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.
27 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
29 PP/02 Make -n do something, by making it not do something.
30 When combined with -bP, the name of an option is not output.
32 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
35 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
36 $sender_host_name and config options to manage this, and basic check
39 PP/05 DSCP support for outbound connections and control modifier for inbound.
41 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
42 (Only plugin which currently uses this is kerberos4, which nobody should
43 be using, but we should make it available and other future plugins might
44 conceivably use it, even though it would break NAT; stuff *should* be
45 using channel bindings instead).
47 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
48 name; added for Sendmail compatibility; requires admin caller.
49 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
50 just ignore it); requires trusted caller.
51 Also parse but ignore: -Ac -Am -X<logfile>
54 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
56 TL/02 Add +smtp_confirmation as a default logging option.
58 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
59 Patch by Magnus Holmgren from 2007-02-20.
61 TL/04 Bugzilla 1281 - Spec typo.
62 Bugzilla 1283 - Spec typo.
63 Bugzilla 1290 - Spec grammar fixes.
65 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
67 TL/06 Add Experimental DMARC support using libopendmarc libraries.
69 TL/07 Fix an out of order global option causing a segfault. Reported to dev
70 mailing list by by Dmitry Isaikin.
72 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
74 JH/02 Support "G" suffix to numbers in ${if comparisons.
76 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
78 NM/01 Bugzilla 1197 - Spec typo
79 Bugzilla 1196 - Spec examples corrections
81 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
83 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
84 gnutls_enable_pkcs11, but renamed to more accurately indicate its
87 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
88 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
90 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
91 "acl {{name}{arg}...}", and optional args on acl condition
94 JH/05 Permit multiple router/transport headers_add/remove lines.
96 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
98 JH/07 Avoid using a waiting database for a single-message-only transport.
99 Performance patch from Paul Fisher. Bugzilla 1262.
101 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
104 JH/09 Add $headers_added variable, with content from use of ACL modifier
105 add_header (but not yet added to the message). Bugzilla 199.
107 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
108 Pulled from Bugzilla 817 by Wolfgang Breyha.
110 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
112 (nb: this is the same fix as in Exim 4.80.1)
114 JH/11 Add A= logging on delivery lines, and a client_set_id option on
117 JH/12 Add optional authenticated_sender logging to A= and a log_selector
120 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
122 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
123 advertise SMTP AUTH mechanism to us, instead of a generic
124 protocol violation error. Also, make Exim more robust to bad
125 data from the Dovecot auth socket.
127 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
129 When a queue runner is handling a message, Exim first routes the
130 recipient addresses, during which it prunes them based on the retry
131 hints database. After that it attempts to deliver the message to
132 any remaining recipients. It then updates the hints database using
135 So if a recipient address works intermittently, it can get repeatedly
136 deferred at routing time. The retry hints record remains fresh so the
137 address never reaches the final cutoff time.
139 This is a fairly common occurrence when a user is bumping up against
140 their storage quota. Exim had some logic in its local delivery code
141 to deal with this. However it did not apply to per-recipient defers
142 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
144 This change adds a proper retry rule check during routing so that the
145 final cutoff time is checked against the message's age. We only do
146 this check if there is an address retry record and there is not a
147 domain retry record; this implies that previous attempts to handle
148 the address had the retry_use_local_parts option turned on. We use
149 this as an approximation for the destination being like a local
150 delivery, as in LMTP.
152 I suspect this new check makes the old local delivery cutoff check
153 redundant, but I have not verified this so I left the code in place.
155 TF/02 Correct gecos expansion when From: is a prefix of the username.
157 Test 0254 submits a message to Exim with the header
161 When I ran the test suite under the user fanf2, Exim expanded
162 the header to contain my full name, whereas it should have added
163 a Resent-Sender: header. It erroneously treats any prefix of the
164 username as equal to the username.
166 This change corrects that bug.
168 GF/01 DCC debug and logging tidyup
169 Error conditions log to paniclog rather than rejectlog.
170 Debug lines prefixed by "DCC: " to remove any ambiguity.
172 TF/03 Avoid unnecessary rebuilds of lookup-related code.
174 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
175 Bug spotted by Jeremy Harris; was flawed since initial commit.
176 Would have resulted in OCSP responses post-SNI triggering an Exim
177 NULL dereference and crash.
179 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
181 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
182 Bug detection, analysis and fix by Samuel Thibault.
183 Bugzilla 1331, Debian bug #698092.
185 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
187 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
188 Server implementation by Todd Lyons, client by JH.
189 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
190 config variable "prdr_enable" controls whether the server
191 advertises the facility. If the client requests PRDR a new
192 acl_data_smtp_prdr ACL is called once for each recipient, after
193 the body content is received and before the acl_smtp_data ACL.
194 The client is controlled by bolth of: a hosts_try_prdr option
195 on the smtp transport, and the server advertisement.
196 Default client logging of deliveries and rejections involving
197 PRDR are flagged with the string "PRDR".
199 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
200 fclose(). Diagnosis by Todd Lyons.
202 PP/17 Update configure.default to handle IPv6 localhost better.
203 Patch by Alain Williams (plus minor tweaks).
206 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
207 This is now consistent with GnuTLS, and is now documented: the
208 previous undocumented portable approach to treating the option as
209 unset was to force an expansion failure. That still works, and
210 an empty string is now equivalent.
212 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
213 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
214 not performing validation itself.
216 PP/20 Added force_command boolean option to pipe transport.
217 Patch from Nick Koston, of cPanel Inc.
219 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
222 TF/04 Added udpsend ACL modifer and hexquote expansion operator
224 PP/21 Fix eximon continuous updating with timestamped log-files.
225 Broken in a format-string cleanup in 4.80, missed when I repaired the
226 other false fix of the same issue.
227 Report and fix from Heiko Schlichting.
230 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
231 Report from Prashanth Katuri.
233 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
234 It's SecureTransport, so affects any MacOS clients which use the
235 system-integrated TLS libraries, including email clients.
237 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
238 using a MIME ACL for non-SMTP local injection.
239 Report and assistance in diagnosis by Warren Baker.
241 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
243 JH/16 Fix comparisons for 64b. Bugzilla 1385.
245 TL/09 Add expansion variable $authenticated_fail_id to keep track of
246 last id that failed so it may be referenced in subsequent ACL's.
248 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
251 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
252 ldap library initialization, allowing self-signed CA's to be
253 used. Also properly sets require_cert option later in code by
254 using NULL (global ldap config) instead of ldap handle (per
255 session). Bug diagnosis and testing by alxgomz.
257 TL/12 Enhanced documentation in the ratelimit.pl script provided in
258 the src/util/ subdirectory.
260 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
261 renamed to Transport Post Delivery Action by Jeremy Harris, as
264 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
265 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
266 redis_servers = needs to be configured which will be used by the redis
267 lookup. Patch from Warren Baker, of The Packet Hub.
269 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
271 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
272 hostname or reverse DNS when processing a host list. Used suggestions
273 from multiple comments on this bug.
275 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
277 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
278 Missed a few lines, added it to make the runtest require no keyboard
281 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
282 contains upper case chars. Make router use caseful_local_part.
284 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
285 support when GnuTLS has been built with p11-kit.
291 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
293 This, or similar/improved, will also be change PP/11 of 4.82.
299 PP/01 Handle short writes when writing local log-files.
300 In practice, only affects FreeBSD (8 onwards).
301 Bugzilla 1053, with thanks to Dmitry Isaikin.
303 NM/01 Bugzilla 949 - Documentation tweak
305 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
308 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
310 PP/02 Implemented gsasl authenticator.
312 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
314 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
315 `pkg-config foo` for cflags/libs.
317 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
318 with rest of GSASL and with heimdal_gssapi.
320 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
321 `pkg-config foo` for cflags/libs for the TLS implementation.
323 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
324 properties get this fed in as external SSF. A number of robustness
325 and debugging improvements to the cyrus_sasl authenticator.
327 PP/08 cyrus_sasl server now expands the server_realm option.
329 PP/09 Bugzilla 1214 - Log authentication information in reject log.
330 Patch by Jeremy Harris.
332 PP/10 Added dbmjz lookup type.
334 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
336 PP/12 MAIL args handles TAB as well as SP, for better interop with
337 non-compliant senders.
338 Analysis and variant patch by Todd Lyons.
340 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
341 Bug report from Lars Müller <lars@samba.org> (via SUSE),
342 Patch from Dirk Mueller <dmueller@suse.com>
344 PP/13 tls_peerdn now print-escaped for spool files.
345 Observed some $tls_peerdn in wild which contained \n, which resulted
346 in spool file corruption.
348 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
349 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
350 or write after TLS renegotiation, which otherwise led to messages
353 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
354 as a tracking header (ie: a signed header comes before the signature).
355 Patch from Wolfgang Breyha.
357 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
358 comma-sep list; embedded commas doubled.
360 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
362 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
364 Report and patch from Dmitry Banschikov.
366 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
367 Removed SSL_clear() after SSL_new() which led to protocol negotiation
368 failures. We appear to now support TLS1.1+ with Exim.
370 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
371 lets Exim select keys and certificates based upon TLS SNI from client.
372 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
373 before an outbound SMTP session. New log_selector, +tls_sni.
375 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
376 NULL dereference. Report and patch from Alun Jones.
378 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
379 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
380 Not seeing resolver debug output on NetBSD, but suspect this is a
381 resolver implementation change.
383 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
384 Left warnings. Added "eximon gdb" invocation mode.
386 PP/21 Defaulting "accept_8bitmime" to true, not false.
388 PP/22 Added -bw for inetd wait mode support.
390 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
391 locate the relevant includes and libraries. Made this the default.
393 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
394 Bugzilla 1246, report and most of solution from Tomasz Kusy.
396 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
397 This may cause build issues on older platforms.
399 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
400 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
401 gnutls_require_mac & gnutls_require_protocols (no longer supported).
402 Added SNI support via GnuTLS too.
403 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
405 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
407 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
408 Applied second patch from Janne, implementing suggestion to default
409 multiple-strings-in-record handling to match SPF spec.
411 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
413 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
414 read-only, out of scope).
415 Patch from Wolfgang Breyha, report from Stuart Northfield.
417 PP/29 Fix three issues highlighted by clang analyser static analysis.
418 Only crash-plausible issue would require the Cambridge-specific
419 iplookup router and a misconfiguration.
420 Report from Marcin Mirosław.
422 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
424 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
425 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
426 As part of this, removing so much warning spew let me fix some minor
427 real issues in debug logging.
429 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
430 assignment on my part. Fixed.
432 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
433 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
434 Janne Snabb (who went above and beyond: thank you).
436 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
437 string otherwise requires a connection and a bunch more work and it's
438 relatively easy to get wrong. Should also expose TLS library linkage
441 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
442 64-bit ${eval} (JH/03).
444 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
445 GNU libc to support some of the 64-bit stuff, should not lead to
446 conflicts. Defined before os.h is pulled in, so if a given platform
447 needs to override this, it can.
449 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
450 protection layer was required, which is not implemented.
451 Bugzilla 1254, patch from Wolfgang Breyha.
453 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
454 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
455 tls_dhparam take prime identifiers. Also unbreak combination of
456 OpenSSL+DH_params+TLSSNI.
458 PP/39 Disable SSLv2 by default in OpenSSL support.
464 PP/01 Solaris build fix for Oracle's LDAP libraries.
465 Bugzilla 1109, patch from Stephen Usher.
467 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
469 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
472 TF/02 Fix a couple more cases where we did not log the error message
473 when unlink() failed. See also change 4.74-TF/03.
475 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
476 lock up or crash if it happened to be inside a call to libc when it
477 got a SIGUSR1 from exiwhat.
479 The SIGUSR1 handler appends the current process status to the process
480 log which is later printed by exiwhat. It used to use the general
481 purpose logging code to do this, but several functions it calls are
482 not safe for signals.
484 The new output code in the SIGUSR1 handler is specific to the process
485 log, and simple enough that it's easy to inspect for signal safety.
486 Removing some special cases also simplifies the general logging code.
487 Removing the spurious timestamps from the process log simplifies
490 TF/04 Improved ratelimit ACL condition.
492 The /noupdate option has been deprecated in favour of /readonly which
493 has clearer semantics. The /leaky, /strict, and /readonly update modes
494 are mutually exclusive. The update mode is no longer included in the
495 database key; it just determines when the database is updated. (This
496 means that when you upgrde Exim will forget old rate measurements.)
498 Exim now checks that the per_* options are used with an update mode that
499 makes sense for the current ACL. For example, when Exim is processing a
500 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
501 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
502 must specify per_mail/readonly. If you omit the update mode it defaults to
503 /leaky where that makes sense (as before) or /readonly where required.
505 The /noupdate option is now undocumented but still supported for
506 backwards compatibility. It is equivalent to /readonly except that in
507 ACLs where /readonly is required you may specify /leaky/noupdate or
508 /strict/noupdate which are treated the same as /readonly.
510 A useful new feature is the /count= option. This is a generalization
511 of the per_byte option, so that you can measure the throughput of other
512 aggregate values. For example, the per_byte option is now equivalent
513 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
515 The per_rcpt option has been generalized using the /count= mechanism
516 (though it's more complicated than the per_byte equivalence). When it is
517 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
518 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
519 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
520 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
521 Note that using per_rcpt with a non-readonly update mode in more than
522 one ACL will cause the recipients to be double-counted. (The per_mail
523 and per_byte options don't have this problem.)
525 The handling of very low rates has changed slightly. If the computed rate
526 is less than the event's count (usually one) then this event is the first
527 after a long gap. In this case the rate is set to the same as this event's
528 count, so that the first message of a spam run is counted properly.
530 The major new feature is a mechanism for counting the rate of unique
531 events. The new per_addr option counts the number of different
532 recipients that someone has sent messages to in the last time period. It
533 behaves like per_rcpt if all the recipient addresses are different, but
534 duplicate recipient addresses do not increase the measured rate. Like
535 the /count= option this is a general mechanism, so the per_addr option
536 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
537 example, measure the rate that a client uses different sender addresses
538 with the options per_mail/unique=$sender_address. There are further
539 details in the main documentation.
541 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
543 TF/06 Removed a few PCRE remnants.
545 TF/07 Automatically extract Exim's version number from tags in the git
546 repository when doing development or release builds.
548 PP/02 Raise smtp_cmd_buffer_size to 16kB.
549 Bugzilla 879. Patch from Paul Fisher.
551 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
552 Heavily based on revision 40f9a89a from Simon Arlott's tree.
555 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
557 PP/05 Variable $av_failed, true if the AV scanner deferred.
558 Bugzilla 1078. Patch from John Horne.
560 PP/06 Stop make process more reliably on build failure.
561 Bugzilla 1087. Patch from Heiko Schlittermann.
563 PP/07 Make maildir_use_size_file an _expandable_ boolean.
564 Bugzilla 1089. Patch from Heiko Schlittermann.
566 PP/08 Handle ${run} returning more data than OS pipe buffer size.
567 Bugzilla 1131. Patch from Holger Weiß.
569 PP/09 Handle IPv6 addresses with SPF.
570 Bugzilla 860. Patch from Wolfgang Breyha.
572 PP/10 GnuTLS: support TLS 1.2 & 1.1.
574 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
577 PP/11 match_* no longer expand right-hand-side by default.
578 New compile-time build option, EXPAND_LISTMATCH_RHS.
579 New expansion conditions, "inlist", "inlisti".
581 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
583 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
585 PP/14 fix log_write() format string regression from TF/03.
586 Bugzilla 1152. Patch from Dmitry Isaikin.
592 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
594 PP/02 Harmonised TLS library version reporting; only show if debugging.
595 Layout now matches that introduced for other libraries in 4.74 PP/03.
597 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
599 PP/04 New "dns_use_edns0" global option.
601 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
604 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
605 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
607 TK/01 Updated PolarSSL code to 0.14.2.
608 Bugzilla 1097. Patch from Andreas Metzler.
610 PP/07 Catch divide-by-zero in ${eval:...}.
613 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
616 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
617 format-string attack -- SECURITY: remote arbitrary code execution.
619 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
620 time unintentionally subject to list matching rules, letting the header
621 cause arbitrary Exim lookups (of items which can occur in lists, *not*
622 arbitrary string expansion). This allowed for information disclosure.
624 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
625 INT_MIN/-1 -- value coerced to INT_MAX.
631 NM/01 Workround for PCRE version dependency in version reporting
634 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
635 This fixes portability to compilers other than gcc, notably
636 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
638 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
639 makefiles for portability to HP-UX and POSIX correctness.
641 PP/01 Permit LOOKUP_foo enabling on the make command-line.
642 Also via indented variable definition in the Makefile.
643 (Debugging by Oliver Heesakkers).
645 PP/02 Restore caching of spamd results with expanded spamd_address.
646 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
648 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
649 Improves build reliability. Fix from: Frank Elsner
651 NM/02 Fix wide character breakage in the rfc2047 coding
652 Fixes bug 1064. Patch from Andrey N. Oktyabrski
654 NM/03 Allow underscore in dnslist lookups
655 Fixes bug 1026. Patch from Graeme Fowler
657 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
658 Code patches from Adam Ciarcinski of NetBSD.
660 NM/04 Fixed exiqgrep to cope with mailq missing size issue
663 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
664 is logged, to avoid truncation. Patch from John Horne.
666 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
667 Patch from Jakob Hirsch.
669 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
670 SQL string expansion failure details.
671 Patch from Andrey Oktyabrski.
673 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
674 Patch from Simon Arlott.
676 PP/09 New lookups functionality failed to compile on old gcc which rejects
677 extern declarations in function scope.
678 Patch from Oliver Fleischmann
680 PP/10 Use sig_atomic_t for flags set from signal handlers.
681 Check getgroups() return and improve debugging.
682 Fixed developed for diagnosis in bug 927 (which turned out to be
685 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
686 Patch from Mark Zealey.
688 PP/12 Bugzilla 1056: Improved spamd server selection.
689 Patch from Mark Zealey.
691 PP/13 Bugzilla 1086: Deal with maildir quota file races.
692 Based on patch from Heiko Schlittermann.
694 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
695 Patch from Uwe Doering, sign-off by Michael Haardt.
697 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
698 variable declaration deep within a block. Bug and patch from
701 PP/15 lookups-Makefile IRIX compatibilty coercion.
703 PP/16 Make DISABLE_DKIM build knob functional.
705 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
706 Patch by Simon Arlott
708 TF/03 Fix valgrind.h portability to C89 compilers that do not support
709 variable argument macros. Our copy now differs from upstream.
715 TF/01 Failure to get a lock on a hints database can have serious
716 consequences so log it to the panic log.
718 TF/02 Log LMTP confirmation messages in the same way as SMTP,
719 controlled using the smtp_confirmation log selector.
721 TF/03 Include the error message when we fail to unlink a spool file.
723 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
724 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
725 for maintaining out-of-tree patches for some time.
727 PP/01 Bugzilla 139: Documentation and portability issues.
728 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
729 Handle per-OS dynamic-module compilation flags.
731 PP/02 Let /dev/null have normal permissions.
732 The 4.73 fixes were a little too stringent and complained about the
733 permissions on /dev/null. Exempt it from some checks.
734 Reported by Andreas M. Kirchwitz.
736 PP/03 Report version information for many libraries, including
737 Exim version information for dynamically loaded libraries. Created
738 version.h, now support a version extension string for distributors
739 who patch heavily. Dynamic module ABI change.
741 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
742 privilege escalation vulnerability whereby the Exim run-time user
743 can cause root to append content of the attacker's choosing to
746 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
749 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
750 If dropping privileges for untrusted macros, we disabled normal logging
751 on the basis that it would fail; for the Exim run-time user, this is not
752 the case, and it resulted in successful deliveries going unlogged.
753 Fixed. Reported by Andreas Metzler.
759 PP/01 Date: & Message-Id: revert to normally being appended to a message,
760 only prepend for the Resent-* case. Fixes regression introduced in
761 Exim 4.70 by NM/22 for Bugzilla 607.
763 PP/02 Include check_rfc2047_length in configure.default because we're seeing
764 increasing numbers of administrators be bitten by this.
766 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
768 PP/03 Bugzilla 994: added openssl_options main configuration option.
770 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
772 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
774 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
776 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
777 without a peer certificate, leading to a segfault because of an
778 assumption that peers always have certificates. Be a little more
779 paranoid. Problem reported by Martin Tscholak.
781 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
782 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
783 NB: ClamAV planning to remove STREAM in "middle of 2010".
784 CL also introduces -bmalware, various -d+acl logging additions and
785 more caution in buffer sizes.
787 PP/09 Implemented reverse_ip expansion operator.
789 PP/10 Bugzilla 937: provide a "debug" ACL control.
791 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
793 PP/12 Bugzilla 973: Implement --version.
795 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
797 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
799 PP/15 Bugzilla 816: support multiple condition rules on Routers.
801 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
802 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
803 ignore trailing whitespace.
805 JJ/02 prevent non-panic DKIM error from being sent to paniclog
807 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
810 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
811 Notification from Dr Andrew Aitchison.
813 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
814 ExtendedDetectionInfo response format.
815 Notification from John Horne.
817 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
820 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
821 XSL and documented dependency on system catalogs, with examples of how
824 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
827 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
828 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
829 configuration file which is writeable by the Exim user or group.
831 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
832 of configuration files to cover files specified with the -C option if
833 they are going to be used with root privileges, not just the default
836 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
837 option (effectively making it always true).
839 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
840 files to be used while preserving root privileges.
842 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
843 that rogue child processes cannot use them.
845 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
846 run-time user, instead of root.
848 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
849 Exim run-time user without dropping privileges.
851 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
852 result string, instead of calling string_vformat() twice with the same
855 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
856 for other users. Others should always drop root privileges if they use
857 -C on the command line, even for a whitelisted configure file.
859 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
861 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
867 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
868 $data_path, and $header_path variables; fixed documentation bugs and
871 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
872 exipick to access non-standard spools, including the "frozen" queue
875 NM/01 Bugzilla 965: Support mysql stored procedures.
876 Patch from Alain Williams
878 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
880 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
881 Patch from Andreas Metzler
883 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
884 Patch from Kirill Miazine
886 NM/05 Bugzilla 671: Added umask to procmail example.
888 JJ/03 installed exipick 20100323.0, fixing doc bug
890 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
891 directory. Notification and patch from Dan Rosenberg.
893 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
895 TK/02 Improve log output when DKIM signing operation fails.
897 MH/01 Treat the transport option dkim_domain as a colon separated
898 list, not as a single string, and sign the message with each element,
899 omitting multiple occurences of the same signer.
901 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
902 Bugzilla 985, 986. Patch by Simon Arlott
904 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
905 Patch by Simon Arlott
907 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
908 MBX locking. Notification from Dan Rosenberg.
914 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
916 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
918 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
920 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
922 NM/04 Bugzilla 915: Flag broken perl installation during build.
928 TK/01 Added patch by Johannes Berg that expands the main option
929 "spamd_address" if it starts with a dollar sign.
931 TK/02 Write list of recipients to X-Envelope-Sender header when building
932 the mbox-format spool file for content scanning (suggested by Jakob
935 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
936 (http://www.dcc-servers.net/) support via dccifd. Activated by
937 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
939 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
940 by Mark Daniel Reidel <mr@df.eu>.
942 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
943 When building exim an external PCRE library is now needed -
944 PCRE is a system library on the majority of modern systems.
945 See entry on PCRE_LIBS in EDITME file.
947 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
948 conversation. Added nologin parameter to request.
949 Patch contributed by Kirill Miazine.
951 TF/01 Do not log submission mode rewrites if they do not change the address.
953 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
955 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
956 log files in place. Contributed by Roberto Lima.
958 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
960 TF/03 Bugzilla 615: When checking the local_parts router precondition
961 after a local_part_suffix or local_part_prefix option, Exim now
962 does not use the address's named list lookup cache, since this
963 contains cached lookups for the whole local part.
965 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
966 Robert Millan. Documentation is in experimental-spec.txt.
968 TF/04 Bugzilla 668: Fix parallel build (make -j).
970 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
972 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
973 Patch provided by Jan Srzednicki.
975 TF/05 Leading white space used to be stripped from $spam_report which
976 wrecked the formatting. Now it is preserved.
978 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
979 that they are available at delivery time.
981 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
983 TF/08 TLS error reporting now respects the incoming_interface and
984 incoming_port log selectors.
986 TF/09 Produce a more useful error message if an SMTP transport's hosts
987 setting expands to an empty string.
989 NM/06 Bugzilla 744: EXPN did not work under TLS.
990 Patch provided by Phil Pennock.
992 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
993 Patch provided by Richard Godbee.
995 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
996 acl_smtp_notquit, added index entry.
998 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
999 Patch provided by Eugene Bujak.
1001 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1002 accept(). Patch provided by Maxim Dounin.
1004 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1005 Patch provided by Phil Pennock.
1007 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1009 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1010 Patch provided by Brad "anomie" Jorsch.
1012 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1013 Patch provided by Dean Brooks.
1015 TK/05 Add native DKIM support (does not depend on external libraries).
1017 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1018 Patch provided by Graeme Fowler.
1020 NM/16 Bugzilla 851: Documentation example syntax fix.
1022 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1024 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1027 NM/19 Bugzilla 745: TLS version reporting.
1028 Patch provided by Phil Pennock.
1030 NM/20 Bugzilla 167: bool: condition support.
1031 Patch provided by Phil Pennock.
1033 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1034 clients. Patch provided by Phil Pennock.
1036 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1037 Patch provided by Brad "anomie" Jorsch.
1039 NM/23 Bugzilla 687: Fix misparses in eximstats.
1040 Patch provided by Heiko Schlittermann.
1042 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1043 Patch provided by Heiko Schlittermann.
1045 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1046 plus update to original patch.
1048 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1050 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1051 Patch provided by David Brownlee.
1053 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1055 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1057 NM/30 Bugzilla 888: TLS documentation bugfixes.
1059 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1061 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1062 Unlike the original bugzilla I have changed all shell scripts in src tree.
1064 NM/33 Bugzilla 898: Transport filter timeout fix.
1065 Patch by Todd Rinaldo.
1067 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1068 Patch by Serge Demonchaux.
1070 NM/35 Bugzilla 39: Base64 decode bug fixes.
1071 Patch by Jakob Hirsch.
1073 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1075 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1077 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1079 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1085 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1086 ALT-N's libdkim that I have put here:
1087 http://duncanthrax.net/exim-experimental/
1089 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1090 were called 'true' and it seems that C99 defines that as a reserved
1091 keyword to be used with 'bool' variable types. That means you could
1092 not include C99-style headers which use bools without triggering
1093 build errors in sieve.c.
1095 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1096 as mailq or other aliases. Changed the --help handling significantly
1097 to do whats expected. exim_usage() emits usage/help information.
1099 SC/01 Added the -bylocaldomain option to eximstats.
1101 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1103 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1105 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1111 PH/01 Another patch from the Sieve maintainer.
1113 PH/02 When an IPv6 address is converted to a string for single-key lookup
1114 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1115 dots are used instead of colons so that keys in lsearch files need not
1116 contain colons. This was done some time before quoting was made available
1117 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1118 (notated using the quote facility) so as to distinguish them from IPv4
1119 keys. This meant that lookups for IP addresses in host lists did not work
1120 for iplsearch lookups.
1122 This has been fixed by arranging for IPv6 addresses to be expressed with
1123 colons if the lookup type is iplsearch. This is not incompatible, because
1124 previously such lookups could never work.
1126 The situation is now rather anomolous, since one *can* have colons in
1127 ordinary lsearch keys. However, making the change in all cases is
1128 incompatible and would probably break a number of configurations.
1130 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1133 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1134 conversion specification without a maximum field width, thereby enabling
1135 a rogue spamd server to cause a buffer overflow. While nobody in their
1136 right mind would setup Exim to query an untrusted spamd server, an
1137 attacker that gains access to a server running spamd could potentially
1138 exploit this vulnerability to run arbitrary code as the Exim user.
1140 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1141 $primary_hostname instead of what libspf2 thinks the hosts name is.
1143 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1144 a directory entry by the name of the lookup key. Previously, if a
1145 symlink pointed to a non-existing file or a file in a directory that
1146 Exim lacked permissions to read, a lookup for a key matching that
1147 symlink would fail. Now it is enough that a matching directory entry
1148 exists, symlink or not. (Bugzilla 503.)
1150 PH/03 The body_linecount and body_zerocount variables are now exported in the
1153 PH/04 Added the $dnslist_matched variable.
1155 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1156 This means they are set thereafter only if the connection becomes
1159 PH/06 Added the client_condition to authenticators so that some can be skipped
1160 by clients under certain conditions.
1162 PH/07 The error message for a badly-placed control=no_multiline_responses left
1163 "_responses" off the end of the name.
1165 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1167 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1168 (without spaces) instead of just copying the configuration text.
1170 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1172 PH/11 Added $max_received_linelength.
1174 PH/12 Added +ignore_defer and +include_defer to host lists.
1176 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1177 way in which PCRE > 7.0 is built.
1179 PH/14 Implemented queue_only_load_latch.
1181 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1182 MAIL command. The effect was to mangle the value on 64-bit systems.
1184 PH/16 Another patch from the Sieve maintainer.
1186 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1188 PH/18 If a system quota error occurred while trying to create the file for
1189 a maildir delivery, the message "Mailbox is full" was not appended to the
1190 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1191 only to a quota excession during the actual writing of the file.
1193 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1194 characters?) which causes problems in log lines. The DN values are now
1195 passed through string_printing() before being added to log lines.
1197 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1198 and InterBase are left for another time.)
1200 PH/21 Added message_body_newlines option.
1202 PH/22 Guard against possible overflow in moan_check_errorcopy().
1204 PH/23 POSIX allows open() to be a macro; guard against that.
1206 PH/24 If the recipient of an error message contained an @ in the local part
1207 (suitably quoted, of course), incorrect values were put in $domain and
1208 $local_part during the evaluation of errors_copy.
1214 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1215 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1218 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1219 issue a MAIL command.
1221 PH/02 In an ACL statement such as
1223 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1225 if a client was not listed at all, or was listed with a value other than
1226 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1227 the condition was not true (as it should be), so access was not denied.
1228 The bug was that the ! inversion was incorrectly passed on to the second
1229 item. This has been fixed.
1231 PH/03 Added additional dnslists conditions == and =& which are different from
1232 = and & when the dns lookup returns more than one IP address.
1234 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1235 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1237 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1238 FSYNC, which compiles an option called disable_fsync that allows for
1239 bypassing fsync(). The documentation is heavily laced with warnings.
1241 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1243 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1244 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1245 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1246 including adding "make clean"; (3) Added -fPIC when compiling the test
1247 dynamically loaded module, to get rid of a warning.
1249 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1250 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1251 = 0s. The bug is otherwise harmless.
1253 PH/07 There was a bug in the dovecot authenticator such that the value of
1254 $auth1 could be overwritten, and so not correctly preserved, after a
1255 successful authentication. This usually meant that the value preserved by
1256 the server_setid option was incorrect.
1258 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1260 PH/09 Installed PCRE release 7.0.
1262 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1263 run for batched SMTP input. It is now run at the start of every message
1264 in the batch. While fixing this I discovered that the process information
1265 (output by running exiwhat) was not always getting set for -bs and -bS
1266 input. This is fixed, and it now also says "batched" for BSMTP.
1268 PH/11 Added control=no_pipelining.
1270 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1271 patch, slightly modified), and move the expansion of helo_data till after
1272 the connection is made in the smtp transport (so it can use these
1275 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1277 PH/14 Added log_selector = +pid.
1279 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1281 PH/16 Add ${if forany and ${if forall.
1283 PH/17 Added dsn_from option to vary the From: line in DSNs.
1285 PH/18 Flush SMTP output before performing a callout, unless control =
1286 no_callout_flush is set.
1288 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1289 was true (the default) a successful delivery failed to delete the retry
1290 item, thus causing premature timeout of the address. The bug is now
1293 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1295 PH/21 Long custom messages for fakedefer and fakereject are now split up
1296 into multiline reponses in the same way that messages for "deny" and
1297 other ACL rejections are.
1299 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1300 with slight modification.
1302 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1303 draft, changing the syntax and factoring some duplicate code".
1305 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1306 for deliveries of the second and subsequent messages over the same SMTP
1309 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1310 ${reduce, with only minor "tidies".
1312 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1314 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1315 expansion side effects.
1317 PH/27 When a message times out after an over-quota error from an Exim-imposed
1318 quota, the bounce message says "mailbox is full". This message was not
1319 being given when it was a system quota that was exceeded. It now should
1322 MH/03 Made $recipients available in local_scan(). local_scan() already has
1323 better access to the recipient list through recipients_list[], but
1324 $recipients can be useful in postmaster-provided expansion strings.
1326 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1327 in the case of a MAIL command with additional options following the
1328 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1329 were accidentally chopped off.
1331 PH/29 SMTP synchronization checks are implemented when a command is read -
1332 there is a check that no more input is waiting when there shouldn't be
1333 any. However, for some commands, a delay in an ACL can mean that it is
1334 some time before the response is written. In this time, more input might
1335 arrive, invalidly. So now there are extra checks after an ACL has run for
1336 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1337 pipelining has not been advertised.
1339 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1341 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1342 returned in the SMTP response when smtp_return_error_details was false.
1343 This has been fixed.
1345 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1346 instead of the C I/O that was originally supplied, because problems were
1347 reported on Solaris.
1349 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1350 Exim which did not show up earlier: it was assuming that a call to
1351 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1352 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1353 was a macro that became an assignment, so it seemed to work. This has
1354 changed to a proper function call with a void return, hence the compile
1355 error. Exim's code has been fixed.
1357 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1360 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1362 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1364 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1365 because it turns out to be too big or there is a timeout) there is no
1366 "Completed" line in the log. When some messages of this type were
1367 selected by exigrep, they were listed as "not completed". Others were
1368 picked up by some special patterns. I have improved the selection
1369 criteria to be more general.
1371 PH/38 The host_find_failed option in the manualroute router can now be set
1372 to "ignore", to completely ignore a host whose IP address cannot be
1373 found. If all hosts are ignored, the behaviour is controlled by the new
1374 host_all_ignored option.
1376 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1377 homing or because of multiple MX records with /mx) generated more than
1378 one IP address, and the following item turned out to be the local host,
1379 all the secondary addresses of the first item were incorrectly removed
1380 from the list, along with the local host and any following hosts (which
1381 is what is supposed to happen).
1383 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1384 whoever called Exim into the -H file. In the case of the daemon it was
1385 behaving confusingly. When first started, it used values for whoever
1386 started the daemon, but after a SIGHUP it used the Exim user (because it
1387 calls itself on a restart). I have changed the code so that it now always
1390 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1391 message are rejected with the same error (e.g. no authentication or bad
1392 sender address), and a DATA command is nevertheless sent (as can happen
1393 with PIPELINING or a stupid MUA), the error message that was given to the
1394 RCPT commands is included in the rejection of the DATA command. This is
1395 intended to be helpful for MUAs that show only the final error to their
1398 PH/42 Another patch from the Sieve maintainer.
1400 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1401 Eximstats - Fixed some broken HTML links and added missing column headers
1403 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1404 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1406 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1408 PH/43 Yet another patch from the Sieve maintainer.
1410 PH/44 I found a way to check for a TCP/IP connection going away before sending
1411 the response to the final '.' that terminates a message, but only in the
1412 case where the client has not sent further data following the '.'
1413 (unfortunately, this is allowed). However, in many cases there won't be
1414 any further data because there won't be any more messages to send. A call
1415 to select() can be used: if it shows that the input is "ready", there is
1416 either input waiting, or the socket has been closed. An attempt to read
1417 the next input character can distinguish the two cases. Previously, Exim
1418 would have sent an OK response which the client would never have see.
1419 This could lead to message repetition. This fix should cure that, at
1420 least in a lot of common cases.
1422 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1423 advertised in response to EHLO.
1429 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1430 fixed by 4.65/MH/01 (is this a record?) are fixed:
1432 (i) An empty string was always treated as zero by the numeric comparison
1433 operators. This behaviour has been restored.
1435 (ii) It is documented that the numeric comparison operators always treat
1436 their arguments as decimal numbers. This was broken in that numbers
1437 starting with 0 were being interpreted as octal.
1439 While fixing these problems I realized that there was another issue that
1440 hadn't been noticed. Values of message_size_limit (both the global option
1441 and the transport option) were treated as octal if they started with 0.
1442 The documentation was vague. These values are now always treated as
1443 decimal, and I will make that clear in the documentation.
1449 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1450 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1453 MH/01 Don't check that the operands of numeric comparison operators are
1454 integers when their expansion is in "skipping" mode (fixes bug
1455 introduced by 4.64-PH/07).
1457 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1458 child addresses, Exim now panics and dies. Previously, because the count
1459 is held in a short int, deliveries were likely to be lost. As such a
1460 large number of recipients for a single message is ridiculous
1461 (performance will be very, very poor), I have chosen to impose a limit
1462 rather than extend the field.
1468 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1469 leftover -K file (the existence of which was triggered by #402).
1470 While we were at it, introduced process PID as part of the -K
1471 filename. This should rule out race conditions when creating
1474 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1475 processing considerably. Previous code took too long for large mails,
1476 triggering a timeout which in turn triggers #401.
1478 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1479 in the DK code in transports.c. sendfile() is not really portable,
1480 hence the _LINUX specificness.
1482 TF/01 In the add_headers option to the mail command in an Exim filter,
1483 there was a bug that Exim would claim a syntax error in any
1484 header after the first one which had an odd number of characters
1487 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1488 callout verification, Exim cached a "reject" for the entire domain. This
1489 is correct for most verifications, but it is not correct for a recipient
1490 verification with use_sender or use_postmaster set, because in that case
1491 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1492 case of MAIL FROM:<> rejection from other early rejections (e.g.
1493 rejection of HELO). When verifying a recipient using a non-null MAIL
1494 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1495 Whatever the result of the callout, the value of the domain cache is
1496 left unchanged (for any other kind of callout, getting as far as trying
1497 RCPT means that the domain itself is ok).
1499 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1502 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1503 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1504 session, the daemon ignores ECONNECT errors and logs others; it now
1505 ignores EPIPE as well.
1507 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1508 (quoted-printable decoding).
1510 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1511 later the small subsequent patch to fix an introduced bug.
1513 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1515 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1517 PH/08 An error is now given if message_size_limit is specified negative.
1519 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1520 to be given (somewhat) arbitrary names.
1522 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1525 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1526 miscellaneous code fixes
1528 PH/10 Added the log_reject_target ACL modifier to specify where to log
1531 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1532 hostname. This is wrong, because it relates to the incoming message (and
1533 probably the interface on which it is arriving) and not to the outgoing
1534 callout (which could be using a different interface). This has been
1535 changed to use the value of the helo_data option from the smtp transport
1536 instead - this is what is used when a message is actually being sent. If
1537 there is no remote transport (possible with a router that sets up host
1538 addresses), $smtp_active_hostname is used.
1540 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1541 tweaks were necessary in order to get it to work (see also 21 below):
1542 (a) The code assumed that strncpy() returns a negative number on buffer
1543 overflow, which isn't the case. Replaced with Exim's string_format()
1545 (b) There were several signed/unsigned issues. I just did the minimum
1546 hacking in of casts. There is scope for a larger refactoring.
1547 (c) The code used strcasecmp() which is not a standard C function.
1548 Replaced with Exim's strcmpic() function.
1549 (d) The code set only $1; it now sets $auth1 as well.
1550 (e) A simple test gave the error "authentication client didn't specify
1551 service in request". It would seem that Dovecot has changed its
1552 interface. Fortunately there's a specification; I followed it and
1553 changed what the client sends and it appears to be working now.
1555 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1558 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1559 address A is aliased to B and C, where B exists and C does not. Without
1560 -v the output is "A verified" because verification stops after a
1561 successful redirection if more than one address is generated. However,
1562 with -v the child addresses are also verified. Exim was outputting "A
1563 failed to verify" and then showing the successful verification for C,
1564 with its parentage. It now outputs "B failed to verify", showing B's
1565 parentage before showing the successful verification of C.
1567 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1568 look up a TXT record in a specific list after matching in a combined
1571 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1572 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1573 they consult the DNS. I had assumed they would set it the way they
1574 wanted; and indeed my experiments on Linux seem to show that in some
1575 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1576 To be on the safe side, however, I have now made the interface to
1577 host_find_byname() similar to host_find_bydns(), with an argument
1578 containing the DNS resolver options. The host_find_byname() function now
1579 sets these options at its start, just as host_find_bydns() does. The smtp
1580 transport options dns_qualify_single and dns_search_parents are passed to
1581 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1582 of host_find_byname() use the default settings of RES_DEFNAMES
1583 (qualify_single) but not RES_DNSRCH (search_parents).
1585 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1586 spool_read_header() do less string testing, by means of a preliminary
1587 switch on the second character of optional "-foo" lines. (This is
1588 overdue, caused by the large number of possibilities that now exist.
1589 Originally there were few.) While I was there, I also converted the
1590 str(n)cmp tests so they don't re-test the leading "-" and the first
1591 character, in the hope this might squeeze out yet more improvement.
1593 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1594 flag allowing group syntax was set by the header_syntax check but not
1595 turned off, possible causing trouble later; (2) The flag was not being
1596 set at all for the header_verify test, causing "group"-style headers to
1597 be rejected. I have now set it in this case, and also caused header_
1598 verify to ignore an empty address taken from a group. While doing this, I
1599 came across some other cases where the code for allowing group syntax
1600 while scanning a header line wasn't quite right (mostly, not resetting
1601 the flag correctly in the right place). These bugs could have caused
1602 trouble for malformed header lines. I hope it is now all correct.
1604 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1605 with the "reply" argument non-NULL. The code, however (which originally
1606 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1607 but it didn't always do it. This confused somebody who was copying the
1608 code for some other use. I have removed all the tests.
1610 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1611 feature that was used to support insecure browsers during the U.S. crypto
1612 embargo. It requires special client support, and Exim is probably the
1613 only MTA that supported it -- and would never use it because real RSA is
1614 always available. This code has been removed, because it had the bad
1615 effect of slowing Exim down by computing (never used) parameters for the
1616 RSA_EXPORT functionality.
1618 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1619 authenticator to fail if there's a tab character in the incoming data
1620 (there should never be unless someone is messing about, as it's supposed
1621 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1622 if the connection is using TLS or if the remote IP is the same as the
1623 local IP, and the "valid-client-cert option" if a client certificate has
1626 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1627 authenticators. This can be used for authorization after authentication
1628 succeeds. (In the case of plaintext, it servers for both authentication
1631 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1632 if any retry times were supplied.
1634 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1635 connection, where there is no client IP address to check. In this
1636 situation, the verify now always succeeds.
1638 PH/25 Applied John Jetmore's -Mset patch.
1640 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1642 PH/27 In a string expansion for a processed (not raw) header when multiple
1643 headers of the same name were present, leading whitespace was being
1644 removed from all of them, but trailing whitespace was being removed only
1645 from the last one. Now trailing whitespace is removed from each header
1646 before concatenation. Completely empty headers in a concatenation (as
1647 before) are ignored.
1649 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1650 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1652 PH/29 [Removed. This was a change that I later backed out, and forgot to
1653 correct the ChangeLog entry (that I had efficiently created) before
1654 committing the later change.]
1656 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1657 address errors (4xx response to RCPT) over the same connection as other
1658 messages routed to the same hosts. Such deliveries are always "forced",
1659 so retry times are not inspected. This resulted in far too many retries
1660 for the affected addresses. The effect occurred only when there were more
1661 hosts than the hosts_max_try setting in the smtp transport when it had
1662 the 4xx errors. Those hosts that it had tried were not added to the list
1663 of hosts for which the message was waiting, so if all were tried, there
1664 was no problem. Two fixes have been applied:
1666 (i) If there are any address or message errors in an SMTP delivery, none
1667 of the hosts (tried or untried) are now added to the list of hosts
1668 for which the message is waiting, so the message should not be a
1669 candidate for sending over the same connection that was used for a
1670 successful delivery of some other message. This seems entirely
1671 reasonable: after all the message is NOT "waiting for some host".
1672 This is so "obvious" that I'm not sure why it wasn't done
1673 previously. Hope I haven't missed anything, but it can't do any
1674 harm, as the worst effect is to miss an optimization.
1676 (ii) If, despite (i), such a delivery is accidentally attempted, the
1677 routing retry time is respected, so at least it doesn't keep
1678 hammering the server.
1680 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1681 in ${readsocket because some servers need this prod.
1683 PH/32 Added some extra debug output when updating a wait-xxx database.
1685 PH/33 The hint "could be header name not terminated by colon", which has been
1686 given for certain expansion errors for a long time, was not being given
1687 for the ${if def:h_colon_omitted{... case.
1689 PH/34 The spec says: "With one important exception, whenever a domain list is
1690 being scanned, $domain contains the subject domain." There was at least
1691 one case where this was not true.
1693 PH/35 The error "getsockname() failed: connection reset by peer" was being
1694 written to the panic log as well as the main log, but it isn't really
1695 panic-worthy as it just means the connection died rather early on. I have
1696 removed the panic log writing for the ECONNRESET error when getsockname()
1699 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1700 runs only) independently of the message's sender address. This meant
1701 that, if the 4xx error was in fact related to the sender, a different
1702 message to the same recipient with a different sender could confuse
1703 things. In particualar, this can happen when sending to a greylisting
1704 server, but other circumstances could also provoke similar problems.
1705 I have changed the default so that the retry time for these errors is now
1706 based a combination of the sender and recipient addresses. This change
1707 can be overridden by setting address_retry_include_sender=false in the
1710 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1711 remote server are returned as part of bounce messages. This was not
1712 happening for LMTP over a pipe (the lmtp transport), but now it is the
1713 same for both kinds of LMTP.
1715 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1716 in header lines that were in fact CNAMEs. This is no longer the case.
1718 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1719 and queue runs started by the daemon processed all messages. This has
1720 been fixed so that -R and -S can now usefully be given with -q<time>.
1722 PH/40 Import PCRE release 6.7 (fixes some bugs).
1724 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1726 PH/42 Give an error if -q is specified more than once.
1728 PH/43 Renamed the variables $interface_address and $interface_port as
1729 $received_ip_address and $received_port, to make it clear that these
1730 values apply to message reception, and not to the outgoing interface when
1731 a message is delivered. (The old names remain recognized, of course.)
1733 PH/44 There was no timeout on the connect() call when using a Unix domain
1734 socket in the ${readsocket expansion. There now is.
1736 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1737 be meaningful with "accept".
1739 SC/01 Eximstats V1.43
1740 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1742 SC/02 Eximstats V1.44
1743 Use a glob alias rather than an array ref in the generated
1744 parser. This improves both readability and performance.
1746 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1747 Collect SpamAssassin and rejection statistics.
1748 Don't display local sender or destination tables unless
1749 there is data to show.
1750 Added average volumes into the top table text output.
1752 SC/04 Eximstats V1.46
1753 Collect data on the number of addresses (recipients)
1754 as well as the number of messages.
1756 SC/05 Eximstats V1.47
1757 Added 'Message too big' to the list of mail rejection
1758 reasons (thanks to Marco Gaiarin).
1760 SC/06 Eximstats V1.48
1761 Mainlog lines which have GMT offsets and are too short to
1762 have a flag are now skipped.
1764 SC/07 Eximstats V1.49 (Alain Williams)
1765 Added the -emptyok flag.
1767 SC/08 Eximstats V1.50
1768 Fixes for obtaining the IP address from reject messages.
1770 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1771 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1772 whitesspace changes from 4.64-PH/27
1774 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1777 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1778 are found, allow negative numbers in numeric criteria)
1780 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1782 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1783 to match changes made in 4.64-PH/43
1785 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1787 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1788 "domain" is actually an IP address, and force a failure. This locks out
1789 those revolvers/nameservers that support "A-for-A" lookups, in
1790 contravention of the specifications.
1792 PH/48 When a host name was looked up from an IP address, and the subsequent
1793 forward lookup of the name timed out, the host name was left in
1794 $sender_host_name, contrary to the specification.
1796 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1797 restricted to single-key lookups, Exim was not diagnosing an error if
1798 * or *@ was used with a query-style lookup.
1800 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1802 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1803 long ago, but noone interested enough thought of it. Let's just say that
1804 the "1.1" means that there are some new functions that weren't there at
1805 some point in the past.
1807 PH/51 Error processing for expansion failure of helo_data from an smtp
1808 transport during callout processing was broken.
1810 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1811 tested/used via the -bh/-bhc/-bs options.
1813 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1814 bug, fixed in subsequent PCRE releases).
1816 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1817 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1819 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1825 SC/01 Use a glob alias rather than an array ref in eximstats generated
1826 parser. This improves both readability and performance.
1828 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1829 Don't display local sender or destination tables in eximstats unless
1830 there is data to show.
1831 Added average volumes into the eximstats top table text output.
1833 SC/03 Collect data on the number of addresses (recipients) as well
1834 as the number of messages in eximstats.
1836 TF/01 Correct an error in the documentation for the redirect router. Exim
1837 does (usually) call initgroups() when daemonizing.
1839 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1840 with consistent privilege compared to when running as a daemon.
1842 TF/03 Note in the spec that $authenticated_id is not set for local
1843 submissions from trusted users.
1845 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1846 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1848 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1849 by adding some example configuration directives to the default
1850 configuration file. A little bit of work is required to uncomment the
1851 directives and define how usernames and passwords are checked, but
1852 there is now a framework to start from.
1854 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1855 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1856 without this. I don't know how relevant this is to other LDAP libraries.
1858 PH/02 Add the verb name to the "unknown ACL verb" error.
1860 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1862 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1864 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1865 directory not expanded when it should be if an expanded home directory
1866 was set for the address (which is overridden by the transport).
1868 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1871 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1872 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1873 because it is too late at that time, and has no effect.
1875 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1876 security issue with \' (bugzilla #107). I could not use the
1877 PQescapeStringConn() function, because it needs a PGconn value as one of
1880 PH/08 When testing addresses using -bt, indicate those final addresses that
1881 are duplicates that would not cause an additional delivery. At least one
1882 person was confused, thinking that -bt output corresponded to deliveries.
1883 (Suppressing duplicates isn't a good idea as you lose the information
1884 about possibly different redirections that led to the duplicates.)
1886 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1887 systems where poll() doesn't work, in particular OS X.
1889 PH/10 Added more information to debugging output for retry time not reached.
1891 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1892 operations in malware.c.
1894 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1897 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1898 syslog_duplication set false, log lines that would normally be written
1899 both the the main log and to the reject log were not written to syslog at
1902 PH/14 In the default configuration, change the use of "message" in ACL warn
1903 statements to "add_header".
1905 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1906 not followed by a command (e.g. "seen endif").
1908 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1909 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1912 PH/17 Added extra conditions to the default value of delay_warning_condition
1916 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1917 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1918 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1921 The Auto-Submitted: and various List- headers are standardised, whereas I
1922 don't think Precedence: ever was.
1924 PH/18 Refactored debugging code in route_finduser() to show more information,
1925 in particular, the error code if getpwnam() issues one.
1927 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1928 This is apparently needed in addition to the PH/07 change above to avoid
1929 any possible encoding problems.
1931 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1932 but not after initializing Perl.
1934 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1935 output them only if debugging. By default they are written stderr,
1936 apparently, which is not desirable.
1938 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1941 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1944 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1946 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1947 authenticated or an ident call has been made. Suppress the default
1948 values for $authenticated_id and $authenticated_sender (but permit -oMai
1949 and -oMas) when testing with -bh.
1951 PH/24 Re-jigged the order of the tests in the default configuration so that the
1952 tests for valid domains and recipients precede the DNS black list and CSA
1953 tests, on the grounds that those ones are more expensive.
1955 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1956 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1957 This bug exists in every version of Exim that I still have, right back to
1960 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1961 However, an attempt to turn on case-sensitivity in a regex key by
1962 including (?-i) didn't work because the subject string was already
1963 lowercased, and the effects were non-intuitive. It turns out that a
1964 one-line patch can be used to allow (?-i) to work as expected.
1970 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1971 other effects) broke the use of negated acl sub-conditions.
1973 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1976 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1977 "Deny" causes Exim to reject the incoming connection with a 554 error.
1978 Unfortunately, if there is a major crisis, such as a disk failure,
1979 tcp-wrappers gives "deny", whereas what one would like would be some
1980 kind of temporary error. A kludge has been added to help with this.
1981 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1982 554 error is used if errno is still zero or contains ENOENT (which occurs
1983 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1986 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1988 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1989 errors. Otherwise a message that provokes a temporary error (when other
1990 messages do not) can cause a whole host to time out.
1992 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1993 addresses were routed directly to files or pipes from a redirect router.
1994 File deliveries just didn't batch; pipe deliveries might have suffered
1997 PH/06 A failure to get a lock for a hints database would erroneously always say
1998 "Failed to get write lock", even when it was really a read lock.
2000 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2001 of 0600. This has been changed to use the value of the lockfile_mode
2002 option (which defaults to 0600).
2004 PH/08 Applied small patch from the Sieve maintainer.
2006 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2007 folder from quota calculations, a direct delivery into this folder messed
2008 up the contents of the maildirsize file. This was because the regex was
2009 used only to exclude .Trash (or whatever) when the size of the mailbox
2010 was calculated. There was no check that a delivery was happening into an
2011 excluded directory. This bug has been fixed by ignoring all quota
2012 processing for deliveries into excluded directories.
2014 PH/10 Added the maildirfolder_create_regex option to appendfile.
2020 PH/01 The code for finding all the local interface addresses on a FreeBSD
2021 system running IPv6 was broken. This may well have applied to all BSD
2022 systems, as well as to others that have similar system calls. The broken
2023 code found IPv4 interfaces correctly, but gave incorrect values for the
2024 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2025 that it would not match correctly against @[] and not recognize the IPv6
2028 PH/02 The ipliteral router was not recognizing addresses of the form user@
2029 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2031 PH/03 Added disable_ipv6.
2033 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2034 decoded form, because it is most often used to construct To: headers
2035 lines in autoreplies, and the decoded form may well be syntactically
2036 invalid. However, $reply_address has leading white space removed, and all
2037 newlines turned into spaces so that the autoreply transport does not
2040 PH/05 If group was specified without a user on a router, and no group or user
2041 was specified on a transport, the group from the router was ignored.
2043 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2044 for visible compile-time settings that can be used to change these
2045 numbers, for those that want even more. Backwards compatibility with old
2046 spool files has been maintained. However, going back to a previous Exim
2047 release will lost any variables that are in spool files.
2049 PH/07 Two small changes when running in the test harness: increase delay when
2050 passing a TCP/IP connection to a new process, in case the original
2051 process has to generate a bounce, and remove special handling of
2052 127.0.0.2 (sic), which is no longer necessary.
2054 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2055 be the same on different OS.
2057 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2060 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2061 whether --show-vars was specified or not
2063 JJ/02 exipick: Added support for new ACL variable spool format introduced
2066 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2067 syntactically invalid From: or Reply-to: line, and a filter used this to
2068 generate an autoreply, and therefore failed to obtain an address for the
2069 autoreply, Exim could try to deliver to a non-existent relative file
2070 name, causing unrelated and misleading errors. What now happens is that
2071 it logs this as a hard delivery error, but does not attempt to create a
2074 PH/11 The exinext utility has a -C option for testing purposes, but although
2075 the given file was scanned by exinext itself; it wasn't being passed on
2076 when Exim was called.
2078 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2079 an end-of-file indication when reading a command response.
2081 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2082 compiled. In many other places in Exim, IPv6 addresses are always
2083 recognized, so I have changed this. It also means that IPv4 domain
2084 literals of the form [IPV4:n.n.n.n] are now always recognized.
2086 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2087 used if the router is not running as root, for example, when verifying at
2088 ACL time, or when using -bh. The debugging output from this situation was
2089 non-existent - all you got was a failure to exec. I have made two
2092 (a) Failures to set uid/gid, the current directory, or a process leader
2093 in a subprocess such as that created by queryprogram now generate
2094 suitable debugging ouput when -d is set.
2096 (b) The queryprogram router detects when it is not running as root,
2097 outputs suitable debugging information if -d is set, and then runs
2098 the subprocess without attempting to change uid/gid.
2100 PH/15 Minor change to Makefile for building test_host (undocumented testing
2103 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2104 additional section of a DNS packet that returns MX or SRV records.
2105 Instead, it always explicitly searches for A/AAAA records. This avoids
2106 major problems that occur when a DNS server includes only records of one
2107 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2108 fixed another bug: if SRV records were looked up and the corresponding
2109 address records were *not* found in the additional section, the port
2110 values from the SRV records were lost.
2112 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2113 using the correct key (the original address) when searching the retry
2114 rules in order to find which one to use for generating the retry hint.
2116 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2117 adding the default one. Similarly, if it contains a Reply-To: header, the
2118 errors_reply_to option, if set, is not used.
2120 PH/19 When calculating a retry time, Exim used to measure the "time since
2121 failure" by looking at the "first failed" field in the retry record. Now
2122 it does not use this if it is later than than the arrival time of the
2123 message. Instead it uses the arrival time. This makes for better
2124 behaviour in cases where some deliveries succeed, thus re-setting the
2125 "first failed" field. An example is a quota failure for a huge message
2126 when small messages continue to be delivered. Without this change, the
2127 "time since failure" will always be short, possible causing more frequent
2128 delivery attempts for the huge message than are intended.
2129 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2131 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2132 $1, $2, $3) because the numerical variables can be reset during some
2133 expansion items (e.g. "match"), thereby losing the authentication data.
2135 PH/21 Make -bV show the size of off_t variables so that the test suite can
2136 decide whether to run tests for quotas > 2G.
2138 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2139 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2140 filecount value is greater than 2G or if a quota value is greater than 2G
2141 on a system where the size of off_t is not greater than 4, a panic error
2144 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2145 never match. The debug and -bh output now contains an explicit error
2146 message indicating a malformed IPv4 address or mask.
2148 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2149 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2150 PH/23 above applies.
2152 PH/25 Do not write to syslog when running in the test harness. The only
2153 occasion when this arises is a failure to open the main or panic logs
2154 (for which there is an explicit test).
2156 PH/26 Added the /no_tell option to "control=freeze".
2158 PH/27 If a host name lookup failed very early in a connection, for example, if
2159 the IP address matched host_lookup and the reverse lookup yielded a name
2160 that did not have a forward lookup, an error message of the form "no IP
2161 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2162 could be logged. Now it outputs the IP address instead of "NULL".
2164 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2165 allows the sender and the authenticated sender to be set when
2166 submitting a message from within Exim. Since child_open_exim() is
2167 documented for local_scan(), the new function should be too.
2169 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2170 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2171 results in an empty string is now treated as unset.
2173 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2175 PH/31 Added sender_verify_fail logging option.
2177 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2178 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2179 tidied the source and removed it altogether.
2181 PH/33 When a queue run was abandoned because the load average was too high, a
2182 log line was always written; now it is written only if the queue_run log
2183 selector is set. In addition, the log line for abandonment now contains
2184 information about the queue run such as the pid. This is always present
2185 in "start" and "stop" lines but was omitted from the "abandon" line.
2187 PH/34 Omit spaces between a header name and the colon in the error message that
2188 is given when verify = headers_syntax fails (if there are lots of them,
2189 the message gets confusing).
2191 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2192 names, as there are now some PTR records that contain slashes. This check
2193 is only to protect against broken name servers that fall over on strange
2194 characters, so the fact that it applies to all lookups doesn't matter.
2196 PH/36 Now that the new test suite is complete, we can remove some of the
2197 special code in Exim that was needed for the old test suite. For example,
2198 sorting DNS records because real resolvers return them in an arbitrary
2199 order. The new test suite's fake resolver always returns records in the
2202 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2203 bounces) except when queue_only is set, to avoid logging races between
2204 the different processes.
2206 PH/38 Panic-die if .include specifies a non-absolute path.
2208 PH/39 A tweak to the "H" retry rule from its user.
2210 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2211 a label. They prevented compilation on older perls.
2213 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2214 a warning to be raised on newish perls.
2216 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2217 on queue. Changes to match documented behaviour of showing count of
2218 messages matching specified criteria.
2220 PH/40 Changed the default ident timeout from 30s to 5s.
2222 PH/41 Added support for the use of login_cap features, on those BSD systems
2223 that have them, for controlling the resources used by pipe deliveries.
2225 PH/42 The content-scanning code uses fopen() to create files in which to put
2226 message data. Previously it was not paying any attention to the mode of
2227 the files. Exim runs with umask(0) because the rest of the code creates
2228 files with open(), and sets the required mode explicitly. Thus, these
2229 files were ending up world-writeable. This was not a big issue, because,
2230 being within the spool directory, they were not world-accessible. I have
2231 created a function called modefopen, which takes an additional mode
2232 argument. It sets umask(777), creates the file, chmods it to the required
2233 mode, then resets the umask. All the relevant calls to fopen() in the
2234 content scanning code have been changed to use this function.
2236 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2237 to 24 hours. This avoids potential overflow problems when processing G
2238 and H retry rules. I suspect nobody ever tinkers with this value.
2240 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2242 PH/45 When the plaintext authenticator is running as a client, the server's
2243 challenges are checked to ensure they are valid base64 strings. By
2244 default, the authentication attempt is cancelled if an invalid string is
2245 received. Setting client_ignore_invalid_base64 true ignores these errors.
2246 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2247 they are received. Thus, the responses can be made to depend on the
2248 challenges. If an invalid string is ignored, an empty string is placed in
2251 PH/46 Messages that are created by the autoreply transport now contains a
2252 References: header, in accordance with RFCs 2822 and 3834.
2254 PH/47 Added authenticated_sender_force to the smtp transport.
2256 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2258 PH/49 Installed latest patch from the Sieve maintainer.
2260 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2261 was also set, the appendfile transport was unnecessarily scanning a
2262 directory of message files (e.g. for maildir delivery) to find the count
2263 of files (along with the size), even though it did not need this
2264 information. It now does the scan only if it needs to find either the
2265 size of the count of files.
2267 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2269 PH/52 Two bugs concerned with error handling when the smtp transport is
2272 (i) Exim was not creating retry information for temporary errors given
2273 for individual recipients after the DATA command when the smtp transport
2274 was used in LMTP mode. This meant that they could be retried too
2275 frequently, and not timed out correctly.
2277 (ii) Exim was setting the flag that allows error details to be returned
2278 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2279 recipients that were returned after the DATA command.
2281 PH/53 This is related to PH/52, but is more general: for any failing address,
2282 when detailed error information was permitted to be returned to the
2283 sender, but the error was temporary, then after the final timeout, only
2284 "retry timeout exceeded" was returned. Now it returns the full error as
2285 well as "retry timeout exceeded".
2287 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2288 do this, and (what is worse) MTAs that accept it.
2290 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2291 will now be deprecated.
2293 PH/56 New os.c-cygwin from the Cygwin maintainer.
2295 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2296 formats (previously only available in exim formats via -bpr, -bpru,
2297 and -bpra. Now also available in native and exiqgrep formats)
2299 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2300 with very large, slow to parse queues
2302 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2304 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2306 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2307 responses to authentication challenges, though it was showing the
2308 challenges; (ii) I've removed the CR characters from the debug output for
2311 PH/58 Allow for the insertion of a newline as well as a space when a string
2312 is turned into more than one encoded-word during RFC 2047 encoding. The
2313 Sieve code now uses this.
2315 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2316 data_4xx, lost_connection, tls_required.
2318 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2319 message was being sent as an SMTP response.
2321 PH/61 Add -l and -k options to exicyclog.
2323 PH/62 When verifying, if an address was redirected to one new address, so that
2324 verification continued, and the new address failed or deferred after
2325 having set something in $address_data, the value of $address_data was not
2326 passed back to the ACL. This was different to the case when no
2327 redirection occurred. The value is now passed back in both cases.
2329 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2330 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2331 use login_cap.h, so on its own it isn't the distinguishing feature. The
2332 new name refers directly to the setclassresources() function.
2334 PH/65 Added configuration files for NetBSD3.
2336 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2338 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2339 is preferred over IPv4.
2341 PH/68 The bounce_return_message and bounce_return_body options were not being
2342 honoured for bounces generated during the reception of non-SMTP messages.
2343 In particular, this applied to messages rejected by the ACL. This bug has
2344 been fixed. However, if bounce_return_message is true and bounce_return_
2345 body is false, the headers that are returned for a non-SMTP message
2346 include only those that have been read before the error was detected.
2347 (In the case of an ACL rejection, they have all been read.)
2349 PH/69 The HTML version of the specification is now built in a directory called
2350 spec_html instead of spec.html, because the latter looks like a path with
2351 a MIME-type, and this confuses some software.
2353 PH/70 Catch two compiler warnings in sieve.c.
2355 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2356 function verify_get_ident() calls ip_connect() to connect a socket, but
2357 if the "connect()" function timed out, ip_connect() used to close the
2358 socket. However, verify_get_ident() also closes the socket later, and in
2359 between Exim writes to the log, which may get opened at this point. When
2360 the socket was closed in ip_connect(), the log could get the same file
2361 descriptor number as the socket. This naturally causes chaos. The fix is
2362 not to close the socket in ip_connect(); the socket should be closed by
2363 the function that creates it. There was only one place in the code where
2364 this was missing, in the iplookup router, which I don't think anybody now
2365 uses, but I've fixed it anyway.
2367 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2368 well as to direct DNS lookups. Otherwise the handling of names in host
2369 lists is inconsistent and therefore confusing.
2375 PH/01 Two changes to the default runtime configuration:
2377 (1) Move the checks for relay_from_hosts and authenticated clients from
2378 after to before the (commented out) DNS black list checks.
2380 (2) Add control=submission to the relay_from_hosts and authenticated
2381 clients checks, on the grounds that messages accepted by these
2382 statements are most likely to be submissions.
2384 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2386 (1) Generate an error if the third argument for the ${prvs expansion is
2389 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2392 (3) Reset the variables that are obtained from the first argument of
2393 ${prvscheck and used in the second argument before leaving the code,
2394 because their memory is reclaimed, so using them afterwards may do
2397 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2398 one (it's much easier than Tom thought :-).
2400 (5) Because of (4), we can now allow for the use of $prvscheck_result
2401 inside the third argument.
2403 PH/03 For some reason, the default setting of PATH when running a command from
2404 a pipe transport was just "/usr/bin". I have changed it to
2407 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2408 anything to be listed in the output from -bV.
2410 PH/05 When a filter generated an autoreply, the entire To: header line was
2411 quoted in the delivery log line, like this:
2413 => >A.N.Other <ano@some.domain> <original@ddress> ...
2415 This has been changed so that it extracts the operative address. There
2416 may be more than one such address. If so, they are comma-separated, like
2419 => >ano@some.domain,ona@other.domain <original@ddress> ...
2421 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2422 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2423 not being looked up in the rDNS to get a host name, Exim was showing the
2424 IP address twice in Received: lines, even though the IP addresses were
2425 identical. For example:
2427 Received: from [1.2.3.4] (helo=[1.2.3.4])
2429 However, if the real host name was known, it was omitting the HELO data
2430 if it matched the actual IP address. This has been tidied up so that it
2431 doesn't show the same IP address twice.
2433 PH/07 When both +timestamp and +memory debugging was on, the value given by
2434 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2435 called by the debug printing, thereby overwriting the timestamp buffer.
2436 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2438 PH/08 When the original message was included in an autoreply transport, it
2439 always said "this is a copy of the message, including all the headers",
2440 even if body_only or headers_only was set. It now gives an appropriate
2443 PH/09 Applied a patch from the Sieve maintainer which:
2445 o fixes some comments
2446 o adds the (disabled) notify extension core
2447 o adds some debug output for the result of if/elsif tests
2448 o points to the current vacation draft in the documentation
2449 and documents the missing references header update
2453 o fixes a bug in processing the envelope test (when testing
2454 multiple envelope elements, the last element determinted the
2457 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2458 Electronic Mail") by including:
2460 Auto-submitted: auto-generated
2462 in the messages that it generates (bounce messages and others, such as
2463 warnings). In the case of bounce messages for non-SMTP mesages, there was
2464 also a typo: it was using "Auto_submitted" (underscore instead of
2465 hyphen). Since every message generated by Exim is necessarily in response
2466 to another message, thes have all been changed to:
2468 Auto-Submitted: auto-replied
2470 in accordance with these statements in the RFC:
2472 The auto-replied keyword:
2474 - SHOULD be used on messages sent in direct response to another
2475 message by an automatic process,
2477 - MUST NOT be used on manually-generated messages,
2479 - MAY be used on Delivery Status Notifications (DSNs) and Message
2480 Disposition Notifications (MDNs),
2482 - MUST NOT be used on messages generated by automatic or periodic
2483 processes, except for messages which are automatic responses to
2486 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2487 to the default Received: header definition.
2489 PH/12 Added log selector acl_warn_skipped (default on).
2491 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2492 variables because (a) they are in the wrong storage pool and (b) even if
2493 they were copied, it wouldn't work properly because of the caching.
2495 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2496 checking when decoding. Apparently there are clients that generate
2497 overlong encoded strings. Why am I not surprised?
2499 PH/15 If the first argument of "${if match_address" was not empty, but did not
2500 contain an "@" character, Exim crashed. Now it writes a panic log message
2501 and treats the condition as false.
2503 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2505 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2506 extension "envelope-auth". The code is finished and in agreement with
2507 other implementations, but there is no documentation so far and in fact,
2508 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2509 not changing the active code.
2511 Print executed "if" and "elsif" statements when debugging is used. This
2512 helps a great deal to understand what a filter does.
2514 Document more things not specified clearly in RFC3028. I had all this
2515 sorted out, when out of a sudden new issues came to my mind. Oops."
2517 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2520 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2521 canonical form (as documented). However, after a host name lookup from
2522 the IP address, check_host() was doing a simple string comparison with
2523 addresses acquired from the DNS when checking that the found name did
2524 have the original IP as one of its addresses. Since any found IPv6
2525 addresses are likely to be in abbreviated form, the comparison could
2526 fail. Luckily, there already exists a function for doing the comparison
2527 by converting both addresses to binary, so now that is used instead of
2528 the text comparison.
2530 PH/20 There was another similar case to PH/19, when a complete host name was
2531 given in a host list; looking up its IP address could give an abbreviated
2532 form, whereas the current host's name might or might not be abbreviated.
2533 The same fix has been applied.
2539 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2540 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2543 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2544 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2546 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2548 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2549 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2550 Also, for "Auto-Submitted", treat anything other than "no" as
2551 non-personal, in accordance with RFC 3834. (Previously it treated
2552 anything starting "auto-" as non-personal.)
2554 TF/01 The control=submission/name=... option had a problem with syntax
2555 errors if the name included a slash character. The /name= option
2556 now slurps the rest of the string, so it can include any characters
2557 but it must come last in the list of options (after /sender_retain
2560 PH/05 Some modifications to the interface to the fake nameserver for the new
2568 TK/01 Added the "success_on_redirect" address verification option. See
2569 NewStuff for rationale and an example.
2571 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2573 PH/02 Patch to exigrep to allow it to work on syslog lines.
2575 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2576 fread() to skip over the body file's header line, because in Cygwin the
2577 header line is locked and is inaccessible.
2579 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2580 co-exist for some time) to make it clear that it is the Exim ID that is
2581 referenced, not the Message-ID: header line.
2583 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2584 string_format() function, because snprintf() does not exist on all
2587 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2588 ${stat: expansion item.
2590 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2591 protocol synchronization error", to keep the pedants happy.
2593 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2594 well as for IRIX systems, when gcc is being used. See the host.c source
2597 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2599 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2602 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2603 transport and to the smtp transport in LMTP mode.
2605 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2607 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2608 run a filter in a subprocess. This could lead to confusion in subsequent
2609 lookups in the parent process. There should also be a search_tidyup() at
2610 the end of the subprocess.
2612 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2613 only if the host matched helo_try_verify_hosts, which caused the
2614 verification to occur when the EHLO/HELO command was issued. The ACL just
2615 tested the remembered result. Now, if a previous verification attempt has
2616 not happened, "verify = helo" does it there and then.
2618 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2620 TK/03 Fix log output including CR from clamd.
2622 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2623 exist provoked a memory error which could cause a segfault.
2625 PH/15 Installed PCRE 6.2
2627 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2629 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2630 of the problem. Specifically, suggested +O2 rather than +O1 for the
2633 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2635 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2636 record, the port was not correctly logged when the outgoing_port log
2637 selector was set (it logged the transort's default port).
2639 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2640 fallback_hosts, and "hosts" in the smtp transport.
2642 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2643 host errors such as "Connection refused".
2645 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2646 authentication with radiusclient 0.4.9:
2648 - Error returned from rc_read_config was caught wrongly
2649 - Username/password not passed on to radius server due to wrong length.
2651 The presumption is that some radiusclient API changes for 4.51/PH/17
2652 were not taken care of correctly. The code is still untested by me (my
2653 Linux distribution still has 0.3.2 of radiusclient), but it was
2654 contributed by a Radius user.
2656 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2657 expanding the "port" option of the smtp transport.
2659 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2660 while reading a MIME header. Thanks to Tom Hughes for a patch.
2662 PH/24 Include config.h inside local_scan.h so that configuration settings are
2665 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2666 that in an ACL for RCPT (for example), you can examine exactly what was
2669 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2670 commands, but it was not correctly comparing the address with the actual
2671 client host address. Thus, it would show the EHLO address in Received:
2672 header lines when this was not necessary.
2674 PH/27 Added the % operator to ${eval:}.
2676 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2677 it should be ignoring failures (because with -C, for example, it has lost
2678 privilege). It wasn't ignoring creation failures other than "already
2681 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2684 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2685 because an input line was too long, either on its own, or by virtue of
2686 too many continuations, the temporary file was not being removed, and the
2687 return code was incorrect.
2689 PH/31 Missing "BOOL" in function definition in filtertest.c.
2691 PH/32 Applied Sieve patches from the maintainer.
2693 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2695 PH/33 Added "verify = not_blind".
2697 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2698 Local/Makefile (with some defaults set). These are used in built scripts
2699 such as exicyclog, but they have never been used in the exim_install
2700 script (though there are many overriding facilities there). I have
2701 arranged that the exim_install script now takes note of these two
2704 PH/35 Installed configuration files for Dragonfly.
2706 PH/36 When a locally submitted message by a trusted user did not contain a
2707 From: header, and the sender address was obtained from -f or from an SMTP
2708 MAIL command, and the trusted user did not use -F to supply a sender
2709 name, $originator_name was incorrectly used when constructing a From:
2710 header. Furthermore, $originator_name was used for submission mode
2711 messages from external hosts without From: headers in a similar way,
2712 which is clearly wrong.
2714 PH/37 Added control=suppress_local_fixups.
2716 PH/38 When log_selector = +received_sender was set, and the addition of the
2717 sender made the log line's construction buffer exactly full, or one byte
2718 less than full, an overflow happened when the terminating "\n" was
2721 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2722 when the result of a list match is failure because a DNS lookup failed.
2724 PH/40 RM_COMMAND is now used in the building process.
2726 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2727 the "build-* directories that it finds.
2729 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2730 address in a domain literal was a prefix of an interface address.
2732 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2733 when verifying a sender address, unless rewrite_headers is false.
2735 PH/44 Wrote a long comment about why errors_to addresses are verified as
2736 recipients, not senders.
2738 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2739 the ratelimit ACL was added.
2741 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2743 PH/46 Added extra information about PostgreSQL errors to the error string.
2745 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2746 avoiding the need to install special zones in a real server. This is
2747 backwards compatible; if it can't find the fake resolver, it drops back.
2748 Thus, both old and new test suites can be run.
2750 TF/02 Added util/ratelimit.pl
2752 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2753 clock is set back in time.
2755 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2756 Candler <B.Candler@pobox.com>.
2758 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2759 OK for addresses that are the result of redirections.
2761 PH/48 A number of further additions for the benefit of the new test suite,
2762 including a fake gethostbyname() that interfaces to the fake DNS resolver
2765 TF/06 The fix for widen_domains has also been applied to qualify_single and
2766 search_parents which are the other dnslookup options that can cause
2769 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2772 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2774 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2775 service" if it encountered an MX record with an empty target hostname.
2776 The message is now "an MX or SRV record indicated no SMTP service".
2778 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2779 if the DNS of the sending site is misconfigured. This is quite a
2780 common situation. This change restores the behaviour of treating a
2781 helo verification defer as a failure.
2783 PH/51 If self=fail was set on a router, the bounce message did not include the
2784 actual error message.
2790 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2792 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2793 command itself ended in error, the underlying message about the transport
2794 filter timeout was being overwritten with the pipe command error. Now the
2795 underlying error message should be appended to the second error message.
2797 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2799 PH/02 Reduce the amount of output that "make" produces by default. Full output
2800 can still be requested.
2802 PH/03 The warning log line about a condition test deferring for a "warn" verb
2803 was being output only once per connection, rather than after each
2804 occurrence (because it was using the same function as for successful
2805 "warn" verbs). This seems wrong, so I have changed it.
2807 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2808 it should not have, which might have caused a crash in the right
2809 circumstances, but probably never did.
2811 PH/04 Installed a modified version of Tony Finch's patch to make submission
2812 mode fix the return path as well as the Sender: header line, and to
2813 add a /name= option so that you can make the user's friendly name appear
2816 TF/03 Added the control = fakedefer ACL modifier.
2818 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2819 Mark Lowes for thorough testing.
2821 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2823 TK/03 Merged latest SRS patch from Miles Wilton.
2825 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2826 with the definition in sysexits.h (which is #included earlier).
2827 Fortunately, Exim does not actually use EX_OK. The code used to try to
2828 preserve the sysexits.h value, by assumimg that macro definitions were
2829 scanned for macro replacements. I have been disabused of this notion,
2830 so now the code just undefines EX_OK before #including unistd.h.
2832 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2833 in the smtp transport. When a block could not be written in a single
2834 write() function, the timeout was being re-applied to each part-write.
2835 This seems wrong - if the receiver was accepting one byte at a time it
2836 would take for ever. The timeout is now adjusted when this happens. It
2837 doesn't have to be particularly precise.
2839 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2840 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2842 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2843 without a domain if the check to <postmaster@domain> fails.
2845 SC/01 Eximstats: added -xls and the ability to specify output files
2846 (patch written by Frank Heydlauf).
2848 SC/02 Eximstats: use FileHandles for outputing results.
2850 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2852 SC/04 Eximstats: fixed display of large numbers with -nvr option
2854 SC/05 Eximstats: fixed merging of reports with empty tables.
2856 SC/06 Eximstats: added the -include_original_destination flag
2858 SC/07 Eximstats: removed tabs and trailing whitespace.
2860 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2862 TK/06 MBOX spool code: Add real "From " MBOX separator line
2863 so the .eml file is really in mbox format (even though
2864 most programs do not really care). Patch from Alex Miller.
2866 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2867 The latter is generated from $received_to and is only set if the
2868 message has one envelope recipient. SA can use these headers,
2869 obviously out-of-the-box. Patch from Alex Miller.
2871 PH/08 The ${def test on a variable was returning false if the variable's
2872 value was "0", contrary to what the specification has always said!
2873 The result should be true unless the variable is empty.
2875 PH/09 The syntax error of a character other than { following "${if
2876 def:variable_name" (after optional whitespace) was not being diagnosed.
2877 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2878 accidental colon was present, for example, could give incorrect results.
2880 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2881 result is used (not including appendfile, where other changes are about
2884 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2885 This involved changing a lot of size variables from int to off_t. It
2886 should work with maildirs and everything.
2888 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2889 spamd dying while we are connected to it.
2891 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2894 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2895 function for BDB 4.3.
2897 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2899 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2900 the sources for PCRE have been re-arranged and more files are now
2903 PH/15 The code I had for printing potentially long long variables in PH/11
2904 above was not the best (it lost precision). The length of off_t variables
2905 is now inspected at build time, and an appropriate printing format (%ld
2906 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2907 to be "long long int" or "long int". This is needed for the internal
2908 formatting function string_vformat().
2910 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2911 the configuration file to be ":syslog", then the script "guesses" where
2912 the logs files are, rather than using the compiled in default. In our
2913 case the guess is not the same as the compiled default, so the script
2914 suddenly stopped working when I started to use syslog. The patch checks
2915 to see if log_file_path is "". If so, it attempts to read it from exim
2916 with no configuration file to get the compiled in version, before it
2917 falls back to the previous guessing code."
2919 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2920 implementing BATV in an Exim configuration. See NewStuff for the gory
2923 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2924 Makefile that are specific to HP-UX.
2926 PH/18 If the "use_postmaster" option was set for a recipient callout together
2927 with the "random" option, the postmaster address was used as the MAIL
2928 FROM address for the random test, but not for the subsequent recipient
2929 test. It is now used for both.
2931 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2932 patch removes a few documentation additions to RFC 3028, because the
2933 latest draft now contains them. It adds the new en;ascii-case comparator
2934 and a new error check for 8bit text in MIME parts. Comparator and
2935 require names are now matched exactly. I enabled the subaddress
2936 extension, but it is not well tested yet (read: it works for me)."
2938 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2939 rework some of the code of TK/09 above to avoid the hardwired use of
2940 "%lld" and "long long". Replaced the call to snprintf() with a call to
2943 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2944 records point to non-existent hosts", "retry timeout exceeded", and
2945 "retry time not reached for any host after a long failure period".
2947 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2948 experimental DomainKeys support:
2950 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2951 (2) On an error such as an illegally used "control", the wrong name for
2952 the control was given.
2954 These problems did NOT occur unless DomainKeys support was compiled.
2956 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2958 PH/24 Added ${if match_ip condition.
2960 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2961 databases so that it will be absolutely obvious if a crash occurs in the
2962 DB library. This is a regular occurrence (often caused by mis-matched
2965 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2966 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2967 picked up on a user's system that detects such things. There doesn't seem
2968 to be a gcc warning option for this - only an attribute that has to be
2969 put on the function's prototype. It seems that in Fedora Core 4 they have
2970 set this on a number of new functions. No doubt there will be more in due
2973 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2974 specify a transport. However, if an address that was verified by such a
2975 router was the subject of a callout, Exim crashed because it tried to
2976 read the rcpt_include_affixes from the non-existent transport. Now it
2977 just assumes that the setting of that option is false. This bug was
2978 introduced by 4.51/PH/31.
2980 PH/28 Changed -d+all to exclude +memory, because that information is very
2981 rarely of interest, but it makes the output a lot bigger. People tend to
2982 do -d+all out of habit.
2984 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2985 code in os-type was giving problems when libc.so lives in lib64, like on
2988 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2989 aren't the modern standard, and it seems that some systems' include files
2990 don't always have them. Exim was already checking for some of the newer
2991 ones like T_AAAA, and defining it itself. I've added checks for all the
2992 record types that Exim uses.
2994 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2995 not automatically generating a new one, as it is supposed to. This
2996 prevented TLS from working. If the file did exist, but contained invalid
2997 data, a new version was generated, as expected. It was only the case of a
2998 non-existent file that was broken.
3000 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3001 with a change in libdomainkeys > 0.64.
3003 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3004 from DNS. If the selector record carries the flag, it now has
3005 precedence over the domain-wide flag.
3007 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3009 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3010 the use of a port name that isn't defined in /etc/services) occurred, the
3011 message was deferred as in a normal delivery, and thus remained on the
3012 spool, instead of being failed because of the mua_wrapper setting. This
3013 is now fixed, and I tidied up some of the mua_wrapper messages at the
3016 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3017 the messages in a hash of arrays rather than using individual hashes.
3018 This is a bit cleaner and results in dramatic memory savings, albeit
3019 at a slight CPU cost.
3021 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3022 as requested by Marc Sherman.
3024 SC/10 Eximstats: added histograms for user specified patterns as requested
3027 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3029 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3030 fopen() in the content-scanning modules that did not already have it.
3036 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3037 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3039 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3041 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3043 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3044 to test Sieve filters that use "vacation".
3046 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3047 that changes the way the GnuTLS parameters are stored in the cache file.
3048 The new format can be generated externally. For backward compatibility,
3049 if the data in the cache doesn't make sense, Exim assumes it has read an
3050 old-format file, and it generates new data and writes a new file. This
3051 means that you can't go back to an older release without removing the
3054 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3055 work if there are any delivery delays because "one_time" forces the
3056 parent to be marked "delivered", so its unseen clone is never tried
3057 again. For this reason, Exim now forbids the simultaneous setting of
3060 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3061 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3062 entry if you want to know the details. The fix, however, neglected to
3063 consider the case when local delivery batching is involved. The test for
3064 "previously delivered" was not happening when checking to see if an
3065 address could be batched with a previous (undelivered) one; under
3066 certain circumstances this could lead to multiple deliveries to the same
3069 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3070 in its include files, and this causes problems building Exim.
3072 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3073 header_syntax) but Exim was just ignoring anything given after a slash.
3074 In particular, this caused confusion with an attempt to use "verify =
3075 reverse_host_lookup/defer_ok". An error is now given when options are
3076 supplied for verify items that do not have them. (Maybe reverse_host_
3077 lookup should have a defer_ok option, but that's a different point.)
3079 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3080 defined by RFC 821) to 2048, because there were problems with some AUTH
3081 commands, and RFC 1869 says the size should be increased for extended
3082 SMTP commands that take arguments.
3084 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3087 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3088 "unknown" error; now it says that the functionality isn't in the binary.
3090 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3091 an address' error message when a string expansion fails (syntax or
3092 whatever). Otherwise the password may appear in the log. Following change
3093 PH/42 below, there is no longer a chance of it appearing in a bounce
3096 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3098 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3099 of its addresses was ever tried. (Bugzilla bug #2.)
3101 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3102 the result incorrectly in the debug output. (It correctly added a newline
3103 to what was transported.)
3105 TF/01 Added $received_time.
3107 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3108 commented out examples of how to interface to a virus scanner and to
3109 SpamAssassin. Also added commented examples of av_scanner and
3110 spamd_address settings.
3112 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3113 and controls are allowed in which ACLs. There were a couple of minor
3114 errors. Some of the entries in the conditions table (which is a table of
3115 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3116 negation of where the condition IS allowed.
3118 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3120 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3121 header file does not have a version number, so I've had to invent a new
3122 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3123 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3124 radiusclient), but it was contributed by a Radius user.
3126 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3127 files or directories, for OpenSSL.
3129 PH/19 When an Exim process that is running as root has to create an Exim log
3130 file, it does so in a subprocess that runs as exim:exim so as to get the
3131 ownership right at creation (otherwise, other Exim processes might see
3132 the file with the wrong ownership). There was no test for failure of this
3133 fork() call, which would lead to the process getting stuck as it waited
3134 for a non-existent subprocess. Forks do occasionally fail when resources
3135 run out. I reviewed all the other calls to fork(); they all seem to check
3138 PH/20 When checking for unexpected SMTP input at connect time (before writing
3139 the banner), Exim was not dealing correctly with a non-positive return
3140 from the read() function. If the client had disconnected by this time,
3141 the result was a log entry for a synchronization error with an empty
3142 string after "input=" when read() returned zero. If read() returned -1
3143 (an event I could not check), uninitialized data bytes were printed.
3144 There were reports of junk text (parts of files, etc) appearing after
3147 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3149 PH/22 Added support for macro redefinition, and (re)definition in between
3150 driver and ACL definitions.
3152 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3153 forgetting to use the resulting value; it was using the unexpanded value.
3155 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3156 hadn't been configured. The fix is from Juergen Kreileder, who
3157 understands it better than I do:
3159 "Here's what I see happening with three configured cyrus_sasl
3160 authenticators configured (plain, login, cram-md5):
3162 On startup auth_cyrus_sasl_init() gets called for each of these.
3163 This means three calls to sasl_listmech() without a specified mech_list.
3164 => SASL tests which mechs of all available mechs actually work
3165 => three warnings about OTP not working
3166 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3168 With the patch, sasl_listmech() also gets called three times. But now
3169 SASL's mech_list option is set to the server_mech specified in the the
3170 authenticator. Or in other words, the answer from sasl_listmech()
3171 gets limited to just the mech you're testing for (which is different
3173 => the return list contains just 'plain' or 'login', 'cram-md5' or
3174 nothing depending on the value of ob->server_mech.
3176 I've just tested the patch: Authentication still works fine,
3177 unavailable mechs specified in the exim configuration are still
3178 caught, and the auth.log warnings about OTP are gone."
3180 PH/25 When debugging is enabled, the contents of the command line are added
3181 to the debugging output, even when log_selector=+arguments is not
3184 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3185 answer is "GNU", and only if the return is "GNU/something" is the answer
3188 PH/27 $acl_verify_message is now set immediately after the failure of a
3189 verification in an ACL, and so is available in subsequent modifiers. In
3190 particular, the message can be preserved by coding like this:
3192 warn !verify = sender
3193 set acl_m0 = $acl_verify_message
3195 Previously, $acl_verify_message was set only while expanding "message"
3196 and "log_message" when a very denied access.
3198 PH/28 Modified OS/os.c-Linux with
3200 -#ifndef OS_LOAD_AVERAGE
3201 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3203 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3204 nomenclature these days.)
3206 PH/29 Installed patch from the Sieve maintainer that adds the options
3207 sieve_useraddress and sieve_subaddress to the redirect router.
3209 PH/30 In these circumstances:
3210 . Two addresses routed to the same list of hosts;
3211 . First host does not offer TLS;
3212 . First host accepts first address;
3213 . First host gives temporary error to second address;
3214 . Second host offers TLS and a TLS session is established;
3215 . Second host accepts second address.
3216 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3217 and peerdn, if requested) that were in fact used only for the second
3220 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3221 attention to any local part prefix or suffix that was matched by the
3222 router that accepted the address. It now behaves in the same way as it
3223 does for delivery: the affixes are removed from the local part unless
3224 rcpt_include_affixes is set on the transport.
3226 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3227 timeout during the DATA phase of an incoming message.
3229 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3230 applied a patch sanctioned by the Sieve maintainer.
3232 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3233 the uid or gid is negative. A case of a negative gid caused this to be
3234 noticed. The fix allows for either to be negative.
3236 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3237 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3240 PH/36 The change PH/12 above was broken. Fixed it.
3242 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3243 the grounds that routing the same address twice would always produce the
3244 same answer. This might have been true once, but it is certainly no
3245 longer true now. Routing a child address may depend on the previous
3246 routing that produced that child. Some complicated redirection strategies
3247 went wrong when messages had multiple recipients, and made Exim's
3248 behaviour dependent on the order in which the addresses were given.
3250 I have moved the duplicate checking until after the routing is complete.
3251 Exim scans the addresses that are assigned to local and remote
3252 transports, and removes any duplicates. This means that more work will be
3253 done, as duplicates will always all be routed, but duplicates are
3254 presumably rare, so I don't expect this is of any significance.
3256 For deliveries to pipes, files, and autoreplies, the duplicate checking
3257 still happens during the routing process, since they are not going to be
3260 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3261 It corrects a timeout issue with spamd. This is Ian's comment: "The
3262 background is that sometimes spamd either never reads data from a
3263 connection it has accepted, or it never writes response data. The exiscan
3264 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3265 blindly assumes that writes won't block so it may never time out."
3267 PH/39 Allow G after quota size as well as K and M.
3269 PH/40 The value set for $authenticated_id in an authenticator may not contain
3270 binary zeroes or newlines because the value is written to log lines and
3271 to spool files. There was no check on this. Now the value is run through
3272 the string_printing() function so that such characters are converted to
3273 printable escape sequences.
3275 PH/41 $message_linecount is a new variable that contains the total number of
3276 lines in the message. Compare $body_linecount, which is the count for the
3279 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3280 bounce and delay warning messages, except in certain special cases, which
3283 (a) An SMTP error message from a remote host;
3284 (b) A message specified in a :fail: redirection;
3285 (c) A message specified in a "fail" command in a system filter;
3286 (d) A message specified in a FAIL return from the queryprogram router;
3287 (e) A message specified by the cannot_route_message router option.
3289 In these cases only, Exim does include the error details in bounce and
3290 warning messages. There are also a few cases where bland messages such
3291 as "unrouteable address" or "local delivery error" are given.
3293 PH/43 $value is now also set for the "else" part of a ${run expansion.
3295 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3296 being worked on, but at least Exim now implements the latest version to
3299 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3300 process to complete was treated as a delivery failure, a timeout while
3301 writing the message to the pipe was logged, but erroneously treated as a
3302 successful delivery. Such timeouts include transport filter timeouts. For
3303 consistency with the overall process timeout, these timeouts are now
3304 treated as errors, giving rise to delivery failures by default. However,
3305 there is now a new Boolean option for the pipe transport called
3306 timeout_defer, which, if set TRUE, converts the failures into defers for
3307 both kinds of timeout. A transport filter timeout is now identified in
3310 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3311 systems where "make" and "gmake" are different, calling "gmake" at top
3312 level broke things. I've arranged for the value of $(MAKE) to be passed
3313 from the Makefile to this script so that it can call the same version of
3317 A note about Exim versions 4.44 and 4.50
3318 ----------------------------------------
3320 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3321 changes of various kinds. As a consequence, a big documentation update was
3322 needed. This delayed the release for rather longer than seemed good, especially
3323 in the light of a couple of (minor) security issues. Therefore, the changes
3324 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3325 release. So 4.44 and 4.50 are in effect two different branches that both start
3328 I have left the 4.50 change log unchanged; it contains all the changes since
3329 4.43. The change log for 4.44 is below; many of its items are identical to
3330 those for 4.50. This seems to be the most sensible way to preserve the
3331 historical information.
3337 1. Minor wording change to the doc/README.SIEVE file.
3339 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3340 computation of the current number of files was incorrect.
3342 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3343 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3346 4. Give more explanation in the error message when the command for a transport
3347 filter fails to execute.
3349 5. There are several places where Exim runs a non-Exim command in a
3350 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3351 was being done only for the command run by the queryprogram router. It is
3352 now done for all such subprocesses. The other cases are: ${run, transport
3353 filters, and the commands run by the lmtp and pipe transports.
3355 6. Added CONFIGURE_GROUP build-time option.
3357 7. Some older OS have a limit of 256 on the maximum number of file
3358 descriptors. Exim was using setrlimit() to set 1000 as a large value
3359 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3360 systems. I've change it so that if it can't get 1000, it tries for 256.
3362 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3363 was an oversight, and furthermore, ever since the addition of extra
3364 controls (e.g. 4.43/32), the checks on when to allow different forms of
3365 "control" were broken. There should now be diagnostics for all cases when a
3366 control that does not make sense is encountered.
3368 9. Added the /retain_sender option to "control=submission".
3370 10. $recipients is now available in the predata ACL (oversight).
3372 11. Tidy the search cache before the fork to do a delivery from a message
3373 received from the command line. Otherwise the child will trigger a lookup
3374 failure and thereby defer the delivery if it tries to use (for example) a
3375 cached ldap connection that the parent has called unbind on.
3377 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3378 of $address_data from the recipient verification was clobbered by the
3379 sender verification.
3381 13. The value of address_data from a sender verification is now available in
3382 $sender_address_data in subsequent conditions in the ACL statement.
3384 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3386 15. Added a new option "connect=<time>" to callout options, to set a different
3389 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3390 was its contents. (It was OK if the option was not defined at all.)
3392 17. A "Completed" log line is now written for messages that are removed from
3393 the spool by the -Mrm option.
3395 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3396 information about exactly what failed.
3398 19. Added -dd to debug only the daemon process.
3400 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3401 handles timeouts, both on the server side and network timeouts. Renamed the
3402 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3404 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3405 It is now set to "smtps".
3407 22. $host_address is now set to the target address during the checking of
3408 ignore_target_hosts.
3410 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3411 being passed; this would have caused $sender_host_name to have been used if
3412 matching the list had actually called for a host name (not very likely,
3413 since this list is usually IP addresses). A host name is now passed as
3416 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3417 code that specifies a non-restarting handler (typically sigaction() in
3418 modern systems) in an attempt to fix a rare and obscure crash bug.
3420 25. Narrowed the window for a race in the daemon that could cause it to ignore
3421 SIGCHLD signals. This is not a major problem, because they are used only to
3422 wake it up if nothing else does.
3424 26. A malformed maildirsize file could cause Exim to calculate negative values
3425 for the mailbox size or file count. Odd effects could occur as a result.
3426 The maildirsize information is now recalculated if the size or filecount
3429 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3430 support for a long time. Removed HAVE_SYS_VFS_H.
3432 28. Installed the latest version of exipick from John Jetmore.
3434 29. In an address list, if the pattern was not a regular expression, an empty
3435 subject address (from a bounce message) matched only if the pattern was an
3436 empty string. Non-empty patterns were not even tested. This was the wrong
3437 because it is perfectly reasonable to use an empty address as part of a
3438 database query. An empty address is now tested by patterns that are
3439 lookups. However, all the other forms of pattern expect the subject to
3440 contain a local part and a domain, and therefore, for them, an empty
3441 address still always fails if the pattern is not itself empty.
3443 30. Exim went into a mad DNS loop when attempting to do a callout where the
3444 host was specified on an smtp transport, and looking it up yielded more
3445 than one IP address.
3447 31. Re-factored the code for checking spool and log partition space into a
3448 function that finds that data and another that does the check. The former
3449 is then used to implement four new variables: $spool_space, $log_space,
3450 $spool_inodes, and $log_inodes.
3452 32. The RFC2047 encoding function was originally intended for short strings
3453 such as real names; it was not keeping to the 75-character limit for
3454 encoded words that the RFC imposes. It now respects the limit, and
3455 generates multiple encoded words if necessary. To be on the safe side, I
3456 have increased the buffer size for the ${rfc2047: expansion operator from
3459 33. It is now permitted to omit both strings after an "if" condition; if the
3460 condition is true, the result is "true". As before, when the second string
3461 is omitted, a false condition yields an empty string. This makes it less
3462 cumbersome to write custom ACL and router conditions.
3464 34. Failure to deliver a bounce message always caused it to be frozen, even if
3465 there was an errors_to setting on the router. The errors_to setting is now
3468 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3469 canonical form (fully expanded) before being placed in
3470 $sender_host_address.
3472 36. The table in the code that translates DNS record types into text (T_A to
3473 "A" for instance) was missing entries for NS and CNAME. It is just possible
3474 that this could have caused confusion if both these types were looked up
3475 for the same domain, because the text type is used as part of Exim's
3476 per-process caching. But the chance of anyone hitting this buglet seems
3479 37. The dnsdb lookup has been extended in a number of ways.
3481 (1) There is a new type, "zns", which walks up the domain tree until it
3482 finds some nameserver records. It should be used with care.
3484 (2) There is a new type, "mxh", which is like "mx" except that it returns
3485 just the host names, not the priorities.
3487 (3) It is now possible to give a list of domains (or IP addresses) to be
3488 looked up. The behaviour when one of the lookups defers can be
3489 controlled by a keyword.
3491 (4) It is now possible to specify the separator character for use when
3492 multiple records are returned.
3494 38. The dnslists ACL condition has been extended: it is now possible to supply
3495 a list of IP addresses and/or domains to be looked up in a particular DNS
3498 39. Added log_selector=+queue_time_overall.
3500 40. When running the queue in the test harness, wait just a tad after forking a
3501 delivery process, to get repeatability of debugging output.
3503 41. Include certificate and key file names in error message when GnuTLS fails
3504 to set them up, because the GnuTLS error message doesn't include the name
3505 of the failing file when there is a problem reading it.
3507 42. Allow both -bf and -bF in the same test run.
3509 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3511 44. The "Exiscan patch" is now merged into the mainline Exim source.
3513 45. Sometimes the final signoff response after QUIT could fail to get
3514 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3515 < 0 before doing a fflush(). This bug looks as though it goes back to the
3516 introduction of TLS in release 3.20, but "sometimes" must have been rare
3517 because the tests only now provoked it.
3519 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3520 (this can affect the format of dates).
3522 47. exim_tidydb, when checking for the continued existence of a message for
3523 which it has found a message-specific retry record, was not finding
3524 messages that were in split spool directories. Consequently, it was
3525 deleting retry records that should have stayed in existence.
3527 48. Steve fixed some bugs in eximstats.
3529 49. The SPA authentication driver was not abandoning authentication and moving
3530 on to the next authenticator when an expansion was forced to fail,
3531 contradicting the general specification for all authenticators. Instead it
3532 was generating a temporary error. It now behaves as specified.
3534 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3535 (the order specifies the preference for clients). The order is now AES256,
3536 AES128, 3DES, ARCFOUR128.
3538 51. Small patch to Sieve code - explicitly set From: when generating an
3541 52. Exim crashed if a remote delivery caused a very long error message to be
3542 recorded - for instance if somebody sent an entire SpamAssassin report back
3543 as a large number of 550 error lines. This bug was coincidentally fixed by
3544 increasing the size of one of Exim's internal buffers (big_buffer) that
3545 happened as part of the Exiscan merge. However, to be on the safe side, I
3546 have made the code more robust (and fixed the comments that describe what
3549 53. Now that there can be additional text after "Completed" in log lines (if
3550 the queue_time_overall log selector is set), a one-byte patch to exigrep
3551 was needed to allow it to recognize "Completed" as not the last thing in
3554 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3555 patch that reportedly fixes this has been added. I am not expert enough to
3556 create a test for it. This is what the patch creator wrote:
3558 "I found a little strange behaviour of ldap code when working with
3559 Windows 2003 AD Domain, where users was placed in more than one
3560 Organization Units. When I tried to give exim partial DN, the exit code
3561 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3562 But simultaneously result of request was absolutely normal ldap result,
3563 so I produce this patch..."
3565 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3566 REFERENCE, so I have modified the code to exclude the patch when that macro
3569 55. Some experimental protocols are using DNS PTR records for new purposes. The
3570 keys for these records are domain names, not reversed IP addresses. The
3571 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3572 leaves it alone. Component reversal etc. now happens only for IP addresses.
3575 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3577 57. Double the size of the debug message buffer (to 2048) so that more of very
3578 long debug lines gets shown.
3580 58. The exicyclog utility now does better if the number of log files to keep
3581 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3583 59. Two changes related to the smtp_active_hostname option:
3585 (1) $smtp_active_hostname is now available as a variable.
3586 (2) The default for smtp_banner uses $smtp_active_hostname instead
3587 of $primary_hostname.
3589 60. The host_aton() function is supposed to be passed a string that is known
3590 to be a valid IP address. However, in the case of IPv6 addresses, it was
3591 not checking this. This is a hostage to fortune. Exim now panics and dies
3592 if the condition is not met. A case was found where this could be provoked
3593 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3594 components; fortuitously, this particular loophole had already been fixed
3595 by change 4.50/55 above.
3597 If there are any other similar loopholes, the new check in host_aton()
3598 itself should stop them being exploited. The report I received stated that
3599 data on the command line could provoke the exploit when Exim was running as
3600 exim, but did not say which command line option was involved. All I could
3601 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3602 running as the user.
3605 61. There was a buffer overflow vulnerability in the SPA authentication code
3606 (which came originally from the Samba project). I have added a test to the
3607 spa_base64_to_bits() function which I hope fixes it.
3610 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3611 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3613 63. The daemon start-up calls getloadavg() while still root for those OS that
3614 need the first call to be done as root, but it missed one case: when
3615 deliver_queue_load_max is set with deliver_drop_privilege. This is
3616 necessary for the benefit of the queue runner, because there is no re-exec
3617 when deliver_drop_privilege is set.
3619 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3620 This has been fixed.
3622 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3623 was in use, was not putting the data itself into the right store pool;
3624 consequently, it could be overwritten for a subsequent message in the same
3625 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3628 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3630 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3631 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3632 the function were treating the return as a boolean value, which happened to
3633 work because 0=false and not-0=true, but is not correct code.
3635 68. The host_aton() function was not handling scoped IPv6 addresses (those
3636 with, for example, "%eth0" on the end) correctly.
3638 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3639 negated items (that is, ~something) in unsigned ints. Some compilers
3640 apparently mutter when there is no cast.
3642 70. If an address verification called from an ACL failed, and did not produce a
3643 user-specific message (i.e. there was only a "system" message), nothing was
3644 put in $acl_verify_message. In this situation, it now puts the system
3647 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3648 session; change 4.31/43 added the unwanted input to the log line - except
3649 that it did not do this in the start of session case. It now does.
3651 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3652 This is wrong and can cause the other end to generate a synchronization
3653 error if it is another Exim or anything else that does the synchronization
3654 check. A QUIT command is no longer sent after a timeout.
3656 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3657 during host lookups.
3659 74. The defer_ok option of callout verification was not working if it was used
3660 when verifying addresses in header lines, that is, for this case:
3662 verify = header_sender/callout=defer_ok
3664 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3665 those file descriptors could be used for SMTP connections. If anything
3666 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3667 could be sent to the SMTP client, causing chaos. The daemon now opens
3668 stdin, stdout, and stderr to /dev/null when it puts itself into the
3671 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3672 log by default. The user can override this with suitable Perl magic.
3674 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3675 the log message when discard triggers, was not working for the DATA ACL or
3676 for the non-SMTP ACL.
3678 78. Error message wording change in sieve.c.
3680 79. If smtp_accept_max_per_host was set, the number of connections could be
3681 restricted to fewer than expected, because the daemon was trying to set up
3682 a new connection before checking whether the processes handling previous
3683 connections had finished. The check for completed processes is now done
3684 earlier. On busy systems, this bug wouldn't be noticed because something
3685 else would have woken the daemon, and it would have reaped the completed
3688 80. If a message was submitted locally by a user whose login name contained one
3689 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3690 It caused a spool format error. I have fixed the spool reading code. A
3691 related problem was that the "from" clause in the Received: line became
3692 illegal because of the space(s). It is now covered by ${quote_local_part.
3694 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3697 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3699 83. Patch from Sieve maintainer to fix unterminated string problem in
3700 "vacation" handling.
3702 84. Some minor changes to the Linux configuration files to help with other
3703 OS variants using glibc.
3705 85. One more patch for Sieve to update vacation handling to latest spec.
3708 ----------------------------------------------------
3709 See the note above about the 4.44 and 4.50 releases.
3710 ----------------------------------------------------
3716 1. Change 4.43/35 introduced a bug that caused file counts to be
3717 incorrectly computed when quota_filecount was set in an appendfile
3720 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3721 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3724 3. Give more explanation in the error message when the command for a transport
3725 filter fails to execute.
3727 4. There are several places where Exim runs a non-Exim command in a
3728 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3729 was being done only for the command run by the queryprogram router. It is
3730 now done for all such subprocesses. The other cases are: ${run, transport
3731 filters, and the commands run by the lmtp and pipe transports.
3733 5. Some older OS have a limit of 256 on the maximum number of file
3734 descriptors. Exim was using setrlimit() to set 1000 as a large value
3735 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3736 systems. I've change it so that if it can't get 1000, it tries for 256.
3738 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3739 was an oversight, and furthermore, ever since the addition of extra
3740 controls (e.g. 4.43/32), the checks on when to allow different forms of
3741 "control" were broken. There should now be diagnostics for all cases when a
3742 control that does not make sense is encountered.
3744 7. $recipients is now available in the predata ACL (oversight).
3746 8. Tidy the search cache before the fork to do a delivery from a message
3747 received from the command line. Otherwise the child will trigger a lookup
3748 failure and thereby defer the delivery if it tries to use (for example) a
3749 cached ldap connection that the parent has called unbind on.
3751 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3752 of $address_data from the recipient verification was clobbered by the
3753 sender verification.
3755 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3756 was its contents. (It was OK if the option was not defined at all.)
3758 11. A "Completed" log line is now written for messages that are removed from
3759 the spool by the -Mrm option.
3761 12. $host_address is now set to the target address during the checking of
3762 ignore_target_hosts.
3764 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3765 being passed; this would have caused $sender_host_name to have been used if
3766 matching the list had actually called for a host name (not very likely,
3767 since this list is usually IP addresses). A host name is now passed as
3770 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3771 code that specifies a non-restarting handler (typically sigaction() in
3772 modern systems) in an attempt to fix a rare and obscure crash bug.
3774 15. Narrowed the window for a race in the daemon that could cause it to ignore
3775 SIGCHLD signals. This is not a major problem, because they are used only to
3776 wake it up if nothing else does.
3778 16. A malformed maildirsize file could cause Exim to calculate negative values
3779 for the mailbox size or file count. Odd effects could occur as a result.
3780 The maildirsize information is now recalculated if the size or filecount
3783 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3784 support for a long time. Removed HAVE_SYS_VFS_H.
3786 18. Updated exipick to current release from John Jetmore.
3788 19. Allow an empty sender to be matched against a lookup in an address list.
3789 Previously the only cases considered were a regular expression, or an
3792 20. Exim went into a mad DNS lookup loop when doing a callout where the
3793 host was specified on the transport, if the DNS lookup yielded more than
3796 21. The RFC2047 encoding function was originally intended for short strings
3797 such as real names; it was not keeping to the 75-character limit for
3798 encoded words that the RFC imposes. It now respects the limit, and
3799 generates multiple encoded words if necessary. To be on the safe side, I
3800 have increased the buffer size for the ${rfc2047: expansion operator from
3803 22. Failure to deliver a bounce message always caused it to be frozen, even if
3804 there was an errors_to setting on the router. The errors_to setting is now
3807 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3808 canonical form (fully expanded) before being placed in
3809 $sender_host_address.
3811 24. Updated eximstats to version 1.33
3813 25. Include certificate and key file names in error message when GnuTLS fails
3814 to set them up, because the GnuTLS error message doesn't include the name
3815 of the failing file when there is a problem reading it.
3817 26. Expand error message when OpenSSL has problems setting up cert/key files.
3820 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3821 (this can affect the format of dates).
3823 28. exim_tidydb, when checking for the continued existence of a message for
3824 which it has found a message-specific retry record, was not finding
3825 messages that were in split spool directories. Consequently, it was
3826 deleting retry records that should have stayed in existence.
3828 29. eximstats updated to version 1.35
3829 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3830 1.35 - bugfix such that pie charts by volume are generated correctly
3832 30. The SPA authentication driver was not abandoning authentication and moving
3833 on to the next authenticator when an expansion was forced to fail,
3834 contradicting the general specification for all authenticators. Instead it
3835 was generating a temporary error. It now behaves as specified.
3837 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3838 (the order specifies the preference for clients). The order is now AES256,
3839 AES128, 3DES, ARCFOUR128.
3841 31. Small patch to Sieve code - explicitly set From: when generating an
3844 32. Exim crashed if a remote delivery caused a very long error message to be
3845 recorded - for instance if somebody sent an entire SpamAssassin report back
3846 as a large number of 550 error lines. This bug was coincidentally fixed by
3847 increasing the size of one of Exim's internal buffers (big_buffer) that
3848 happened as part of the Exiscan merge. However, to be on the safe side, I
3849 have made the code more robust (and fixed the comments that describe what
3852 33. Some experimental protocols are using DNS PTR records for new purposes. The
3853 keys for these records are domain names, not reversed IP addresses. The
3854 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3855 leaves it alone. Component reversal etc. now happens only for IP addresses.
3858 34. The host_aton() function is supposed to be passed a string that is known
3859 to be a valid IP address. However, in the case of IPv6 addresses, it was
3860 not checking this. This is a hostage to fortune. Exim now panics and dies
3861 if the condition is not met. A case was found where this could be provoked
3862 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3863 components; fortuitously, this particular loophole had already been fixed
3864 by change 4.50/55 or 4.44/33 above.
3866 If there are any other similar loopholes, the new check in host_aton()
3867 itself should stop them being exploited. The report I received stated that
3868 data on the command line could provoke the exploit when Exim was running as
3869 exim, but did not say which command line option was involved. All I could
3870 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3871 running as the user.
3874 35. There was a buffer overflow vulnerability in the SPA authentication code
3875 (which came originally from the Samba project). I have added a test to the
3876 spa_base64_to_bits() function which I hope fixes it.
3879 36. The daemon start-up calls getloadavg() while still root for those OS that
3880 need the first call to be done as root, but it missed one case: when
3881 deliver_queue_load_max is set with deliver_drop_privilege. This is
3882 necessary for the benefit of the queue runner, because there is no re-exec
3883 when deliver_drop_privilege is set.
3885 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3886 was in use, was not putting the data itself into the right store pool;
3887 consequently, it could be overwritten for a subsequent message in the same
3888 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3891 38. Sometimes the final signoff response after QUIT could fail to get
3892 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3893 < 0 before doing a fflush(). This bug looks as though it goes back to the
3894 introduction of TLS in release 3.20, but "sometimes" must have been rare
3895 because the tests only now provoked it.
3901 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3902 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3903 Now it returns an integer. A number of calls were still expecting a T/F
3904 return. Fortuitously, in all cases, the tests worked in OK situations,
3905 which is the norm. However, things would have gone wrong on any write
3906 failures on the smtp file descriptor. This function is used when sending
3907 messages over SMTP and also when doing verify callouts.
3909 2. When Exim is called to do synchronous delivery of a locally submitted
3910 message (the -odf or -odi options), it no longer closes stderr before doing
3913 3. Implemented the mua_wrapper option.
3915 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3917 5. Implemented the functions header_remove(), header_testname(),
3918 header_add_at_position(), and receive_remove_recipient(), and exported them
3921 6. If an ACL "warn" statement specified the addition of headers, Exim already
3922 inserted X-ACL-Warn: at the start if there was no header name. However, it
3923 was not making this test for the second and subsequent header lines if
3924 there were newlines in the string. This meant that an invalid header could
3925 be inserted if Exim was badly configured.
3927 7. Allow an ACL "warn" statement to add header lines at the start or after all
3928 the Received: headers, as well as at the end.
3930 8. Added the rcpt_4xx retry error code.
3932 9. Added postmaster_mailfrom=xxx to callout verification option.
3934 10. Added mailfrom=xxxx to the callout verification option, for verify=
3937 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3938 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3940 12. Inserted some casts to stop certain compilers warning when using pointer
3941 differences as field lengths or precisions in printf-type calls (mostly
3942 affecting debugging statements).
3944 13. Added optional readline() support for -be (dynamically loaded).
3946 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3947 same clock tick as a message's arrival, so that its received time was the
3948 same as the "first fail" time on the retry record, and that message
3949 remained on the queue past the ultimate address timeout, every queue runner
3950 would try a delivery (because it was past the ultimate address timeout) but
3951 after another failure, the ultimate address timeout, which should have then
3952 bounced the address, did not kick in. This was a "< instead of <=" error;
3953 in most cases the first failure would have been in the next clock tick
3954 after the received time, and all would be well.
3956 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3957 being recognized when the domain list was tested by the match_domain
3958 condition in an expansion string.
3960 16. Added the ${str2b64: operator.
3962 17. Exim was always calling setrlimit() to set a large limit for the number of
3963 processes, without checking whether the existing limit was already
3964 adequate. (It did check for the limit on file descriptors.) Furthermore,
3965 errors from getrlimit() and setrlimit() were being ignored. Now they are
3966 logged to the main and panic logs, but Exim does carry on, to try to do its
3967 job under whatever limits there are.
3969 18. Imported PCRE 5.0.
3971 19. Trivial typo in log message " temporarily refused connection" (the leading
3974 20. If the log selector return_path_on_delivery was set and an address was
3975 redirected to /dev/null, the delivery process crashed because it assumed
3976 that a return path would always be set for a "successful" delivery. In this
3977 case, the whole delivery is bypassed as an optimization, and therefore no
3980 21. Internal re-arrangement: the function for sending a challenge and reading
3981 a response while authentication was assuming a zero-terminated challenge
3982 string. It's now changed to take a pointer and a length, to allow for
3983 binary data in such strings.
3985 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3987 23. Exim was not respecting finduser_retries when seeking the login of the
3988 uid under which it was called; it was always trying 10 times. (The default
3989 setting of finduser_retries is zero.) Also, it was sleeping after the final
3990 failure, which is pointless.
3992 24. Implemented tls_on_connect_ports.
3994 25. Implemented acl_smtp_predata.
3996 26. If the domain in control=submission is set empty, Exim assumes that the
3997 authenticated id is a complete email address when it generates From: or
3998 Sender: header lines.
4000 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4001 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4002 chown and chgrp in /bin and hostname in /usr/bin.
4004 28. Exim was keeping the "process log" file open after each use, just as it
4005 does for the main log. This opens the possibility of it remaining open for
4006 long periods when the USR1 signal hits a daemon. Occasional processlog
4007 errors were reported, that could have been caused by this. Anyway, it seems
4008 much more sensible not to leave this file open at all, so that is what now
4011 29. The long-running daemon process does not normally write to the log once it
4012 has entered its main loop, and it closes the log before doing so. This is
4013 so that log files can straightforwardly be renamed and moved. However,
4014 there are a couple of unusual error situations where the daemon does write
4015 log entries, and I had neglected to close the log afterwards.
4017 30. The text of an SMTP error response that was received during a remote
4018 delivery was being truncated at 512 bytes. This is too short for some of
4019 the long messages that one sometimes sees. I've increased the limit to
4022 31. It is now possible to make retry rules that apply only when a message has a
4023 specific sender, in particular, an empty sender.
4025 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4026 it possible to be selective about when SMTP synchronization is enforced.
4028 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4030 32. Implemented hosts_connection_nolog.
4032 33. Added an ACL for QUIT.
4034 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4037 35. Added mailbox_size and mailbox_filecount to appendfile.
4039 36. Added control = no_multiline_responses to ACLs.
4041 37. There was a bug in the logic of the code that waits for the clock to tick
4042 in the case where the clock went backwards by a substantial amount such
4043 that the microsecond fraction of "now" was more than the microsecond
4044 fraction of "then" (but the whole seconds number was less).
4046 38. Added support for the libradius Radius client library this is found on
4047 FreeBSD (previously only the radiusclient library was supported).
4053 1. When certain lookups returned multiple values in the form name=value, the
4054 quoting of the values was not always being done properly. Specifically:
4055 (a) If the value started with a double quote, but contained no whitespace,
4057 (b) If the value contained whitespace other than a space character (i.e.
4058 tabs or newlines or carriage returns) it was not quoted.
4059 This fix has been applied to the mysql and pgsql lookups by writing a
4060 separate quoting function and calling it from the lookup code. The fix
4061 should probably also be applied to nisplus, ibase and oracle lookups, but
4062 since I cannot test any of those, I have not disturbed their existing code.
4064 2. A hit in the callout cache for a specific address caused a log line with no
4065 reason for rejecting RCPT. Now it says "Previous (cached) callout
4066 verification failure".
4068 3. There was an off-by-one bug in the queryprogram router. An over-long
4069 return line was truncated at 256 instead of 255 characters, thereby
4070 overflowing its buffer with the terminating zero. As well as fixing this, I
4071 have increased the buffer size to 1024 (and made a note to document this).
4073 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4074 when Exim is waiting for an SMTP response from a remote server, Exim
4075 restarts its select() call on the socket, thereby resetting its timeout.
4076 This is not a problem when such interrupts are rare. Somebody set up a cron
4077 job to run exiwhat every 2 minutes, which is less than the normal select()
4078 timeout (5 or 10 minutes). This meant that the select() timeout never
4079 kicked in because it was always reset. I have fixed this by comparing the
4080 time when an interrupt arrives with the time at the start of the first call
4081 to select(). If more time than the timeout has elapsed, the interrupt is
4082 treated as a timeout.
4084 5. Some internal re-factoring in preparation for the addition of Sieve
4085 extensions (by MH). In particular, the "personal" test is moved to a
4086 separate function, and given an option for scanning Cc: and Bcc: (which is
4087 not set for Exim filters).
4089 6. When Exim created an email address using the login of the caller as the
4090 local part (e.g. when creating a From: or Sender: header line), it was not
4091 quoting the local part when it contained special characters such as @.
4093 7. Installed new OpenBSD configuration files.
4095 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4096 try to make them clearer.
4098 9. Callout options, other than the timeout value, were being ignored when
4099 verifying sender addresses in header lines. For example, when using
4101 verify = header_sender/callout=no_cache
4103 the cache was (incorrectly) being used.
4105 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4106 only the Cygwin environment.
4108 11. When return_path_on_delivery was set as a log selector, if different remote
4109 addresses in the same message used different return paths and parallel
4110 remote delivery occurred, the wrong values would sometimes be logged.
4111 (Whenever a remote delivery process finished, the return path value from
4112 the most recently started remote delivery process was logged.)
4114 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4115 lines when AUTH and/or TLS are in use. This is the "received protocol"
4116 field. Exim used to use "asmtp" for authenticated SMTP, without any
4117 indication (in the protocol name) for TLS use. Now it follows the RFC and
4118 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4119 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4120 names appear in log lines as well as in Received: header lines.
4122 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4123 extensions, and comparison tests, and to fix some bugs.
4125 14. Changes to the "personal" filter test:
4127 (1) The test was buggy in that it was just doing the equivalent of
4128 "contains" tests on header lines. For example, if a user's address was
4129 anne@some.where, the "personal" test would incorrectly be true for
4131 To: susanne@some.where
4133 This test is now done by extracting each address from the header in turn,
4134 and checking the entire address. Other tests that are part of "personal"
4135 are now done using regular expressions (for example, to check local parts
4136 of addresses in From: header lines).
4138 (2) The list of non-personal local parts in From: addresses has been
4139 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4140 taken from the Sieve specification recommendations.
4142 (3) If the message contains any header line starting with "List-" it is
4143 treated as non-personal.
4145 (4) The test for "circular" in the Subject: header line has been removed
4146 because it now seems ill-conceived.
4148 15. Minor typos in src/EDITME comments corrected.
4150 16. Installed latest exipick from John Jetmore.
4152 17. If headers_add on a router specified a text string that was too long for
4153 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4154 of string_sprintf() is now avoided.
4156 18. $message_body_size was not set (it was always zero) when running the DATA
4157 ACL and the local_scan() function.
4159 19. For the "mail" command in an Exim filter, no default was being set for
4160 the once_repeat time, causing a random time value to be used if "once" was
4161 specified. (If the value happened to be <= 0, no repeat happened.) The
4162 default is now 0s, meaning "never repeat". The "vacation" command was OK
4163 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4164 (I found it when inspecting the code).
4166 20. There is now an overall timeout for performing a callout verification. It
4167 defaults to 4 times the callout timeout, which applies to individual SMTP
4168 commands during the callout. The overall timeout applies when there is more
4169 than one host that can be tried. The timeout is checked before trying the
4170 next host. This prevents very long delays if there are a large number of
4171 hosts and all are timing out (e.g. when the network connections are timing
4172 out). The value of the overall timeout can be changed by specifying an
4173 additional sub-option for "callout", called "maxwait". For example:
4175 verify = sender/callout=5s,maxwait=20s
4177 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4178 to the end before writing, but this should make it even safer).
4180 22. Exim was forgetting that it had advertised PIPELINING for the second and
4181 subsequent messages on an SMTP connection. It was also not resetting its
4182 memory on STARTTLS and an internal HELO.
4184 23. When Exim logs an SMTP synchronization error within a session, it now
4185 records whether PIPELINING has been advertised or not.
4187 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4188 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4189 rather than long int.
4191 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4193 26. Added the never_mail option to autoreply.
4199 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4200 crash if the getsockname() call failed; this can happen if a connection is
4201 closed very soon after it is established. The problem was simply in the
4202 order in which certain operations were done, causing Exim to try to write
4203 to the SMTP stream before it had set up the file descriptor. The bug has
4204 been fixed by making things happen in the correct order.
4210 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4211 before the connection was closed, thus losing the rejection response.
4213 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4214 some early Solaris releases, but causes trouble in current releases where
4215 socklen_t is defined.
4217 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4220 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4223 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4224 that releasing the top of it at the end releases what was used for sub-
4225 expansions (unless the block got too big). However, discard this block if
4226 the first thing is a variable or header, so that we can use its block when
4227 it is dynamic (useful for very large $message_headers, for example).
4229 6. Lookups now cache *every* query, not just the most recent. A new, separate
4230 store pool is used for this. It can be recovered when all lookup caches are
4231 flushed. Lookups now release memory at the end of their result strings.
4232 This has involved some general refactoring of the lookup sources.
4234 7. Some code has been added to the store_xxx() functions to reduce the amount
4235 of flapping under certain conditions.
4237 8. log_incoming_interface used to affect only the <= reception log lines. Now
4238 it causes the local interface and port to be added to several more SMTP log
4239 lines, for example "SMTP connection from", and rejection lines.
4241 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4243 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4245 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4246 could be overwritten at the end of the current message (or the start of a
4247 new message if it was set in a HELO ACL). The value is now preserved for
4248 the duration of the SMTP connection.
4250 12. If a transport had a headers_rewrite setting, and a matching header line
4251 contained an unqualified address, that address was qualified, even if it
4252 did not match any rewriting rules. The underlying bug was that the values
4253 of the flags that permit the existence of unqualified sender and recipient
4254 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4255 non-local messages, and by -bnq for local messages) were not being
4256 preserved with the message after it was received.
4258 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4259 "next input=" as part of the text comprising the host identity instead of
4260 the correct text. The code was using the same buffer for two different
4261 strings. However, depending on which order the printing function evaluated
4262 its arguments, the bug did not always show up. Under Linux, for example, my
4263 test suite worked just fine.
4265 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4266 doesn't work with some older versions of Perl. It has been changed to "my",
4267 which in any case is probably the better facility to use.
4269 15. A really picky compiler found some instances of statements for creating
4270 error messages that either had too many or two few arguments for the format
4273 16. The size of the buffer for calls to the DNS resolver has been increased
4274 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4275 for addresses that have a lot of PTR records. This alleviates a problem; it
4276 does not fully solve it.
4278 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4279 buffer now truncates the list and logs the incident, which is the same
4280 action as happens when Exim is looking up a host name and its aliases.
4281 Previously in this situation something unpredictable would happen;
4282 sometimes it was "internal error: store_reset failed".
4284 18. If a server dropped the connection unexpectedly when an Exim client was
4285 using GnuTLS and trying to read a response, the client delivery process
4286 crashed while trying to generate an error log message.
4288 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4289 string, for example:
4291 warn message = H1: something\nH2: something
4293 the text was added as a single header line from Exim's point of view
4294 though it ended up OK in the delivered message. However, searching for the
4295 second and subsequent header lines using $h_h2: did not work. This has been
4296 fixed. Similarly, if a system filter added multiple headers in this way,
4297 the routers could not see them.
4299 20. Expanded the error message when iplsearch is called with an invalid key to
4300 suggest using net-iplsearch in a host list.
4302 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4303 ACLs are no longer actually imposed (and a message to that effect is
4306 22. If a "gecos" field in a passwd entry contained escaped characters, in
4307 particular, if it contained a \" sequence, Exim got it wrong when building
4308 a From: or a Sender: header from that name. A second bug also caused
4309 incorrect handling when an unquoted " was present following a character
4310 that needed quoting.
4312 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4313 was not being matched caselessly.
4315 24. Arranged for all hyphens in the exim.8 source to be escaped with
4318 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4319 properly. Recipient callout cache records were still being keyed to include
4320 the sender, even when use_sender was set false. This led to far more
4321 callouts that were necessary. The sender is no longer included in the key
4322 when use_sender is false.
4324 26. Added "control = submission" modifier to ACLs.
4326 27. Added the ${base62d: operator to decode base 62 numbers.
4328 28. dnsdb lookups can now access SRV records.
4330 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4331 the configuration file.
4333 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4334 (-v) mode. This makes the output for a verbose queue run more intelligible.
4336 31. Added a use_postmaster feature to recipient callouts.
4338 32. Added the $body_zerocount variable, containing the number of binary zero
4339 bytes in the message body.
4341 33. The time of last modification of the "new" subdirectory is now used as the
4342 "mailbox time last read" when there is a quota error for a maildir
4345 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4347 35. Added +ignore_unknown as a special item in host lists.
4349 36. Code for decoding IPv6 addresses in host lists is now included, even if
4350 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4351 address was recognized as an IP address, but was then not correctly decoded
4352 into binary, causing unexpected and incorrect effects when compared with
4359 1. Very minor rewording of debugging text in manualroute to say "list of
4360 hosts" instead of "hostlist".
4362 2. If verify=header_syntax was set, and a header line with an unqualified
4363 address (no domain) and a large number of spaces between the end of the
4364 name and the colon was received, the reception process suffered a buffer
4365 overflow, and (when I tested it) crashed. This was caused by some obsolete
4366 code that should have been removed. The fix is to remove it!
4368 3. When running in the test harness, delay a bit after writing a bounce
4369 message to get a bit more predictability in the log output.
4371 4. Added a call to search_tidyup() just before forking a reception process. In
4372 theory, someone could use a lookup in the expansion of smtp_accept_max_
4373 per_host which, without the tidyup, could leave open a database connection.
4375 5. Added the variables $recipient_data and $sender_data which get set from a
4376 lookup success in an ACL "recipients" or "senders" condition, or a router
4377 "senders" option, similar to $domain_data and $local_part_data.
4379 6. Moved the writing of debug_print from before to after the "senders" test
4382 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4383 problems for message scanning, either using a data ACL, or using
4384 local_scan() because the Received: header was not generated till after they
4385 were called (in order to set the time as the time of reception completion).
4386 I have revised the way this works. The header is now generated after the
4387 body is received, but before the ACL or local_scan() are called. After they
4388 are run, the timestamp in the header is updated.
4394 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4395 before starting a queue runner without re-exec. This happened only when
4396 deliver_drop_privilege was set or when the Exim user was set to root. The
4397 effect of the bug was that timeouts during subsequent deliveries caused
4398 crashes instead of being properly handled. The handler is now left at its
4399 default (and expected) setting.
4401 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4402 message, again when deliver_drop_privilege is set or Exim is run as root.
4403 The bug described in (1) was not present in this case, but the tidying up
4404 of the other signals was missing. I have made the two cases consistent.
4406 3. The ignore_target_hosts setting on a manualroute router was being ignored
4407 for hosts that were looked up using the /MX notation.
4409 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4412 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4413 operated on the sender address. After changing the $sender_address to <>
4414 for the sender address verify, Exim was re-instated it as the original
4415 (before rewriting) address, but remembering that it had rewritten it, so it
4416 wasn't rewriting it again. This bug also had the effect of breaking the
4417 sender address verification caching when the sender address was rewritten.
4419 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4420 This has been changed so that if the ip literal address matches
4421 ignore_target_hosts, the router declines.
4423 7. Added expansion conditions match_domain, match_address, and match_local_
4424 part (NOT match_host).
4426 8. The placeholder for the Received: header didn't have a length field set.
4428 9. Added code to Exim itself and to exim_lock to test for a specific race
4429 condition that could lead to file corruption when using MBX delivery. The
4430 issue is with the lockfile that is created in /tmp. If this file is removed
4431 after a process has opened it but before that process has acquired a lock,
4432 there is the potential for a second process to recreate the file and also
4433 acquire a lock. This could lead to two Exim processes writing to the file
4434 at the same time. The added code performs the same test as UW imapd; it
4435 checks after acquiring the lock that its file descriptor still refers to
4436 the same named file.
4438 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4439 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4442 11. Added the smtp_active_hostname option. If used, this will typically be made
4443 to depend on the incoming interface address. Because $interface_address is
4444 not set up until the daemon has forked a reception process, error responses
4445 that can happen earlier (such as "too many connections") no longer contain
4448 12. If an expansion in a condition on a "warn" statement fails because a lookup
4449 defers, the "warn" statement is abandoned, and the next ACL statement is
4450 processed. Previously this caused the whole ACL to be aborted.
4452 13. Added the iplsearch lookup type.
4454 14. Added ident_timeout as a log selector.
4456 15. Added tls_certificate_verified as a log selector.
4458 16. Added a global option tls_require_ciphers (compare the smtp transport
4459 option of the same name). This controls incoming TLS connections.
4461 17. I finally figured out how to make tls_require_ciphers do a similar thing
4462 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4463 before starting the TLS session.
4465 18. Tabs are now shown as \t in -bP output.
4467 19. If the log selector return_path_on_delivery was set, Exim crashed when
4468 bouncing a message because it had too many Received: header lines.
4470 20. If two routers both had headers_remove settings, and the first one included
4471 a superfluous trailing colon, the final name in the first list and the
4472 first name in the second list were incorrectly joined into one item (with a
4473 colon in the middle).
4479 1. Added -C and -D options to the exinext utility, mainly to make it easier
4480 to include in the automated testing, but these could be helpful when
4481 multiple configurations are in use.
4483 2. The exinext utility was not formatting the output nicely when there was
4484 an alternate port involved in the retry record key, nor when there was a
4485 message id as well (for retries that were specific to a specific message
4486 and a specific host). It was also confused by IPv6 addresses, because of
4487 the additional colons they contain. I have fixed the IPv4 problem, and
4488 patched it up to do a reasonable job for IPv6.
4490 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4491 delivery, the log line now contains "pipelined" if PIPELINING was used.
4493 4. An SMTP transport process used to panic and die if the bind() call to set
4494 an explicit outgoing interface failed. This has been changed; it is now
4495 treated in the same way as a connect() failure.
4497 5. A reference to $sender_host_name in the part of a conditional expansion
4498 that was being skipped was still causing a DNS lookup. This no longer
4501 6. The def: expansion condition was not recognizing references to header lines
4502 that used bh_ and bheader_.
4504 7. Added the _cache feature to named lists.
4506 8. The code for checking quota_filecount in the appendfile transport was
4507 allowing one more file than it should have been.
4509 9. For compatibility with Sendmail, the command line option
4517 and sets the incoming protocol and host name (for trusted callers). The
4518 host name and its colon can be omitted when only the protocol is to be set.
4519 Note the Exim already has two private options, -pd and -ps, that refer to
4520 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4521 "s", but I don't think that's a major issue.
4523 10. A number of refactoring changes to the code, none of which should affect
4526 (a) The number of logging options was getting close to filling up the
4527 32-bit word that was used as a bit map. I have split them into two classes:
4528 those that are passed in the argument to log_write(), and those that are
4529 only ever tested independently outside of that function. These are now in
4530 separate 32-bit words, so there is plenty of room for expansion again.
4531 There is no change in the user interface or the logging behaviour.
4533 (b) When building, for example, log lines, the code previously used a
4534 macro that called string_cat() twice, in order to add two strings. This is
4535 not really sufficiently general. Furthermore, there was one instance where
4536 it was actually wrong because one of the argument was used twice, and in
4537 one call a function was used. (As it happened, calling the function twice
4538 did not affect the overall behaviour.) The macro has been replaced by a
4539 function that can join an arbitrary number of extra strings onto a growing
4542 (c) The code for expansion conditions now uses a table and a binary chop
4543 instead of a serial search (which was left over from when there were very
4544 few conditions). Also, it now recognizes conditions like "pam" even when
4545 the relevant support is not compiled in: a suitably worded error message is
4546 given if an attempt is made to use such a condition.
4548 11. Added ${time_interval:xxxxx}.
4550 12. A bug was causing one of the ddress fields not to be passed back correctly
4551 from remote delivery subprocesses. The field in question was not being
4552 subsequently used, so this caused to problems in practice.
4554 13. Added new log selectors queue_time and deliver_time.
4556 14. Might have fixed a bug in maildirsizefile handling that threw up
4557 "unexpected character" debug warnings, and recalculated the data
4558 unnecessarily. In any case, I expanded the warning message to give more
4561 15. Added the message "Restricted characters in address" to the statements in
4562 the default ACL that block characters like @ and % in local parts.
4564 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4565 Three changes have been made:
4567 (a) There was a serious bug; a negative response to MAIL caused the whole
4568 recipient domain to be cached as invalid, thereby blocking all messages
4569 to all local parts at the same domain, from all senders. This bug has
4570 been fixed. The domain is no longer cached after a negative response to
4571 MAIL if the sender used is not empty.
4573 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4576 (c) A new callout option, "use_sender" has been added for people who want
4577 the modified behaviour.
4583 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4586 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4587 indeed breaks things for older releases.
4589 3. Added additional logging to the case where there is a problem reading data
4590 from a filter that is running in a subprocess using a pipe, in order to
4591 try to track down a specific problem.
4593 4. Testing facility fudge: when running in the test harness and attempting
4594 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4595 getting "No route to host". Convert this to a timeout.
4597 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4600 6. Some OS don't have socklen_t but use size_t instead. This affects the
4601 fifth argument of getsockopt() amongst other things. This is now
4602 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4603 can be set for individual OS. I have set it for SunOS5, OSF1, and
4604 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4605 some earlier ones do not.
4607 7. Change 4.30/15 was not doing the test caselessly.
4609 8. The standard form for an IPv6 address literal was being rejected by address
4610 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4611 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4612 this, as well as the form without the "IPv6" on the front (but only when
4613 address literals are enabled, of course).
4615 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4617 10. Exim crashed if a message with an empty sender address specified by -f
4618 encountered a router with an errors_to setting. This could be provoked only
4619 by a command such as
4623 where an empty string was supplied; "<>" did not hit this bug.
4625 11. Installed PCRE release 4.5.
4627 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4628 remained set. It is now erased.
4630 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4631 times from message ids (which are base 36 rather than the normal 62).
4633 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4634 were being counted as actual protocol errors, and logged if the log
4635 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4636 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4637 and DATA following a set of rejected RCPTs do not count as protocol errors.
4638 In other words, Exim assumes they were pipelined, though this may not
4639 actually be the case. Of course, in all cases the client gets an
4640 appropriate error code.
4642 15. If a lookup fails in an ACL condition, a message about the failure may
4643 be available; it is used if testing the ACL cannot continue, because most
4644 such messages specify what the cause of the deferral is. However, some
4645 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4646 that caused an old message to be retained and used if a later statement
4647 caused a defer, replacing the real cause of the deferral.
4649 16. If an IP address had so many PTR records that the DNS lookup buffer
4650 was not large enough to hold them, Exim could crash while trying to process
4651 the truncated data. It now detects and logs this case.
4653 17. Further to 4.21/58, another change has been made: if (and only if) the
4654 first line of a message (the first header line) ends with CRLF, a bare LF
4655 in a subsequent header line has a space inserted after it, so as not to
4656 terminate the header.
4658 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4659 unnecessarily, used atoi() instead of strtol(), and didn't check the
4660 termination when getting file sizes from file names by regex.
4662 19. Completely re-implemented the support for maildirsize files, in the light
4663 of a number of problems with the previous contributed implementation
4664 (4.30/29). In particular:
4666 . If the quota is zero, the maildirsize file is maintained, but no quota is
4669 . If the maildir directory does not exist, it is created before any attempt
4670 to write a maildirsize file.
4672 . The quota value in the file is just a cache; if the quota is changed in
4673 the transport, the new value overrides.
4675 . A regular expression is available for excluding directories from the
4678 20. The autoreply transport checks the characters in options that define the
4679 message's headers; it allows continued headers, but it was checking with
4680 isspace() after an embedded newline instead of explicitly looking for a
4683 21. If all the "regular" hosts to which an address was routed had passed their
4684 expiry times, and had not reached their retry times, the address was
4685 bounced, even if fallback hosts were defined. Now Exim should go on to try
4688 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4689 equivalent code in the SMTP transport. Some hosts send humungous responses
4690 to HELO/EHLO, more than 1024 it seems.
4692 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4693 gives compiler warnings in some environments. I've now done it "properly",
4696 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4697 (because of problems with the built-in one) was declared to return uschar *
4698 instead of char *, causing compiler failure.
4700 25. Fixed a file descriptor leak when processing alias/forward files.
4702 26. Fixed a minor format string issue in dbfn.c.
4704 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4706 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4707 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4710 29. When a sender address is verified, it is cached, to save repeating the test
4711 when there is more than one recipient in a message. However, when the
4712 verification involves a callout, it is possible for different callout
4713 options to be set for different recipients. It is too complicated to keep
4714 track of this in the cache, so now Exim always runs a verification when a
4715 callout is required, relying on the callout cache for the optimization.
4716 The overhead is duplication of the address routing, but this should not be
4719 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4720 to be verified with callout=postmaster, and the main callout worked but the
4721 postmaster check failed, the verification correctly failed. However, if a
4722 subsequent RCPT command asked for sender verification *without* the
4723 postmaster check, incorrect caching caused this verification also to fail,
4726 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4727 it was not caching the DNS options (qualify_single, search_parents) that
4728 were used when the lookup failed. A subsequent lookup with different
4729 options therefore always gave the same answer, though there were cases
4730 where it should not have. (Example: a "domains = !$mx_any" option on a
4731 dnslookup router: the "domains" option is always processed without any
4732 widening, but the router might have qualify_single set.) Now Exim uses the
4733 cached value only when the same options are set.
4735 32. Added John Jetmore's "exipick" utility to the distribution.
4737 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4738 than a timeout (e.g. a certificate is required, and is not provided), an
4739 Exim server now closes the connection immediately. Previously it waited for
4740 the client to close - but if the client is SSL, it seems that they each
4741 wait for each other, leading to a delay before one of them times out.
4743 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4744 maintained 0.8.x compatibility because I don't think many are using it, and
4745 it is clearly obsolete.
4747 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4750 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4751 client certificate was expired. A simple patch fixes this, though I don't
4752 understand the full logic of why the verify callback is called multiple
4755 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4756 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4757 which causes problems with some clients (such as the Certicom SSL Plus
4758 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4759 disables the coutermeasure allowing Eudora to connect."
4761 38. Exim was not checking that a write() to a log file succeeded. This could
4762 lead to Bad Things if a log got too big, in particular if it hit a file
4763 size limit. Exim now panics and dies if it cannot write to a log file, just
4764 as it does if it cannot open a log file.
4766 39. Modified OS/Makefile-Linux so that it now contains
4768 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4770 The two -D definitions ensure that Exim is compiled with large file
4771 support, which makes it possible to handle log files that are bigger than
4774 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4775 instance) a domain was checked against a named list that involved a lookup,
4776 causing $domain_data to be set, then another domain was checked against the
4777 same list, then the first domain was re-checked, the value of $domain_data
4778 after the final check could be wrong. In particular, if the second check
4779 failed, it could be set empty. This bug probably also applied to
4782 41. The strip_trailing_dot option was not being applied to the address given
4783 with the -f command-line option.
4785 42. The code for reading a message's header from the spool was incrementing
4786 $received_count, but never initializing it. This meant that the value was
4787 incorrect (doubled) while delivering a message in the same process in which
4788 it was received. In the most common configuration of Exim, this never
4789 happens - a fresh exec is done - but it can happen when
4790 deliver_drop_privilege is set.
4792 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4793 it now includes up to 150 characters of the unexpected data in the log
4796 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4797 and building data strings. The size of both of these buffers was 10 000
4798 bytes - far larger than anybody would *ever* want, thought I. Needless to
4799 say, somebody hit the limit. I have increased the maximum line length to
4800 20 000 and the maximum data length of concatenated lines to 100 000. I have
4801 also fixed two bugs, because there was no checking on these buffers. Tsk,
4802 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4803 buffer is too small.
4805 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4806 lsearch lookups. Now it does.
4808 46. When parsing a route_list item in a manualroute router, a fixed-length
4809 buffer was used for the list of hosts. I made this 1024 bytes long,
4810 thinking that nobody would ever have a list of hosts that long. Wrong.
4811 Somebody had a whole pile of complicated expansion conditions, and the
4812 string was silently truncated, leading to an expansion error. It turns out
4813 that it is easier to change to an unlimited length (owing to other changes
4814 that have happened since this code was originally written) than to build
4815 structure for giving a limitation error. The length of the item that
4816 expands into the list of hosts is now unlimited.
4818 47. The lsearch lookup could not handle data where the length of text line was
4819 more than 4095 characters. Such lines were truncated, leading to shortened
4820 data being returned. It should now handle lines of any length.
4822 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4823 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4826 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4827 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4828 no longer happens in comment lines. A list of replacements is now placed
4829 at the head of all of the source files, except those whose only change is
4830 to replace PERL_COMMAND in the very first #! line.
4832 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4833 messages on the queue, with a bottom-up merge sort, using code contributed
4834 by Michael Haardt. This should make operations like -bp somewhat faster on
4835 large queues. It won't affect queue runners, except when queue_run_in_order
4838 51. Installed eximstats 1.31 in the distribution.
4840 52. Added support for SRV lookups to the dnslookup router.
4842 53. If an ACL referred to $message_body or $message_body_end, the value was not
4843 reset for any messages that followed in the same SMTP session.
4845 54. The store-handling optimization for building very long strings was not
4846 differentiating between the different store pools. I don't think this
4847 actually made any difference in practice, but I've tidied it.
4849 55. While running the routers to verify a sender address, $sender_address
4850 was still set to the sender address. This is wrong, because when routing to
4851 send a bounce to the sender, it would be empty. Therefore, I have changed
4852 it so that, while verifying a sender address, $sender_address is set to <>.
4853 (There is no change to what happens when verifying a recipient address.)
4855 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4856 target A or AAAA records (if not already returned) without resetting the
4857 qualify_single or search_parents options of the DNS resolver. These are
4858 inappropriate in this case because the targets of MX and SRV records must
4859 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4860 target that, when qualified, matched something in the local domain. These
4861 two options are now turned off when doing these lookups.
4863 57. It seems that at least some releases of Reiserfs (which does not have the
4864 concept of a fixed number of inodes) returns zero and not -1 for the
4865 number of available inodes. This interacted badly with check_spool_inodes,
4866 which assumed that -1 was the "no such thing" setting. What I have done is
4867 to check that the total number of inodes is greater than zero before doing
4868 the test of how many are available.
4870 58. When a "warn" ACL statement has a log_message modifier, the message is
4871 remembered, and not repeated. This is to avoid a lot of repetition when a
4872 message has many recipients that cause the same warning to be written.
4873 Howewer, Exim was preserving the list of already written lines for an
4874 entire SMTP session, which doesn't seem right. The memory is now reset if a
4875 new message is started.
4877 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4878 debugging output unless log_rewrite was also set.
4880 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4881 of (int)(handle) when we know that handle contains (void *)(-1).
4883 61. The Exim daemon panic-logs an error return when it closes the incoming
4884 connection. However "connection reset by peer" seems to be common, and
4885 isn't really an error worthy of noting specially, so that particular error
4888 62. When Exim is trying to find all the local interfaces, it used to panic and
4889 die if the ioctl to get the interface flags failed. However, it seems that
4890 on at least one OS (Solaris 9) it is possible to have an interface that is
4891 included in the list of interfaces, but for which you get a failure error
4892 for this call. This happens when the interface is not "plumbed" into a
4893 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4894 failure of the "get flags" call assumes that the interface is down.
4896 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4897 makes life easier for people who are doing arithmetic on fields extracted
4898 from dates, where you often get leading zeros that should not be
4899 interpreted as octal.
4901 64. Added qualify_domain to the redirect router, to override the global
4904 65. If a pathologically long header line contained very many addresses (the
4905 report of this problem mentioned 10 000) and each of them was rewritten,
4906 Exim could use up a very large amount of memory. (It kept on making new
4907 copies of the header line as it rewrote, and never released the old ones.)
4908 At the expense of a bit more processing, the header rewriting function has
4909 been changed so that it no longer eats memory in this way.
4911 66. The generation of the Received: header has been moved from the time that a
4912 message starts to be received, to the time that it finishes. The timestamp
4913 in the Received: header should now be very close to that of the <= log
4914 line. There are two side-effects of this change:
4916 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4917 logged header lines no longer include the local Received: line, because
4918 it has not yet been created. The same applies to a copy of the message
4919 that is returned to a non-SMTP sender when a message is rejected.
4921 (b) When a filter file is tested using -bf, no additional Received: header
4922 is added to the test message. After some thought, I decided that this
4925 This change does not affect the value of $received_for. It is still set
4926 after address rewriting, but before local_scan() is called.
4928 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4930 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4931 gave an unhelpful panic error message, and a defer error. I have managed to
4932 change this behaviour so that it now rejects any supplied certificate,
4933 which seems right, as the list of acceptable certificates is empty.
4935 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4936 gave an unhelpful defer error. I have not managed to make this reject any
4937 supplied certificates, but the error message it gives is "no certificate
4938 supplied", which is not helpful.
4940 70. exigrep's output now also includes lines that are not associated with any
4941 message, but which match the given pattern. Implemented by a patch from
4942 Martin Sluka, which also tidied up the Perl a bit.
4944 71. Recipient callout verification, like sender verification, was using <> in
4945 the MAIL FROM command. This isn't really the right thing, since the actual
4946 sender may affect whether the remote host accepts the recipient or not. I
4947 have changed it to use the actual sender in the callout; this means that
4948 the cache record is now keyed on a recipient/sender pair, not just the
4949 recipient address. There doesn't seem to be a real danger of callout loops,
4950 since a callout by the remote host to check the sender would use <>.
4951 [SEE ABOVE: changed after hitting problems.]
4953 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4954 temporary errors. However, in the case of such a code being given after
4955 the end of a data transmission (i.e. after ".") Exim was failing to write
4956 a retry record for the message. (Yes, there was some broken host that was
4957 actually sending 8xx at this point.)
4959 73. An unknown lookup type in a host list could cause Exim to panic-die when
4960 the list was checked. (An example that provoked this was putting <; in the
4961 middle of a list instead of at the start.) If this happened during a DATA
4962 ACL check, a -D file could be left lying around. This kind of configuration
4963 error no longer causes Exim to die; instead it causes a defer errror. The
4964 incident is still logged to the main and panic logs.
4966 74. Buglet left over from Exim 3 conversion. The message "too many messages
4967 in one connection" was written to the rejectlog but not the mainlog, except
4968 when address rewriting (yes!) was being logged.
4970 75. Added write_rejectlog option.
4972 76. When a system filter was run not as root (that is, when system_filter_user
4973 was set), the values of the $n variables were not being returned to the
4974 main process; thus, they were not subsequently available in the $sn
4977 77. Added +return_path_on_delivery log selector.
4979 78. A connection timeout was being treated differently from recipients deferred
4980 when testing hosts_max_try with a message that was older than the host's
4981 retry timeout. (The host should not be counted, thus allowing all hosts to
4982 be tried at least once before bouncing.) This may have been the cause of an
4983 occasionally reported bug whereby a message would remain on the queue
4984 longer than the retry timeout, but would be bounced if a delivery was
4985 forced. I say "may" because I never totally pinned down the problem;
4986 setting up timeout/retry tests is difficult. See also the next item.
4988 79. The ultimate address timeout was not being applied to errors that involved
4989 a combination of host plus message (for example, a timeout on a MAIL
4990 command). When an address resolved to a number of possible hosts, and they
4991 were not all tried for each delivery (e.g. because of hosts_max_try), a
4992 message could remain on the queue longer than the retry timeout.
4994 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4997 81. Fixed an obscure SMTP outgoing bug which required at least the following
4998 conditions: (a) there was another message waiting for the same server;
4999 (b) the server returned 5xx to all RCPT commands in the first message so
5000 that the message was not completed; (c) the server dropped the connection
5001 or gave a negative response to the RSET that Exim sends to abort the
5002 transaction. The observed case was a dropped connection after DATA that had
5003 been sent in pipelining mode. That is, the server had advertised PIPELINING
5004 but was not implementing it correctly. The effect of the bug was incorrect
5005 behaviour, such as trying another host, and this could lead to a crash.
5011 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5012 and daemon.c were passed as pointers to ints; they should have been
5013 pointers to socklen_t variables (which are typically unsigned ints).
5015 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5018 3. Fixed a really odd bug that affected only the testing scheme; patching a
5019 certain fixed string in the binary changed the value of another string that
5020 happened to be identical to the end of the original first string.
5022 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5023 name", it returns that address as the IP address. On some operating
5024 systems (e.g. Solaris), it also passes back the IP address string as the
5025 "host name". However, on others (e.g. Linux), it passes back an empty
5026 string. Exim wasn't checking for this, and was changing the host name to an
5027 empty string, assuming it had been canonicized.
5029 5. Although rare, it is permitted to have more than one PTR record for a given
5030 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5031 all the names associated with an address, because they do in Solaris.
5032 However, it seems that they do not in Linux for data that comes from the
5033 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5034 I found this out when I moved to a new Linux workstation and tried to run
5035 the Exim test suite.
5037 To get round this problem I have changed the code so that it now does its
5038 own call to the DNS to look up PTR records when searching for a host name.
5039 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5040 addresses that are only in /etc/hosts are still found.
5042 This behaviour is, however, controlled by an option called host_lookup_
5043 order, which defaults to "bydns:byaddr". If people want to use the other
5044 order, or indeed, just use one or the other means of lookup, they can
5045 specify it in this variable.
5047 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5048 some operating systems, this comes back from gethostbyaddr() as an empty
5049 string, and this is what Exim used to test for. However, it seems that in
5050 other systems, "." is yielded. Exim now tests for this case too.
5052 7. The values of check_spool_space and check_log_space are now held internally
5053 as a number of kilobytes instead of an absolute number of bytes. If a
5054 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5055 kilobyte. This means that much larger values can be stored.
5057 8. Exim monitor: an attempt to get the action menu when not actually pointing
5058 at a message produces an empty menu entitled "No message selected". This
5059 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5060 no entries in it ("Shell widget menu has zero width and/or height"). So I
5061 have added a single, blank menu entry in this case.
5063 9. Added ${quote_local_part.
5065 10. MIME decoding is now applied to the contents of Subject: header lines when
5068 11. Now that a reference to $sender_host_address automatically causes a reverse
5069 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5070 host lookup before query-style lookups in lists that might use this
5071 variable. This has therefore been abolished, and the "net-" prefix is no
5072 longer necessary for query-style lookups.
5074 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5075 have been a typo for LFLAGS, so it has been changed.
5077 13. The install script calls Exim with "-C /dev/null" in order to find the
5078 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5079 to be output. Howeve, since Exim outputs its version number before the
5080 error, it didn't break the script. It just looked ugly. I fixed this by
5081 always allowing "-C /dev/null" if the caller is root.
5083 14. Ignore overlarge ACL variable number when reading spool file - insurance
5084 against a later release with more variables having written the file.
5086 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5087 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5088 this, as well as the form without the "IPv6" on the front.
5090 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5091 OS/Makefile-Darwin file.
5093 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5094 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5096 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5097 variable. However, some OS use ints for inodes. I've added cast to long int
5098 to get rid of the compiler warning.
5100 19. I had forgotten to lock out "/../" in configuration file names when
5101 ALT_CONFIG_PREFIX was set.
5103 20. Routers used for verification do not need to specify transports. However,
5104 if such a router generated a host list, and callout was configured, Exim
5105 crashed, because it could not find a port number from the (non-existent)
5106 transport. It now assumes port 25 in this circumstance.
5108 21. Added the -t option to exigrep.
5110 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5111 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5112 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5113 them is set, LOOKUP_LSEARCH is forced.
5115 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5117 24. Added sender and host information to the "rejected by local_scan()" log
5118 line; previously there was no indication of these.
5120 25. Added .include_if_exists.
5122 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5123 Linux. It turns out that not all file systems support this. Apparently some
5124 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5125 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5126 not supported on the file descriptor, is now ignored when Exim is trying to
5127 sync a directory. This applies only to Linux.
5129 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5131 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5132 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5135 29. Added support for maildirsize files from supplied patch (modified a bit).
5137 30. The use of :fail: followed by an empty string could lead Exim to respond to
5138 sender verification failures with (e.g.):
5140 550 Verification failed for <xxx>
5141 550 Sender verify failed
5143 where the first response line was missing the '-' that indicates it is not
5144 the final line of the response.
5146 31. The loop for finding the name of the user that called Exim had a hardwired
5147 limit of 10; it now uses the value of finduser_retries, which is used for
5148 all other user lookups.
5150 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5153 33. Exim was neglecting to zero errno before one call of strtol() when
5154 expanding a string and expecting an integer value. On some systems this
5155 resulted in spurious "integer overflow" errors. Also, it was casting the
5156 result into an int without checking.
5158 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5159 did not work. The code looks as if it has *never* worked, though it appears
5160 to have been documented since at least releast 1.62. I have made it work.
5162 35. The "timeout_DNS" error in retry rules, also documented since at least
5163 1.62, also never worked. As it isn't clear exactly what this means, and
5164 clearly it isn't a major issue, I have abolished the feature by treating it
5165 as "timeout", and writing a warning to the main and panic logs.
5167 36. The display of retry rules for -brt wasn't always showing the error code
5170 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5171 timeout_connect_A, timeout_connect_MX.
5173 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5174 to the empty sender.
5176 39. The daemon was not analysing the content of -oX till after it had closed
5177 stderr and disconnected from the controlling terminal. This meant that any
5178 syntax errors were only noted on the panic log, and the return code from
5179 the command was 0. By re-arranging the code a little, I've made the
5180 decoding happen first, so such errors now appear on stderr, and the return
5181 code is 1. However, the actual setting up of the sockets still happens in
5182 the disconnected process, so errors there are still only recorded on the
5185 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5186 connections (as happens on some IP stacks) was logged at start up time as
5187 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5188 it from "IPv6 and IPv4", which means that two separate sockets are being
5191 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5192 says whether AF_INET or AF_INET6 was passed as an argument.
5194 42. Exiwhat output was messed up when time zones were included in log
5197 43. Exiwhat now gives more information about the daemon's listening ports,
5198 and whether -tls-on-connect was used.
5200 44. The "port" option of the smtp transport is now expanded.
5202 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5203 silently ignored. Now an error message is written to the main and panic
5206 46. There's a new ACL modifier called "logwrite" which writes to a log file
5207 as soon as it is encountered.
5209 47. Added $local_user_uid and $local_user_gid at routing time.
5211 48. Exim crashed when trying to verify a sender address that was being
5214 49. Exim was recognizing only a space character after ".include". It now also
5215 recognizes a tab character.
5217 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5218 extracting the relevant information from the specification. The man page no
5219 longer contains scrambled data for the -d option, and I've added a section
5220 at the front about calling Exim under different names.
5222 51. Added "extra_headers" argument to the "mail" command in filter files.
5224 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5227 53. Installed eximstats 1.29.
5229 54. Added transport_filter_timeout as a generic transport option.
5231 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5232 Cc: header lines. This was required by RFC 822, but it not required by RFC
5235 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5236 incoming messages that did not have them. Now it does so only if the
5237 message originates locally, that is, if there is no associated remote host
5238 address. When Resent- header lines are present, this applies to the Resent-
5239 lines rather than the non-Resent- lines.
5241 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5242 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5244 58. Messages for configuration errors now include the name of the main
5245 configuration file - useful now that there may be more than one file in a
5246 list (.included file names were always shown).
5248 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5249 for those rare installations that do not start the daemon as root or run it
5250 setuid root. I've cut out the call to initgroups() if the daemon is not
5253 60. The Exim user and group can now be bound into the binary as text strings
5254 that are looked up at the start of Exim's processing.
5256 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5258 62. Added $mailstore_basename variable.
5260 63. Installed patch to sieve.c from Michael Haardt.
5262 64. When Exim failed to open the panic log after failing to open the main log,
5263 the original message it was trying to log was written to stderr and debug
5264 output, but if they were not available (the usual case in production), it
5265 was lost. Now it is written to syslog before the two lines that record the
5266 failures to open the logs.
5268 65. Users' Exim filters run in subprocesses under the user's uid. It is
5269 possible for a "deliver" command or an alias in a "personal" command to
5270 provoke an address rewrite. If logging of address rewriting is configured,
5271 this fails because the process is not running as root or exim. There may be
5272 a better way of dealing with this, but for the moment (because 4.30 needs
5273 to be released), I have disabled address rewrite logging when running a
5274 filter in a non-root, non-exim process.
5280 1. The buildconfig auxiliary program wasn't quoting the value set for
5281 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5282 not defined. This bug was masked in 4.22 by the effect that was fixed in
5285 2. Some messages that were rejected after a message id was allocated were
5286 shown as "incomplete" by exigrep. It no longer does this for messages that
5287 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5289 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5290 have allow_domain_literals set, the ID did not get logged in the <= line.
5291 Domain literals are now always recognized in Message-ID: header lines.
5293 4. The first argument for a ${extract expansion item is the key name or field
5294 number. Leading and trailing spaces in this item were not being ignored,
5295 causing some misleading effects.
5297 5. When deliver_drop_privilege was set, single queue runner processes started
5298 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5299 same command in the process it spins off) were not dropping privilege.
5301 6. When the daemon running as "exim" started a queue runner, it always
5302 re-executed Exim in the spun-off process. This is a waste of effort when
5303 deliver_drop_privilege is set. The new process now just calls the
5304 queue-runner function directly.
5310 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5313 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5314 ignored. Though the use of -oP was forcing the writing of a pid file, it
5315 was always written to the default place.
5317 3. If the message "no IP address found for host xxxx" is generated during
5318 incoming verification, it is now followed by identification of the incoming
5319 connection (so you can more easily find what provoked it).
5321 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5323 5. Added some features to "harden" Exim a bit more against certain attacks:
5325 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5326 be put in Local/Makefile. This is like the never_users runtime option,
5327 but it cannot be overridden. The default setting is "root".
5329 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5330 prefix string with which any file named in a -C command line option
5333 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5334 is retained for -C and -D only if the caller of Exim is root. Without
5335 it, the exim user may also use -C and -D and retain privilege.
5337 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5338 command line option is disabled.
5340 6. Macro names set by the -D option must start with an upper case letter, just
5341 like macro names defined in the configuration file.
5343 7. Added "dereference=" facility to LDAP.
5345 8. Two instances of the typo "uknown" in the source files are fixed.
5347 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5348 the Configure-Makefile script screwed up while processing it.
5350 10. Incorporated PCRE 4.4.
5352 11. The SMTP synchronization check was not operating right at the start of an
5353 SMTP session. For example, it could not catch a HELO sent before the client
5354 waited for the greeting. There is now a check for outstanding input at the
5355 point when the greeting is written. Because of the duplex, asynchronous
5356 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5357 way, but not yet received, when the check is performed.
5359 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5360 on TCP/IP sockets, because this apparently causes some broken clients to
5363 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5364 unchanged) from the Cygwin maintainer.
5366 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5367 was supported instead of testing for mbx. Effectively a typo.
5369 15. The spa authenticator server code was not checking that the input it
5370 received was valid base64.
5372 16. The debug output line for the "set" modifier in ACLs was not showing the
5373 name of the variable that was being set.
5375 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5377 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5378 lookup on its own. Something else was needed to trigger the lookup. For
5379 example, a match in host_lookup or the need for a host name in a host list.
5380 Now, if $sender_host_name is referenced and the host name has not yet been
5381 looked up, a lookup is performed. If the lookup fails, the variable remains
5382 empty, and $host_lookup_failed is set to "1".
5384 19. Added "eqi" as a case-independent comparison operator.
5386 20. The saslauthd authentication condition could segfault if neither service
5387 nor realm was specified.
5389 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5390 error message that was logged was misleading, and incoming SMTP
5391 connections were dropped. The message is now more accurate, and temporary
5392 errors are given to SMTP connections.
5394 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5395 (except RSET) with a 421 error, until QUIT is received. However, it was
5396 failing to send a response to QUIT.
5398 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5399 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5400 ACL is now run after the helo_try_verify_hosts code.
5402 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5403 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5404 case-independent, so other case variants are also recognized). Apparently
5405 some systems use these upper case variants.
5407 25. If more than two messages were waiting for the same host, and a transport
5408 filter was specified for the transport, Exim sent two messages over the
5409 same TCP/IP connection, and then failed with "socket operation on non-
5410 socket" when it tried to send the third.
5412 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5414 27. The extern definition of crypt16() in expand.c was not being excluded when
5415 the OS had its own crypt16() function.
5417 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5418 as a preferred synonym for return_size_limit, both as an option and as an
5421 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5423 30. Changed the default configuration ACL to relax the local part checking rule
5424 for addresses that are not in any local domains. For these addresses,
5425 slashes and pipe symbols are allowed within local parts, but the sequence
5426 /../ is explicitly forbidden.
5428 31. SPA server authentication was not clearing the challenge buffer before
5431 32. log_message in a "warn" ACL statement was writing to the reject log as
5432 well as to the main log, which contradicts the documentation and doesn't
5433 seem right (because no rejection is happening). So I have stopped it.
5435 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5436 However, I am unable to do any testing of this.
5438 34. Fixed an infelicity in the appendfile transport. When checking directories
5439 for a mailbox, to see if any needed to be created, it was accidentally
5440 using path names with one or more superfluous leading slashes; tracing
5441 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5443 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5444 contents are added to the log line that is written for every discarded
5445 recipient. (Previously a log_message setting was ignored.)
5447 36. The ${quote: operator now quotes the string if it is empty.
5449 37. The install script runs exim in order to find its version number. If for
5450 some reason other than non-existence or emptiness, which it checks, it
5451 could not run './exim', it was installing it with an empty version number,
5452 i.e. as "exim-". This error state is now caught, and the installation is
5455 38. An argument was missing from the function that creates an error message
5456 when Exim fails to connect to the socket for saslauthd authentication.
5457 This could cause Exim to crash, or give a corrupted message.
5459 39. Added isip, isip4, and isip6 to ${if conditions.
5461 40. The ACL variables $acl_xx are now saved with the message, and can be
5462 accessed later in routers, transports, and filters.
5464 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5465 strings in the file are not string-expanded.
5467 42. If a MAIL command specified a SIZE value that was too large to fit into an
5468 int variable, the check against message_size_limit failed. Such values are
5469 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5470 day this will have to be increased, but I don't think I want to be around
5471 when emails are that large.
5478 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5479 iconv() is not standard in FreeBSD.
5481 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5482 IPv6 enabled. The observed symptom was a segmentation fault on return from
5483 the function os_common_find_running_interfaces() in src/os.c.
5485 3. In the check_special_case() function in daemon.c I had used "errno" as an
5486 argument name, which causes warnings on some systems. This was basically a
5487 typo, since it was named "eno" in the comments!
5489 4. The code that waits for the clock to tick (at a resolution of some fraction
5490 of a second) so as to ensure message-id uniqueness was always waiting for
5491 at least one whole tick, when it could have waited for less. [This is
5492 almost certainly not relevant at current processor speeds, where it is
5493 unlikely to ever wait at all. But we try to future-proof.]
5495 5. The function that sleeps for a time interval that includes fractions of a
5496 second contained a race. It did not block SIGALRM between setting the
5497 timer, and suspending (a couple of lines later). If the interval was short
5498 and the sigsuspend() was delayed until after it had expired, the suspension
5499 never ended. On busy systems this could lead to processes getting stuck for
5502 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5503 process, before it forks any delivery processes. The open lookup caching
5504 mechanism meant that the open file or database connection was passed into
5505 the delivery process. The problem was that delivery processes always tidy
5506 up cached lookup data. This could cause a problem for the next delivery
5507 process started by the queue runner, because the external queue runner
5508 process does not know about the closure. So the next delivery process
5509 still has data in the lookup cache. In the case of a file lookup, there was
5510 no problem because closing a file descriptor in a subprocess doesn't affect
5511 the parent. However, if the lookup was caching a connection to a database,
5512 the connection was closed, and the second delivery process was likely to
5513 see errors such as "PGSQL: query failed: server closed the connection
5514 unexpectedly". The problem has been fixed by closing all cached lookups
5515 in a queue runner before running a delivery process.
5517 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5518 seem to have the "const" qualifier which it has on other OS. I've
5521 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5522 *of the same type* (client or server) with the same public name that an
5523 error should be diagnosed.
5525 9. When Exim looked up a host name for an IP address, but failed to find the
5526 original IP address when looking up the host name (a safety check), it
5527 output the message "<ip address> does not match any IP for NULL", which was
5528 confusing, to say the least. The bug was that the host name should have
5529 appeared instead of "NULL".
5531 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5532 user that is built into the binary, and the -C or -D options is used, root
5533 privilege is dropped before the configuration file is read. In addition,
5534 logging is switched to stderr instead of the normal log files. If the
5535 configuration then re-defines the Exim user, the unprivileged environment
5536 is probably not what is expected, so Exim logs a panic warning message (but
5539 However, if deliver_drop_privilege is set, the unprivileged state may well
5540 be exactly what is intended, so the warning has been cut out in that case,
5541 and Exim is allowed to try to write to its normal log files.
5547 1. smtp_return_error_details was not giving details for temporary sender
5548 or receiver verification errors.
5550 2. Diagnose a configuration error if two authenticators have the same public
5553 3. Exim used not to create the message log file for a message until the first
5554 delivery attempt. This could be confusing when incoming messages were held
5555 for policy or load reasons. The message log file is now created at the time
5556 the message is received, and an initial "Received" line is written to it.
5558 4. The automatically generated man page for command line options had a minor
5559 bug that caused no ill effects; however, a more serious problem was that
5560 the procedure for building the man page automatically didn't always
5561 operate. Consequently, release 4.20 contains an out-of-date version. This
5562 shouldn't happen again.
5564 5. When building Exim with embedded Perl support, the script that builds the
5565 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5566 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5568 6. The freeze_tell option was not being used for messages that were frozen on
5569 arrival, either by an ACL or by local_scan().
5571 7. Added the smtp_incomplete_transaction log selector.
5573 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5574 was accepting AUTH without a new EHLO.
5576 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5577 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5580 10. Logging of TCP/IP connections (when configured) now happens in the main
5581 daemon process instead of the child process, so that the TCP/IP connection
5582 count is more accurate (but it can never be perfect).
5584 11. The use of "drop" in a nested ACL was not being handled correctly in the
5585 outer ACL. Now, if condition failure induced by the nested "drop" causes
5586 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5587 or "require"), the connection is dropped.
5589 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5590 that yield "discard" can now be used with an "accept" or a "discard" verb,
5591 but an error is generated for any others (because I can't see a useful way
5592 to define what should happen).
5594 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5595 are now processed in the same way as lines in the Exim configuration file.
5596 In particular, continuation lines are supported.
5598 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5600 15. Added -ti meaning -t -i.
5602 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5603 domains, and warn by logging if others are found.
5605 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5606 returned from GIFCONF when Exim is trying to find the list of interfaces on
5607 a host. The code in os.c has been modified to copy each ifreq to an aligned
5608 structure in all cases.
5610 Also, in some cases, the returned ifreq's were being copied to a 'struct
5611 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5612 means the last couple of bytes of an IPv6 address could be chopped if the
5613 ifreq contained only a normal sockaddr (14 bytes storage).
5615 18. Named domain lists were not supported in the hosts_treat_as_local option.
5616 An entry such as +xxxx was not recognized, and was treated as a literal
5619 19. Ensure that header lines added by a DATA ACL are included in the reject log
5620 if the ACL subsequently rejects the message.
5622 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5623 MD5 (which is deprecated).
5625 21. When testing a filter file using -bf, Exim was writing a message when it
5626 took the sender from a "From " line in the message, but it was not doing so
5627 when it took $return_path from a Return-Path: header line. It now does.
5629 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5630 with a valid header line field name (a series of printing characters
5631 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5633 23. Changed "disc" in the source to "disk" to conform to the documentation and
5634 the book and for uniformity.
5636 24. Ignore Sendmail's -Ooption=value command line item.
5638 25. When execve() failed while trying to run a command in a pipe transport,
5639 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5640 could be confused with a return value of 69 from the command itself. This
5641 has been changed to 127, the value the shell returns if it is asked to run
5642 a non-existent command. The wording for the related log line suggests a
5643 non-existent command as the problem.
5645 26. If received_header_text expands to an empty string, do not add a Received:
5646 header line to the message. (Well, it adds a token one on the spool, but
5647 marks it "old" so that it doesn't get used or transmitted.)
5649 27. Installed eximstats 1.28 (addition of -nt option).
5651 28. There was no check for failure on the call to getsockname() in the daemon
5652 code. This can fail if there is a shortage of resources on the system, with
5653 ENOMEM, for example. A temporary error is now given on failure.
5655 29. Contrary to the C standard, it seems that in some environments, the
5656 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5657 program. Exim now does this explicitly; it affects the formatting of
5658 timestamps using strftime().
5660 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5661 complaints. I've now initialized all the variables, to avoid this.
5663 32. Header lines added by a system filter were not being "seen" during
5664 transport-time rewrites.
5666 33. The info_callback() function passed to OpenSSL is set up with type void
5667 (*)(SSL *, int, int), as described somewhere. However, when calling the
5668 function (actually a macro) that sets it up, the type void(*)() is
5669 expected. I've put in a cast to prevent warnings from picky compilers.
5671 34. If a DNS black list lookup found a CNAME record, but there were no A
5672 records associated with the domain it pointed at, Exim crashed.
5674 35. If a DNS black list lookup returned more than one A record, Exim ignored
5675 all but the first. It now scans all returned addresses if a particular IP
5676 value is being sought. In this situation, the contents of the
5677 $dnslist_value variable are a list of all the addresses, separated by a
5680 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5681 to accept a host name and all its aliases if the forward lookup for any of
5682 them yielded the IP address of the incoming connection. Now it accepts only
5683 those names whose forward lookup yields the correct IP address. Any other
5684 names are discarded. This closes a loophole whereby a rogue DNS
5685 administrator could create reverse DNS records to break through a
5686 wildcarded host restriction in an ACL.
5688 37. If a user filter or a system filter that ran in a subprocess used any of
5689 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5690 the wrong values were passed to the pipe command ($thisaddress had the
5691 value of $0, $0 had the value of $1, etc). This bug was introduced by
5692 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5694 38. Improved the line breaking for long SMTP error messages from ACLs.
5695 Previously, if there was no break point between 40 and 75 characters, Exim
5696 left the rest of the message alone. Two changes have been made: (a) I've
5697 reduced the minimum length to 35 characters; (b) if it can't find a break
5698 point between 35 and 75 characters, it looks ahead and uses the first one
5699 that it finds. This may give the occasional overlong line, but at least the
5700 remaining text gets split now.
5702 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5703 file descriptors might be low, and that setting 1000 would always raise it.
5704 It turns out that in some environments, the limit is already over 1000 and
5705 that lowering it causes trouble. So now Exim takes care not to decrease it.
5707 40. When delivering a message, the value of $return_path is set to $sender_
5708 address at the start of routing (routers may change the value). By an
5709 oversight, this default was not being set up when an address was tested by
5710 -bt or -bv, which affected the outcome if any router or filter referred to
5713 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5714 log only when "message" or "log_message" are set. However, if one of the
5715 conditions was an address verification, or a call to a nested ACL, the
5716 messages generated by the underlying test were being passed through. This
5717 no longer happens. The underlying message is available in $acl_verify_
5718 message for both "message" and "log_message" expansions, so it can be
5719 passed through if needed.
5721 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5722 new expansion $bh_ to give the encoded byte string without charset
5723 translation. Translation happens only if iconv() is available; HAVE_ICONV
5724 indicates this at build time. HEADERS_CHARSET gives the charset to
5725 translate to; headers_charset can change it in the configuration, and
5726 "headers charset" can change it in an individual filter file.
5728 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5729 that creates RFC 2047 encoded "words" labels them as that charset instead
5730 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5731 expansion operator; (ii) when Exim creates a From: line for a local
5732 message; (iii) when a header line is rewritten to include a "phrase" part.
5734 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5735 buggy, causing it to skip the first lines of messages whose message ID
5736 ended in 'D'. This would not have bitten before Exim release 4.14, because
5737 message IDs were unlikely to end in 'D' before then. The effect was to have
5738 incorrect size information for certain domains.
5740 45. #include "config.h" was missing at the start of the crypt16.c module. This
5741 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5744 46. If there was a timeout during a "random" callout check, Exim treated it as
5745 a failure of the random address, and carried on sending RSET and the real
5746 address. If the delay was just some slowness somewhere, the response to the
5747 original RCPT would be taken as a response to RSET and so on, causing
5748 mayhem of various kinds.
5750 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5751 when I implemented it. It didn't allow for the fact that some option values
5752 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5753 the right test for positive values.
5755 48. Domain names in DNS records are case-independent. Exim always looks them up
5756 in lower case. Some resolvers return domain names in exactly the case they
5757 appear in the zone file, that is, they may contain uppercase letters. Not
5758 all resolvers do this - some return always lower case. Exim was treating a
5759 change of case by a resolver as a change of domain, similar to a widening
5760 of a domain abbreviation. This triggered its re-routing code and so it was
5761 trying to route what was effectively the same domain again. This normally
5762 caused routing to fail (because the router wouldn't handle the domain
5763 twice). Now Exim checks for this case specially, and just changes the
5764 casing of the domain that it ultimately uses when it transmits the message
5767 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5770 50. If a filter generated a file delivery with a non-absolute name (possible if
5771 no home directory exists for the router), the forbid_file option was not
5774 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5775 the existing equality matching.
5777 52. Exim was using ints instead of ino_t variables in some places where it was
5778 dealing with inode numbers.
5780 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5781 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5782 called TMPDIR, and if it finds it is different, it changes its value.
5784 54. The smtp_printf() function is now made available to local_scan() so
5785 additional output lines can be written before returning. There is also an
5786 smtp_fflush() function to enable the detection of a dropped connection.
5787 The variables smtp_input and smtp_batched_input are exported to
5790 55. Changed the default runtime configuration: the message "Unknown user"
5791 has been removed from the ACL, and instead placed on the localuser router,
5792 using the cannot_route_message feature. This means that any verification
5793 failures that generate their own messages won't get overridden. Similarly,
5794 the "Unrouteable address" message that was in the ACL for unverifiable
5795 relay addresses has also been removed.
5797 56. Added hosts_avoid_esmtp to the smtp transport.
5799 57. The exicyclog script was not checking for the esoteric option
5800 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5801 will work only if exicyclog is run under the appropriate euid.
5803 58. Following a discussion on the list, the rules by which Exim recognises line
5804 endings on incoming messages have been changed. The -dropcr and drop_cr
5805 options are now no-ops, retained only for backwards compatibility. The
5806 following line terminators are recognized: LF CRLF CR. However, special
5807 processing applies to CR:
5809 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5810 nor a local message in the state where . is a terminator.
5812 (ii) If a bare CR is encountered in a header line, an extra space is added
5813 after the line terminator so as not to end the header. The reasoning
5814 behind this is that bare CRs in header lines are most likely either
5815 to be mistakes, or people trying to play silly games.
5817 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5818 was being incorrectly given as 18 bytes larger than it should have been.
5819 This is a VOB (very old bug).
5821 60. This may never have affected anything current, but just in case it has:
5822 When the local host is found other than at the start of a list of hosts,
5823 the local host, those with the same MX, and any that follow, are discarded.
5824 When the list in question was part of a longer list of hosts, the following
5825 hosts (not currently being processed) were also being discarded. This no
5826 longer happens. I'm not sure if this situation could ever has previously
5829 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5832 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5833 header. This is something that is recommended in a new Internet Draft, and
5834 is something that is documented as being done by Sendmail. There are two
5835 possible values. For messages generated by the autoreply transport, Exim
5838 Auto-Submitted: auto-replied
5840 whereas for all other generated messages (e.g. bounces) it adds
5842 Auto-Submitted: auto-generated
5844 63. The "personal" condition in filters now includes a test for the
5845 Auto-Submitted: header. If it contains the string "auto-" the message it
5846 not considered personal.
5848 64. Added rcpt_include_affixes as a generic transport option.
5850 65. Added queue_only_override (default true).
5852 66. Added the syslog_duplication option.
5854 67. If what should have been the first header line of a message consisted of
5855 a space followed by a colon, Exim was mis-interpreting it as a header line.
5856 It isn't of course - it is syntactically invalid and should therefore be
5857 treated as the start of the message body. The misbehaviour could have
5858 caused a number of strange effects, including loss of data in subsequent
5859 header lines, and spool format errors.
5861 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5862 client host had authenticated. This control can now be exercised by an ACL
5863 for more flexibility.
5865 69. By default, callouts do not happen when testing with -bh. There is now a
5866 variant, -bhc, which does actually run the callout code, including
5867 consulting and updating the callout cache.
5869 70. Added support for saslauthd authentication, courtesy of Alexander
5872 71. If statvfs() failed on the spool or log directories while checking their
5873 size for availability, Exim confusingly gave the error "space shortage".
5874 Furthermore, in debugging mode it crashed with a floating point exception.
5875 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5876 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5877 serious problem, Exim now writes to the main and panic logs when this
5878 happens, with details of the failure. It then refuses to accept the
5879 incoming message, giving the message "spool directory problem" or "log
5880 directory problem" with a 421 code for SMTP messages.
5882 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5883 0, 1, and 2 exist, because some OS complain for execs without them (see
5884 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5885 descriptors. However, the code omitted to check that the open succeeded,
5886 causing mysterious errors if for some reason the permissions on /dev/null
5887 got screwed. Now Exim writes a message to the main and panic logs, and
5888 bombs out if it can't open /dev/null.
5890 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5891 interact so that it is all more flexible. It is supposed to remain
5892 backwards compatible. Also added extra_local_interfaces.
5894 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5895 to bomb out with an assertion failure - to the client this appears as a
5896 connection drop. This problem occurs in the part of the code that was taken
5897 from the Samba project. Fortunately, the assertion is in a very simple
5898 function, so I have fixed this by reproducing the function inline in the
5899 one place where it is called, and arranging for authentication to fail
5900 instead of killing the process with assert().
5902 75. The SPA client code was not working when the server requested OEM rather
5903 than Unicode encoding.
5905 76. Added code to make require_files with a specific uid setting more usable in
5906 the case where statting the file as root fails - usually a non-root-mounted
5907 NFS file system. When this happens and the failure is EACCES, Exim now
5908 forks a subprocess and does the per-uid checking as the relevant uid.
5910 77. Added process_log_path.
5912 78. If log_file_path was not explicitly set, a setting of check_log_space or
5913 check_log_inodes was ignored.
5915 79. If a space check for the spool or log partitions fails, the incident is now
5916 logged. Of course, in the latter case the data may get lost...
5918 80. Added the %p formatting code to string_format() so that it can be used to
5919 print addresses in debug_print(). Adjusted all the address printing in the
5920 debugging in store.c to use %p rather than %d.
5922 81. There was a concern that a line of code in smtp_in.c could overflow a
5923 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5924 initially expressed, the concern was not well-founded, because trailing
5925 spaces are removed early. However, if the trailing spaces were followed by
5926 a NULL, they did not get removed, so the overflow was possible. Two fixes
5929 (a) I re-wrote the offending code in a cleaner fashion.
5930 (b) If an incoming SMTP command contains a NULL character, it is rejected
5933 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5934 runs initgroups(), so that if the Exim user is in any additional groups,
5935 they will be used during message reception.
5941 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob