the configuration text.
1215 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1217 PH/11 Added $max_received_linelength.
1219 PH/12 Added +ignore_defer and +include_defer to host lists.
1221 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1222 way in which PCRE > 7.0 is built.
1224 PH/14 Implemented queue_only_load_latch.
1226 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1227 MAIL command. The effect was to mangle the value on 64-bit systems.
1229 PH/16 Another patch from the Sieve maintainer.
1231 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1233 PH/18 If a system quota error occurred while trying to create the file for
1234 a maildir delivery, the message "Mailbox is full" was not appended to the
1235 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1236 only to a quota excession during the actual writing of the file.
1238 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1239 characters?) which causes problems in log lines. The DN values are now
1240 passed through string_printing() before being added to log lines.
1242 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1243 and InterBase are left for another time.)
1245 PH/21 Added message_body_newlines option.
1247 PH/22 Guard against possible overflow in moan_check_errorcopy().
1249 PH/23 POSIX allows open() to be a macro; guard against that.
1251 PH/24 If the recipient of an error message contained an @ in the local part
1252 (suitably quoted, of course), incorrect values were put in $domain and
1253 $local_part during the evaluation of errors_copy.
1259 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1260 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1263 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1264 issue a MAIL command.
1266 PH/02 In an ACL statement such as
1268 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1270 if a client was not listed at all, or was listed with a value other than
1271 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1272 the condition was not true (as it should be), so access was not denied.
1273 The bug was that the ! inversion was incorrectly passed on to the second
1274 item. This has been fixed.
1276 PH/03 Added additional dnslists conditions == and =& which are different from
1277 = and & when the dns lookup returns more than one IP address.
1279 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1280 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1282 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1283 FSYNC, which compiles an option called disable_fsync that allows for
1284 bypassing fsync(). The documentation is heavily laced with warnings.
1286 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1288 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1289 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1290 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1291 including adding "make clean"; (3) Added -fPIC when compiling the test
1292 dynamically loaded module, to get rid of a warning.
1294 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1295 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1296 = 0s. The bug is otherwise harmless.
1298 PH/07 There was a bug in the dovecot authenticator such that the value of
1299 $auth1 could be overwritten, and so not correctly preserved, after a
1300 successful authentication. This usually meant that the value preserved by
1301 the server_setid option was incorrect.
1303 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1305 PH/09 Installed PCRE release 7.0.
1307 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1308 run for batched SMTP input. It is now run at the start of every message
1309 in the batch. While fixing this I discovered that the process information
1310 (output by running exiwhat) was not always getting set for -bs and -bS
1311 input. This is fixed, and it now also says "batched" for BSMTP.
1313 PH/11 Added control=no_pipelining.
1315 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1316 patch, slightly modified), and move the expansion of helo_data till after
1317 the connection is made in the smtp transport (so it can use these
1320 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1322 PH/14 Added log_selector = +pid.
1324 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1326 PH/16 Add ${if forany and ${if forall.
1328 PH/17 Added dsn_from option to vary the From: line in DSNs.
1330 PH/18 Flush SMTP output before performing a callout, unless control =
1331 no_callout_flush is set.
1333 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1334 was true (the default) a successful delivery failed to delete the retry
1335 item, thus causing premature timeout of the address. The bug is now
1338 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1340 PH/21 Long custom messages for fakedefer and fakereject are now split up
1341 into multiline reponses in the same way that messages for "deny" and
1342 other ACL rejections are.
1344 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1345 with slight modification.
1347 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1348 draft, changing the syntax and factoring some duplicate code".
1350 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1351 for deliveries of the second and subsequent messages over the same SMTP
1354 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1355 ${reduce, with only minor "tidies".
1357 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1359 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1360 expansion side effects.
1362 PH/27 When a message times out after an over-quota error from an Exim-imposed
1363 quota, the bounce message says "mailbox is full". This message was not
1364 being given when it was a system quota that was exceeded. It now should
1367 MH/03 Made $recipients available in local_scan(). local_scan() already has
1368 better access to the recipient list through recipients_list[], but
1369 $recipients can be useful in postmaster-provided expansion strings.
1371 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1372 in the case of a MAIL command with additional options following the
1373 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1374 were accidentally chopped off.
1376 PH/29 SMTP synchronization checks are implemented when a command is read -
1377 there is a check that no more input is waiting when there shouldn't be
1378 any. However, for some commands, a delay in an ACL can mean that it is
1379 some time before the response is written. In this time, more input might
1380 arrive, invalidly. So now there are extra checks after an ACL has run for
1381 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1382 pipelining has not been advertised.
1384 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1386 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1387 returned in the SMTP response when smtp_return_error_details was false.
1388 This has been fixed.
1390 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1391 instead of the C I/O that was originally supplied, because problems were
1392 reported on Solaris.
1394 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1395 Exim which did not show up earlier: it was assuming that a call to
1396 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1397 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1398 was a macro that became an assignment, so it seemed to work. This has
1399 changed to a proper function call with a void return, hence the compile
1400 error. Exim's code has been fixed.
1402 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1405 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1407 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1409 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1410 because it turns out to be too big or there is a timeout) there is no
1411 "Completed" line in the log. When some messages of this type were
1412 selected by exigrep, they were listed as "not completed". Others were
1413 picked up by some special patterns. I have improved the selection
1414 criteria to be more general.
1416 PH/38 The host_find_failed option in the manualroute router can now be set
1417 to "ignore", to completely ignore a host whose IP address cannot be
1418 found. If all hosts are ignored, the behaviour is controlled by the new
1419 host_all_ignored option.
1421 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1422 homing or because of multiple MX records with /mx) generated more than
1423 one IP address, and the following item turned out to be the local host,
1424 all the secondary addresses of the first item were incorrectly removed
1425 from the list, along with the local host and any following hosts (which
1426 is what is supposed to happen).
1428 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1429 whoever called Exim into the -H file. In the case of the daemon it was
1430 behaving confusingly. When first started, it used values for whoever
1431 started the daemon, but after a SIGHUP it used the Exim user (because it
1432 calls itself on a restart). I have changed the code so that it now always
1435 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1436 message are rejected with the same error (e.g. no authentication or bad
1437 sender address), and a DATA command is nevertheless sent (as can happen
1438 with PIPELINING or a stupid MUA), the error message that was given to the
1439 RCPT commands is included in the rejection of the DATA command. This is
1440 intended to be helpful for MUAs that show only the final error to their
1443 PH/42 Another patch from the Sieve maintainer.
1445 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1446 Eximstats - Fixed some broken HTML links and added missing column headers
1448 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1449 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1451 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1453 PH/43 Yet another patch from the Sieve maintainer.
1455 PH/44 I found a way to check for a TCP/IP connection going away before sending
1456 the response to the final '.' that terminates a message, but only in the
1457 case where the client has not sent further data following the '.'
1458 (unfortunately, this is allowed). However, in many cases there won't be
1459 any further data because there won't be any more messages to send. A call
1460 to select() can be used: if it shows that the input is "ready", there is
1461 either input waiting, or the socket has been closed. An attempt to read
1462 the next input character can distinguish the two cases. Previously, Exim
1463 would have sent an OK response which the client would never have see.
1464 This could lead to message repetition. This fix should cure that, at
1465 least in a lot of common cases.
1467 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1468 advertised in response to EHLO.
1474 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1475 fixed by 4.65/MH/01 (is this a record?) are fixed:
1477 (i) An empty string was always treated as zero by the numeric comparison
1478 operators. This behaviour has been restored.
1480 (ii) It is documented that the numeric comparison operators always treat
1481 their arguments as decimal numbers. This was broken in that numbers
1482 starting with 0 were being interpreted as octal.
1484 While fixing these problems I realized that there was another issue that
1485 hadn't been noticed. Values of message_size_limit (both the global option
1486 and the transport option) were treated as octal if they started with 0.
1487 The documentation was vague. These values are now always treated as
1488 decimal, and I will make that clear in the documentation.
1494 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1495 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1498 MH/01 Don't check that the operands of numeric comparison operators are
1499 integers when their expansion is in "skipping" mode (fixes bug
1500 introduced by 4.64-PH/07).
1502 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1503 child addresses, Exim now panics and dies. Previously, because the count
1504 is held in a short int, deliveries were likely to be lost. As such a
1505 large number of recipients for a single message is ridiculous
1506 (performance will be very, very poor), I have chosen to impose a limit
1507 rather than extend the field.
1513 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1514 leftover -K file (the existence of which was triggered by #402).
1515 While we were at it, introduced process PID as part of the -K
1516 filename. This should rule out race conditions when creating
1519 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1520 processing considerably. Previous code took too long for large mails,
1521 triggering a timeout which in turn triggers #401.
1523 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1524 in the DK code in transports.c. sendfile() is not really portable,
1525 hence the _LINUX specificness.
1527 TF/01 In the add_headers option to the mail command in an Exim filter,
1528 there was a bug that Exim would claim a syntax error in any
1529 header after the first one which had an odd number of characters
1532 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1533 callout verification, Exim cached a "reject" for the entire domain. This
1534 is correct for most verifications, but it is not correct for a recipient
1535 verification with use_sender or use_postmaster set, because in that case
1536 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1537 case of MAIL FROM:<> rejection from other early rejections (e.g.
1538 rejection of HELO). When verifying a recipient using a non-null MAIL
1539 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1540 Whatever the result of the callout, the value of the domain cache is
1541 left unchanged (for any other kind of callout, getting as far as trying
1542 RCPT means that the domain itself is ok).
1544 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1547 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1548 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1549 session, the daemon ignores ECONNECT errors and logs others; it now
1550 ignores EPIPE as well.
1552 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1553 (quoted-printable decoding).
1555 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1556 later the small subsequent patch to fix an introduced bug.
1558 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1560 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1562 PH/08 An error is now given if message_size_limit is specified negative.
1564 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1565 to be given (somewhat) arbitrary names.
1567 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1570 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1571 miscellaneous code fixes
1573 PH/10 Added the log_reject_target ACL modifier to specify where to log
1576 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1577 hostname. This is wrong, because it relates to the incoming message (and
1578 probably the interface on which it is arriving) and not to the outgoing
1579 callout (which could be using a different interface). This has been
1580 changed to use the value of the helo_data option from the smtp transport
1581 instead - this is what is used when a message is actually being sent. If
1582 there is no remote transport (possible with a router that sets up host
1583 addresses), $smtp_active_hostname is used.
1585 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1586 tweaks were necessary in order to get it to work (see also 21 below):
1587 (a) The code assumed that strncpy() returns a negative number on buffer
1588 overflow, which isn't the case. Replaced with Exim's string_format()
1590 (b) There were several signed/unsigned issues. I just did the minimum
1591 hacking in of casts. There is scope for a larger refactoring.
1592 (c) The code used strcasecmp() which is not a standard C function.
1593 Replaced with Exim's strcmpic() function.
1594 (d) The code set only $1; it now sets $auth1 as well.
1595 (e) A simple test gave the error "authentication client didn't specify
1596 service in request". It would seem that Dovecot has changed its
1597 interface. Fortunately there's a specification; I followed it and
1598 changed what the client sends and it appears to be working now.
1600 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1603 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1604 address A is aliased to B and C, where B exists and C does not. Without
1605 -v the output is "A verified" because verification stops after a
1606 successful redirection if more than one address is generated. However,
1607 with -v the child addresses are also verified. Exim was outputting "A
1608 failed to verify" and then showing the successful verification for C,
1609 with its parentage. It now outputs "B failed to verify", showing B's
1610 parentage before showing the successful verification of C.
1612 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1613 look up a TXT record in a specific list after matching in a combined
1616 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1617 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1618 they consult the DNS. I had assumed they would set it the way they
1619 wanted; and indeed my experiments on Linux seem to show that in some
1620 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1621 To be on the safe side, however, I have now made the interface to
1622 host_find_byname() similar to host_find_bydns(), with an argument
1623 containing the DNS resolver options. The host_find_byname() function now
1624 sets these options at its start, just as host_find_bydns() does. The smtp
1625 transport options dns_qualify_single and dns_search_parents are passed to
1626 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1627 of host_find_byname() use the default settings of RES_DEFNAMES
1628 (qualify_single) but not RES_DNSRCH (search_parents).
1630 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1631 spool_read_header() do less string testing, by means of a preliminary
1632 switch on the second character of optional "-foo" lines. (This is
1633 overdue, caused by the large number of possibilities that now exist.
1634 Originally there were few.) While I was there, I also converted the
1635 str(n)cmp tests so they don't re-test the leading "-" and the first
1636 character, in the hope this might squeeze out yet more improvement.
1638 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1639 flag allowing group syntax was set by the header_syntax check but not
1640 turned off, possible causing trouble later; (2) The flag was not being
1641 set at all for the header_verify test, causing "group"-style headers to
1642 be rejected. I have now set it in this case, and also caused header_
1643 verify to ignore an empty address taken from a group. While doing this, I
1644 came across some other cases where the code for allowing group syntax
1645 while scanning a header line wasn't quite right (mostly, not resetting
1646 the flag correctly in the right place). These bugs could have caused
1647 trouble for malformed header lines. I hope it is now all correct.
1649 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1650 with the "reply" argument non-NULL. The code, however (which originally
1651 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1652 but it didn't always do it. This confused somebody who was copying the
1653 code for some other use. I have removed all the tests.
1655 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1656 feature that was used to support insecure browsers during the U.S. crypto
1657 embargo. It requires special client support, and Exim is probably the
1658 only MTA that supported it -- and would never use it because real RSA is
1659 always available. This code has been removed, because it had the bad
1660 effect of slowing Exim down by computing (never used) parameters for the
1661 RSA_EXPORT functionality.
1663 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1664 authenticator to fail if there's a tab character in the incoming data
1665 (there should never be unless someone is messing about, as it's supposed
1666 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1667 if the connection is using TLS or if the remote IP is the same as the
1668 local IP, and the "valid-client-cert option" if a client certificate has
1671 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1672 authenticators. This can be used for authorization after authentication
1673 succeeds. (In the case of plaintext, it servers for both authentication
1676 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1677 if any retry times were supplied.
1679 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1680 connection, where there is no client IP address to check. In this
1681 situation, the verify now always succeeds.
1683 PH/25 Applied John Jetmore's -Mset patch.
1685 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1687 PH/27 In a string expansion for a processed (not raw) header when multiple
1688 headers of the same name were present, leading whitespace was being
1689 removed from all of them, but trailing whitespace was being removed only
1690 from the last one. Now trailing whitespace is removed from each header
1691 before concatenation. Completely empty headers in a concatenation (as
1692 before) are ignored.
1694 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1695 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1697 PH/29 [Removed. This was a change that I later backed out, and forgot to
1698 correct the ChangeLog entry (that I had efficiently created) before
1699 committing the later change.]
1701 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1702 address errors (4xx response to RCPT) over the same connection as other
1703 messages routed to the same hosts. Such deliveries are always "forced",
1704 so retry times are not inspected. This resulted in far too many retries
1705 for the affected addresses. The effect occurred only when there were more
1706 hosts than the hosts_max_try setting in the smtp transport when it had
1707 the 4xx errors. Those hosts that it had tried were not added to the list
1708 of hosts for which the message was waiting, so if all were tried, there
1709 was no problem. Two fixes have been applied:
1711 (i) If there are any address or message errors in an SMTP delivery, none
1712 of the hosts (tried or untried) are now added to the list of hosts
1713 for which the message is waiting, so the message should not be a
1714 candidate for sending over the same connection that was used for a
1715 successful delivery of some other message. This seems entirely
1716 reasonable: after all the message is NOT "waiting for some host".
1717 This is so "obvious" that I'm not sure why it wasn't done
1718 previously. Hope I haven't missed anything, but it can't do any
1719 harm, as the worst effect is to miss an optimization.
1721 (ii) If, despite (i), such a delivery is accidentally attempted, the
1722 routing retry time is respected, so at least it doesn't keep
1723 hammering the server.
1725 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1726 in ${readsocket because some servers need this prod.
1728 PH/32 Added some extra debug output when updating a wait-xxx database.
1730 PH/33 The hint "could be header name not terminated by colon", which has been
1731 given for certain expansion errors for a long time, was not being given
1732 for the ${if def:h_colon_omitted{... case.
1734 PH/34 The spec says: "With one important exception, whenever a domain list is
1735 being scanned, $domain contains the subject domain." There was at least
1736 one case where this was not true.
1738 PH/35 The error "getsockname() failed: connection reset by peer" was being
1739 written to the panic log as well as the main log, but it isn't really
1740 panic-worthy as it just means the connection died rather early on. I have
1741 removed the panic log writing for the ECONNRESET error when getsockname()
1744 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1745 runs only) independently of the message's sender address. This meant
1746 that, if the 4xx error was in fact related to the sender, a different
1747 message to the same recipient with a different sender could confuse
1748 things. In particualar, this can happen when sending to a greylisting
1749 server, but other circumstances could also provoke similar problems.
1750 I have changed the default so that the retry time for these errors is now
1751 based a combination of the sender and recipient addresses. This change
1752 can be overridden by setting address_retry_include_sender=false in the
1755 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1756 remote server are returned as part of bounce messages. This was not
1757 happening for LMTP over a pipe (the lmtp transport), but now it is the
1758 same for both kinds of LMTP.
1760 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1761 in header lines that were in fact CNAMEs. This is no longer the case.
1763 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1764 and queue runs started by the daemon processed all messages. This has
1765 been fixed so that -R and -S can now usefully be given with -q<time>.
1767 PH/40 Import PCRE release 6.7 (fixes some bugs).
1769 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1771 PH/42 Give an error if -q is specified more than once.
1773 PH/43 Renamed the variables $interface_address and $interface_port as
1774 $received_ip_address and $received_port, to make it clear that these
1775 values apply to message reception, and not to the outgoing interface when
1776 a message is delivered. (The old names remain recognized, of course.)
1778 PH/44 There was no timeout on the connect() call when using a Unix domain
1779 socket in the ${readsocket expansion. There now is.
1781 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1782 be meaningful with "accept".
1784 SC/01 Eximstats V1.43
1785 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1787 SC/02 Eximstats V1.44
1788 Use a glob alias rather than an array ref in the generated
1789 parser. This improves both readability and performance.
1791 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1792 Collect SpamAssassin and rejection statistics.
1793 Don't display local sender or destination tables unless
1794 there is data to show.
1795 Added average volumes into the top table text output.
1797 SC/04 Eximstats V1.46
1798 Collect data on the number of addresses (recipients)
1799 as well as the number of messages.
1801 SC/05 Eximstats V1.47
1802 Added 'Message too big' to the list of mail rejection
1803 reasons (thanks to Marco Gaiarin).
1805 SC/06 Eximstats V1.48
1806 Mainlog lines which have GMT offsets and are too short to
1807 have a flag are now skipped.
1809 SC/07 Eximstats V1.49 (Alain Williams)
1810 Added the -emptyok flag.
1812 SC/08 Eximstats V1.50
1813 Fixes for obtaining the IP address from reject messages.
1815 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1816 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1817 whitesspace changes from 4.64-PH/27
1819 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1822 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1823 are found, allow negative numbers in numeric criteria)
1825 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1827 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1828 to match changes made in 4.64-PH/43
1830 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1832 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1833 "domain" is actually an IP address, and force a failure. This locks out
1834 those revolvers/nameservers that support "A-for-A" lookups, in
1835 contravention of the specifications.
1837 PH/48 When a host name was looked up from an IP address, and the subsequent
1838 forward lookup of the name timed out, the host name was left in
1839 $sender_host_name, contrary to the specification.
1841 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1842 restricted to single-key lookups, Exim was not diagnosing an error if
1843 * or *@ was used with a query-style lookup.
1845 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1847 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1848 long ago, but noone interested enough thought of it. Let's just say that
1849 the "1.1" means that there are some new functions that weren't there at
1850 some point in the past.
1852 PH/51 Error processing for expansion failure of helo_data from an smtp
1853 transport during callout processing was broken.
1855 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1856 tested/used via the -bh/-bhc/-bs options.
1858 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1859 bug, fixed in subsequent PCRE releases).
1861 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1862 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1864 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1870 SC/01 Use a glob alias rather than an array ref in eximstats generated
1871 parser. This improves both readability and performance.
1873 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1874 Don't display local sender or destination tables in eximstats unless
1875 there is data to show.
1876 Added average volumes into the eximstats top table text output.
1878 SC/03 Collect data on the number of addresses (recipients) as well
1879 as the number of messages in eximstats.
1881 TF/01 Correct an error in the documentation for the redirect router. Exim
1882 does (usually) call initgroups() when daemonizing.
1884 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1885 with consistent privilege compared to when running as a daemon.
1887 TF/03 Note in the spec that $authenticated_id is not set for local
1888 submissions from trusted users.
1890 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1891 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1893 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1894 by adding some example configuration directives to the default
1895 configuration file. A little bit of work is required to uncomment the
1896 directives and define how usernames and passwords are checked, but
1897 there is now a framework to start from.
1899 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1900 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1901 without this. I don't know how relevant this is to other LDAP libraries.
1903 PH/02 Add the verb name to the "unknown ACL verb" error.
1905 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1907 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1909 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1910 directory not expanded when it should be if an expanded home directory
1911 was set for the address (which is overridden by the transport).
1913 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1916 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1917 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1918 because it is too late at that time, and has no effect.
1920 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1921 security issue with \' (bugzilla #107). I could not use the
1922 PQescapeStringConn() function, because it needs a PGconn value as one of
1925 PH/08 When testing addresses using -bt, indicate those final addresses that
1926 are duplicates that would not cause an additional delivery. At least one
1927 person was confused, thinking that -bt output corresponded to deliveries.
1928 (Suppressing duplicates isn't a good idea as you lose the information
1929 about possibly different redirections that led to the duplicates.)
1931 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1932 systems where poll() doesn't work, in particular OS X.
1934 PH/10 Added more information to debugging output for retry time not reached.
1936 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1937 operations in malware.c.
1939 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1942 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1943 syslog_duplication set false, log lines that would normally be written
1944 both the the main log and to the reject log were not written to syslog at
1947 PH/14 In the default configuration, change the use of "message" in ACL warn
1948 statements to "add_header".
1950 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1951 not followed by a command (e.g. "seen endif").
1953 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1954 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1957 PH/17 Added extra conditions to the default value of delay_warning_condition
1961 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1962 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1963 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1966 The Auto-Submitted: and various List- headers are standardised, whereas I
1967 don't think Precedence: ever was.
1969 PH/18 Refactored debugging code in route_finduser() to show more information,
1970 in particular, the error code if getpwnam() issues one.
1972 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1973 This is apparently needed in addition to the PH/07 change above to avoid
1974 any possible encoding problems.
1976 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1977 but not after initializing Perl.
1979 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1980 output them only if debugging. By default they are written stderr,
1981 apparently, which is not desirable.
1983 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1986 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1989 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1991 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1992 authenticated or an ident call has been made. Suppress the default
1993 values for $authenticated_id and $authenticated_sender (but permit -oMai
1994 and -oMas) when testing with -bh.
1996 PH/24 Re-jigged the order of the tests in the default configuration so that the
1997 tests for valid domains and recipients precede the DNS black list and CSA
1998 tests, on the grounds that those ones are more expensive.
2000 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2001 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2002 This bug exists in every version of Exim that I still have, right back to
2005 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2006 However, an attempt to turn on case-sensitivity in a regex key by
2007 including (?-i) didn't work because the subject string was already
2008 lowercased, and the effects were non-intuitive. It turns out that a
2009 one-line patch can be used to allow (?-i) to work as expected.
2015 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2016 other effects) broke the use of negated acl sub-conditions.
2018 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2021 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2022 "Deny" causes Exim to reject the incoming connection with a 554 error.
2023 Unfortunately, if there is a major crisis, such as a disk failure,
2024 tcp-wrappers gives "deny", whereas what one would like would be some
2025 kind of temporary error. A kludge has been added to help with this.
2026 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2027 554 error is used if errno is still zero or contains ENOENT (which occurs
2028 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2031 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2033 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2034 errors. Otherwise a message that provokes a temporary error (when other
2035 messages do not) can cause a whole host to time out.
2037 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2038 addresses were routed directly to files or pipes from a redirect router.
2039 File deliveries just didn't batch; pipe deliveries might have suffered
2042 PH/06 A failure to get a lock for a hints database would erroneously always say
2043 "Failed to get write lock", even when it was really a read lock.
2045 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2046 of 0600. This has been changed to use the value of the lockfile_mode
2047 option (which defaults to 0600).
2049 PH/08 Applied small patch from the Sieve maintainer.
2051 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2052 folder from quota calculations, a direct delivery into this folder messed
2053 up the contents of the maildirsize file. This was because the regex was
2054 used only to exclude .Trash (or whatever) when the size of the mailbox
2055 was calculated. There was no check that a delivery was happening into an
2056 excluded directory. This bug has been fixed by ignoring all quota
2057 processing for deliveries into excluded directories.
2059 PH/10 Added the maildirfolder_create_regex option to appendfile.
2065 PH/01 The code for finding all the local interface addresses on a FreeBSD
2066 system running IPv6 was broken. This may well have applied to all BSD
2067 systems, as well as to others that have similar system calls. The broken
2068 code found IPv4 interfaces correctly, but gave incorrect values for the
2069 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2070 that it would not match correctly against @[] and not recognize the IPv6
2073 PH/02 The ipliteral router was not recognizing addresses of the form user@
2074 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2076 PH/03 Added disable_ipv6.
2078 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2079 decoded form, because it is most often used to construct To: headers
2080 lines in autoreplies, and the decoded form may well be syntactically
2081 invalid. However, $reply_address has leading white space removed, and all
2082 newlines turned into spaces so that the autoreply transport does not
2085 PH/05 If group was specified without a user on a router, and no group or user
2086 was specified on a transport, the group from the router was ignored.
2088 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2089 for visible compile-time settings that can be used to change these
2090 numbers, for those that want even more. Backwards compatibility with old
2091 spool files has been maintained. However, going back to a previous Exim
2092 release will lost any variables that are in spool files.
2094 PH/07 Two small changes when running in the test harness: increase delay when
2095 passing a TCP/IP connection to a new process, in case the original
2096 process has to generate a bounce, and remove special handling of
2097 127.0.0.2 (sic), which is no longer necessary.
2099 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2100 be the same on different OS.
2102 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2105 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2106 whether --show-vars was specified or not
2108 JJ/02 exipick: Added support for new ACL variable spool format introduced
2111 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2112 syntactically invalid From: or Reply-to: line, and a filter used this to
2113 generate an autoreply, and therefore failed to obtain an address for the
2114 autoreply, Exim could try to deliver to a non-existent relative file
2115 name, causing unrelated and misleading errors. What now happens is that
2116 it logs this as a hard delivery error, but does not attempt to create a
2119 PH/11 The exinext utility has a -C option for testing purposes, but although
2120 the given file was scanned by exinext itself; it wasn't being passed on
2121 when Exim was called.
2123 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2124 an end-of-file indication when reading a command response.
2126 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2127 compiled. In many other places in Exim, IPv6 addresses are always
2128 recognized, so I have changed this. It also means that IPv4 domain
2129 literals of the form [IPV4:n.n.n.n] are now always recognized.
2131 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2132 used if the router is not running as root, for example, when verifying at
2133 ACL time, or when using -bh. The debugging output from this situation was
2134 non-existent - all you got was a failure to exec. I have made two
2137 (a) Failures to set uid/gid, the current directory, or a process leader
2138 in a subprocess such as that created by queryprogram now generate
2139 suitable debugging ouput when -d is set.
2141 (b) The queryprogram router detects when it is not running as root,
2142 outputs suitable debugging information if -d is set, and then runs
2143 the subprocess without attempting to change uid/gid.
2145 PH/15 Minor change to Makefile for building test_host (undocumented testing
2148 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2149 additional section of a DNS packet that returns MX or SRV records.
2150 Instead, it always explicitly searches for A/AAAA records. This avoids
2151 major problems that occur when a DNS server includes only records of one
2152 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2153 fixed another bug: if SRV records were looked up and the corresponding
2154 address records were *not* found in the additional section, the port
2155 values from the SRV records were lost.
2157 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2158 using the correct key (the original address) when searching the retry
2159 rules in order to find which one to use for generating the retry hint.
2161 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2162 adding the default one. Similarly, if it contains a Reply-To: header, the
2163 errors_reply_to option, if set, is not used.
2165 PH/19 When calculating a retry time, Exim used to measure the "time since
2166 failure" by looking at the "first failed" field in the retry record. Now
2167 it does not use this if it is later than than the arrival time of the
2168 message. Instead it uses the arrival time. This makes for better
2169 behaviour in cases where some deliveries succeed, thus re-setting the
2170 "first failed" field. An example is a quota failure for a huge message
2171 when small messages continue to be delivered. Without this change, the
2172 "time since failure" will always be short, possible causing more frequent
2173 delivery attempts for the huge message than are intended.
2174 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2176 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2177 $1, $2, $3) because the numerical variables can be reset during some
2178 expansion items (e.g. "match"), thereby losing the authentication data.
2180 PH/21 Make -bV show the size of off_t variables so that the test suite can
2181 decide whether to run tests for quotas > 2G.
2183 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2184 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2185 filecount value is greater than 2G or if a quota value is greater than 2G
2186 on a system where the size of off_t is not greater than 4, a panic error
2189 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2190 never match. The debug and -bh output now contains an explicit error
2191 message indicating a malformed IPv4 address or mask.
2193 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2194 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2195 PH/23 above applies.
2197 PH/25 Do not write to syslog when running in the test harness. The only
2198 occasion when this arises is a failure to open the main or panic logs
2199 (for which there is an explicit test).
2201 PH/26 Added the /no_tell option to "control=freeze".
2203 PH/27 If a host name lookup failed very early in a connection, for example, if
2204 the IP address matched host_lookup and the reverse lookup yielded a name
2205 that did not have a forward lookup, an error message of the form "no IP
2206 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2207 could be logged. Now it outputs the IP address instead of "NULL".
2209 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2210 allows the sender and the authenticated sender to be set when
2211 submitting a message from within Exim. Since child_open_exim() is
2212 documented for local_scan(), the new function should be too.
2214 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2215 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2216 results in an empty string is now treated as unset.
2218 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2220 PH/31 Added sender_verify_fail logging option.
2222 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2223 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2224 tidied the source and removed it altogether.
2226 PH/33 When a queue run was abandoned because the load average was too high, a
2227 log line was always written; now it is written only if the queue_run log
2228 selector is set. In addition, the log line for abandonment now contains
2229 information about the queue run such as the pid. This is always present
2230 in "start" and "stop" lines but was omitted from the "abandon" line.
2232 PH/34 Omit spaces between a header name and the colon in the error message that
2233 is given when verify = headers_syntax fails (if there are lots of them,
2234 the message gets confusing).
2236 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2237 names, as there are now some PTR records that contain slashes. This check
2238 is only to protect against broken name servers that fall over on strange
2239 characters, so the fact that it applies to all lookups doesn't matter.
2241 PH/36 Now that the new test suite is complete, we can remove some of the
2242 special code in Exim that was needed for the old test suite. For example,
2243 sorting DNS records because real resolvers return them in an arbitrary
2244 order. The new test suite's fake resolver always returns records in the
2247 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2248 bounces) except when queue_only is set, to avoid logging races between
2249 the different processes.
2251 PH/38 Panic-die if .include specifies a non-absolute path.
2253 PH/39 A tweak to the "H" retry rule from its user.
2255 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2256 a label. They prevented compilation on older perls.
2258 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2259 a warning to be raised on newish perls.
2261 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2262 on queue. Changes to match documented behaviour of showing count of
2263 messages matching specified criteria.
2265 PH/40 Changed the default ident timeout from 30s to 5s.
2267 PH/41 Added support for the use of login_cap features, on those BSD systems
2268 that have them, for controlling the resources used by pipe deliveries.
2270 PH/42 The content-scanning code uses fopen() to create files in which to put
2271 message data. Previously it was not paying any attention to the mode of
2272 the files. Exim runs with umask(0) because the rest of the code creates
2273 files with open(), and sets the required mode explicitly. Thus, these
2274 files were ending up world-writeable. This was not a big issue, because,
2275 being within the spool directory, they were not world-accessible. I have
2276 created a function called modefopen, which takes an additional mode
2277 argument. It sets umask(777), creates the file, chmods it to the required
2278 mode, then resets the umask. All the relevant calls to fopen() in the
2279 content scanning code have been changed to use this function.
2281 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2282 to 24 hours. This avoids potential overflow problems when processing G
2283 and H retry rules. I suspect nobody ever tinkers with this value.
2285 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2287 PH/45 When the plaintext authenticator is running as a client, the server's
2288 challenges are checked to ensure they are valid base64 strings. By
2289 default, the authentication attempt is cancelled if an invalid string is
2290 received. Setting client_ignore_invalid_base64 true ignores these errors.
2291 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2292 they are received. Thus, the responses can be made to depend on the
2293 challenges. If an invalid string is ignored, an empty string is placed in
2296 PH/46 Messages that are created by the autoreply transport now contains a
2297 References: header, in accordance with RFCs 2822 and 3834.
2299 PH/47 Added authenticated_sender_force to the smtp transport.
2301 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2303 PH/49 Installed latest patch from the Sieve maintainer.
2305 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2306 was also set, the appendfile transport was unnecessarily scanning a
2307 directory of message files (e.g. for maildir delivery) to find the count
2308 of files (along with the size), even though it did not need this
2309 information. It now does the scan only if it needs to find either the
2310 size of the count of files.
2312 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2314 PH/52 Two bugs concerned with error handling when the smtp transport is
2317 (i) Exim was not creating retry information for temporary errors given
2318 for individual recipients after the DATA command when the smtp transport
2319 was used in LMTP mode. This meant that they could be retried too
2320 frequently, and not timed out correctly.
2322 (ii) Exim was setting the flag that allows error details to be returned
2323 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2324 recipients that were returned after the DATA command.
2326 PH/53 This is related to PH/52, but is more general: for any failing address,
2327 when detailed error information was permitted to be returned to the
2328 sender, but the error was temporary, then after the final timeout, only
2329 "retry timeout exceeded" was returned. Now it returns the full error as
2330 well as "retry timeout exceeded".
2332 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2333 do this, and (what is worse) MTAs that accept it.
2335 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2336 will now be deprecated.
2338 PH/56 New os.c-cygwin from the Cygwin maintainer.
2340 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2341 formats (previously only available in exim formats via -bpr, -bpru,
2342 and -bpra. Now also available in native and exiqgrep formats)
2344 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2345 with very large, slow to parse queues
2347 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2349 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2351 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2352 responses to authentication challenges, though it was showing the
2353 challenges; (ii) I've removed the CR characters from the debug output for
2356 PH/58 Allow for the insertion of a newline as well as a space when a string
2357 is turned into more than one encoded-word during RFC 2047 encoding. The
2358 Sieve code now uses this.
2360 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2361 data_4xx, lost_connection, tls_required.
2363 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2364 message was being sent as an SMTP response.
2366 PH/61 Add -l and -k options to exicyclog.
2368 PH/62 When verifying, if an address was redirected to one new address, so that
2369 verification continued, and the new address failed or deferred after
2370 having set something in $address_data, the value of $address_data was not
2371 passed back to the ACL. This was different to the case when no
2372 redirection occurred. The value is now passed back in both cases.
2374 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2375 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2376 use login_cap.h, so on its own it isn't the distinguishing feature. The
2377 new name refers directly to the setclassresources() function.
2379 PH/65 Added configuration files for NetBSD3.
2381 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2383 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2384 is preferred over IPv4.
2386 PH/68 The bounce_return_message and bounce_return_body options were not being
2387 honoured for bounces generated during the reception of non-SMTP messages.
2388 In particular, this applied to messages rejected by the ACL. This bug has
2389 been fixed. However, if bounce_return_message is true and bounce_return_
2390 body is false, the headers that are returned for a non-SMTP message
2391 include only those that have been read before the error was detected.
2392 (In the case of an ACL rejection, they have all been read.)
2394 PH/69 The HTML version of the specification is now built in a directory called
2395 spec_html instead of spec.html, because the latter looks like a path with
2396 a MIME-type, and this confuses some software.
2398 PH/70 Catch two compiler warnings in sieve.c.
2400 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2401 function verify_get_ident() calls ip_connect() to connect a socket, but
2402 if the "connect()" function timed out, ip_connect() used to close the
2403 socket. However, verify_get_ident() also closes the socket later, and in
2404 between Exim writes to the log, which may get opened at this point. When
2405 the socket was closed in ip_connect(), the log could get the same file
2406 descriptor number as the socket. This naturally causes chaos. The fix is
2407 not to close the socket in ip_connect(); the socket should be closed by
2408 the function that creates it. There was only one place in the code where
2409 this was missing, in the iplookup router, which I don't think anybody now
2410 uses, but I've fixed it anyway.
2412 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2413 well as to direct DNS lookups. Otherwise the handling of names in host
2414 lists is inconsistent and therefore confusing.
2420 PH/01 Two changes to the default runtime configuration:
2422 (1) Move the checks for relay_from_hosts and authenticated clients from
2423 after to before the (commented out) DNS black list checks.
2425 (2) Add control=submission to the relay_from_hosts and authenticated
2426 clients checks, on the grounds that messages accepted by these
2427 statements are most likely to be submissions.
2429 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2431 (1) Generate an error if the third argument for the ${prvs expansion is
2434 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2437 (3) Reset the variables that are obtained from the first argument of
2438 ${prvscheck and used in the second argument before leaving the code,
2439 because their memory is reclaimed, so using them afterwards may do
2442 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2443 one (it's much easier than Tom thought :-).
2445 (5) Because of (4), we can now allow for the use of $prvscheck_result
2446 inside the third argument.
2448 PH/03 For some reason, the default setting of PATH when running a command from
2449 a pipe transport was just "/usr/bin". I have changed it to
2452 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2453 anything to be listed in the output from -bV.
2455 PH/05 When a filter generated an autoreply, the entire To: header line was
2456 quoted in the delivery log line, like this:
2458 => >A.N.Other <ano@some.domain> <original@ddress> ...
2460 This has been changed so that it extracts the operative address. There
2461 may be more than one such address. If so, they are comma-separated, like
2464 => >ano@some.domain,ona@other.domain <original@ddress> ...
2466 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2467 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2468 not being looked up in the rDNS to get a host name, Exim was showing the
2469 IP address twice in Received: lines, even though the IP addresses were
2470 identical. For example:
2472 Received: from [1.2.3.4] (helo=[1.2.3.4])
2474 However, if the real host name was known, it was omitting the HELO data
2475 if it matched the actual IP address. This has been tidied up so that it
2476 doesn't show the same IP address twice.
2478 PH/07 When both +timestamp and +memory debugging was on, the value given by
2479 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2480 called by the debug printing, thereby overwriting the timestamp buffer.
2481 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2483 PH/08 When the original message was included in an autoreply transport, it
2484 always said "this is a copy of the message, including all the headers",
2485 even if body_only or headers_only was set. It now gives an appropriate
2488 PH/09 Applied a patch from the Sieve maintainer which:
2490 o fixes some comments
2491 o adds the (disabled) notify extension core
2492 o adds some debug output for the result of if/elsif tests
2493 o points to the current vacation draft in the documentation
2494 and documents the missing references header update
2498 o fixes a bug in processing the envelope test (when testing
2499 multiple envelope elements, the last element determinted the
2502 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2503 Electronic Mail") by including:
2505 Auto-submitted: auto-generated
2507 in the messages that it generates (bounce messages and others, such as
2508 warnings). In the case of bounce messages for non-SMTP mesages, there was
2509 also a typo: it was using "Auto_submitted" (underscore instead of
2510 hyphen). Since every message generated by Exim is necessarily in response
2511 to another message, thes have all been changed to:
2513 Auto-Submitted: auto-replied
2515 in accordance with these statements in the RFC:
2517 The auto-replied keyword:
2519 - SHOULD be used on messages sent in direct response to another
2520 message by an automatic process,
2522 - MUST NOT be used on manually-generated messages,
2524 - MAY be used on Delivery Status Notifications (DSNs) and Message
2525 Disposition Notifications (MDNs),
2527 - MUST NOT be used on messages generated by automatic or periodic
2528 processes, except for messages which are automatic responses to
2531 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2532 to the default Received: header definition.
2534 PH/12 Added log selector acl_warn_skipped (default on).
2536 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2537 variables because (a) they are in the wrong storage pool and (b) even if
2538 they were copied, it wouldn't work properly because of the caching.
2540 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2541 checking when decoding. Apparently there are clients that generate
2542 overlong encoded strings. Why am I not surprised?
2544 PH/15 If the first argument of "${if match_address" was not empty, but did not
2545 contain an "@" character, Exim crashed. Now it writes a panic log message
2546 and treats the condition as false.
2548 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2550 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2551 extension "envelope-auth". The code is finished and in agreement with
2552 other implementations, but there is no documentation so far and in fact,
2553 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2554 not changing the active code.
2556 Print executed "if" and "elsif" statements when debugging is used. This
2557 helps a great deal to understand what a filter does.
2559 Document more things not specified clearly in RFC3028. I had all this
2560 sorted out, when out of a sudden new issues came to my mind. Oops."
2562 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2565 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2566 canonical form (as documented). However, after a host name lookup from
2567 the IP address, check_host() was doing a simple string comparison with
2568 addresses acquired from the DNS when checking that the found name did
2569 have the original IP as one of its addresses. Since any found IPv6
2570 addresses are likely to be in abbreviated form, the comparison could
2571 fail. Luckily, there already exists a function for doing the comparison
2572 by converting both addresses to binary, so now that is used instead of
2573 the text comparison.
2575 PH/20 There was another similar case to PH/19, when a complete host name was
2576 given in a host list; looking up its IP address could give an abbreviated
2577 form, whereas the current host's name might or might not be abbreviated.
2578 The same fix has been applied.
2584 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2585 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2588 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2589 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2591 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2593 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2594 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2595 Also, for "Auto-Submitted", treat anything other than "no" as
2596 non-personal, in accordance with RFC 3834. (Previously it treated
2597 anything starting "auto-" as non-personal.)
2599 TF/01 The control=submission/name=... option had a problem with syntax
2600 errors if the name included a slash character. The /name= option
2601 now slurps the rest of the string, so it can include any characters
2602 but it must come last in the list of options (after /sender_retain
2605 PH/05 Some modifications to the interface to the fake nameserver for the new
2613 TK/01 Added the "success_on_redirect" address verification option. See
2614 NewStuff for rationale and an example.
2616 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2618 PH/02 Patch to exigrep to allow it to work on syslog lines.
2620 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2621 fread() to skip over the body file's header line, because in Cygwin the
2622 header line is locked and is inaccessible.
2624 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2625 co-exist for some time) to make it clear that it is the Exim ID that is
2626 referenced, not the Message-ID: header line.
2628 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2629 string_format() function, because snprintf() does not exist on all
2632 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2633 ${stat: expansion item.
2635 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2636 protocol synchronization error", to keep the pedants happy.
2638 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2639 well as for IRIX systems, when gcc is being used. See the host.c source
2642 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2644 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2647 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2648 transport and to the smtp transport in LMTP mode.
2650 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2652 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2653 run a filter in a subprocess. This could lead to confusion in subsequent
2654 lookups in the parent process. There should also be a search_tidyup() at
2655 the end of the subprocess.
2657 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2658 only if the host matched helo_try_verify_hosts, which caused the
2659 verification to occur when the EHLO/HELO command was issued. The ACL just
2660 tested the remembered result. Now, if a previous verification attempt has
2661 not happened, "verify = helo" does it there and then.
2663 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2665 TK/03 Fix log output including CR from clamd.
2667 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2668 exist provoked a memory error which could cause a segfault.
2670 PH/15 Installed PCRE 6.2
2672 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2674 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2675 of the problem. Specifically, suggested +O2 rather than +O1 for the
2678 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2680 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2681 record, the port was not correctly logged when the outgoing_port log
2682 selector was set (it logged the transort's default port).
2684 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2685 fallback_hosts, and "hosts" in the smtp transport.
2687 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2688 host errors such as "Connection refused".
2690 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2691 authentication with radiusclient 0.4.9:
2693 - Error returned from rc_read_config was caught wrongly
2694 - Username/password not passed on to radius server due to wrong length.
2696 The presumption is that some radiusclient API changes for 4.51/PH/17
2697 were not taken care of correctly. The code is still untested by me (my
2698 Linux distribution still has 0.3.2 of radiusclient), but it was
2699 contributed by a Radius user.
2701 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2702 expanding the "port" option of the smtp transport.
2704 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2705 while reading a MIME header. Thanks to Tom Hughes for a patch.
2707 PH/24 Include config.h inside local_scan.h so that configuration settings are
2710 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2711 that in an ACL for RCPT (for example), you can examine exactly what was
2714 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2715 commands, but it was not correctly comparing the address with the actual
2716 client host address. Thus, it would show the EHLO address in Received:
2717 header lines when this was not necessary.
2719 PH/27 Added the % operator to ${eval:}.
2721 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2722 it should be ignoring failures (because with -C, for example, it has lost
2723 privilege). It wasn't ignoring creation failures other than "already
2726 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2729 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2730 because an input line was too long, either on its own, or by virtue of
2731 too many continuations, the temporary file was not being removed, and the
2732 return code was incorrect.
2734 PH/31 Missing "BOOL" in function definition in filtertest.c.
2736 PH/32 Applied Sieve patches from the maintainer.
2738 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2740 PH/33 Added "verify = not_blind".
2742 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2743 Local/Makefile (with some defaults set). These are used in built scripts
2744 such as exicyclog, but they have never been used in the exim_install
2745 script (though there are many overriding facilities there). I have
2746 arranged that the exim_install script now takes note of these two
2749 PH/35 Installed configuration files for Dragonfly.
2751 PH/36 When a locally submitted message by a trusted user did not contain a
2752 From: header, and the sender address was obtained from -f or from an SMTP
2753 MAIL command, and the trusted user did not use -F to supply a sender
2754 name, $originator_name was incorrectly used when constructing a From:
2755 header. Furthermore, $originator_name was used for submission mode
2756 messages from external hosts without From: headers in a similar way,
2757 which is clearly wrong.
2759 PH/37 Added control=suppress_local_fixups.
2761 PH/38 When log_selector = +received_sender was set, and the addition of the
2762 sender made the log line's construction buffer exactly full, or one byte
2763 less than full, an overflow happened when the terminating "\n" was
2766 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2767 when the result of a list match is failure because a DNS lookup failed.
2769 PH/40 RM_COMMAND is now used in the building process.
2771 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2772 the "build-* directories that it finds.
2774 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2775 address in a domain literal was a prefix of an interface address.
2777 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2778 when verifying a sender address, unless rewrite_headers is false.
2780 PH/44 Wrote a long comment about why errors_to addresses are verified as
2781 recipients, not senders.
2783 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2784 the ratelimit ACL was added.
2786 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2788 PH/46 Added extra information about PostgreSQL errors to the error string.
2790 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2791 avoiding the need to install special zones in a real server. This is
2792 backwards compatible; if it can't find the fake resolver, it drops back.
2793 Thus, both old and new test suites can be run.
2795 TF/02 Added util/ratelimit.pl
2797 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2798 clock is set back in time.
2800 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2801 Candler <B.Candler@pobox.com>.
2803 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2804 OK for addresses that are the result of redirections.
2806 PH/48 A number of further additions for the benefit of the new test suite,
2807 including a fake gethostbyname() that interfaces to the fake DNS resolver
2810 TF/06 The fix for widen_domains has also been applied to qualify_single and
2811 search_parents which are the other dnslookup options that can cause
2814 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2817 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2819 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2820 service" if it encountered an MX record with an empty target hostname.
2821 The message is now "an MX or SRV record indicated no SMTP service".
2823 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2824 if the DNS of the sending site is misconfigured. This is quite a
2825 common situation. This change restores the behaviour of treating a
2826 helo verification defer as a failure.
2828 PH/51 If self=fail was set on a router, the bounce message did not include the
2829 actual error message.
2835 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2837 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2838 command itself ended in error, the underlying message about the transport
2839 filter timeout was being overwritten with the pipe command error. Now the
2840 underlying error message should be appended to the second error message.
2842 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2844 PH/02 Reduce the amount of output that "make" produces by default. Full output
2845 can still be requested.
2847 PH/03 The warning log line about a condition test deferring for a "warn" verb
2848 was being output only once per connection, rather than after each
2849 occurrence (because it was using the same function as for successful
2850 "warn" verbs). This seems wrong, so I have changed it.
2852 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2853 it should not have, which might have caused a crash in the right
2854 circumstances, but probably never did.
2856 PH/04 Installed a modified version of Tony Finch's patch to make submission
2857 mode fix the return path as well as the Sender: header line, and to
2858 add a /name= option so that you can make the user's friendly name appear
2861 TF/03 Added the control = fakedefer ACL modifier.
2863 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2864 Mark Lowes for thorough testing.
2866 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2868 TK/03 Merged latest SRS patch from Miles Wilton.
2870 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2871 with the definition in sysexits.h (which is #included earlier).
2872 Fortunately, Exim does not actually use EX_OK. The code used to try to
2873 preserve the sysexits.h value, by assumimg that macro definitions were
2874 scanned for macro replacements. I have been disabused of this notion,
2875 so now the code just undefines EX_OK before #including unistd.h.
2877 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2878 in the smtp transport. When a block could not be written in a single
2879 write() function, the timeout was being re-applied to each part-write.
2880 This seems wrong - if the receiver was accepting one byte at a time it
2881 would take for ever. The timeout is now adjusted when this happens. It
2882 doesn't have to be particularly precise.
2884 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2885 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2887 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2888 without a domain if the check to <postmaster@domain> fails.
2890 SC/01 Eximstats: added -xls and the ability to specify output files
2891 (patch written by Frank Heydlauf).
2893 SC/02 Eximstats: use FileHandles for outputing results.
2895 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2897 SC/04 Eximstats: fixed display of large numbers with -nvr option
2899 SC/05 Eximstats: fixed merging of reports with empty tables.
2901 SC/06 Eximstats: added the -include_original_destination flag
2903 SC/07 Eximstats: removed tabs and trailing whitespace.
2905 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2907 TK/06 MBOX spool code: Add real "From " MBOX separator line
2908 so the .eml file is really in mbox format (even though
2909 most programs do not really care). Patch from Alex Miller.
2911 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2912 The latter is generated from $received_to and is only set if the
2913 message has one envelope recipient. SA can use these headers,
2914 obviously out-of-the-box. Patch from Alex Miller.
2916 PH/08 The ${def test on a variable was returning false if the variable's
2917 value was "0", contrary to what the specification has always said!
2918 The result should be true unless the variable is empty.
2920 PH/09 The syntax error of a character other than { following "${if
2921 def:variable_name" (after optional whitespace) was not being diagnosed.
2922 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2923 accidental colon was present, for example, could give incorrect results.
2925 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2926 result is used (not including appendfile, where other changes are about
2929 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2930 This involved changing a lot of size variables from int to off_t. It
2931 should work with maildirs and everything.
2933 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2934 spamd dying while we are connected to it.
2936 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2939 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2940 function for BDB 4.3.
2942 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2944 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2945 the sources for PCRE have been re-arranged and more files are now
2948 PH/15 The code I had for printing potentially long long variables in PH/11
2949 above was not the best (it lost precision). The length of off_t variables
2950 is now inspected at build time, and an appropriate printing format (%ld
2951 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2952 to be "long long int" or "long int". This is needed for the internal
2953 formatting function string_vformat().
2955 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2956 the configuration file to be ":syslog", then the script "guesses" where
2957 the logs files are, rather than using the compiled in default. In our
2958 case the guess is not the same as the compiled default, so the script
2959 suddenly stopped working when I started to use syslog. The patch checks
2960 to see if log_file_path is "". If so, it attempts to read it from exim
2961 with no configuration file to get the compiled in version, before it
2962 falls back to the previous guessing code."
2964 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2965 implementing BATV in an Exim configuration. See NewStuff for the gory
2968 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2969 Makefile that are specific to HP-UX.
2971 PH/18 If the "use_postmaster" option was set for a recipient callout together
2972 with the "random" option, the postmaster address was used as the MAIL
2973 FROM address for the random test, but not for the subsequent recipient
2974 test. It is now used for both.
2976 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2977 patch removes a few documentation additions to RFC 3028, because the
2978 latest draft now contains them. It adds the new en;ascii-case comparator
2979 and a new error check for 8bit text in MIME parts. Comparator and
2980 require names are now matched exactly. I enabled the subaddress
2981 extension, but it is not well tested yet (read: it works for me)."
2983 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2984 rework some of the code of TK/09 above to avoid the hardwired use of
2985 "%lld" and "long long". Replaced the call to snprintf() with a call to
2988 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2989 records point to non-existent hosts", "retry timeout exceeded", and
2990 "retry time not reached for any host after a long failure period".
2992 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2993 experimental DomainKeys support:
2995 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2996 (2) On an error such as an illegally used "control", the wrong name for
2997 the control was given.
2999 These problems did NOT occur unless DomainKeys support was compiled.
3001 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3003 PH/24 Added ${if match_ip condition.
3005 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3006 databases so that it will be absolutely obvious if a crash occurs in the
3007 DB library. This is a regular occurrence (often caused by mis-matched
3010 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3011 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3012 picked up on a user's system that detects such things. There doesn't seem
3013 to be a gcc warning option for this - only an attribute that has to be
3014 put on the function's prototype. It seems that in Fedora Core 4 they have
3015 set this on a number of new functions. No doubt there will be more in due
3018 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3019 specify a transport. However, if an address that was verified by such a
3020 router was the subject of a callout, Exim crashed because it tried to
3021 read the rcpt_include_affixes from the non-existent transport. Now it
3022 just assumes that the setting of that option is false. This bug was
3023 introduced by 4.51/PH/31.
3025 PH/28 Changed -d+all to exclude +memory, because that information is very
3026 rarely of interest, but it makes the output a lot bigger. People tend to
3027 do -d+all out of habit.
3029 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3030 code in os-type was giving problems when libc.so lives in lib64, like on
3033 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3034 aren't the modern standard, and it seems that some systems' include files
3035 don't always have them. Exim was already checking for some of the newer
3036 ones like T_AAAA, and defining it itself. I've added checks for all the
3037 record types that Exim uses.
3039 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3040 not automatically generating a new one, as it is supposed to. This
3041 prevented TLS from working. If the file did exist, but contained invalid
3042 data, a new version was generated, as expected. It was only the case of a
3043 non-existent file that was broken.
3045 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3046 with a change in libdomainkeys > 0.64.
3048 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3049 from DNS. If the selector record carries the flag, it now has
3050 precedence over the domain-wide flag.
3052 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3054 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3055 the use of a port name that isn't defined in /etc/services) occurred, the
3056 message was deferred as in a normal delivery, and thus remained on the
3057 spool, instead of being failed because of the mua_wrapper setting. This
3058 is now fixed, and I tidied up some of the mua_wrapper messages at the
3061 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3062 the messages in a hash of arrays rather than using individual hashes.
3063 This is a bit cleaner and results in dramatic memory savings, albeit
3064 at a slight CPU cost.
3066 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3067 as requested by Marc Sherman.
3069 SC/10 Eximstats: added histograms for user specified patterns as requested
3072 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3074 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3075 fopen() in the content-scanning modules that did not already have it.
3081 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3082 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3084 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3086 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3088 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3089 to test Sieve filters that use "vacation".
3091 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3092 that changes the way the GnuTLS parameters are stored in the cache file.
3093 The new format can be generated externally. For backward compatibility,
3094 if the data in the cache doesn't make sense, Exim assumes it has read an
3095 old-format file, and it generates new data and writes a new file. This
3096 means that you can't go back to an older release without removing the
3099 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3100 work if there are any delivery delays because "one_time" forces the
3101 parent to be marked "delivered", so its unseen clone is never tried
3102 again. For this reason, Exim now forbids the simultaneous setting of
3105 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3106 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3107 entry if you want to know the details. The fix, however, neglected to
3108 consider the case when local delivery batching is involved. The test for
3109 "previously delivered" was not happening when checking to see if an
3110 address could be batched with a previous (undelivered) one; under
3111 certain circumstances this could lead to multiple deliveries to the same
3114 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3115 in its include files, and this causes problems building Exim.
3117 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3118 header_syntax) but Exim was just ignoring anything given after a slash.
3119 In particular, this caused confusion with an attempt to use "verify =
3120 reverse_host_lookup/defer_ok". An error is now given when options are
3121 supplied for verify items that do not have them. (Maybe reverse_host_
3122 lookup should have a defer_ok option, but that's a different point.)
3124 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3125 defined by RFC 821) to 2048, because there were problems with some AUTH
3126 commands, and RFC 1869 says the size should be increased for extended
3127 SMTP commands that take arguments.
3129 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3132 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3133 "unknown" error; now it says that the functionality isn't in the binary.
3135 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3136 an address' error message when a string expansion fails (syntax or
3137 whatever). Otherwise the password may appear in the log. Following change
3138 PH/42 below, there is no longer a chance of it appearing in a bounce
3141 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3143 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3144 of its addresses was ever tried. (Bugzilla bug #2.)
3146 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3147 the result incorrectly in the debug output. (It correctly added a newline
3148 to what was transported.)
3150 TF/01 Added $received_time.
3152 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3153 commented out examples of how to interface to a virus scanner and to
3154 SpamAssassin. Also added commented examples of av_scanner and
3155 spamd_address settings.
3157 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3158 and controls are allowed in which ACLs. There were a couple of minor
3159 errors. Some of the entries in the conditions table (which is a table of
3160 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3161 negation of where the condition IS allowed.
3163 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3165 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3166 header file does not have a version number, so I've had to invent a new
3167 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3168 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3169 radiusclient), but it was contributed by a Radius user.
3171 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3172 files or directories, for OpenSSL.
3174 PH/19 When an Exim process that is running as root has to create an Exim log
3175 file, it does so in a subprocess that runs as exim:exim so as to get the
3176 ownership right at creation (otherwise, other Exim processes might see
3177 the file with the wrong ownership). There was no test for failure of this
3178 fork() call, which would lead to the process getting stuck as it waited
3179 for a non-existent subprocess. Forks do occasionally fail when resources
3180 run out. I reviewed all the other calls to fork(); they all seem to check
3183 PH/20 When checking for unexpected SMTP input at connect time (before writing
3184 the banner), Exim was not dealing correctly with a non-positive return
3185 from the read() function. If the client had disconnected by this time,
3186 the result was a log entry for a synchronization error with an empty
3187 string after "input=" when read() returned zero. If read() returned -1
3188 (an event I could not check), uninitialized data bytes were printed.
3189 There were reports of junk text (parts of files, etc) appearing after
3192 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3194 PH/22 Added support for macro redefinition, and (re)definition in between
3195 driver and ACL definitions.
3197 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3198 forgetting to use the resulting value; it was using the unexpanded value.
3200 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3201 hadn't been configured. The fix is from Juergen Kreileder, who
3202 understands it better than I do:
3204 "Here's what I see happening with three configured cyrus_sasl
3205 authenticators configured (plain, login, cram-md5):
3207 On startup auth_cyrus_sasl_init() gets called for each of these.
3208 This means three calls to sasl_listmech() without a specified mech_list.
3209 => SASL tests which mechs of all available mechs actually work
3210 => three warnings about OTP not working
3211 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3213 With the patch, sasl_listmech() also gets called three times. But now
3214 SASL's mech_list option is set to the server_mech specified in the the
3215 authenticator. Or in other words, the answer from sasl_listmech()
3216 gets limited to just the mech you're testing for (which is different
3218 => the return list contains just 'plain' or 'login', 'cram-md5' or
3219 nothing depending on the value of ob->server_mech.
3221 I've just tested the patch: Authentication still works fine,
3222 unavailable mechs specified in the exim configuration are still
3223 caught, and the auth.log warnings about OTP are gone."
3225 PH/25 When debugging is enabled, the contents of the command line are added
3226 to the debugging output, even when log_selector=+arguments is not
3229 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3230 answer is "GNU", and only if the return is "GNU/something" is the answer
3233 PH/27 $acl_verify_message is now set immediately after the failure of a
3234 verification in an ACL, and so is available in subsequent modifiers. In
3235 particular, the message can be preserved by coding like this:
3237 warn !verify = sender
3238 set acl_m0 = $acl_verify_message
3240 Previously, $acl_verify_message was set only while expanding "message"
3241 and "log_message" when a very denied access.
3243 PH/28 Modified OS/os.c-Linux with
3245 -#ifndef OS_LOAD_AVERAGE
3246 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3248 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3249 nomenclature these days.)
3251 PH/29 Installed patch from the Sieve maintainer that adds the options
3252 sieve_useraddress and sieve_subaddress to the redirect router.
3254 PH/30 In these circumstances:
3255 . Two addresses routed to the same list of hosts;
3256 . First host does not offer TLS;
3257 . First host accepts first address;
3258 . First host gives temporary error to second address;
3259 . Second host offers TLS and a TLS session is established;
3260 . Second host accepts second address.
3261 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3262 and peerdn, if requested) that were in fact used only for the second
3265 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3266 attention to any local part prefix or suffix that was matched by the
3267 router that accepted the address. It now behaves in the same way as it
3268 does for delivery: the affixes are removed from the local part unless
3269 rcpt_include_affixes is set on the transport.
3271 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3272 timeout during the DATA phase of an incoming message.
3274 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3275 applied a patch sanctioned by the Sieve maintainer.
3277 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3278 the uid or gid is negative. A case of a negative gid caused this to be
3279 noticed. The fix allows for either to be negative.
3281 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3282 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3285 PH/36 The change PH/12 above was broken. Fixed it.
3287 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3288 the grounds that routing the same address twice would always produce the
3289 same answer. This might have been true once, but it is certainly no
3290 longer true now. Routing a child address may depend on the previous
3291 routing that produced that child. Some complicated redirection strategies
3292 went wrong when messages had multiple recipients, and made Exim's
3293 behaviour dependent on the order in which the addresses were given.
3295 I have moved the duplicate checking until after the routing is complete.
3296 Exim scans the addresses that are assigned to local and remote
3297 transports, and removes any duplicates. This means that more work will be
3298 done, as duplicates will always all be routed, but duplicates are
3299 presumably rare, so I don't expect this is of any significance.
3301 For deliveries to pipes, files, and autoreplies, the duplicate checking
3302 still happens during the routing process, since they are not going to be
3305 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3306 It corrects a timeout issue with spamd. This is Ian's comment: "The
3307 background is that sometimes spamd either never reads data from a
3308 connection it has accepted, or it never writes response data. The exiscan
3309 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3310 blindly assumes that writes won't block so it may never time out."
3312 PH/39 Allow G after quota size as well as K and M.
3314 PH/40 The value set for $authenticated_id in an authenticator may not contain
3315 binary zeroes or newlines because the value is written to log lines and
3316 to spool files. There was no check on this. Now the value is run through
3317 the string_printing() function so that such characters are converted to
3318 printable escape sequences.
3320 PH/41 $message_linecount is a new variable that contains the total number of
3321 lines in the message. Compare $body_linecount, which is the count for the
3324 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3325 bounce and delay warning messages, except in certain special cases, which
3328 (a) An SMTP error message from a remote host;
3329 (b) A message specified in a :fail: redirection;
3330 (c) A message specified in a "fail" command in a system filter;
3331 (d) A message specified in a FAIL return from the queryprogram router;
3332 (e) A message specified by the cannot_route_message router option.
3334 In these cases only, Exim does include the error details in bounce and
3335 warning messages. There are also a few cases where bland messages such
3336 as "unrouteable address" or "local delivery error" are given.
3338 PH/43 $value is now also set for the "else" part of a ${run expansion.
3340 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3341 being worked on, but at least Exim now implements the latest version to
3344 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3345 process to complete was treated as a delivery failure, a timeout while
3346 writing the message to the pipe was logged, but erroneously treated as a
3347 successful delivery. Such timeouts include transport filter timeouts. For
3348 consistency with the overall process timeout, these timeouts are now
3349 treated as errors, giving rise to delivery failures by default. However,
3350 there is now a new Boolean option for the pipe transport called
3351 timeout_defer, which, if set TRUE, converts the failures into defers for
3352 both kinds of timeout. A transport filter timeout is now identified in
3355 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3356 systems where "make" and "gmake" are different, calling "gmake" at top
3357 level broke things. I've arranged for the value of $(MAKE) to be passed
3358 from the Makefile to this script so that it can call the same version of
3362 A note about Exim versions 4.44 and 4.50
3363 ----------------------------------------
3365 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3366 changes of various kinds. As a consequence, a big documentation update was
3367 needed. This delayed the release for rather longer than seemed good, especially
3368 in the light of a couple of (minor) security issues. Therefore, the changes
3369 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3370 release. So 4.44 and 4.50 are in effect two different branches that both start
3373 I have left the 4.50 change log unchanged; it contains all the changes since
3374 4.43. The change log for 4.44 is below; many of its items are identical to
3375 those for 4.50. This seems to be the most sensible way to preserve the
3376 historical information.
3382 1. Minor wording change to the doc/README.SIEVE file.
3384 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3385 computation of the current number of files was incorrect.
3387 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3388 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3391 4. Give more explanation in the error message when the command for a transport
3392 filter fails to execute.
3394 5. There are several places where Exim runs a non-Exim command in a
3395 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3396 was being done only for the command run by the queryprogram router. It is
3397 now done for all such subprocesses. The other cases are: ${run, transport
3398 filters, and the commands run by the lmtp and pipe transports.
3400 6. Added CONFIGURE_GROUP build-time option.
3402 7. Some older OS have a limit of 256 on the maximum number of file
3403 descriptors. Exim was using setrlimit() to set 1000 as a large value
3404 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3405 systems. I've change it so that if it can't get 1000, it tries for 256.
3407 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3408 was an oversight, and furthermore, ever since the addition of extra
3409 controls (e.g. 4.43/32), the checks on when to allow different forms of
3410 "control" were broken. There should now be diagnostics for all cases when a
3411 control that does not make sense is encountered.
3413 9. Added the /retain_sender option to "control=submission".
3415 10. $recipients is now available in the predata ACL (oversight).
3417 11. Tidy the search cache before the fork to do a delivery from a message
3418 received from the command line. Otherwise the child will trigger a lookup
3419 failure and thereby defer the delivery if it tries to use (for example) a
3420 cached ldap connection that the parent has called unbind on.
3422 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3423 of $address_data from the recipient verification was clobbered by the
3424 sender verification.
3426 13. The value of address_data from a sender verification is now available in
3427 $sender_address_data in subsequent conditions in the ACL statement.
3429 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3431 15. Added a new option "connect=<time>" to callout options, to set a different
3434 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3435 was its contents. (It was OK if the option was not defined at all.)
3437 17. A "Completed" log line is now written for messages that are removed from
3438 the spool by the -Mrm option.
3440 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3441 information about exactly what failed.
3443 19. Added -dd to debug only the daemon process.
3445 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3446 handles timeouts, both on the server side and network timeouts. Renamed the
3447 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3449 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3450 It is now set to "smtps".
3452 22. $host_address is now set to the target address during the checking of
3453 ignore_target_hosts.
3455 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3456 being passed; this would have caused $sender_host_name to have been used if
3457 matching the list had actually called for a host name (not very likely,
3458 since this list is usually IP addresses). A host name is now passed as
3461 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3462 code that specifies a non-restarting handler (typically sigaction() in
3463 modern systems) in an attempt to fix a rare and obscure crash bug.
3465 25. Narrowed the window for a race in the daemon that could cause it to ignore
3466 SIGCHLD signals. This is not a major problem, because they are used only to
3467 wake it up if nothing else does.
3469 26. A malformed maildirsize file could cause Exim to calculate negative values
3470 for the mailbox size or file count. Odd effects could occur as a result.
3471 The maildirsize information is now recalculated if the size or filecount
3474 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3475 support for a long time. Removed HAVE_SYS_VFS_H.
3477 28. Installed the latest version of exipick from John Jetmore.
3479 29. In an address list, if the pattern was not a regular expression, an empty
3480 subject address (from a bounce message) matched only if the pattern was an
3481 empty string. Non-empty patterns were not even tested. This was the wrong
3482 because it is perfectly reasonable to use an empty address as part of a
3483 database query. An empty address is now tested by patterns that are
3484 lookups. However, all the other forms of pattern expect the subject to
3485 contain a local part and a domain, and therefore, for them, an empty
3486 address still always fails if the pattern is not itself empty.
3488 30. Exim went into a mad DNS loop when attempting to do a callout where the
3489 host was specified on an smtp transport, and looking it up yielded more
3490 than one IP address.
3492 31. Re-factored the code for checking spool and log partition space into a
3493 function that finds that data and another that does the check. The former
3494 is then used to implement four new variables: $spool_space, $log_space,
3495 $spool_inodes, and $log_inodes.
3497 32. The RFC2047 encoding function was originally intended for short strings
3498 such as real names; it was not keeping to the 75-character limit for
3499 encoded words that the RFC imposes. It now respects the limit, and
3500 generates multiple encoded words if necessary. To be on the safe side, I
3501 have increased the buffer size for the ${rfc2047: expansion operator from
3504 33. It is now permitted to omit both strings after an "if" condition; if the
3505 condition is true, the result is "true". As before, when the second string
3506 is omitted, a false condition yields an empty string. This makes it less
3507 cumbersome to write custom ACL and router conditions.
3509 34. Failure to deliver a bounce message always caused it to be frozen, even if
3510 there was an errors_to setting on the router. The errors_to setting is now
3513 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3514 canonical form (fully expanded) before being placed in
3515 $sender_host_address.
3517 36. The table in the code that translates DNS record types into text (T_A to
3518 "A" for instance) was missing entries for NS and CNAME. It is just possible
3519 that this could have caused confusion if both these types were looked up
3520 for the same domain, because the text type is used as part of Exim's
3521 per-process caching. But the chance of anyone hitting this buglet seems
3524 37. The dnsdb lookup has been extended in a number of ways.
3526 (1) There is a new type, "zns", which walks up the domain tree until it
3527 finds some nameserver records. It should be used with care.
3529 (2) There is a new type, "mxh", which is like "mx" except that it returns
3530 just the host names, not the priorities.
3532 (3) It is now possible to give a list of domains (or IP addresses) to be
3533 looked up. The behaviour when one of the lookups defers can be
3534 controlled by a keyword.
3536 (4) It is now possible to specify the separator character for use when
3537 multiple records are returned.
3539 38. The dnslists ACL condition has been extended: it is now possible to supply
3540 a list of IP addresses and/or domains to be looked up in a particular DNS
3543 39. Added log_selector=+queue_time_overall.
3545 40. When running the queue in the test harness, wait just a tad after forking a
3546 delivery process, to get repeatability of debugging output.
3548 41. Include certificate and key file names in error message when GnuTLS fails
3549 to set them up, because the GnuTLS error message doesn't include the name
3550 of the failing file when there is a problem reading it.
3552 42. Allow both -bf and -bF in the same test run.
3554 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3556 44. The "Exiscan patch" is now merged into the mainline Exim source.
3558 45. Sometimes the final signoff response after QUIT could fail to get
3559 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3560 < 0 before doing a fflush(). This bug looks as though it goes back to the
3561 introduction of TLS in release 3.20, but "sometimes" must have been rare
3562 because the tests only now provoked it.
3564 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3565 (this can affect the format of dates).
3567 47. exim_tidydb, when checking for the continued existence of a message for
3568 which it has found a message-specific retry record, was not finding
3569 messages that were in split spool directories. Consequently, it was
3570 deleting retry records that should have stayed in existence.
3572 48. Steve fixed some bugs in eximstats.
3574 49. The SPA authentication driver was not abandoning authentication and moving
3575 on to the next authenticator when an expansion was forced to fail,
3576 contradicting the general specification for all authenticators. Instead it
3577 was generating a temporary error. It now behaves as specified.
3579 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3580 (the order specifies the preference for clients). The order is now AES256,
3581 AES128, 3DES, ARCFOUR128.
3583 51. Small patch to Sieve code - explicitly set From: when generating an
3586 52. Exim crashed if a remote delivery caused a very long error message to be
3587 recorded - for instance if somebody sent an entire SpamAssassin report back
3588 as a large number of 550 error lines. This bug was coincidentally fixed by
3589 increasing the size of one of Exim's internal buffers (big_buffer) that
3590 happened as part of the Exiscan merge. However, to be on the safe side, I
3591 have made the code more robust (and fixed the comments that describe what
3594 53. Now that there can be additional text after "Completed" in log lines (if
3595 the queue_time_overall log selector is set), a one-byte patch to exigrep
3596 was needed to allow it to recognize "Completed" as not the last thing in
3599 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3600 patch that reportedly fixes this has been added. I am not expert enough to
3601 create a test for it. This is what the patch creator wrote:
3603 "I found a little strange behaviour of ldap code when working with
3604 Windows 2003 AD Domain, where users was placed in more than one
3605 Organization Units. When I tried to give exim partial DN, the exit code
3606 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3607 But simultaneously result of request was absolutely normal ldap result,
3608 so I produce this patch..."
3610 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3611 REFERENCE, so I have modified the code to exclude the patch when that macro
3614 55. Some experimental protocols are using DNS PTR records for new purposes. The
3615 keys for these records are domain names, not reversed IP addresses. The
3616 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3617 leaves it alone. Component reversal etc. now happens only for IP addresses.
3620 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3622 57. Double the size of the debug message buffer (to 2048) so that more of very
3623 long debug lines gets shown.
3625 58. The exicyclog utility now does better if the number of log files to keep
3626 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3628 59. Two changes related to the smtp_active_hostname option:
3630 (1) $smtp_active_hostname is now available as a variable.
3631 (2) The default for smtp_banner uses $smtp_active_hostname instead
3632 of $primary_hostname.
3634 60. The host_aton() function is supposed to be passed a string that is known
3635 to be a valid IP address. However, in the case of IPv6 addresses, it was
3636 not checking this. This is a hostage to fortune. Exim now panics and dies
3637 if the condition is not met. A case was found where this could be provoked
3638 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3639 components; fortuitously, this particular loophole had already been fixed
3640 by change 4.50/55 above.
3642 If there are any other similar loopholes, the new check in host_aton()
3643 itself should stop them being exploited. The report I received stated that
3644 data on the command line could provoke the exploit when Exim was running as
3645 exim, but did not say which command line option was involved. All I could
3646 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3647 running as the user.
3650 61. There was a buffer overflow vulnerability in the SPA authentication code
3651 (which came originally from the Samba project). I have added a test to the
3652 spa_base64_to_bits() function which I hope fixes it.
3655 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3656 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3658 63. The daemon start-up calls getloadavg() while still root for those OS that
3659 need the first call to be done as root, but it missed one case: when
3660 deliver_queue_load_max is set with deliver_drop_privilege. This is
3661 necessary for the benefit of the queue runner, because there is no re-exec
3662 when deliver_drop_privilege is set.
3664 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3665 This has been fixed.
3667 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3668 was in use, was not putting the data itself into the right store pool;
3669 consequently, it could be overwritten for a subsequent message in the same
3670 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3673 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3675 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3676 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3677 the function were treating the return as a boolean value, which happened to
3678 work because 0=false and not-0=true, but is not correct code.
3680 68. The host_aton() function was not handling scoped IPv6 addresses (those
3681 with, for example, "%eth0" on the end) correctly.
3683 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3684 negated items (that is, ~something) in unsigned ints. Some compilers
3685 apparently mutter when there is no cast.
3687 70. If an address verification called from an ACL failed, and did not produce a
3688 user-specific message (i.e. there was only a "system" message), nothing was
3689 put in $acl_verify_message. In this situation, it now puts the system
3692 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3693 session; change 4.31/43 added the unwanted input to the log line - except
3694 that it did not do this in the start of session case. It now does.
3696 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3697 This is wrong and can cause the other end to generate a synchronization
3698 error if it is another Exim or anything else that does the synchronization
3699 check. A QUIT command is no longer sent after a timeout.
3701 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3702 during host lookups.
3704 74. The defer_ok option of callout verification was not working if it was used
3705 when verifying addresses in header lines, that is, for this case:
3707 verify = header_sender/callout=defer_ok
3709 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3710 those file descriptors could be used for SMTP connections. If anything
3711 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3712 could be sent to the SMTP client, causing chaos. The daemon now opens
3713 stdin, stdout, and stderr to /dev/null when it puts itself into the
3716 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3717 log by default. The user can override this with suitable Perl magic.
3719 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3720 the log message when discard triggers, was not working for the DATA ACL or
3721 for the non-SMTP ACL.
3723 78. Error message wording change in sieve.c.
3725 79. If smtp_accept_max_per_host was set, the number of connections could be
3726 restricted to fewer than expected, because the daemon was trying to set up
3727 a new connection before checking whether the processes handling previous
3728 connections had finished. The check for completed processes is now done
3729 earlier. On busy systems, this bug wouldn't be noticed because something
3730 else would have woken the daemon, and it would have reaped the completed
3733 80. If a message was submitted locally by a user whose login name contained one
3734 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3735 It caused a spool format error. I have fixed the spool reading code. A
3736 related problem was that the "from" clause in the Received: line became
3737 illegal because of the space(s). It is now covered by ${quote_local_part.
3739 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3742 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3744 83. Patch from Sieve maintainer to fix unterminated string problem in
3745 "vacation" handling.
3747 84. Some minor changes to the Linux configuration files to help with other
3748 OS variants using glibc.
3750 85. One more patch for Sieve to update vacation handling to latest spec.
3753 ----------------------------------------------------
3754 See the note above about the 4.44 and 4.50 releases.
3755 ----------------------------------------------------
3761 1. Change 4.43/35 introduced a bug that caused file counts to be
3762 incorrectly computed when quota_filecount was set in an appendfile
3765 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3766 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3769 3. Give more explanation in the error message when the command for a transport
3770 filter fails to execute.
3772 4. There are several places where Exim runs a non-Exim command in a
3773 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3774 was being done only for the command run by the queryprogram router. It is
3775 now done for all such subprocesses. The other cases are: ${run, transport
3776 filters, and the commands run by the lmtp and pipe transports.
3778 5. Some older OS have a limit of 256 on the maximum number of file
3779 descriptors. Exim was using setrlimit() to set 1000 as a large value
3780 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3781 systems. I've change it so that if it can't get 1000, it tries for 256.
3783 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3784 was an oversight, and furthermore, ever since the addition of extra
3785 controls (e.g. 4.43/32), the checks on when to allow different forms of
3786 "control" were broken. There should now be diagnostics for all cases when a
3787 control that does not make sense is encountered.
3789 7. $recipients is now available in the predata ACL (oversight).
3791 8. Tidy the search cache before the fork to do a delivery from a message
3792 received from the command line. Otherwise the child will trigger a lookup
3793 failure and thereby defer the delivery if it tries to use (for example) a
3794 cached ldap connection that the parent has called unbind on.
3796 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3797 of $address_data from the recipient verification was clobbered by the
3798 sender verification.
3800 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3801 was its contents. (It was OK if the option was not defined at all.)
3803 11. A "Completed" log line is now written for messages that are removed from
3804 the spool by the -Mrm option.
3806 12. $host_address is now set to the target address during the checking of
3807 ignore_target_hosts.
3809 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3810 being passed; this would have caused $sender_host_name to have been used if
3811 matching the list had actually called for a host name (not very likely,
3812 since this list is usually IP addresses). A host name is now passed as
3815 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3816 code that specifies a non-restarting handler (typically sigaction() in
3817 modern systems) in an attempt to fix a rare and obscure crash bug.
3819 15. Narrowed the window for a race in the daemon that could cause it to ignore
3820 SIGCHLD signals. This is not a major problem, because they are used only to
3821 wake it up if nothing else does.
3823 16. A malformed maildirsize file could cause Exim to calculate negative values
3824 for the mailbox size or file count. Odd effects could occur as a result.
3825 The maildirsize information is now recalculated if the size or filecount
3828 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3829 support for a long time. Removed HAVE_SYS_VFS_H.
3831 18. Updated exipick to current release from John Jetmore.
3833 19. Allow an empty sender to be matched against a lookup in an address list.
3834 Previously the only cases considered were a regular expression, or an
3837 20. Exim went into a mad DNS lookup loop when doing a callout where the
3838 host was specified on the transport, if the DNS lookup yielded more than
3841 21. The RFC2047 encoding function was originally intended for short strings
3842 such as real names; it was not keeping to the 75-character limit for
3843 encoded words that the RFC imposes. It now respects the limit, and
3844 generates multiple encoded words if necessary. To be on the safe side, I
3845 have increased the buffer size for the ${rfc2047: expansion operator from
3848 22. Failure to deliver a bounce message always caused it to be frozen, even if
3849 there was an errors_to setting on the router. The errors_to setting is now
3852 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3853 canonical form (fully expanded) before being placed in
3854 $sender_host_address.
3856 24. Updated eximstats to version 1.33
3858 25. Include certificate and key file names in error message when GnuTLS fails
3859 to set them up, because the GnuTLS error message doesn't include the name
3860 of the failing file when there is a problem reading it.
3862 26. Expand error message when OpenSSL has problems setting up cert/key files.
3865 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3866 (this can affect the format of dates).
3868 28. exim_tidydb, when checking for the continued existence of a message for
3869 which it has found a message-specific retry record, was not finding
3870 messages that were in split spool directories. Consequently, it was
3871 deleting retry records that should have stayed in existence.
3873 29. eximstats updated to version 1.35
3874 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3875 1.35 - bugfix such that pie charts by volume are generated correctly
3877 30. The SPA authentication driver was not abandoning authentication and moving
3878 on to the next authenticator when an expansion was forced to fail,
3879 contradicting the general specification for all authenticators. Instead it
3880 was generating a temporary error. It now behaves as specified.
3882 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3883 (the order specifies the preference for clients). The order is now AES256,
3884 AES128, 3DES, ARCFOUR128.
3886 31. Small patch to Sieve code - explicitly set From: when generating an
3889 32. Exim crashed if a remote delivery caused a very long error message to be
3890 recorded - for instance if somebody sent an entire SpamAssassin report back
3891 as a large number of 550 error lines. This bug was coincidentally fixed by
3892 increasing the size of one of Exim's internal buffers (big_buffer) that
3893 happened as part of the Exiscan merge. However, to be on the safe side, I
3894 have made the code more robust (and fixed the comments that describe what
3897 33. Some experimental protocols are using DNS PTR records for new purposes. The
3898 keys for these records are domain names, not reversed IP addresses. The
3899 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3900 leaves it alone. Component reversal etc. now happens only for IP addresses.
3903 34. The host_aton() function is supposed to be passed a string that is known
3904 to be a valid IP address. However, in the case of IPv6 addresses, it was
3905 not checking this. This is a hostage to fortune. Exim now panics and dies
3906 if the condition is not met. A case was found where this could be provoked
3907 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3908 components; fortuitously, this particular loophole had already been fixed
3909 by change 4.50/55 or 4.44/33 above.
3911 If there are any other similar loopholes, the new check in host_aton()
3912 itself should stop them being exploited. The report I received stated that
3913 data on the command line could provoke the exploit when Exim was running as
3914 exim, but did not say which command line option was involved. All I could
3915 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3916 running as the user.
3919 35. There was a buffer overflow vulnerability in the SPA authentication code
3920 (which came originally from the Samba project). I have added a test to the
3921 spa_base64_to_bits() function which I hope fixes it.
3924 36. The daemon start-up calls getloadavg() while still root for those OS that
3925 need the first call to be done as root, but it missed one case: when
3926 deliver_queue_load_max is set with deliver_drop_privilege. This is
3927 necessary for the benefit of the queue runner, because there is no re-exec
3928 when deliver_drop_privilege is set.
3930 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3931 was in use, was not putting the data itself into the right store pool;
3932 consequently, it could be overwritten for a subsequent message in the same
3933 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3936 38. Sometimes the final signoff response after QUIT could fail to get
3937 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3938 < 0 before doing a fflush(). This bug looks as though it goes back to the
3939 introduction of TLS in release 3.20, but "sometimes" must have been rare
3940 because the tests only now provoked it.
3946 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3947 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3948 Now it returns an integer. A number of calls were still expecting a T/F
3949 return. Fortuitously, in all cases, the tests worked in OK situations,
3950 which is the norm. However, things would have gone wrong on any write
3951 failures on the smtp file descriptor. This function is used when sending
3952 messages over SMTP and also when doing verify callouts.
3954 2. When Exim is called to do synchronous delivery of a locally submitted
3955 message (the -odf or -odi options), it no longer closes stderr before doing
3958 3. Implemented the mua_wrapper option.
3960 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3962 5. Implemented the functions header_remove(), header_testname(),
3963 header_add_at_position(), and receive_remove_recipient(), and exported them
3966 6. If an ACL "warn" statement specified the addition of headers, Exim already
3967 inserted X-ACL-Warn: at the start if there was no header name. However, it
3968 was not making this test for the second and subsequent header lines if
3969 there were newlines in the string. This meant that an invalid header could
3970 be inserted if Exim was badly configured.
3972 7. Allow an ACL "warn" statement to add header lines at the start or after all
3973 the Received: headers, as well as at the end.
3975 8. Added the rcpt_4xx retry error code.
3977 9. Added postmaster_mailfrom=xxx to callout verification option.
3979 10. Added mailfrom=xxxx to the callout verification option, for verify=
3982 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3983 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3985 12. Inserted some casts to stop certain compilers warning when using pointer
3986 differences as field lengths or precisions in printf-type calls (mostly
3987 affecting debugging statements).
3989 13. Added optional readline() support for -be (dynamically loaded).
3991 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3992 same clock tick as a message's arrival, so that its received time was the
3993 same as the "first fail" time on the retry record, and that message
3994 remained on the queue past the ultimate address timeout, every queue runner
3995 would try a delivery (because it was past the ultimate address timeout) but
3996 after another failure, the ultimate address timeout, which should have then
3997 bounced the address, did not kick in. This was a "< instead of <=" error;
3998 in most cases the first failure would have been in the next clock tick
3999 after the received time, and all would be well.
4001 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4002 being recognized when the domain list was tested by the match_domain
4003 condition in an expansion string.
4005 16. Added the ${str2b64: operator.
4007 17. Exim was always calling setrlimit() to set a large limit for the number of
4008 processes, without checking whether the existing limit was already
4009 adequate. (It did check for the limit on file descriptors.) Furthermore,
4010 errors from getrlimit() and setrlimit() were being ignored. Now they are
4011 logged to the main and panic logs, but Exim does carry on, to try to do its
4012 job under whatever limits there are.
4014 18. Imported PCRE 5.0.
4016 19. Trivial typo in log message " temporarily refused connection" (the leading
4019 20. If the log selector return_path_on_delivery was set and an address was
4020 redirected to /dev/null, the delivery process crashed because it assumed
4021 that a return path would always be set for a "successful" delivery. In this
4022 case, the whole delivery is bypassed as an optimization, and therefore no
4025 21. Internal re-arrangement: the function for sending a challenge and reading
4026 a response while authentication was assuming a zero-terminated challenge
4027 string. It's now changed to take a pointer and a length, to allow for
4028 binary data in such strings.
4030 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4032 23. Exim was not respecting finduser_retries when seeking the login of the
4033 uid under which it was called; it was always trying 10 times. (The default
4034 setting of finduser_retries is zero.) Also, it was sleeping after the final
4035 failure, which is pointless.
4037 24. Implemented tls_on_connect_ports.
4039 25. Implemented acl_smtp_predata.
4041 26. If the domain in control=submission is set empty, Exim assumes that the
4042 authenticated id is a complete email address when it generates From: or
4043 Sender: header lines.
4045 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4046 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4047 chown and chgrp in /bin and hostname in /usr/bin.
4049 28. Exim was keeping the "process log" file open after each use, just as it
4050 does for the main log. This opens the possibility of it remaining open for
4051 long periods when the USR1 signal hits a daemon. Occasional processlog
4052 errors were reported, that could have been caused by this. Anyway, it seems
4053 much more sensible not to leave this file open at all, so that is what now
4056 29. The long-running daemon process does not normally write to the log once it
4057 has entered its main loop, and it closes the log before doing so. This is
4058 so that log files can straightforwardly be renamed and moved. However,
4059 there are a couple of unusual error situations where the daemon does write
4060 log entries, and I had neglected to close the log afterwards.
4062 30. The text of an SMTP error response that was received during a remote
4063 delivery was being truncated at 512 bytes. This is too short for some of
4064 the long messages that one sometimes sees. I've increased the limit to
4067 31. It is now possible to make retry rules that apply only when a message has a
4068 specific sender, in particular, an empty sender.
4070 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4071 it possible to be selective about when SMTP synchronization is enforced.
4073 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4075 32. Implemented hosts_connection_nolog.
4077 33. Added an ACL for QUIT.
4079 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4082 35. Added mailbox_size and mailbox_filecount to appendfile.
4084 36. Added control = no_multiline_responses to ACLs.
4086 37. There was a bug in the logic of the code that waits for the clock to tick
4087 in the case where the clock went backwards by a substantial amount such
4088 that the microsecond fraction of "now" was more than the microsecond
4089 fraction of "then" (but the whole seconds number was less).
4091 38. Added support for the libradius Radius client library this is found on
4092 FreeBSD (previously only the radiusclient library was supported).
4098 1. When certain lookups returned multiple values in the form name=value, the
4099 quoting of the values was not always being done properly. Specifically:
4100 (a) If the value started with a double quote, but contained no whitespace,
4102 (b) If the value contained whitespace other than a space character (i.e.
4103 tabs or newlines or carriage returns) it was not quoted.
4104 This fix has been applied to the mysql and pgsql lookups by writing a
4105 separate quoting function and calling it from the lookup code. The fix
4106 should probably also be applied to nisplus, ibase and oracle lookups, but
4107 since I cannot test any of those, I have not disturbed their existing code.
4109 2. A hit in the callout cache for a specific address caused a log line with no
4110 reason for rejecting RCPT. Now it says "Previous (cached) callout
4111 verification failure".
4113 3. There was an off-by-one bug in the queryprogram router. An over-long
4114 return line was truncated at 256 instead of 255 characters, thereby
4115 overflowing its buffer with the terminating zero. As well as fixing this, I
4116 have increased the buffer size to 1024 (and made a note to document this).
4118 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4119 when Exim is waiting for an SMTP response from a remote server, Exim
4120 restarts its select() call on the socket, thereby resetting its timeout.
4121 This is not a problem when such interrupts are rare. Somebody set up a cron
4122 job to run exiwhat every 2 minutes, which is less than the normal select()
4123 timeout (5 or 10 minutes). This meant that the select() timeout never
4124 kicked in because it was always reset. I have fixed this by comparing the
4125 time when an interrupt arrives with the time at the start of the first call
4126 to select(). If more time than the timeout has elapsed, the interrupt is
4127 treated as a timeout.
4129 5. Some internal re-factoring in preparation for the addition of Sieve
4130 extensions (by MH). In particular, the "personal" test is moved to a
4131 separate function, and given an option for scanning Cc: and Bcc: (which is
4132 not set for Exim filters).
4134 6. When Exim created an email address using the login of the caller as the
4135 local part (e.g. when creating a From: or Sender: header line), it was not
4136 quoting the local part when it contained special characters such as @.
4138 7. Installed new OpenBSD configuration files.
4140 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4141 try to make them clearer.
4143 9. Callout options, other than the timeout value, were being ignored when
4144 verifying sender addresses in header lines. For example, when using
4146 verify = header_sender/callout=no_cache
4148 the cache was (incorrectly) being used.
4150 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4151 only the Cygwin environment.
4153 11. When return_path_on_delivery was set as a log selector, if different remote
4154 addresses in the same message used different return paths and parallel
4155 remote delivery occurred, the wrong values would sometimes be logged.
4156 (Whenever a remote delivery process finished, the return path value from
4157 the most recently started remote delivery process was logged.)
4159 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4160 lines when AUTH and/or TLS are in use. This is the "received protocol"
4161 field. Exim used to use "asmtp" for authenticated SMTP, without any
4162 indication (in the protocol name) for TLS use. Now it follows the RFC and
4163 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4164 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4165 names appear in log lines as well as in Received: header lines.
4167 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4168 extensions, and comparison tests, and to fix some bugs.
4170 14. Changes to the "personal" filter test:
4172 (1) The test was buggy in that it was just doing the equivalent of
4173 "contains" tests on header lines. For example, if a user's address was
4174 anne@some.where, the "personal" test would incorrectly be true for
4176 To: susanne@some.where
4178 This test is now done by extracting each address from the header in turn,
4179 and checking the entire address. Other tests that are part of "personal"
4180 are now done using regular expressions (for example, to check local parts
4181 of addresses in From: header lines).
4183 (2) The list of non-personal local parts in From: addresses has been
4184 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4185 taken from the Sieve specification recommendations.
4187 (3) If the message contains any header line starting with "List-" it is
4188 treated as non-personal.
4190 (4) The test for "circular" in the Subject: header line has been removed
4191 because it now seems ill-conceived.
4193 15. Minor typos in src/EDITME comments corrected.
4195 16. Installed latest exipick from John Jetmore.
4197 17. If headers_add on a router specified a text string that was too long for
4198 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4199 of string_sprintf() is now avoided.
4201 18. $message_body_size was not set (it was always zero) when running the DATA
4202 ACL and the local_scan() function.
4204 19. For the "mail" command in an Exim filter, no default was being set for
4205 the once_repeat time, causing a random time value to be used if "once" was
4206 specified. (If the value happened to be <= 0, no repeat happened.) The
4207 default is now 0s, meaning "never repeat". The "vacation" command was OK
4208 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4209 (I found it when inspecting the code).
4211 20. There is now an overall timeout for performing a callout verification. It
4212 defaults to 4 times the callout timeout, which applies to individual SMTP
4213 commands during the callout. The overall timeout applies when there is more
4214 than one host that can be tried. The timeout is checked before trying the
4215 next host. This prevents very long delays if there are a large number of
4216 hosts and all are timing out (e.g. when the network connections are timing
4217 out). The value of the overall timeout can be changed by specifying an
4218 additional sub-option for "callout", called "maxwait". For example:
4220 verify = sender/callout=5s,maxwait=20s
4222 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4223 to the end before writing, but this should make it even safer).
4225 22. Exim was forgetting that it had advertised PIPELINING for the second and
4226 subsequent messages on an SMTP connection. It was also not resetting its
4227 memory on STARTTLS and an internal HELO.
4229 23. When Exim logs an SMTP synchronization error within a session, it now
4230 records whether PIPELINING has been advertised or not.
4232 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4233 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4234 rather than long int.
4236 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4238 26. Added the never_mail option to autoreply.
4244 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4245 crash if the getsockname() call failed; this can happen if a connection is
4246 closed very soon after it is established. The problem was simply in the
4247 order in which certain operations were done, causing Exim to try to write
4248 to the SMTP stream before it had set up the file descriptor. The bug has
4249 been fixed by making things happen in the correct order.
4255 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4256 before the connection was closed, thus losing the rejection response.
4258 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4259 some early Solaris releases, but causes trouble in current releases where
4260 socklen_t is defined.
4262 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4265 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4268 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4269 that releasing the top of it at the end releases what was used for sub-
4270 expansions (unless the block got too big). However, discard this block if
4271 the first thing is a variable or header, so that we can use its block when
4272 it is dynamic (useful for very large $message_headers, for example).
4274 6. Lookups now cache *every* query, not just the most recent. A new, separate
4275 store pool is used for this. It can be recovered when all lookup caches are
4276 flushed. Lookups now release memory at the end of their result strings.
4277 This has involved some general refactoring of the lookup sources.
4279 7. Some code has been added to the store_xxx() functions to reduce the amount
4280 of flapping under certain conditions.
4282 8. log_incoming_interface used to affect only the <= reception log lines. Now
4283 it causes the local interface and port to be added to several more SMTP log
4284 lines, for example "SMTP connection from", and rejection lines.
4286 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4288 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4290 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4291 could be overwritten at the end of the current message (or the start of a
4292 new message if it was set in a HELO ACL). The value is now preserved for
4293 the duration of the SMTP connection.
4295 12. If a transport had a headers_rewrite setting, and a matching header line
4296 contained an unqualified address, that address was qualified, even if it
4297 did not match any rewriting rules. The underlying bug was that the values
4298 of the flags that permit the existence of unqualified sender and recipient
4299 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4300 non-local messages, and by -bnq for local messages) were not being
4301 preserved with the message after it was received.
4303 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4304 "next input=" as part of the text comprising the host identity instead of
4305 the correct text. The code was using the same buffer for two different
4306 strings. However, depending on which order the printing function evaluated
4307 its arguments, the bug did not always show up. Under Linux, for example, my
4308 test suite worked just fine.
4310 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4311 doesn't work with some older versions of Perl. It has been changed to "my",
4312 which in any case is probably the better facility to use.
4314 15. A really picky compiler found some instances of statements for creating
4315 error messages that either had too many or two few arguments for the format
4318 16. The size of the buffer for calls to the DNS resolver has been increased
4319 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4320 for addresses that have a lot of PTR records. This alleviates a problem; it
4321 does not fully solve it.
4323 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4324 buffer now truncates the list and logs the incident, which is the same
4325 action as happens when Exim is looking up a host name and its aliases.
4326 Previously in this situation something unpredictable would happen;
4327 sometimes it was "internal error: store_reset failed".
4329 18. If a server dropped the connection unexpectedly when an Exim client was
4330 using GnuTLS and trying to read a response, the client delivery process
4331 crashed while trying to generate an error log message.
4333 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4334 string, for example:
4336 warn message = H1: something\nH2: something
4338 the text was added as a single header line from Exim's point of view
4339 though it ended up OK in the delivered message. However, searching for the
4340 second and subsequent header lines using $h_h2: did not work. This has been
4341 fixed. Similarly, if a system filter added multiple headers in this way,
4342 the routers could not see them.
4344 20. Expanded the error message when iplsearch is called with an invalid key to
4345 suggest using net-iplsearch in a host list.
4347 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4348 ACLs are no longer actually imposed (and a message to that effect is
4351 22. If a "gecos" field in a passwd entry contained escaped characters, in
4352 particular, if it contained a \" sequence, Exim got it wrong when building
4353 a From: or a Sender: header from that name. A second bug also caused
4354 incorrect handling when an unquoted " was present following a character
4355 that needed quoting.
4357 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4358 was not being matched caselessly.
4360 24. Arranged for all hyphens in the exim.8 source to be escaped with
4363 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4364 properly. Recipient callout cache records were still being keyed to include
4365 the sender, even when use_sender was set false. This led to far more
4366 callouts that were necessary. The sender is no longer included in the key
4367 when use_sender is false.
4369 26. Added "control = submission" modifier to ACLs.
4371 27. Added the ${base62d: operator to decode base 62 numbers.
4373 28. dnsdb lookups can now access SRV records.
4375 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4376 the configuration file.
4378 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4379 (-v) mode. This makes the output for a verbose queue run more intelligible.
4381 31. Added a use_postmaster feature to recipient callouts.
4383 32. Added the $body_zerocount variable, containing the number of binary zero
4384 bytes in the message body.
4386 33. The time of last modification of the "new" subdirectory is now used as the
4387 "mailbox time last read" when there is a quota error for a maildir
4390 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4392 35. Added +ignore_unknown as a special item in host lists.
4394 36. Code for decoding IPv6 addresses in host lists is now included, even if
4395 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4396 address was recognized as an IP address, but was then not correctly decoded
4397 into binary, causing unexpected and incorrect effects when compared with
4404 1. Very minor rewording of debugging text in manualroute to say "list of
4405 hosts" instead of "hostlist".
4407 2. If verify=header_syntax was set, and a header line with an unqualified
4408 address (no domain) and a large number of spaces between the end of the
4409 name and the colon was received, the reception process suffered a buffer
4410 overflow, and (when I tested it) crashed. This was caused by some obsolete
4411 code that should have been removed. The fix is to remove it!
4413 3. When running in the test harness, delay a bit after writing a bounce
4414 message to get a bit more predictability in the log output.
4416 4. Added a call to search_tidyup() just before forking a reception process. In
4417 theory, someone could use a lookup in the expansion of smtp_accept_max_
4418 per_host which, without the tidyup, could leave open a database connection.
4420 5. Added the variables $recipient_data and $sender_data which get set from a
4421 lookup success in an ACL "recipients" or "senders" condition, or a router
4422 "senders" option, similar to $domain_data and $local_part_data.
4424 6. Moved the writing of debug_print from before to after the "senders" test
4427 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4428 problems for message scanning, either using a data ACL, or using
4429 local_scan() because the Received: header was not generated till after they
4430 were called (in order to set the time as the time of reception completion).
4431 I have revised the way this works. The header is now generated after the
4432 body is received, but before the ACL or local_scan() are called. After they
4433 are run, the timestamp in the header is updated.
4439 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4440 before starting a queue runner without re-exec. This happened only when
4441 deliver_drop_privilege was set or when the Exim user was set to root. The
4442 effect of the bug was that timeouts during subsequent deliveries caused
4443 crashes instead of being properly handled. The handler is now left at its
4444 default (and expected) setting.
4446 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4447 message, again when deliver_drop_privilege is set or Exim is run as root.
4448 The bug described in (1) was not present in this case, but the tidying up
4449 of the other signals was missing. I have made the two cases consistent.
4451 3. The ignore_target_hosts setting on a manualroute router was being ignored
4452 for hosts that were looked up using the /MX notation.
4454 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4457 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4458 operated on the sender address. After changing the $sender_address to <>
4459 for the sender address verify, Exim was re-instated it as the original
4460 (before rewriting) address, but remembering that it had rewritten it, so it
4461 wasn't rewriting it again. This bug also had the effect of breaking the
4462 sender address verification caching when the sender address was rewritten.
4464 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4465 This has been changed so that if the ip literal address matches
4466 ignore_target_hosts, the router declines.
4468 7. Added expansion conditions match_domain, match_address, and match_local_
4469 part (NOT match_host).
4471 8. The placeholder for the Received: header didn't have a length field set.
4473 9. Added code to Exim itself and to exim_lock to test for a specific race
4474 condition that could lead to file corruption when using MBX delivery. The
4475 issue is with the lockfile that is created in /tmp. If this file is removed
4476 after a process has opened it but before that process has acquired a lock,
4477 there is the potential for a second process to recreate the file and also
4478 acquire a lock. This could lead to two Exim processes writing to the file
4479 at the same time. The added code performs the same test as UW imapd; it
4480 checks after acquiring the lock that its file descriptor still refers to
4481 the same named file.
4483 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4484 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4487 11. Added the smtp_active_hostname option. If used, this will typically be made
4488 to depend on the incoming interface address. Because $interface_address is
4489 not set up until the daemon has forked a reception process, error responses
4490 that can happen earlier (such as "too many connections") no longer contain
4493 12. If an expansion in a condition on a "warn" statement fails because a lookup
4494 defers, the "warn" statement is abandoned, and the next ACL statement is
4495 processed. Previously this caused the whole ACL to be aborted.
4497 13. Added the iplsearch lookup type.
4499 14. Added ident_timeout as a log selector.
4501 15. Added tls_certificate_verified as a log selector.
4503 16. Added a global option tls_require_ciphers (compare the smtp transport
4504 option of the same name). This controls incoming TLS connections.
4506 17. I finally figured out how to make tls_require_ciphers do a similar thing
4507 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4508 before starting the TLS session.
4510 18. Tabs are now shown as \t in -bP output.
4512 19. If the log selector return_path_on_delivery was set, Exim crashed when
4513 bouncing a message because it had too many Received: header lines.
4515 20. If two routers both had headers_remove settings, and the first one included
4516 a superfluous trailing colon, the final name in the first list and the
4517 first name in the second list were incorrectly joined into one item (with a
4518 colon in the middle).
4524 1. Added -C and -D options to the exinext utility, mainly to make it easier
4525 to include in the automated testing, but these could be helpful when
4526 multiple configurations are in use.
4528 2. The exinext utility was not formatting the output nicely when there was
4529 an alternate port involved in the retry record key, nor when there was a
4530 message id as well (for retries that were specific to a specific message
4531 and a specific host). It was also confused by IPv6 addresses, because of
4532 the additional colons they contain. I have fixed the IPv4 problem, and
4533 patched it up to do a reasonable job for IPv6.
4535 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4536 delivery, the log line now contains "pipelined" if PIPELINING was used.
4538 4. An SMTP transport process used to panic and die if the bind() call to set
4539 an explicit outgoing interface failed. This has been changed; it is now
4540 treated in the same way as a connect() failure.
4542 5. A reference to $sender_host_name in the part of a conditional expansion
4543 that was being skipped was still causing a DNS lookup. This no longer
4546 6. The def: expansion condition was not recognizing references to header lines
4547 that used bh_ and bheader_.
4549 7. Added the _cache feature to named lists.
4551 8. The code for checking quota_filecount in the appendfile transport was
4552 allowing one more file than it should have been.
4554 9. For compatibility with Sendmail, the command line option
4562 and sets the incoming protocol and host name (for trusted callers). The
4563 host name and its colon can be omitted when only the protocol is to be set.
4564 Note the Exim already has two private options, -pd and -ps, that refer to
4565 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4566 "s", but I don't think that's a major issue.
4568 10. A number of refactoring changes to the code, none of which should affect
4571 (a) The number of logging options was getting close to filling up the
4572 32-bit word that was used as a bit map. I have split them into two classes:
4573 those that are passed in the argument to log_write(), and those that are
4574 only ever tested independently outside of that function. These are now in
4575 separate 32-bit words, so there is plenty of room for expansion again.
4576 There is no change in the user interface or the logging behaviour.
4578 (b) When building, for example, log lines, the code previously used a
4579 macro that called string_cat() twice, in order to add two strings. This is
4580 not really sufficiently general. Furthermore, there was one instance where
4581 it was actually wrong because one of the argument was used twice, and in
4582 one call a function was used. (As it happened, calling the function twice
4583 did not affect the overall behaviour.) The macro has been replaced by a
4584 function that can join an arbitrary number of extra strings onto a growing
4587 (c) The code for expansion conditions now uses a table and a binary chop
4588 instead of a serial search (which was left over from when there were very
4589 few conditions). Also, it now recognizes conditions like "pam" even when
4590 the relevant support is not compiled in: a suitably worded error message is
4591 given if an attempt is made to use such a condition.
4593 11. Added ${time_interval:xxxxx}.
4595 12. A bug was causing one of the ddress fields not to be passed back correctly
4596 from remote delivery subprocesses. The field in question was not being
4597 subsequently used, so this caused to problems in practice.
4599 13. Added new log selectors queue_time and deliver_time.
4601 14. Might have fixed a bug in maildirsizefile handling that threw up
4602 "unexpected character" debug warnings, and recalculated the data
4603 unnecessarily. In any case, I expanded the warning message to give more
4606 15. Added the message "Restricted characters in address" to the statements in
4607 the default ACL that block characters like @ and % in local parts.
4609 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4610 Three changes have been made:
4612 (a) There was a serious bug; a negative response to MAIL caused the whole
4613 recipient domain to be cached as invalid, thereby blocking all messages
4614 to all local parts at the same domain, from all senders. This bug has
4615 been fixed. The domain is no longer cached after a negative response to
4616 MAIL if the sender used is not empty.
4618 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4621 (c) A new callout option, "use_sender" has been added for people who want
4622 the modified behaviour.
4628 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4631 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4632 indeed breaks things for older releases.
4634 3. Added additional logging to the case where there is a problem reading data
4635 from a filter that is running in a subprocess using a pipe, in order to
4636 try to track down a specific problem.
4638 4. Testing facility fudge: when running in the test harness and attempting
4639 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4640 getting "No route to host". Convert this to a timeout.
4642 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4645 6. Some OS don't have socklen_t but use size_t instead. This affects the
4646 fifth argument of getsockopt() amongst other things. This is now
4647 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4648 can be set for individual OS. I have set it for SunOS5, OSF1, and
4649 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4650 some earlier ones do not.
4652 7. Change 4.30/15 was not doing the test caselessly.
4654 8. The standard form for an IPv6 address literal was being rejected by address
4655 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4656 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4657 this, as well as the form without the "IPv6" on the front (but only when
4658 address literals are enabled, of course).
4660 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4662 10. Exim crashed if a message with an empty sender address specified by -f
4663 encountered a router with an errors_to setting. This could be provoked only
4664 by a command such as
4668 where an empty string was supplied; "<>" did not hit this bug.
4670 11. Installed PCRE release 4.5.
4672 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4673 remained set. It is now erased.
4675 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4676 times from message ids (which are base 36 rather than the normal 62).
4678 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4679 were being counted as actual protocol errors, and logged if the log
4680 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4681 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4682 and DATA following a set of rejected RCPTs do not count as protocol errors.
4683 In other words, Exim assumes they were pipelined, though this may not
4684 actually be the case. Of course, in all cases the client gets an
4685 appropriate error code.
4687 15. If a lookup fails in an ACL condition, a message about the failure may
4688 be available; it is used if testing the ACL cannot continue, because most
4689 such messages specify what the cause of the deferral is. However, some
4690 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4691 that caused an old message to be retained and used if a later statement
4692 caused a defer, replacing the real cause of the deferral.
4694 16. If an IP address had so many PTR records that the DNS lookup buffer
4695 was not large enough to hold them, Exim could crash while trying to process
4696 the truncated data. It now detects and logs this case.
4698 17. Further to 4.21/58, another change has been made: if (and only if) the
4699 first line of a message (the first header line) ends with CRLF, a bare LF
4700 in a subsequent header line has a space inserted after it, so as not to
4701 terminate the header.
4703 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4704 unnecessarily, used atoi() instead of strtol(), and didn't check the
4705 termination when getting file sizes from file names by regex.
4707 19. Completely re-implemented the support for maildirsize files, in the light
4708 of a number of problems with the previous contributed implementation
4709 (4.30/29). In particular:
4711 . If the quota is zero, the maildirsize file is maintained, but no quota is
4714 . If the maildir directory does not exist, it is created before any attempt
4715 to write a maildirsize file.
4717 . The quota value in the file is just a cache; if the quota is changed in
4718 the transport, the new value overrides.
4720 . A regular expression is available for excluding directories from the
4723 20. The autoreply transport checks the characters in options that define the
4724 message's headers; it allows continued headers, but it was checking with
4725 isspace() after an embedded newline instead of explicitly looking for a
4728 21. If all the "regular" hosts to which an address was routed had passed their
4729 expiry times, and had not reached their retry times, the address was
4730 bounced, even if fallback hosts were defined. Now Exim should go on to try
4733 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4734 equivalent code in the SMTP transport. Some hosts send humungous responses
4735 to HELO/EHLO, more than 1024 it seems.
4737 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4738 gives compiler warnings in some environments. I've now done it "properly",
4741 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4742 (because of problems with the built-in one) was declared to return uschar *
4743 instead of char *, causing compiler failure.
4745 25. Fixed a file descriptor leak when processing alias/forward files.
4747 26. Fixed a minor format string issue in dbfn.c.
4749 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4751 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4752 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4755 29. When a sender address is verified, it is cached, to save repeating the test
4756 when there is more than one recipient in a message. However, when the
4757 verification involves a callout, it is possible for different callout
4758 options to be set for different recipients. It is too complicated to keep
4759 track of this in the cache, so now Exim always runs a verification when a
4760 callout is required, relying on the callout cache for the optimization.
4761 The overhead is duplication of the address routing, but this should not be
4764 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4765 to be verified with callout=postmaster, and the main callout worked but the
4766 postmaster check failed, the verification correctly failed. However, if a
4767 subsequent RCPT command asked for sender verification *without* the
4768 postmaster check, incorrect caching caused this verification also to fail,
4771 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4772 it was not caching the DNS options (qualify_single, search_parents) that
4773 were used when the lookup failed. A subsequent lookup with different
4774 options therefore always gave the same answer, though there were cases
4775 where it should not have. (Example: a "domains = !$mx_any" option on a
4776 dnslookup router: the "domains" option is always processed without any
4777 widening, but the router might have qualify_single set.) Now Exim uses the
4778 cached value only when the same options are set.
4780 32. Added John Jetmore's "exipick" utility to the distribution.
4782 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4783 than a timeout (e.g. a certificate is required, and is not provided), an
4784 Exim server now closes the connection immediately. Previously it waited for
4785 the client to close - but if the client is SSL, it seems that they each
4786 wait for each other, leading to a delay before one of them times out.
4788 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4789 maintained 0.8.x compatibility because I don't think many are using it, and
4790 it is clearly obsolete.
4792 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4795 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4796 client certificate was expired. A simple patch fixes this, though I don't
4797 understand the full logic of why the verify callback is called multiple
4800 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4801 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4802 which causes problems with some clients (such as the Certicom SSL Plus
4803 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4804 disables the coutermeasure allowing Eudora to connect."
4806 38. Exim was not checking that a write() to a log file succeeded. This could
4807 lead to Bad Things if a log got too big, in particular if it hit a file
4808 size limit. Exim now panics and dies if it cannot write to a log file, just
4809 as it does if it cannot open a log file.
4811 39. Modified OS/Makefile-Linux so that it now contains
4813 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4815 The two -D definitions ensure that Exim is compiled with large file
4816 support, which makes it possible to handle log files that are bigger than
4819 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4820 instance) a domain was checked against a named list that involved a lookup,
4821 causing $domain_data to be set, then another domain was checked against the
4822 same list, then the first domain was re-checked, the value of $domain_data
4823 after the final check could be wrong. In particular, if the second check
4824 failed, it could be set empty. This bug probably also applied to
4827 41. The strip_trailing_dot option was not being applied to the address given
4828 with the -f command-line option.
4830 42. The code for reading a message's header from the spool was incrementing
4831 $received_count, but never initializing it. This meant that the value was
4832 incorrect (doubled) while delivering a message in the same process in which
4833 it was received. In the most common configuration of Exim, this never
4834 happens - a fresh exec is done - but it can happen when
4835 deliver_drop_privilege is set.
4837 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4838 it now includes up to 150 characters of the unexpected data in the log
4841 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4842 and building data strings. The size of both of these buffers was 10 000
4843 bytes - far larger than anybody would *ever* want, thought I. Needless to
4844 say, somebody hit the limit. I have increased the maximum line length to
4845 20 000 and the maximum data length of concatenated lines to 100 000. I have
4846 also fixed two bugs, because there was no checking on these buffers. Tsk,
4847 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4848 buffer is too small.
4850 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4851 lsearch lookups. Now it does.
4853 46. When parsing a route_list item in a manualroute router, a fixed-length
4854 buffer was used for the list of hosts. I made this 1024 bytes long,
4855 thinking that nobody would ever have a list of hosts that long. Wrong.
4856 Somebody had a whole pile of complicated expansion conditions, and the
4857 string was silently truncated, leading to an expansion error. It turns out
4858 that it is easier to change to an unlimited length (owing to other changes
4859 that have happened since this code was originally written) than to build
4860 structure for giving a limitation error. The length of the item that
4861 expands into the list of hosts is now unlimited.
4863 47. The lsearch lookup could not handle data where the length of text line was
4864 more than 4095 characters. Such lines were truncated, leading to shortened
4865 data being returned. It should now handle lines of any length.
4867 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4868 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4871 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4872 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4873 no longer happens in comment lines. A list of replacements is now placed
4874 at the head of all of the source files, except those whose only change is
4875 to replace PERL_COMMAND in the very first #! line.
4877 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4878 messages on the queue, with a bottom-up merge sort, using code contributed
4879 by Michael Haardt. This should make operations like -bp somewhat faster on
4880 large queues. It won't affect queue runners, except when queue_run_in_order
4883 51. Installed eximstats 1.31 in the distribution.
4885 52. Added support for SRV lookups to the dnslookup router.
4887 53. If an ACL referred to $message_body or $message_body_end, the value was not
4888 reset for any messages that followed in the same SMTP session.
4890 54. The store-handling optimization for building very long strings was not
4891 differentiating between the different store pools. I don't think this
4892 actually made any difference in practice, but I've tidied it.
4894 55. While running the routers to verify a sender address, $sender_address
4895 was still set to the sender address. This is wrong, because when routing to
4896 send a bounce to the sender, it would be empty. Therefore, I have changed
4897 it so that, while verifying a sender address, $sender_address is set to <>.
4898 (There is no change to what happens when verifying a recipient address.)
4900 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4901 target A or AAAA records (if not already returned) without resetting the
4902 qualify_single or search_parents options of the DNS resolver. These are
4903 inappropriate in this case because the targets of MX and SRV records must
4904 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4905 target that, when qualified, matched something in the local domain. These
4906 two options are now turned off when doing these lookups.
4908 57. It seems that at least some releases of Reiserfs (which does not have the
4909 concept of a fixed number of inodes) returns zero and not -1 for the
4910 number of available inodes. This interacted badly with check_spool_inodes,
4911 which assumed that -1 was the "no such thing" setting. What I have done is
4912 to check that the total number of inodes is greater than zero before doing
4913 the test of how many are available.
4915 58. When a "warn" ACL statement has a log_message modifier, the message is
4916 remembered, and not repeated. This is to avoid a lot of repetition when a
4917 message has many recipients that cause the same warning to be written.
4918 Howewer, Exim was preserving the list of already written lines for an
4919 entire SMTP session, which doesn't seem right. The memory is now reset if a
4920 new message is started.
4922 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4923 debugging output unless log_rewrite was also set.
4925 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4926 of (int)(handle) when we know that handle contains (void *)(-1).
4928 61. The Exim daemon panic-logs an error return when it closes the incoming
4929 connection. However "connection reset by peer" seems to be common, and
4930 isn't really an error worthy of noting specially, so that particular error
4933 62. When Exim is trying to find all the local interfaces, it used to panic and
4934 die if the ioctl to get the interface flags failed. However, it seems that
4935 on at least one OS (Solaris 9) it is possible to have an interface that is
4936 included in the list of interfaces, but for which you get a failure error
4937 for this call. This happens when the interface is not "plumbed" into a
4938 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4939 failure of the "get flags" call assumes that the interface is down.
4941 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4942 makes life easier for people who are doing arithmetic on fields extracted
4943 from dates, where you often get leading zeros that should not be
4944 interpreted as octal.
4946 64. Added qualify_domain to the redirect router, to override the global
4949 65. If a pathologically long header line contained very many addresses (the
4950 report of this problem mentioned 10 000) and each of them was rewritten,
4951 Exim could use up a very large amount of memory. (It kept on making new
4952 copies of the header line as it rewrote, and never released the old ones.)
4953 At the expense of a bit more processing, the header rewriting function has
4954 been changed so that it no longer eats memory in this way.
4956 66. The generation of the Received: header has been moved from the time that a
4957 message starts to be received, to the time that it finishes. The timestamp
4958 in the Received: header should now be very close to that of the <= log
4959 line. There are two side-effects of this change:
4961 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4962 logged header lines no longer include the local Received: line, because
4963 it has not yet been created. The same applies to a copy of the message
4964 that is returned to a non-SMTP sender when a message is rejected.
4966 (b) When a filter file is tested using -bf, no additional Received: header
4967 is added to the test message. After some thought, I decided that this
4970 This change does not affect the value of $received_for. It is still set
4971 after address rewriting, but before local_scan() is called.
4973 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4975 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4976 gave an unhelpful panic error message, and a defer error. I have managed to
4977 change this behaviour so that it now rejects any supplied certificate,
4978 which seems right, as the list of acceptable certificates is empty.
4980 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4981 gave an unhelpful defer error. I have not managed to make this reject any
4982 supplied certificates, but the error message it gives is "no certificate
4983 supplied", which is not helpful.
4985 70. exigrep's output now also includes lines that are not associated with any
4986 message, but which match the given pattern. Implemented by a patch from
4987 Martin Sluka, which also tidied up the Perl a bit.
4989 71. Recipient callout verification, like sender verification, was using <> in
4990 the MAIL FROM command. This isn't really the right thing, since the actual
4991 sender may affect whether the remote host accepts the recipient or not. I
4992 have changed it to use the actual sender in the callout; this means that
4993 the cache record is now keyed on a recipient/sender pair, not just the
4994 recipient address. There doesn't seem to be a real danger of callout loops,
4995 since a callout by the remote host to check the sender would use <>.
4996 [SEE ABOVE: changed after hitting problems.]
4998 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4999 temporary errors. However, in the case of such a code being given after
5000 the end of a data transmission (i.e. after ".") Exim was failing to write
5001 a retry record for the message. (Yes, there was some broken host that was
5002 actually sending 8xx at this point.)
5004 73. An unknown lookup type in a host list could cause Exim to panic-die when
5005 the list was checked. (An example that provoked this was putting <; in the
5006 middle of a list instead of at the start.) If this happened during a DATA
5007 ACL check, a -D file could be left lying around. This kind of configuration
5008 error no longer causes Exim to die; instead it causes a defer errror. The
5009 incident is still logged to the main and panic logs.
5011 74. Buglet left over from Exim 3 conversion. The message "too many messages
5012 in one connection" was written to the rejectlog but not the mainlog, except
5013 when address rewriting (yes!) was being logged.
5015 75. Added write_rejectlog option.
5017 76. When a system filter was run not as root (that is, when system_filter_user
5018 was set), the values of the $n variables were not being returned to the
5019 main process; thus, they were not subsequently available in the $sn
5022 77. Added +return_path_on_delivery log selector.
5024 78. A connection timeout was being treated differently from recipients deferred
5025 when testing hosts_max_try with a message that was older than the host's
5026 retry timeout. (The host should not be counted, thus allowing all hosts to
5027 be tried at least once before bouncing.) This may have been the cause of an
5028 occasionally reported bug whereby a message would remain on the queue
5029 longer than the retry timeout, but would be bounced if a delivery was
5030 forced. I say "may" because I never totally pinned down the problem;
5031 setting up timeout/retry tests is difficult. See also the next item.
5033 79. The ultimate address timeout was not being applied to errors that involved
5034 a combination of host plus message (for example, a timeout on a MAIL
5035 command). When an address resolved to a number of possible hosts, and they
5036 were not all tried for each delivery (e.g. because of hosts_max_try), a
5037 message could remain on the queue longer than the retry timeout.
5039 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5042 81. Fixed an obscure SMTP outgoing bug which required at least the following
5043 conditions: (a) there was another message waiting for the same server;
5044 (b) the server returned 5xx to all RCPT commands in the first message so
5045 that the message was not completed; (c) the server dropped the connection
5046 or gave a negative response to the RSET that Exim sends to abort the
5047 transaction. The observed case was a dropped connection after DATA that had
5048 been sent in pipelining mode. That is, the server had advertised PIPELINING
5049 but was not implementing it correctly. The effect of the bug was incorrect
5050 behaviour, such as trying another host, and this could lead to a crash.
5056 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5057 and daemon.c were passed as pointers to ints; they should have been
5058 pointers to socklen_t variables (which are typically unsigned ints).
5060 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5063 3. Fixed a really odd bug that affected only the testing scheme; patching a
5064 certain fixed string in the binary changed the value of another string that
5065 happened to be identical to the end of the original first string.
5067 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5068 name", it returns that address as the IP address. On some operating
5069 systems (e.g. Solaris), it also passes back the IP address string as the
5070 "host name". However, on others (e.g. Linux), it passes back an empty
5071 string. Exim wasn't checking for this, and was changing the host name to an
5072 empty string, assuming it had been canonicized.
5074 5. Although rare, it is permitted to have more than one PTR record for a given
5075 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5076 all the names associated with an address, because they do in Solaris.
5077 However, it seems that they do not in Linux for data that comes from the
5078 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5079 I found this out when I moved to a new Linux workstation and tried to run
5080 the Exim test suite.
5082 To get round this problem I have changed the code so that it now does its
5083 own call to the DNS to look up PTR records when searching for a host name.
5084 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5085 addresses that are only in /etc/hosts are still found.
5087 This behaviour is, however, controlled by an option called host_lookup_
5088 order, which defaults to "bydns:byaddr". If people want to use the other
5089 order, or indeed, just use one or the other means of lookup, they can
5090 specify it in this variable.
5092 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5093 some operating systems, this comes back from gethostbyaddr() as an empty
5094 string, and this is what Exim used to test for. However, it seems that in
5095 other systems, "." is yielded. Exim now tests for this case too.
5097 7. The values of check_spool_space and check_log_space are now held internally
5098 as a number of kilobytes instead of an absolute number of bytes. If a
5099 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5100 kilobyte. This means that much larger values can be stored.
5102 8. Exim monitor: an attempt to get the action menu when not actually pointing
5103 at a message produces an empty menu entitled "No message selected". This
5104 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5105 no entries in it ("Shell widget menu has zero width and/or height"). So I
5106 have added a single, blank menu entry in this case.
5108 9. Added ${quote_local_part.
5110 10. MIME decoding is now applied to the contents of Subject: header lines when
5113 11. Now that a reference to $sender_host_address automatically causes a reverse
5114 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5115 host lookup before query-style lookups in lists that might use this
5116 variable. This has therefore been abolished, and the "net-" prefix is no
5117 longer necessary for query-style lookups.
5119 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5120 have been a typo for LFLAGS, so it has been changed.
5122 13. The install script calls Exim with "-C /dev/null" in order to find the
5123 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5124 to be output. Howeve, since Exim outputs its version number before the
5125 error, it didn't break the script. It just looked ugly. I fixed this by
5126 always allowing "-C /dev/null" if the caller is root.
5128 14. Ignore overlarge ACL variable number when reading spool file - insurance
5129 against a later release with more variables having written the file.
5131 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5132 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5133 this, as well as the form without the "IPv6" on the front.
5135 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5136 OS/Makefile-Darwin file.
5138 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5139 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5141 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5142 variable. However, some OS use ints for inodes. I've added cast to long int
5143 to get rid of the compiler warning.
5145 19. I had forgotten to lock out "/../" in configuration file names when
5146 ALT_CONFIG_PREFIX was set.
5148 20. Routers used for verification do not need to specify transports. However,
5149 if such a router generated a host list, and callout was configured, Exim
5150 crashed, because it could not find a port number from the (non-existent)
5151 transport. It now assumes port 25 in this circumstance.
5153 21. Added the -t option to exigrep.
5155 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5156 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5157 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5158 them is set, LOOKUP_LSEARCH is forced.
5160 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5162 24. Added sender and host information to the "rejected by local_scan()" log
5163 line; previously there was no indication of these.
5165 25. Added .include_if_exists.
5167 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5168 Linux. It turns out that not all file systems support this. Apparently some
5169 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5170 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5171 not supported on the file descriptor, is now ignored when Exim is trying to
5172 sync a directory. This applies only to Linux.
5174 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5176 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5177 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5180 29. Added support for maildirsize files from supplied patch (modified a bit).
5182 30. The use of :fail: followed by an empty string could lead Exim to respond to
5183 sender verification failures with (e.g.):
5185 550 Verification failed for <xxx>
5186 550 Sender verify failed
5188 where the first response line was missing the '-' that indicates it is not
5189 the final line of the response.
5191 31. The loop for finding the name of the user that called Exim had a hardwired
5192 limit of 10; it now uses the value of finduser_retries, which is used for
5193 all other user lookups.
5195 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5198 33. Exim was neglecting to zero errno before one call of strtol() when
5199 expanding a string and expecting an integer value. On some systems this
5200 resulted in spurious "integer overflow" errors. Also, it was casting the
5201 result into an int without checking.
5203 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5204 did not work. The code looks as if it has *never* worked, though it appears
5205 to have been documented since at least releast 1.62. I have made it work.
5207 35. The "timeout_DNS" error in retry rules, also documented since at least
5208 1.62, also never worked. As it isn't clear exactly what this means, and
5209 clearly it isn't a major issue, I have abolished the feature by treating it
5210 as "timeout", and writing a warning to the main and panic logs.
5212 36. The display of retry rules for -brt wasn't always showing the error code
5215 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5216 timeout_connect_A, timeout_connect_MX.
5218 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5219 to the empty sender.
5221 39. The daemon was not analysing the content of -oX till after it had closed
5222 stderr and disconnected from the controlling terminal. This meant that any
5223 syntax errors were only noted on the panic log, and the return code from
5224 the command was 0. By re-arranging the code a little, I've made the
5225 decoding happen first, so such errors now appear on stderr, and the return
5226 code is 1. However, the actual setting up of the sockets still happens in
5227 the disconnected process, so errors there are still only recorded on the
5230 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5231 connections (as happens on some IP stacks) was logged at start up time as
5232 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5233 it from "IPv6 and IPv4", which means that two separate sockets are being
5236 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5237 says whether AF_INET or AF_INET6 was passed as an argument.
5239 42. Exiwhat output was messed up when time zones were included in log
5242 43. Exiwhat now gives more information about the daemon's listening ports,
5243 and whether -tls-on-connect was used.
5245 44. The "port" option of the smtp transport is now expanded.
5247 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5248 silently ignored. Now an error message is written to the main and panic
5251 46. There's a new ACL modifier called "logwrite" which writes to a log file
5252 as soon as it is encountered.
5254 47. Added $local_user_uid and $local_user_gid at routing time.
5256 48. Exim crashed when trying to verify a sender address that was being
5259 49. Exim was recognizing only a space character after ".include". It now also
5260 recognizes a tab character.
5262 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5263 extracting the relevant information from the specification. The man page no
5264 longer contains scrambled data for the -d option, and I've added a section
5265 at the front about calling Exim under different names.
5267 51. Added "extra_headers" argument to the "mail" command in filter files.
5269 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5272 53. Installed eximstats 1.29.
5274 54. Added transport_filter_timeout as a generic transport option.
5276 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5277 Cc: header lines. This was required by RFC 822, but it not required by RFC
5280 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5281 incoming messages that did not have them. Now it does so only if the
5282 message originates locally, that is, if there is no associated remote host
5283 address. When Resent- header lines are present, this applies to the Resent-
5284 lines rather than the non-Resent- lines.
5286 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5287 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5289 58. Messages for configuration errors now include the name of the main
5290 configuration file - useful now that there may be more than one file in a
5291 list (.included file names were always shown).
5293 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5294 for those rare installations that do not start the daemon as root or run it
5295 setuid root. I've cut out the call to initgroups() if the daemon is not
5298 60. The Exim user and group can now be bound into the binary as text strings
5299 that are looked up at the start of Exim's processing.
5301 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5303 62. Added $mailstore_basename variable.
5305 63. Installed patch to sieve.c from Michael Haardt.
5307 64. When Exim failed to open the panic log after failing to open the main log,
5308 the original message it was trying to log was written to stderr and debug
5309 output, but if they were not available (the usual case in production), it
5310 was lost. Now it is written to syslog before the two lines that record the
5311 failures to open the logs.
5313 65. Users' Exim filters run in subprocesses under the user's uid. It is
5314 possible for a "deliver" command or an alias in a "personal" command to
5315 provoke an address rewrite. If logging of address rewriting is configured,
5316 this fails because the process is not running as root or exim. There may be
5317 a better way of dealing with this, but for the moment (because 4.30 needs
5318 to be released), I have disabled address rewrite logging when running a
5319 filter in a non-root, non-exim process.
5325 1. The buildconfig auxiliary program wasn't quoting the value set for
5326 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5327 not defined. This bug was masked in 4.22 by the effect that was fixed in
5330 2. Some messages that were rejected after a message id was allocated were
5331 shown as "incomplete" by exigrep. It no longer does this for messages that
5332 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5334 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5335 have allow_domain_literals set, the ID did not get logged in the <= line.
5336 Domain literals are now always recognized in Message-ID: header lines.
5338 4. The first argument for a ${extract expansion item is the key name or field
5339 number. Leading and trailing spaces in this item were not being ignored,
5340 causing some misleading effects.
5342 5. When deliver_drop_privilege was set, single queue runner processes started
5343 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5344 same command in the process it spins off) were not dropping privilege.
5346 6. When the daemon running as "exim" started a queue runner, it always
5347 re-executed Exim in the spun-off process. This is a waste of effort when
5348 deliver_drop_privilege is set. The new process now just calls the
5349 queue-runner function directly.
5355 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5358 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5359 ignored. Though the use of -oP was forcing the writing of a pid file, it
5360 was always written to the default place.
5362 3. If the message "no IP address found for host xxxx" is generated during
5363 incoming verification, it is now followed by identification of the incoming
5364 connection (so you can more easily find what provoked it).
5366 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5368 5. Added some features to "harden" Exim a bit more against certain attacks:
5370 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5371 be put in Local/Makefile. This is like the never_users runtime option,
5372 but it cannot be overridden. The default setting is "root".
5374 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5375 prefix string with which any file named in a -C command line option
5378 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5379 is retained for -C and -D only if the caller of Exim is root. Without
5380 it, the exim user may also use -C and -D and retain privilege.
5382 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5383 command line option is disabled.
5385 6. Macro names set by the -D option must start with an upper case letter, just
5386 like macro names defined in the configuration file.
5388 7. Added "dereference=" facility to LDAP.
5390 8. Two instances of the typo "uknown" in the source files are fixed.
5392 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5393 the Configure-Makefile script screwed up while processing it.
5395 10. Incorporated PCRE 4.4.
5397 11. The SMTP synchronization check was not operating right at the start of an
5398 SMTP session. For example, it could not catch a HELO sent before the client
5399 waited for the greeting. There is now a check for outstanding input at the
5400 point when the greeting is written. Because of the duplex, asynchronous
5401 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5402 way, but not yet received, when the check is performed.
5404 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5405 on TCP/IP sockets, because this apparently causes some broken clients to
5408 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5409 unchanged) from the Cygwin maintainer.
5411 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5412 was supported instead of testing for mbx. Effectively a typo.
5414 15. The spa authenticator server code was not checking that the input it
5415 received was valid base64.
5417 16. The debug output line for the "set" modifier in ACLs was not showing the
5418 name of the variable that was being set.
5420 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5422 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5423 lookup on its own. Something else was needed to trigger the lookup. For
5424 example, a match in host_lookup or the need for a host name in a host list.
5425 Now, if $sender_host_name is referenced and the host name has not yet been
5426 looked up, a lookup is performed. If the lookup fails, the variable remains
5427 empty, and $host_lookup_failed is set to "1".
5429 19. Added "eqi" as a case-independent comparison operator.
5431 20. The saslauthd authentication condition could segfault if neither service
5432 nor realm was specified.
5434 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5435 error message that was logged was misleading, and incoming SMTP
5436 connections were dropped. The message is now more accurate, and temporary
5437 errors are given to SMTP connections.
5439 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5440 (except RSET) with a 421 error, until QUIT is received. However, it was
5441 failing to send a response to QUIT.
5443 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5444 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5445 ACL is now run after the helo_try_verify_hosts code.
5447 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5448 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5449 case-independent, so other case variants are also recognized). Apparently
5450 some systems use these upper case variants.
5452 25. If more than two messages were waiting for the same host, and a transport
5453 filter was specified for the transport, Exim sent two messages over the
5454 same TCP/IP connection, and then failed with "socket operation on non-
5455 socket" when it tried to send the third.
5457 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5459 27. The extern definition of crypt16() in expand.c was not being excluded when
5460 the OS had its own crypt16() function.
5462 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5463 as a preferred synonym for return_size_limit, both as an option and as an
5466 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5468 30. Changed the default configuration ACL to relax the local part checking rule
5469 for addresses that are not in any local domains. For these addresses,
5470 slashes and pipe symbols are allowed within local parts, but the sequence
5471 /../ is explicitly forbidden.
5473 31. SPA server authentication was not clearing the challenge buffer before
5476 32. log_message in a "warn" ACL statement was writing to the reject log as
5477 well as to the main log, which contradicts the documentation and doesn't
5478 seem right (because no rejection is happening). So I have stopped it.
5480 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5481 However, I am unable to do any testing of this.
5483 34. Fixed an infelicity in the appendfile transport. When checking directories
5484 for a mailbox, to see if any needed to be created, it was accidentally
5485 using path names with one or more superfluous leading slashes; tracing
5486 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5488 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5489 contents are added to the log line that is written for every discarded
5490 recipient. (Previously a log_message setting was ignored.)
5492 36. The ${quote: operator now quotes the string if it is empty.
5494 37. The install script runs exim in order to find its version number. If for
5495 some reason other than non-existence or emptiness, which it checks, it
5496 could not run './exim', it was installing it with an empty version number,
5497 i.e. as "exim-". This error state is now caught, and the installation is
5500 38. An argument was missing from the function that creates an error message
5501 when Exim fails to connect to the socket for saslauthd authentication.
5502 This could cause Exim to crash, or give a corrupted message.
5504 39. Added isip, isip4, and isip6 to ${if conditions.
5506 40. The ACL variables $acl_xx are now saved with the message, and can be
5507 accessed later in routers, transports, and filters.
5509 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5510 strings in the file are not string-expanded.
5512 42. If a MAIL command specified a SIZE value that was too large to fit into an
5513 int variable, the check against message_size_limit failed. Such values are
5514 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5515 day this will have to be increased, but I don't think I want to be around
5516 when emails are that large.
5523 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5524 iconv() is not standard in FreeBSD.
5526 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5527 IPv6 enabled. The observed symptom was a segmentation fault on return from
5528 the function os_common_find_running_interfaces() in src/os.c.
5530 3. In the check_special_case() function in daemon.c I had used "errno" as an
5531 argument name, which causes warnings on some systems. This was basically a
5532 typo, since it was named "eno" in the comments!
5534 4. The code that waits for the clock to tick (at a resolution of some fraction
5535 of a second) so as to ensure message-id uniqueness was always waiting for
5536 at least one whole tick, when it could have waited for less. [This is
5537 almost certainly not relevant at current processor speeds, where it is
5538 unlikely to ever wait at all. But we try to future-proof.]
5540 5. The function that sleeps for a time interval that includes fractions of a
5541 second contained a race. It did not block SIGALRM between setting the
5542 timer, and suspending (a couple of lines later). If the interval was short
5543 and the sigsuspend() was delayed until after it had expired, the suspension
5544 never ended. On busy systems this could lead to processes getting stuck for
5547 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5548 process, before it forks any delivery processes. The open lookup caching
5549 mechanism meant that the open file or database connection was passed into
5550 the delivery process. The problem was that delivery processes always tidy
5551 up cached lookup data. This could cause a problem for the next delivery
5552 process started by the queue runner, because the external queue runner
5553 process does not know about the closure. So the next delivery process
5554 still has data in the lookup cache. In the case of a file lookup, there was
5555 no problem because closing a file descriptor in a subprocess doesn't affect
5556 the parent. However, if the lookup was caching a connection to a database,
5557 the connection was closed, and the second delivery process was likely to
5558 see errors such as "PGSQL: query failed: server closed the connection
5559 unexpectedly". The problem has been fixed by closing all cached lookups
5560 in a queue runner before running a delivery process.
5562 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5563 seem to have the "const" qualifier which it has on other OS. I've
5566 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5567 *of the same type* (client or server) with the same public name that an
5568 error should be diagnosed.
5570 9. When Exim looked up a host name for an IP address, but failed to find the
5571 original IP address when looking up the host name (a safety check), it
5572 output the message "<ip address> does not match any IP for NULL", which was
5573 confusing, to say the least. The bug was that the host name should have
5574 appeared instead of "NULL".
5576 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5577 user that is built into the binary, and the -C or -D options is used, root
5578 privilege is dropped before the configuration file is read. In addition,
5579 logging is switched to stderr instead of the normal log files. If the
5580 configuration then re-defines the Exim user, the unprivileged environment
5581 is probably not what is expected, so Exim logs a panic warning message (but
5584 However, if deliver_drop_privilege is set, the unprivileged state may well
5585 be exactly what is intended, so the warning has been cut out in that case,
5586 and Exim is allowed to try to write to its normal log files.
5592 1. smtp_return_error_details was not giving details for temporary sender
5593 or receiver verification errors.
5595 2. Diagnose a configuration error if two authenticators have the same public
5598 3. Exim used not to create the message log file for a message until the first
5599 delivery attempt. This could be confusing when incoming messages were held
5600 for policy or load reasons. The message log file is now created at the time
5601 the message is received, and an initial "Received" line is written to it.
5603 4. The automatically generated man page for command line options had a minor
5604 bug that caused no ill effects; however, a more serious problem was that
5605 the procedure for building the man page automatically didn't always
5606 operate. Consequently, release 4.20 contains an out-of-date version. This
5607 shouldn't happen again.
5609 5. When building Exim with embedded Perl support, the script that builds the
5610 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5611 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5613 6. The freeze_tell option was not being used for messages that were frozen on
5614 arrival, either by an ACL or by local_scan().
5616 7. Added the smtp_incomplete_transaction log selector.
5618 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5619 was accepting AUTH without a new EHLO.
5621 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5622 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5625 10. Logging of TCP/IP connections (when configured) now happens in the main
5626 daemon process instead of the child process, so that the TCP/IP connection
5627 count is more accurate (but it can never be perfect).
5629 11. The use of "drop" in a nested ACL was not being handled correctly in the
5630 outer ACL. Now, if condition failure induced by the nested "drop" causes
5631 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5632 or "require"), the connection is dropped.
5634 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5635 that yield "discard" can now be used with an "accept" or a "discard" verb,
5636 but an error is generated for any others (because I can't see a useful way
5637 to define what should happen).
5639 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5640 are now processed in the same way as lines in the Exim configuration file.
5641 In particular, continuation lines are supported.
5643 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5645 15. Added -ti meaning -t -i.
5647 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5648 domains, and warn by logging if others are found.
5650 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5651 returned from GIFCONF when Exim is trying to find the list of interfaces on
5652 a host. The code in os.c has been modified to copy each ifreq to an aligned
5653 structure in all cases.
5655 Also, in some cases, the returned ifreq's were being copied to a 'struct
5656 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5657 means the last couple of bytes of an IPv6 address could be chopped if the
5658 ifreq contained only a normal sockaddr (14 bytes storage).
5660 18. Named domain lists were not supported in the hosts_treat_as_local option.
5661 An entry such as +xxxx was not recognized, and was treated as a literal
5664 19. Ensure that header lines added by a DATA ACL are included in the reject log
5665 if the ACL subsequently rejects the message.
5667 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5668 MD5 (which is deprecated).
5670 21. When testing a filter file using -bf, Exim was writing a message when it
5671 took the sender from a "From " line in the message, but it was not doing so
5672 when it took $return_path from a Return-Path: header line. It now does.
5674 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5675 with a valid header line field name (a series of printing characters
5676 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5678 23. Changed "disc" in the source to "disk" to conform to the documentation and
5679 the book and for uniformity.
5681 24. Ignore Sendmail's -Ooption=value command line item.
5683 25. When execve() failed while trying to run a command in a pipe transport,
5684 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5685 could be confused with a return value of 69 from the command itself. This
5686 has been changed to 127, the value the shell returns if it is asked to run
5687 a non-existent command. The wording for the related log line suggests a
5688 non-existent command as the problem.
5690 26. If received_header_text expands to an empty string, do not add a Received:
5691 header line to the message. (Well, it adds a token one on the spool, but
5692 marks it "old" so that it doesn't get used or transmitted.)
5694 27. Installed eximstats 1.28 (addition of -nt option).
5696 28. There was no check for failure on the call to getsockname() in the daemon
5697 code. This can fail if there is a shortage of resources on the system, with
5698 ENOMEM, for example. A temporary error is now given on failure.
5700 29. Contrary to the C standard, it seems that in some environments, the
5701 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5702 program. Exim now does this explicitly; it affects the formatting of
5703 timestamps using strftime().
5705 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5706 complaints. I've now initialized all the variables, to avoid this.
5708 32. Header lines added by a system filter were not being "seen" during
5709 transport-time rewrites.
5711 33. The info_callback() function passed to OpenSSL is set up with type void
5712 (*)(SSL *, int, int), as described somewhere. However, when calling the
5713 function (actually a macro) that sets it up, the type void(*)() is
5714 expected. I've put in a cast to prevent warnings from picky compilers.
5716 34. If a DNS black list lookup found a CNAME record, but there were no A
5717 records associated with the domain it pointed at, Exim crashed.
5719 35. If a DNS black list lookup returned more than one A record, Exim ignored
5720 all but the first. It now scans all returned addresses if a particular IP
5721 value is being sought. In this situation, the contents of the
5722 $dnslist_value variable are a list of all the addresses, separated by a
5725 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5726 to accept a host name and all its aliases if the forward lookup for any of
5727 them yielded the IP address of the incoming connection. Now it accepts only
5728 those names whose forward lookup yields the correct IP address. Any other
5729 names are discarded. This closes a loophole whereby a rogue DNS
5730 administrator could create reverse DNS records to break through a
5731 wildcarded host restriction in an ACL.
5733 37. If a user filter or a system filter that ran in a subprocess used any of
5734 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5735 the wrong values were passed to the pipe command ($thisaddress had the
5736 value of $0, $0 had the value of $1, etc). This bug was introduced by
5737 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5739 38. Improved the line breaking for long SMTP error messages from ACLs.
5740 Previously, if there was no break point between 40 and 75 characters, Exim
5741 left the rest of the message alone. Two changes have been made: (a) I've
5742 reduced the minimum length to 35 characters; (b) if it can't find a break
5743 point between 35 and 75 characters, it looks ahead and uses the first one
5744 that it finds. This may give the occasional overlong line, but at least the
5745 remaining text gets split now.
5747 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5748 file descriptors might be low, and that setting 1000 would always raise it.
5749 It turns out that in some environments, the limit is already over 1000 and
5750 that lowering it causes trouble. So now Exim takes care not to decrease it.
5752 40. When delivering a message, the value of $return_path is set to $sender_
5753 address at the start of routing (routers may change the value). By an
5754 oversight, this default was not being set up when an address was tested by
5755 -bt or -bv, which affected the outcome if any router or filter referred to
5758 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5759 log only when "message" or "log_message" are set. However, if one of the
5760 conditions was an address verification, or a call to a nested ACL, the
5761 messages generated by the underlying test were being passed through. This
5762 no longer happens. The underlying message is available in $acl_verify_
5763 message for both "message" and "log_message" expansions, so it can be
5764 passed through if needed.
5766 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5767 new expansion $bh_ to give the encoded byte string without charset
5768 translation. Translation happens only if iconv() is available; HAVE_ICONV
5769 indicates this at build time. HEADERS_CHARSET gives the charset to
5770 translate to; headers_charset can change it in the configuration, and
5771 "headers charset" can change it in an individual filter file.
5773 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5774 that creates RFC 2047 encoded "words" labels them as that charset instead
5775 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5776 expansion operator; (ii) when Exim creates a From: line for a local
5777 message; (iii) when a header line is rewritten to include a "phrase" part.
5779 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5780 buggy, causing it to skip the first lines of messages whose message ID
5781 ended in 'D'. This would not have bitten before Exim release 4.14, because
5782 message IDs were unlikely to end in 'D' before then. The effect was to have
5783 incorrect size information for certain domains.
5785 45. #include "config.h" was missing at the start of the crypt16.c module. This
5786 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5789 46. If there was a timeout during a "random" callout check, Exim treated it as
5790 a failure of the random address, and carried on sending RSET and the real
5791 address. If the delay was just some slowness somewhere, the response to the
5792 original RCPT would be taken as a response to RSET and so on, causing
5793 mayhem of various kinds.
5795 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5796 when I implemented it. It didn't allow for the fact that some option values
5797 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5798 the right test for positive values.
5800 48. Domain names in DNS records are case-independent. Exim always looks them up
5801 in lower case. Some resolvers return domain names in exactly the case they
5802 appear in the zone file, that is, they may contain uppercase letters. Not
5803 all resolvers do this - some return always lower case. Exim was treating a
5804 change of case by a resolver as a change of domain, similar to a widening
5805 of a domain abbreviation. This triggered its re-routing code and so it was
5806 trying to route what was effectively the same domain again. This normally
5807 caused routing to fail (because the router wouldn't handle the domain
5808 twice). Now Exim checks for this case specially, and just changes the
5809 casing of the domain that it ultimately uses when it transmits the message
5812 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5815 50. If a filter generated a file delivery with a non-absolute name (possible if
5816 no home directory exists for the router), the forbid_file option was not
5819 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5820 the existing equality matching.
5822 52. Exim was using ints instead of ino_t variables in some places where it was
5823 dealing with inode numbers.
5825 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5826 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5827 called TMPDIR, and if it finds it is different, it changes its value.
5829 54. The smtp_printf() function is now made available to local_scan() so
5830 additional output lines can be written before returning. There is also an
5831 smtp_fflush() function to enable the detection of a dropped connection.
5832 The variables smtp_input and smtp_batched_input are exported to
5835 55. Changed the default runtime configuration: the message "Unknown user"
5836 has been removed from the ACL, and instead placed on the localuser router,
5837 using the cannot_route_message feature. This means that any verification
5838 failures that generate their own messages won't get overridden. Similarly,
5839 the "Unrouteable address" message that was in the ACL for unverifiable
5840 relay addresses has also been removed.
5842 56. Added hosts_avoid_esmtp to the smtp transport.
5844 57. The exicyclog script was not checking for the esoteric option
5845 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5846 will work only if exicyclog is run under the appropriate euid.
5848 58. Following a discussion on the list, the rules by which Exim recognises line
5849 endings on incoming messages have been changed. The -dropcr and drop_cr
5850 options are now no-ops, retained only for backwards compatibility. The
5851 following line terminators are recognized: LF CRLF CR. However, special
5852 processing applies to CR:
5854 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5855 nor a local message in the state where . is a terminator.
5857 (ii) If a bare CR is encountered in a header line, an extra space is added
5858 after the line terminator so as not to end the header. The reasoning
5859 behind this is that bare CRs in header lines are most likely either
5860 to be mistakes, or people trying to play silly games.
5862 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5863 was being incorrectly given as 18 bytes larger than it should have been.
5864 This is a VOB (very old bug).
5866 60. This may never have affected anything current, but just in case it has:
5867 When the local host is found other than at the start of a list of hosts,
5868 the local host, those with the same MX, and any that follow, are discarded.
5869 When the list in question was part of a longer list of hosts, the following
5870 hosts (not currently being processed) were also being discarded. This no
5871 longer happens. I'm not sure if this situation could ever has previously
5874 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5877 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5878 header. This is something that is recommended in a new Internet Draft, and
5879 is something that is documented as being done by Sendmail. There are two
5880 possible values. For messages generated by the autoreply transport, Exim
5883 Auto-Submitted: auto-replied
5885 whereas for all other generated messages (e.g. bounces) it adds
5887 Auto-Submitted: auto-generated
5889 63. The "personal" condition in filters now includes a test for the
5890 Auto-Submitted: header. If it contains the string "auto-" the message it
5891 not considered personal.
5893 64. Added rcpt_include_affixes as a generic transport option.
5895 65. Added queue_only_override (default true).
5897 66. Added the syslog_duplication option.
5899 67. If what should have been the first header line of a message consisted of
5900 a space followed by a colon, Exim was mis-interpreting it as a header line.
5901 It isn't of course - it is syntactically invalid and should therefore be
5902 treated as the start of the message body. The misbehaviour could have
5903 caused a number of strange effects, including loss of data in subsequent
5904 header lines, and spool format errors.
5906 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5907 client host had authenticated. This control can now be exercised by an ACL
5908 for more flexibility.
5910 69. By default, callouts do not happen when testing with -bh. There is now a
5911 variant, -bhc, which does actually run the callout code, including
5912 consulting and updating the callout cache.
5914 70. Added support for saslauthd authentication, courtesy of Alexander
5917 71. If statvfs() failed on the spool or log directories while checking their
5918 size for availability, Exim confusingly gave the error "space shortage".
5919 Furthermore, in debugging mode it crashed with a floating point exception.
5920 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5921 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5922 serious problem, Exim now writes to the main and panic logs when this
5923 happens, with details of the failure. It then refuses to accept the
5924 incoming message, giving the message "spool directory problem" or "log
5925 directory problem" with a 421 code for SMTP messages.
5927 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5928 0, 1, and 2 exist, because some OS complain for execs without them (see
5929 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5930 descriptors. However, the code omitted to check that the open succeeded,
5931 causing mysterious errors if for some reason the permissions on /dev/null
5932 got screwed. Now Exim writes a message to the main and panic logs, and
5933 bombs out if it can't open /dev/null.
5935 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5936 interact so that it is all more flexible. It is supposed to remain
5937 backwards compatible. Also added extra_local_interfaces.
5939 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5940 to bomb out with an assertion failure - to the client this appears as a
5941 connection drop. This problem occurs in the part of the code that was taken
5942 from the Samba project. Fortunately, the assertion is in a very simple
5943 function, so I have fixed this by reproducing the function inline in the
5944 one place where it is called, and arranging for authentication to fail
5945 instead of killing the process with assert().
5947 75. The SPA client code was not working when the server requested OEM rather
5948 than Unicode encoding.
5950 76. Added code to make require_files with a specific uid setting more usable in
5951 the case where statting the file as root fails - usually a non-root-mounted
5952 NFS file system. When this happens and the failure is EACCES, Exim now
5953 forks a subprocess and does the per-uid checking as the relevant uid.
5955 77. Added process_log_path.
5957 78. If log_file_path was not explicitly set, a setting of check_log_space or
5958 check_log_inodes was ignored.
5960 79. If a space check for the spool or log partitions fails, the incident is now
5961 logged. Of course, in the latter case the data may get lost...
5963 80. Added the %p formatting code to string_format() so that it can be used to
5964 print addresses in debug_print(). Adjusted all the address printing in the
5965 debugging in store.c to use %p rather than %d.
5967 81. There was a concern that a line of code in smtp_in.c could overflow a
5968 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5969 initially expressed, the concern was not well-founded, because trailing
5970 spaces are removed early. However, if the trailing spaces were followed by
5971 a NULL, they did not get removed, so the overflow was possible. Two fixes
5974 (a) I re-wrote the offending code in a cleaner fashion.
5975 (b) If an incoming SMTP command contains a NULL character, it is rejected
5978 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5979 runs initgroups(), so that if the Exim user is in any additional groups,
5980 they will be used during message reception.
5986 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob