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.
21 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
23 PP/02 Make -n do something, by making it not do something.
24 When combined with -bP, the name of an option is not output.
26 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
29 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
30 $sender_host_name and config options to manage this, and basic check
33 PP/05 DSCP support for outbound connections and control modifier for inbound.
35 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
36 (Only plugin which currently uses this is kerberos4, which nobody should
37 be using, but we should make it available and other future plugins might
38 conceivably use it, even though it would break NAT; stuff *should* be
39 using channel bindings instead).
41 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
42 name; added for Sendmail compatibility; requires admin caller.
43 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
44 just ignore it); requires trusted caller.
45 Also parse but ignore: -Ac -Am -X<logfile>
48 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
50 TL/02 Add +smtp_confirmation as a default logging option.
52 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
53 Patch by Magnus Holmgren from 2007-02-20.
55 TL/04 Bugzilla 1281 - Spec typo.
56 Bugzilla 1283 - Spec typo.
57 Bugzilla 1290 - Spec grammar fixes.
59 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
61 TL/06 Add Experimental DMARC support using libopendmarc libraries.
63 TL/07 Fix an out of order global option causing a segfault. Reported to dev
64 mailing list by by Dmitry Isaikin.
66 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
68 JH/02 Support "G" suffix to numbers in ${if comparisons.
70 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
72 NM/01 Bugzilla 1197 - Spec typo
73 Bugzilla 1196 - Spec examples corrections
75 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
77 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
78 gnutls_enable_pkcs11, but renamed to more accurately indicate its
81 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
82 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
84 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
85 "acl {{name}{arg}...}", and optional args on acl condition
88 JH/05 Permit multiple router/transport headers_add/remove lines.
90 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
92 JH/07 Avoid using a waiting database for a single-message-only transport.
93 Performance patch from Paul Fisher. Bugzilla 1262.
95 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
98 JH/09 Add $headers_added variable, with content from use of ACL modifier
99 add_header (but not yet added to the message). Bugzilla 199.
101 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
102 Pulled from Bugzilla 817 by Wolfgang Breyha.
104 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
106 (nb: this is the same fix as in Exim 4.80.1)
108 JH/11 Add A= logging on delivery lines, and a client_set_id option on
111 JH/12 Add optional authenticated_sender logging to A= and a log_selector
114 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
116 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
117 advertise SMTP AUTH mechanism to us, instead of a generic
118 protocol violation error. Also, make Exim more robust to bad
119 data from the Dovecot auth socket.
121 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
123 When a queue runner is handling a message, Exim first routes the
124 recipient addresses, during which it prunes them based on the retry
125 hints database. After that it attempts to deliver the message to
126 any remaining recipients. It then updates the hints database using
129 So if a recipient address works intermittently, it can get repeatedly
130 deferred at routing time. The retry hints record remains fresh so the
131 address never reaches the final cutoff time.
133 This is a fairly common occurrence when a user is bumping up against
134 their storage quota. Exim had some logic in its local delivery code
135 to deal with this. However it did not apply to per-recipient defers
136 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
138 This change adds a proper retry rule check during routing so that the
139 final cutoff time is checked against the message's age. We only do
140 this check if there is an address retry record and there is not a
141 domain retry record; this implies that previous attempts to handle
142 the address had the retry_use_local_parts option turned on. We use
143 this as an approximation for the destination being like a local
144 delivery, as in LMTP.
146 I suspect this new check makes the old local delivery cutoff check
147 redundant, but I have not verified this so I left the code in place.
149 TF/02 Correct gecos expansion when From: is a prefix of the username.
151 Test 0254 submits a message to Exim with the header
155 When I ran the test suite under the user fanf2, Exim expanded
156 the header to contain my full name, whereas it should have added
157 a Resent-Sender: header. It erroneously treats any prefix of the
158 username as equal to the username.
160 This change corrects that bug.
162 GF/01 DCC debug and logging tidyup
163 Error conditions log to paniclog rather than rejectlog.
164 Debug lines prefixed by "DCC: " to remove any ambiguity.
166 TF/03 Avoid unnecessary rebuilds of lookup-related code.
168 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
169 Bug spotted by Jeremy Harris; was flawed since initial commit.
170 Would have resulted in OCSP responses post-SNI triggering an Exim
171 NULL dereference and crash.
173 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
175 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
176 Bug detection, analysis and fix by Samuel Thibault.
177 Bugzilla 1331, Debian bug #698092.
179 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
181 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
182 Server implementation by Todd Lyons, client by JH.
183 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
184 config variable "prdr_enable" controls whether the server
185 advertises the facility. If the client requests PRDR a new
186 acl_data_smtp_prdr ACL is called once for each recipient, after
187 the body content is received and before the acl_smtp_data ACL.
188 The client is controlled by bolth of: a hosts_try_prdr option
189 on the smtp transport, and the server advertisement.
190 Default client logging of deliveries and rejections involving
191 PRDR are flagged with the string "PRDR".
193 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
194 fclose(). Diagnosis by Todd Lyons.
196 PP/17 Update configure.default to handle IPv6 localhost better.
197 Patch by Alain Williams (plus minor tweaks).
200 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
201 This is now consistent with GnuTLS, and is now documented: the
202 previous undocumented portable approach to treating the option as
203 unset was to force an expansion failure. That still works, and
204 an empty string is now equivalent.
206 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
207 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
208 not performing validation itself.
210 PP/20 Added force_command boolean option to pipe transport.
211 Patch from Nick Koston, of cPanel Inc.
213 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
216 TF/04 Added udpsend ACL modifer and hexquote expansion operator
218 PP/21 Fix eximon continuous updating with timestamped log-files.
219 Broken in a format-string cleanup in 4.80, missed when I repaired the
220 other false fix of the same issue.
221 Report and fix from Heiko Schlichting.
224 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
225 Report from Prashanth Katuri.
227 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
228 It's SecureTransport, so affects any MacOS clients which use the
229 system-integrated TLS libraries, including email clients.
231 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
232 using a MIME ACL for non-SMTP local injection.
233 Report and assistance in diagnosis by Warren Baker.
235 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
237 JH/16 Fix comparisons for 64b. Bugzilla 1385.
239 TL/09 Add expansion variable $authenticated_fail_id to keep track of
240 last id that failed so it may be referenced in subsequent ACL's.
242 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
245 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
246 ldap library initialization, allowing self-signed CA's to be
247 used. Also properly sets require_cert option later in code by
248 using NULL (global ldap config) instead of ldap handle (per
249 session). Bug diagnosis and testing by alxgomz.
251 TL/12 Enhanced documentation in the ratelimit.pl script provided in
252 the src/util/ subdirectory.
254 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
255 renamed to Transport Post Delivery Action by Jeremy Harris, as
258 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
259 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
260 redis_servers = needs to be configured which will be used by the redis
261 lookup. Patch from Warren Baker, of The Packet Hub.
263 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
265 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
266 hostname or reverse DNS when processing a host list. Used suggestions
267 from multiple comments on this bug.
269 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
271 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
272 Missed a few lines, added it to make the runtest require no keyboard
275 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
276 contains upper case chars. Make router use caseful_local_part.
278 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
279 support when GnuTLS has been built with p11-kit.
285 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
287 This, or similar/improved, will also be change PP/11 of 4.82.
293 PP/01 Handle short writes when writing local log-files.
294 In practice, only affects FreeBSD (8 onwards).
295 Bugzilla 1053, with thanks to Dmitry Isaikin.
297 NM/01 Bugzilla 949 - Documentation tweak
299 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
302 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
304 PP/02 Implemented gsasl authenticator.
306 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
308 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
309 `pkg-config foo` for cflags/libs.
311 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
312 with rest of GSASL and with heimdal_gssapi.
314 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
315 `pkg-config foo` for cflags/libs for the TLS implementation.
317 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
318 properties get this fed in as external SSF. A number of robustness
319 and debugging improvements to the cyrus_sasl authenticator.
321 PP/08 cyrus_sasl server now expands the server_realm option.
323 PP/09 Bugzilla 1214 - Log authentication information in reject log.
324 Patch by Jeremy Harris.
326 PP/10 Added dbmjz lookup type.
328 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
330 PP/12 MAIL args handles TAB as well as SP, for better interop with
331 non-compliant senders.
332 Analysis and variant patch by Todd Lyons.
334 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
335 Bug report from Lars Müller <lars@samba.org> (via SUSE),
336 Patch from Dirk Mueller <dmueller@suse.com>
338 PP/13 tls_peerdn now print-escaped for spool files.
339 Observed some $tls_peerdn in wild which contained \n, which resulted
340 in spool file corruption.
342 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
343 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
344 or write after TLS renegotiation, which otherwise led to messages
347 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
348 as a tracking header (ie: a signed header comes before the signature).
349 Patch from Wolfgang Breyha.
351 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
352 comma-sep list; embedded commas doubled.
354 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
356 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
358 Report and patch from Dmitry Banschikov.
360 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
361 Removed SSL_clear() after SSL_new() which led to protocol negotiation
362 failures. We appear to now support TLS1.1+ with Exim.
364 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
365 lets Exim select keys and certificates based upon TLS SNI from client.
366 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
367 before an outbound SMTP session. New log_selector, +tls_sni.
369 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
370 NULL dereference. Report and patch from Alun Jones.
372 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
373 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
374 Not seeing resolver debug output on NetBSD, but suspect this is a
375 resolver implementation change.
377 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
378 Left warnings. Added "eximon gdb" invocation mode.
380 PP/21 Defaulting "accept_8bitmime" to true, not false.
382 PP/22 Added -bw for inetd wait mode support.
384 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
385 locate the relevant includes and libraries. Made this the default.
387 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
388 Bugzilla 1246, report and most of solution from Tomasz Kusy.
390 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
391 This may cause build issues on older platforms.
393 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
394 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
395 gnutls_require_mac & gnutls_require_protocols (no longer supported).
396 Added SNI support via GnuTLS too.
397 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
399 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
401 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
402 Applied second patch from Janne, implementing suggestion to default
403 multiple-strings-in-record handling to match SPF spec.
405 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
407 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
408 read-only, out of scope).
409 Patch from Wolfgang Breyha, report from Stuart Northfield.
411 PP/29 Fix three issues highlighted by clang analyser static analysis.
412 Only crash-plausible issue would require the Cambridge-specific
413 iplookup router and a misconfiguration.
414 Report from Marcin Mirosław.
416 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
418 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
419 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
420 As part of this, removing so much warning spew let me fix some minor
421 real issues in debug logging.
423 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
424 assignment on my part. Fixed.
426 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
427 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
428 Janne Snabb (who went above and beyond: thank you).
430 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
431 string otherwise requires a connection and a bunch more work and it's
432 relatively easy to get wrong. Should also expose TLS library linkage
435 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
436 64-bit ${eval} (JH/03).
438 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
439 GNU libc to support some of the 64-bit stuff, should not lead to
440 conflicts. Defined before os.h is pulled in, so if a given platform
441 needs to override this, it can.
443 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
444 protection layer was required, which is not implemented.
445 Bugzilla 1254, patch from Wolfgang Breyha.
447 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
448 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
449 tls_dhparam take prime identifiers. Also unbreak combination of
450 OpenSSL+DH_params+TLSSNI.
452 PP/39 Disable SSLv2 by default in OpenSSL support.
458 PP/01 Solaris build fix for Oracle's LDAP libraries.
459 Bugzilla 1109, patch from Stephen Usher.
461 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
463 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
466 TF/02 Fix a couple more cases where we did not log the error message
467 when unlink() failed. See also change 4.74-TF/03.
469 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
470 lock up or crash if it happened to be inside a call to libc when it
471 got a SIGUSR1 from exiwhat.
473 The SIGUSR1 handler appends the current process status to the process
474 log which is later printed by exiwhat. It used to use the general
475 purpose logging code to do this, but several functions it calls are
476 not safe for signals.
478 The new output code in the SIGUSR1 handler is specific to the process
479 log, and simple enough that it's easy to inspect for signal safety.
480 Removing some special cases also simplifies the general logging code.
481 Removing the spurious timestamps from the process log simplifies
484 TF/04 Improved ratelimit ACL condition.
486 The /noupdate option has been deprecated in favour of /readonly which
487 has clearer semantics. The /leaky, /strict, and /readonly update modes
488 are mutually exclusive. The update mode is no longer included in the
489 database key; it just determines when the database is updated. (This
490 means that when you upgrde Exim will forget old rate measurements.)
492 Exim now checks that the per_* options are used with an update mode that
493 makes sense for the current ACL. For example, when Exim is processing a
494 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
495 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
496 must specify per_mail/readonly. If you omit the update mode it defaults to
497 /leaky where that makes sense (as before) or /readonly where required.
499 The /noupdate option is now undocumented but still supported for
500 backwards compatibility. It is equivalent to /readonly except that in
501 ACLs where /readonly is required you may specify /leaky/noupdate or
502 /strict/noupdate which are treated the same as /readonly.
504 A useful new feature is the /count= option. This is a generalization
505 of the per_byte option, so that you can measure the throughput of other
506 aggregate values. For example, the per_byte option is now equivalent
507 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
509 The per_rcpt option has been generalized using the /count= mechanism
510 (though it's more complicated than the per_byte equivalence). When it is
511 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
512 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
513 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
514 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
515 Note that using per_rcpt with a non-readonly update mode in more than
516 one ACL will cause the recipients to be double-counted. (The per_mail
517 and per_byte options don't have this problem.)
519 The handling of very low rates has changed slightly. If the computed rate
520 is less than the event's count (usually one) then this event is the first
521 after a long gap. In this case the rate is set to the same as this event's
522 count, so that the first message of a spam run is counted properly.
524 The major new feature is a mechanism for counting the rate of unique
525 events. The new per_addr option counts the number of different
526 recipients that someone has sent messages to in the last time period. It
527 behaves like per_rcpt if all the recipient addresses are different, but
528 duplicate recipient addresses do not increase the measured rate. Like
529 the /count= option this is a general mechanism, so the per_addr option
530 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
531 example, measure the rate that a client uses different sender addresses
532 with the options per_mail/unique=$sender_address. There are further
533 details in the main documentation.
535 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
537 TF/06 Removed a few PCRE remnants.
539 TF/07 Automatically extract Exim's version number from tags in the git
540 repository when doing development or release builds.
542 PP/02 Raise smtp_cmd_buffer_size to 16kB.
543 Bugzilla 879. Patch from Paul Fisher.
545 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
546 Heavily based on revision 40f9a89a from Simon Arlott's tree.
549 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
551 PP/05 Variable $av_failed, true if the AV scanner deferred.
552 Bugzilla 1078. Patch from John Horne.
554 PP/06 Stop make process more reliably on build failure.
555 Bugzilla 1087. Patch from Heiko Schlittermann.
557 PP/07 Make maildir_use_size_file an _expandable_ boolean.
558 Bugzilla 1089. Patch from Heiko Schlittermann.
560 PP/08 Handle ${run} returning more data than OS pipe buffer size.
561 Bugzilla 1131. Patch from Holger Weiß.
563 PP/09 Handle IPv6 addresses with SPF.
564 Bugzilla 860. Patch from Wolfgang Breyha.
566 PP/10 GnuTLS: support TLS 1.2 & 1.1.
568 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
571 PP/11 match_* no longer expand right-hand-side by default.
572 New compile-time build option, EXPAND_LISTMATCH_RHS.
573 New expansion conditions, "inlist", "inlisti".
575 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
577 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
579 PP/14 fix log_write() format string regression from TF/03.
580 Bugzilla 1152. Patch from Dmitry Isaikin.
586 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
588 PP/02 Harmonised TLS library version reporting; only show if debugging.
589 Layout now matches that introduced for other libraries in 4.74 PP/03.
591 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
593 PP/04 New "dns_use_edns0" global option.
595 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
598 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
599 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
601 TK/01 Updated PolarSSL code to 0.14.2.
602 Bugzilla 1097. Patch from Andreas Metzler.
604 PP/07 Catch divide-by-zero in ${eval:...}.
607 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
610 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
611 format-string attack -- SECURITY: remote arbitrary code execution.
613 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
614 time unintentionally subject to list matching rules, letting the header
615 cause arbitrary Exim lookups (of items which can occur in lists, *not*
616 arbitrary string expansion). This allowed for information disclosure.
618 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
619 INT_MIN/-1 -- value coerced to INT_MAX.
625 NM/01 Workround for PCRE version dependency in version reporting
628 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
629 This fixes portability to compilers other than gcc, notably
630 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
632 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
633 makefiles for portability to HP-UX and POSIX correctness.
635 PP/01 Permit LOOKUP_foo enabling on the make command-line.
636 Also via indented variable definition in the Makefile.
637 (Debugging by Oliver Heesakkers).
639 PP/02 Restore caching of spamd results with expanded spamd_address.
640 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
642 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
643 Improves build reliability. Fix from: Frank Elsner
645 NM/02 Fix wide character breakage in the rfc2047 coding
646 Fixes bug 1064. Patch from Andrey N. Oktyabrski
648 NM/03 Allow underscore in dnslist lookups
649 Fixes bug 1026. Patch from Graeme Fowler
651 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
652 Code patches from Adam Ciarcinski of NetBSD.
654 NM/04 Fixed exiqgrep to cope with mailq missing size issue
657 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
658 is logged, to avoid truncation. Patch from John Horne.
660 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
661 Patch from Jakob Hirsch.
663 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
664 SQL string expansion failure details.
665 Patch from Andrey Oktyabrski.
667 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
668 Patch from Simon Arlott.
670 PP/09 New lookups functionality failed to compile on old gcc which rejects
671 extern declarations in function scope.
672 Patch from Oliver Fleischmann
674 PP/10 Use sig_atomic_t for flags set from signal handlers.
675 Check getgroups() return and improve debugging.
676 Fixed developed for diagnosis in bug 927 (which turned out to be
679 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
680 Patch from Mark Zealey.
682 PP/12 Bugzilla 1056: Improved spamd server selection.
683 Patch from Mark Zealey.
685 PP/13 Bugzilla 1086: Deal with maildir quota file races.
686 Based on patch from Heiko Schlittermann.
688 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
689 Patch from Uwe Doering, sign-off by Michael Haardt.
691 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
692 variable declaration deep within a block. Bug and patch from
695 PP/15 lookups-Makefile IRIX compatibilty coercion.
697 PP/16 Make DISABLE_DKIM build knob functional.
699 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
700 Patch by Simon Arlott
702 TF/03 Fix valgrind.h portability to C89 compilers that do not support
703 variable argument macros. Our copy now differs from upstream.
709 TF/01 Failure to get a lock on a hints database can have serious
710 consequences so log it to the panic log.
712 TF/02 Log LMTP confirmation messages in the same way as SMTP,
713 controlled using the smtp_confirmation log selector.
715 TF/03 Include the error message when we fail to unlink a spool file.
717 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
718 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
719 for maintaining out-of-tree patches for some time.
721 PP/01 Bugzilla 139: Documentation and portability issues.
722 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
723 Handle per-OS dynamic-module compilation flags.
725 PP/02 Let /dev/null have normal permissions.
726 The 4.73 fixes were a little too stringent and complained about the
727 permissions on /dev/null. Exempt it from some checks.
728 Reported by Andreas M. Kirchwitz.
730 PP/03 Report version information for many libraries, including
731 Exim version information for dynamically loaded libraries. Created
732 version.h, now support a version extension string for distributors
733 who patch heavily. Dynamic module ABI change.
735 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
736 privilege escalation vulnerability whereby the Exim run-time user
737 can cause root to append content of the attacker's choosing to
740 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
743 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
744 If dropping privileges for untrusted macros, we disabled normal logging
745 on the basis that it would fail; for the Exim run-time user, this is not
746 the case, and it resulted in successful deliveries going unlogged.
747 Fixed. Reported by Andreas Metzler.
753 PP/01 Date: & Message-Id: revert to normally being appended to a message,
754 only prepend for the Resent-* case. Fixes regression introduced in
755 Exim 4.70 by NM/22 for Bugzilla 607.
757 PP/02 Include check_rfc2047_length in configure.default because we're seeing
758 increasing numbers of administrators be bitten by this.
760 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
762 PP/03 Bugzilla 994: added openssl_options main configuration option.
764 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
766 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
768 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
770 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
771 without a peer certificate, leading to a segfault because of an
772 assumption that peers always have certificates. Be a little more
773 paranoid. Problem reported by Martin Tscholak.
775 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
776 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
777 NB: ClamAV planning to remove STREAM in "middle of 2010".
778 CL also introduces -bmalware, various -d+acl logging additions and
779 more caution in buffer sizes.
781 PP/09 Implemented reverse_ip expansion operator.
783 PP/10 Bugzilla 937: provide a "debug" ACL control.
785 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
787 PP/12 Bugzilla 973: Implement --version.
789 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
791 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
793 PP/15 Bugzilla 816: support multiple condition rules on Routers.
795 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
796 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
797 ignore trailing whitespace.
799 JJ/02 prevent non-panic DKIM error from being sent to paniclog
801 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
804 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
805 Notification from Dr Andrew Aitchison.
807 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
808 ExtendedDetectionInfo response format.
809 Notification from John Horne.
811 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
814 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
815 XSL and documented dependency on system catalogs, with examples of how
818 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
821 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
822 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
823 configuration file which is writeable by the Exim user or group.
825 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
826 of configuration files to cover files specified with the -C option if
827 they are going to be used with root privileges, not just the default
830 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
831 option (effectively making it always true).
833 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
834 files to be used while preserving root privileges.
836 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
837 that rogue child processes cannot use them.
839 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
840 run-time user, instead of root.
842 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
843 Exim run-time user without dropping privileges.
845 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
846 result string, instead of calling string_vformat() twice with the same
849 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
850 for other users. Others should always drop root privileges if they use
851 -C on the command line, even for a whitelisted configure file.
853 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
855 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
861 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
862 $data_path, and $header_path variables; fixed documentation bugs and
865 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
866 exipick to access non-standard spools, including the "frozen" queue
869 NM/01 Bugzilla 965: Support mysql stored procedures.
870 Patch from Alain Williams
872 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
874 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
875 Patch from Andreas Metzler
877 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
878 Patch from Kirill Miazine
880 NM/05 Bugzilla 671: Added umask to procmail example.
882 JJ/03 installed exipick 20100323.0, fixing doc bug
884 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
885 directory. Notification and patch from Dan Rosenberg.
887 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
889 TK/02 Improve log output when DKIM signing operation fails.
891 MH/01 Treat the transport option dkim_domain as a colon separated
892 list, not as a single string, and sign the message with each element,
893 omitting multiple occurences of the same signer.
895 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
896 Bugzilla 985, 986. Patch by Simon Arlott
898 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
899 Patch by Simon Arlott
901 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
902 MBX locking. Notification from Dan Rosenberg.
908 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
910 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
912 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
914 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
916 NM/04 Bugzilla 915: Flag broken perl installation during build.
922 TK/01 Added patch by Johannes Berg that expands the main option
923 "spamd_address" if it starts with a dollar sign.
925 TK/02 Write list of recipients to X-Envelope-Sender header when building
926 the mbox-format spool file for content scanning (suggested by Jakob
929 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
930 (http://www.dcc-servers.net/) support via dccifd. Activated by
931 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
933 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
934 by Mark Daniel Reidel <mr@df.eu>.
936 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
937 When building exim an external PCRE library is now needed -
938 PCRE is a system library on the majority of modern systems.
939 See entry on PCRE_LIBS in EDITME file.
941 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
942 conversation. Added nologin parameter to request.
943 Patch contributed by Kirill Miazine.
945 TF/01 Do not log submission mode rewrites if they do not change the address.
947 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
949 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
950 log files in place. Contributed by Roberto Lima.
952 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
954 TF/03 Bugzilla 615: When checking the local_parts router precondition
955 after a local_part_suffix or local_part_prefix option, Exim now
956 does not use the address's named list lookup cache, since this
957 contains cached lookups for the whole local part.
959 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
960 Robert Millan. Documentation is in experimental-spec.txt.
962 TF/04 Bugzilla 668: Fix parallel build (make -j).
964 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
966 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
967 Patch provided by Jan Srzednicki.
969 TF/05 Leading white space used to be stripped from $spam_report which
970 wrecked the formatting. Now it is preserved.
972 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
973 that they are available at delivery time.
975 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
977 TF/08 TLS error reporting now respects the incoming_interface and
978 incoming_port log selectors.
980 TF/09 Produce a more useful error message if an SMTP transport's hosts
981 setting expands to an empty string.
983 NM/06 Bugzilla 744: EXPN did not work under TLS.
984 Patch provided by Phil Pennock.
986 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
987 Patch provided by Richard Godbee.
989 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
990 acl_smtp_notquit, added index entry.
992 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
993 Patch provided by Eugene Bujak.
995 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
996 accept(). Patch provided by Maxim Dounin.
998 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
999 Patch provided by Phil Pennock.
1001 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1003 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1004 Patch provided by Brad "anomie" Jorsch.
1006 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1007 Patch provided by Dean Brooks.
1009 TK/05 Add native DKIM support (does not depend on external libraries).
1011 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1012 Patch provided by Graeme Fowler.
1014 NM/16 Bugzilla 851: Documentation example syntax fix.
1016 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1018 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1021 NM/19 Bugzilla 745: TLS version reporting.
1022 Patch provided by Phil Pennock.
1024 NM/20 Bugzilla 167: bool: condition support.
1025 Patch provided by Phil Pennock.
1027 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1028 clients. Patch provided by Phil Pennock.
1030 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1031 Patch provided by Brad "anomie" Jorsch.
1033 NM/23 Bugzilla 687: Fix misparses in eximstats.
1034 Patch provided by Heiko Schlittermann.
1036 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1037 Patch provided by Heiko Schlittermann.
1039 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1040 plus update to original patch.
1042 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1044 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1045 Patch provided by David Brownlee.
1047 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1049 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1051 NM/30 Bugzilla 888: TLS documentation bugfixes.
1053 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1055 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1056 Unlike the original bugzilla I have changed all shell scripts in src tree.
1058 NM/33 Bugzilla 898: Transport filter timeout fix.
1059 Patch by Todd Rinaldo.
1061 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1062 Patch by Serge Demonchaux.
1064 NM/35 Bugzilla 39: Base64 decode bug fixes.
1065 Patch by Jakob Hirsch.
1067 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1069 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1071 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1073 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1079 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1080 ALT-N's libdkim that I have put here:
1081 http://duncanthrax.net/exim-experimental/
1083 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1084 were called 'true' and it seems that C99 defines that as a reserved
1085 keyword to be used with 'bool' variable types. That means you could
1086 not include C99-style headers which use bools without triggering
1087 build errors in sieve.c.
1089 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1090 as mailq or other aliases. Changed the --help handling significantly
1091 to do whats expected. exim_usage() emits usage/help information.
1093 SC/01 Added the -bylocaldomain option to eximstats.
1095 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1097 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1099 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1105 PH/01 Another patch from the Sieve maintainer.
1107 PH/02 When an IPv6 address is converted to a string for single-key lookup
1108 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1109 dots are used instead of colons so that keys in lsearch files need not
1110 contain colons. This was done some time before quoting was made available
1111 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1112 (notated using the quote facility) so as to distinguish them from IPv4
1113 keys. This meant that lookups for IP addresses in host lists did not work
1114 for iplsearch lookups.
1116 This has been fixed by arranging for IPv6 addresses to be expressed with
1117 colons if the lookup type is iplsearch. This is not incompatible, because
1118 previously such lookups could never work.
1120 The situation is now rather anomolous, since one *can* have colons in
1121 ordinary lsearch keys. However, making the change in all cases is
1122 incompatible and would probably break a number of configurations.
1124 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1127 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1128 conversion specification without a maximum field width, thereby enabling
1129 a rogue spamd server to cause a buffer overflow. While nobody in their
1130 right mind would setup Exim to query an untrusted spamd server, an
1131 attacker that gains access to a server running spamd could potentially
1132 exploit this vulnerability to run arbitrary code as the Exim user.
1134 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1135 $primary_hostname instead of what libspf2 thinks the hosts name is.
1137 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1138 a directory entry by the name of the lookup key. Previously, if a
1139 symlink pointed to a non-existing file or a file in a directory that
1140 Exim lacked permissions to read, a lookup for a key matching that
1141 symlink would fail. Now it is enough that a matching directory entry
1142 exists, symlink or not. (Bugzilla 503.)
1144 PH/03 The body_linecount and body_zerocount variables are now exported in the
1147 PH/04 Added the $dnslist_matched variable.
1149 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1150 This means they are set thereafter only if the connection becomes
1153 PH/06 Added the client_condition to authenticators so that some can be skipped
1154 by clients under certain conditions.
1156 PH/07 The error message for a badly-placed control=no_multiline_responses left
1157 "_responses" off the end of the name.
1159 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1161 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1162 (without spaces) instead of just copying the configuration text.
1164 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1166 PH/11 Added $max_received_linelength.
1168 PH/12 Added +ignore_defer and +include_defer to host lists.
1170 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1171 way in which PCRE > 7.0 is built.
1173 PH/14 Implemented queue_only_load_latch.
1175 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1176 MAIL command. The effect was to mangle the value on 64-bit systems.
1178 PH/16 Another patch from the Sieve maintainer.
1180 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1182 PH/18 If a system quota error occurred while trying to create the file for
1183 a maildir delivery, the message "Mailbox is full" was not appended to the
1184 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1185 only to a quota excession during the actual writing of the file.
1187 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1188 characters?) which causes problems in log lines. The DN values are now
1189 passed through string_printing() before being added to log lines.
1191 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1192 and InterBase are left for another time.)
1194 PH/21 Added message_body_newlines option.
1196 PH/22 Guard against possible overflow in moan_check_errorcopy().
1198 PH/23 POSIX allows open() to be a macro; guard against that.
1200 PH/24 If the recipient of an error message contained an @ in the local part
1201 (suitably quoted, of course), incorrect values were put in $domain and
1202 $local_part during the evaluation of errors_copy.
1208 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1209 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1212 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1213 issue a MAIL command.
1215 PH/02 In an ACL statement such as
1217 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1219 if a client was not listed at all, or was listed with a value other than
1220 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1221 the condition was not true (as it should be), so access was not denied.
1222 The bug was that the ! inversion was incorrectly passed on to the second
1223 item. This has been fixed.
1225 PH/03 Added additional dnslists conditions == and =& which are different from
1226 = and & when the dns lookup returns more than one IP address.
1228 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1229 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1231 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1232 FSYNC, which compiles an option called disable_fsync that allows for
1233 bypassing fsync(). The documentation is heavily laced with warnings.
1235 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1237 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1238 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1239 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1240 including adding "make clean"; (3) Added -fPIC when compiling the test
1241 dynamically loaded module, to get rid of a warning.
1243 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1244 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1245 = 0s. The bug is otherwise harmless.
1247 PH/07 There was a bug in the dovecot authenticator such that the value of
1248 $auth1 could be overwritten, and so not correctly preserved, after a
1249 successful authentication. This usually meant that the value preserved by
1250 the server_setid option was incorrect.
1252 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1254 PH/09 Installed PCRE release 7.0.
1256 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1257 run for batched SMTP input. It is now run at the start of every message
1258 in the batch. While fixing this I discovered that the process information
1259 (output by running exiwhat) was not always getting set for -bs and -bS
1260 input. This is fixed, and it now also says "batched" for BSMTP.
1262 PH/11 Added control=no_pipelining.
1264 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1265 patch, slightly modified), and move the expansion of helo_data till after
1266 the connection is made in the smtp transport (so it can use these
1269 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1271 PH/14 Added log_selector = +pid.
1273 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1275 PH/16 Add ${if forany and ${if forall.
1277 PH/17 Added dsn_from option to vary the From: line in DSNs.
1279 PH/18 Flush SMTP output before performing a callout, unless control =
1280 no_callout_flush is set.
1282 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1283 was true (the default) a successful delivery failed to delete the retry
1284 item, thus causing premature timeout of the address. The bug is now
1287 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1289 PH/21 Long custom messages for fakedefer and fakereject are now split up
1290 into multiline reponses in the same way that messages for "deny" and
1291 other ACL rejections are.
1293 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1294 with slight modification.
1296 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1297 draft, changing the syntax and factoring some duplicate code".
1299 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1300 for deliveries of the second and subsequent messages over the same SMTP
1303 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1304 ${reduce, with only minor "tidies".
1306 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1308 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1309 expansion side effects.
1311 PH/27 When a message times out after an over-quota error from an Exim-imposed
1312 quota, the bounce message says "mailbox is full". This message was not
1313 being given when it was a system quota that was exceeded. It now should
1316 MH/03 Made $recipients available in local_scan(). local_scan() already has
1317 better access to the recipient list through recipients_list[], but
1318 $recipients can be useful in postmaster-provided expansion strings.
1320 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1321 in the case of a MAIL command with additional options following the
1322 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1323 were accidentally chopped off.
1325 PH/29 SMTP synchronization checks are implemented when a command is read -
1326 there is a check that no more input is waiting when there shouldn't be
1327 any. However, for some commands, a delay in an ACL can mean that it is
1328 some time before the response is written. In this time, more input might
1329 arrive, invalidly. So now there are extra checks after an ACL has run for
1330 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1331 pipelining has not been advertised.
1333 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1335 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1336 returned in the SMTP response when smtp_return_error_details was false.
1337 This has been fixed.
1339 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1340 instead of the C I/O that was originally supplied, because problems were
1341 reported on Solaris.
1343 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1344 Exim which did not show up earlier: it was assuming that a call to
1345 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1346 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1347 was a macro that became an assignment, so it seemed to work. This has
1348 changed to a proper function call with a void return, hence the compile
1349 error. Exim's code has been fixed.
1351 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1354 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1356 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1358 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1359 because it turns out to be too big or there is a timeout) there is no
1360 "Completed" line in the log. When some messages of this type were
1361 selected by exigrep, they were listed as "not completed". Others were
1362 picked up by some special patterns. I have improved the selection
1363 criteria to be more general.
1365 PH/38 The host_find_failed option in the manualroute router can now be set
1366 to "ignore", to completely ignore a host whose IP address cannot be
1367 found. If all hosts are ignored, the behaviour is controlled by the new
1368 host_all_ignored option.
1370 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1371 homing or because of multiple MX records with /mx) generated more than
1372 one IP address, and the following item turned out to be the local host,
1373 all the secondary addresses of the first item were incorrectly removed
1374 from the list, along with the local host and any following hosts (which
1375 is what is supposed to happen).
1377 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1378 whoever called Exim into the -H file. In the case of the daemon it was
1379 behaving confusingly. When first started, it used values for whoever
1380 started the daemon, but after a SIGHUP it used the Exim user (because it
1381 calls itself on a restart). I have changed the code so that it now always
1384 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1385 message are rejected with the same error (e.g. no authentication or bad
1386 sender address), and a DATA command is nevertheless sent (as can happen
1387 with PIPELINING or a stupid MUA), the error message that was given to the
1388 RCPT commands is included in the rejection of the DATA command. This is
1389 intended to be helpful for MUAs that show only the final error to their
1392 PH/42 Another patch from the Sieve maintainer.
1394 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1395 Eximstats - Fixed some broken HTML links and added missing column headers
1397 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1398 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1400 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1402 PH/43 Yet another patch from the Sieve maintainer.
1404 PH/44 I found a way to check for a TCP/IP connection going away before sending
1405 the response to the final '.' that terminates a message, but only in the
1406 case where the client has not sent further data following the '.'
1407 (unfortunately, this is allowed). However, in many cases there won't be
1408 any further data because there won't be any more messages to send. A call
1409 to select() can be used: if it shows that the input is "ready", there is
1410 either input waiting, or the socket has been closed. An attempt to read
1411 the next input character can distinguish the two cases. Previously, Exim
1412 would have sent an OK response which the client would never have see.
1413 This could lead to message repetition. This fix should cure that, at
1414 least in a lot of common cases.
1416 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1417 advertised in response to EHLO.
1423 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1424 fixed by 4.65/MH/01 (is this a record?) are fixed:
1426 (i) An empty string was always treated as zero by the numeric comparison
1427 operators. This behaviour has been restored.
1429 (ii) It is documented that the numeric comparison operators always treat
1430 their arguments as decimal numbers. This was broken in that numbers
1431 starting with 0 were being interpreted as octal.
1433 While fixing these problems I realized that there was another issue that
1434 hadn't been noticed. Values of message_size_limit (both the global option
1435 and the transport option) were treated as octal if they started with 0.
1436 The documentation was vague. These values are now always treated as
1437 decimal, and I will make that clear in the documentation.
1443 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1444 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1447 MH/01 Don't check that the operands of numeric comparison operators are
1448 integers when their expansion is in "skipping" mode (fixes bug
1449 introduced by 4.64-PH/07).
1451 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1452 child addresses, Exim now panics and dies. Previously, because the count
1453 is held in a short int, deliveries were likely to be lost. As such a
1454 large number of recipients for a single message is ridiculous
1455 (performance will be very, very poor), I have chosen to impose a limit
1456 rather than extend the field.
1462 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1463 leftover -K file (the existence of which was triggered by #402).
1464 While we were at it, introduced process PID as part of the -K
1465 filename. This should rule out race conditions when creating
1468 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1469 processing considerably. Previous code took too long for large mails,
1470 triggering a timeout which in turn triggers #401.
1472 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1473 in the DK code in transports.c. sendfile() is not really portable,
1474 hence the _LINUX specificness.
1476 TF/01 In the add_headers option to the mail command in an Exim filter,
1477 there was a bug that Exim would claim a syntax error in any
1478 header after the first one which had an odd number of characters
1481 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1482 callout verification, Exim cached a "reject" for the entire domain. This
1483 is correct for most verifications, but it is not correct for a recipient
1484 verification with use_sender or use_postmaster set, because in that case
1485 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1486 case of MAIL FROM:<> rejection from other early rejections (e.g.
1487 rejection of HELO). When verifying a recipient using a non-null MAIL
1488 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1489 Whatever the result of the callout, the value of the domain cache is
1490 left unchanged (for any other kind of callout, getting as far as trying
1491 RCPT means that the domain itself is ok).
1493 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1496 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1497 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1498 session, the daemon ignores ECONNECT errors and logs others; it now
1499 ignores EPIPE as well.
1501 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1502 (quoted-printable decoding).
1504 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1505 later the small subsequent patch to fix an introduced bug.
1507 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1509 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1511 PH/08 An error is now given if message_size_limit is specified negative.
1513 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1514 to be given (somewhat) arbitrary names.
1516 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1519 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1520 miscellaneous code fixes
1522 PH/10 Added the log_reject_target ACL modifier to specify where to log
1525 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1526 hostname. This is wrong, because it relates to the incoming message (and
1527 probably the interface on which it is arriving) and not to the outgoing
1528 callout (which could be using a different interface). This has been
1529 changed to use the value of the helo_data option from the smtp transport
1530 instead - this is what is used when a message is actually being sent. If
1531 there is no remote transport (possible with a router that sets up host
1532 addresses), $smtp_active_hostname is used.
1534 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1535 tweaks were necessary in order to get it to work (see also 21 below):
1536 (a) The code assumed that strncpy() returns a negative number on buffer
1537 overflow, which isn't the case. Replaced with Exim's string_format()
1539 (b) There were several signed/unsigned issues. I just did the minimum
1540 hacking in of casts. There is scope for a larger refactoring.
1541 (c) The code used strcasecmp() which is not a standard C function.
1542 Replaced with Exim's strcmpic() function.
1543 (d) The code set only $1; it now sets $auth1 as well.
1544 (e) A simple test gave the error "authentication client didn't specify
1545 service in request". It would seem that Dovecot has changed its
1546 interface. Fortunately there's a specification; I followed it and
1547 changed what the client sends and it appears to be working now.
1549 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1552 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1553 address A is aliased to B and C, where B exists and C does not. Without
1554 -v the output is "A verified" because verification stops after a
1555 successful redirection if more than one address is generated. However,
1556 with -v the child addresses are also verified. Exim was outputting "A
1557 failed to verify" and then showing the successful verification for C,
1558 with its parentage. It now outputs "B failed to verify", showing B's
1559 parentage before showing the successful verification of C.
1561 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1562 look up a TXT record in a specific list after matching in a combined
1565 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1566 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1567 they consult the DNS. I had assumed they would set it the way they
1568 wanted; and indeed my experiments on Linux seem to show that in some
1569 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1570 To be on the safe side, however, I have now made the interface to
1571 host_find_byname() similar to host_find_bydns(), with an argument
1572 containing the DNS resolver options. The host_find_byname() function now
1573 sets these options at its start, just as host_find_bydns() does. The smtp
1574 transport options dns_qualify_single and dns_search_parents are passed to
1575 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1576 of host_find_byname() use the default settings of RES_DEFNAMES
1577 (qualify_single) but not RES_DNSRCH (search_parents).
1579 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1580 spool_read_header() do less string testing, by means of a preliminary
1581 switch on the second character of optional "-foo" lines. (This is
1582 overdue, caused by the large number of possibilities that now exist.
1583 Originally there were few.) While I was there, I also converted the
1584 str(n)cmp tests so they don't re-test the leading "-" and the first
1585 character, in the hope this might squeeze out yet more improvement.
1587 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1588 flag allowing group syntax was set by the header_syntax check but not
1589 turned off, possible causing trouble later; (2) The flag was not being
1590 set at all for the header_verify test, causing "group"-style headers to
1591 be rejected. I have now set it in this case, and also caused header_
1592 verify to ignore an empty address taken from a group. While doing this, I
1593 came across some other cases where the code for allowing group syntax
1594 while scanning a header line wasn't quite right (mostly, not resetting
1595 the flag correctly in the right place). These bugs could have caused
1596 trouble for malformed header lines. I hope it is now all correct.
1598 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1599 with the "reply" argument non-NULL. The code, however (which originally
1600 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1601 but it didn't always do it. This confused somebody who was copying the
1602 code for some other use. I have removed all the tests.
1604 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1605 feature that was used to support insecure browsers during the U.S. crypto
1606 embargo. It requires special client support, and Exim is probably the
1607 only MTA that supported it -- and would never use it because real RSA is
1608 always available. This code has been removed, because it had the bad
1609 effect of slowing Exim down by computing (never used) parameters for the
1610 RSA_EXPORT functionality.
1612 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1613 authenticator to fail if there's a tab character in the incoming data
1614 (there should never be unless someone is messing about, as it's supposed
1615 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1616 if the connection is using TLS or if the remote IP is the same as the
1617 local IP, and the "valid-client-cert option" if a client certificate has
1620 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1621 authenticators. This can be used for authorization after authentication
1622 succeeds. (In the case of plaintext, it servers for both authentication
1625 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1626 if any retry times were supplied.
1628 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1629 connection, where there is no client IP address to check. In this
1630 situation, the verify now always succeeds.
1632 PH/25 Applied John Jetmore's -Mset patch.
1634 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1636 PH/27 In a string expansion for a processed (not raw) header when multiple
1637 headers of the same name were present, leading whitespace was being
1638 removed from all of them, but trailing whitespace was being removed only
1639 from the last one. Now trailing whitespace is removed from each header
1640 before concatenation. Completely empty headers in a concatenation (as
1641 before) are ignored.
1643 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1644 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1646 PH/29 [Removed. This was a change that I later backed out, and forgot to
1647 correct the ChangeLog entry (that I had efficiently created) before
1648 committing the later change.]
1650 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1651 address errors (4xx response to RCPT) over the same connection as other
1652 messages routed to the same hosts. Such deliveries are always "forced",
1653 so retry times are not inspected. This resulted in far too many retries
1654 for the affected addresses. The effect occurred only when there were more
1655 hosts than the hosts_max_try setting in the smtp transport when it had
1656 the 4xx errors. Those hosts that it had tried were not added to the list
1657 of hosts for which the message was waiting, so if all were tried, there
1658 was no problem. Two fixes have been applied:
1660 (i) If there are any address or message errors in an SMTP delivery, none
1661 of the hosts (tried or untried) are now added to the list of hosts
1662 for which the message is waiting, so the message should not be a
1663 candidate for sending over the same connection that was used for a
1664 successful delivery of some other message. This seems entirely
1665 reasonable: after all the message is NOT "waiting for some host".
1666 This is so "obvious" that I'm not sure why it wasn't done
1667 previously. Hope I haven't missed anything, but it can't do any
1668 harm, as the worst effect is to miss an optimization.
1670 (ii) If, despite (i), such a delivery is accidentally attempted, the
1671 routing retry time is respected, so at least it doesn't keep
1672 hammering the server.
1674 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1675 in ${readsocket because some servers need this prod.
1677 PH/32 Added some extra debug output when updating a wait-xxx database.
1679 PH/33 The hint "could be header name not terminated by colon", which has been
1680 given for certain expansion errors for a long time, was not being given
1681 for the ${if def:h_colon_omitted{... case.
1683 PH/34 The spec says: "With one important exception, whenever a domain list is
1684 being scanned, $domain contains the subject domain." There was at least
1685 one case where this was not true.
1687 PH/35 The error "getsockname() failed: connection reset by peer" was being
1688 written to the panic log as well as the main log, but it isn't really
1689 panic-worthy as it just means the connection died rather early on. I have
1690 removed the panic log writing for the ECONNRESET error when getsockname()
1693 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1694 runs only) independently of the message's sender address. This meant
1695 that, if the 4xx error was in fact related to the sender, a different
1696 message to the same recipient with a different sender could confuse
1697 things. In particualar, this can happen when sending to a greylisting
1698 server, but other circumstances could also provoke similar problems.
1699 I have changed the default so that the retry time for these errors is now
1700 based a combination of the sender and recipient addresses. This change
1701 can be overridden by setting address_retry_include_sender=false in the
1704 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1705 remote server are returned as part of bounce messages. This was not
1706 happening for LMTP over a pipe (the lmtp transport), but now it is the
1707 same for both kinds of LMTP.
1709 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1710 in header lines that were in fact CNAMEs. This is no longer the case.
1712 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1713 and queue runs started by the daemon processed all messages. This has
1714 been fixed so that -R and -S can now usefully be given with -q<time>.
1716 PH/40 Import PCRE release 6.7 (fixes some bugs).
1718 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1720 PH/42 Give an error if -q is specified more than once.
1722 PH/43 Renamed the variables $interface_address and $interface_port as
1723 $received_ip_address and $received_port, to make it clear that these
1724 values apply to message reception, and not to the outgoing interface when
1725 a message is delivered. (The old names remain recognized, of course.)
1727 PH/44 There was no timeout on the connect() call when using a Unix domain
1728 socket in the ${readsocket expansion. There now is.
1730 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1731 be meaningful with "accept".
1733 SC/01 Eximstats V1.43
1734 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1736 SC/02 Eximstats V1.44
1737 Use a glob alias rather than an array ref in the generated
1738 parser. This improves both readability and performance.
1740 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1741 Collect SpamAssassin and rejection statistics.
1742 Don't display local sender or destination tables unless
1743 there is data to show.
1744 Added average volumes into the top table text output.
1746 SC/04 Eximstats V1.46
1747 Collect data on the number of addresses (recipients)
1748 as well as the number of messages.
1750 SC/05 Eximstats V1.47
1751 Added 'Message too big' to the list of mail rejection
1752 reasons (thanks to Marco Gaiarin).
1754 SC/06 Eximstats V1.48
1755 Mainlog lines which have GMT offsets and are too short to
1756 have a flag are now skipped.
1758 SC/07 Eximstats V1.49 (Alain Williams)
1759 Added the -emptyok flag.
1761 SC/08 Eximstats V1.50
1762 Fixes for obtaining the IP address from reject messages.
1764 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1765 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1766 whitesspace changes from 4.64-PH/27
1768 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1771 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1772 are found, allow negative numbers in numeric criteria)
1774 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1776 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1777 to match changes made in 4.64-PH/43
1779 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1781 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1782 "domain" is actually an IP address, and force a failure. This locks out
1783 those revolvers/nameservers that support "A-for-A" lookups, in
1784 contravention of the specifications.
1786 PH/48 When a host name was looked up from an IP address, and the subsequent
1787 forward lookup of the name timed out, the host name was left in
1788 $sender_host_name, contrary to the specification.
1790 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1791 restricted to single-key lookups, Exim was not diagnosing an error if
1792 * or *@ was used with a query-style lookup.
1794 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1796 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1797 long ago, but noone interested enough thought of it. Let's just say that
1798 the "1.1" means that there are some new functions that weren't there at
1799 some point in the past.
1801 PH/51 Error processing for expansion failure of helo_data from an smtp
1802 transport during callout processing was broken.
1804 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1805 tested/used via the -bh/-bhc/-bs options.
1807 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1808 bug, fixed in subsequent PCRE releases).
1810 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1811 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1813 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1819 SC/01 Use a glob alias rather than an array ref in eximstats generated
1820 parser. This improves both readability and performance.
1822 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1823 Don't display local sender or destination tables in eximstats unless
1824 there is data to show.
1825 Added average volumes into the eximstats top table text output.
1827 SC/03 Collect data on the number of addresses (recipients) as well
1828 as the number of messages in eximstats.
1830 TF/01 Correct an error in the documentation for the redirect router. Exim
1831 does (usually) call initgroups() when daemonizing.
1833 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1834 with consistent privilege compared to when running as a daemon.
1836 TF/03 Note in the spec that $authenticated_id is not set for local
1837 submissions from trusted users.
1839 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1840 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1842 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1843 by adding some example configuration directives to the default
1844 configuration file. A little bit of work is required to uncomment the
1845 directives and define how usernames and passwords are checked, but
1846 there is now a framework to start from.
1848 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1849 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1850 without this. I don't know how relevant this is to other LDAP libraries.
1852 PH/02 Add the verb name to the "unknown ACL verb" error.
1854 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1856 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1858 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1859 directory not expanded when it should be if an expanded home directory
1860 was set for the address (which is overridden by the transport).
1862 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1865 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1866 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1867 because it is too late at that time, and has no effect.
1869 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1870 security issue with \' (bugzilla #107). I could not use the
1871 PQescapeStringConn() function, because it needs a PGconn value as one of
1874 PH/08 When testing addresses using -bt, indicate those final addresses that
1875 are duplicates that would not cause an additional delivery. At least one
1876 person was confused, thinking that -bt output corresponded to deliveries.
1877 (Suppressing duplicates isn't a good idea as you lose the information
1878 about possibly different redirections that led to the duplicates.)
1880 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1881 systems where poll() doesn't work, in particular OS X.
1883 PH/10 Added more information to debugging output for retry time not reached.
1885 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1886 operations in malware.c.
1888 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1891 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1892 syslog_duplication set false, log lines that would normally be written
1893 both the the main log and to the reject log were not written to syslog at
1896 PH/14 In the default configuration, change the use of "message" in ACL warn
1897 statements to "add_header".
1899 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1900 not followed by a command (e.g. "seen endif").
1902 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1903 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1906 PH/17 Added extra conditions to the default value of delay_warning_condition
1910 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1911 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1912 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1915 The Auto-Submitted: and various List- headers are standardised, whereas I
1916 don't think Precedence: ever was.
1918 PH/18 Refactored debugging code in route_finduser() to show more information,
1919 in particular, the error code if getpwnam() issues one.
1921 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1922 This is apparently needed in addition to the PH/07 change above to avoid
1923 any possible encoding problems.
1925 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1926 but not after initializing Perl.
1928 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1929 output them only if debugging. By default they are written stderr,
1930 apparently, which is not desirable.
1932 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1935 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1938 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1940 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1941 authenticated or an ident call has been made. Suppress the default
1942 values for $authenticated_id and $authenticated_sender (but permit -oMai
1943 and -oMas) when testing with -bh.
1945 PH/24 Re-jigged the order of the tests in the default configuration so that the
1946 tests for valid domains and recipients precede the DNS black list and CSA
1947 tests, on the grounds that those ones are more expensive.
1949 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1950 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1951 This bug exists in every version of Exim that I still have, right back to
1954 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1955 However, an attempt to turn on case-sensitivity in a regex key by
1956 including (?-i) didn't work because the subject string was already
1957 lowercased, and the effects were non-intuitive. It turns out that a
1958 one-line patch can be used to allow (?-i) to work as expected.
1964 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1965 other effects) broke the use of negated acl sub-conditions.
1967 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1970 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1971 "Deny" causes Exim to reject the incoming connection with a 554 error.
1972 Unfortunately, if there is a major crisis, such as a disk failure,
1973 tcp-wrappers gives "deny", whereas what one would like would be some
1974 kind of temporary error. A kludge has been added to help with this.
1975 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1976 554 error is used if errno is still zero or contains ENOENT (which occurs
1977 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1980 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1982 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1983 errors. Otherwise a message that provokes a temporary error (when other
1984 messages do not) can cause a whole host to time out.
1986 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1987 addresses were routed directly to files or pipes from a redirect router.
1988 File deliveries just didn't batch; pipe deliveries might have suffered
1991 PH/06 A failure to get a lock for a hints database would erroneously always say
1992 "Failed to get write lock", even when it was really a read lock.
1994 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1995 of 0600. This has been changed to use the value of the lockfile_mode
1996 option (which defaults to 0600).
1998 PH/08 Applied small patch from the Sieve maintainer.
2000 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2001 folder from quota calculations, a direct delivery into this folder messed
2002 up the contents of the maildirsize file. This was because the regex was
2003 used only to exclude .Trash (or whatever) when the size of the mailbox
2004 was calculated. There was no check that a delivery was happening into an
2005 excluded directory. This bug has been fixed by ignoring all quota
2006 processing for deliveries into excluded directories.
2008 PH/10 Added the maildirfolder_create_regex option to appendfile.
2014 PH/01 The code for finding all the local interface addresses on a FreeBSD
2015 system running IPv6 was broken. This may well have applied to all BSD
2016 systems, as well as to others that have similar system calls. The broken
2017 code found IPv4 interfaces correctly, but gave incorrect values for the
2018 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2019 that it would not match correctly against @[] and not recognize the IPv6
2022 PH/02 The ipliteral router was not recognizing addresses of the form user@
2023 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2025 PH/03 Added disable_ipv6.
2027 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2028 decoded form, because it is most often used to construct To: headers
2029 lines in autoreplies, and the decoded form may well be syntactically
2030 invalid. However, $reply_address has leading white space removed, and all
2031 newlines turned into spaces so that the autoreply transport does not
2034 PH/05 If group was specified without a user on a router, and no group or user
2035 was specified on a transport, the group from the router was ignored.
2037 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2038 for visible compile-time settings that can be used to change these
2039 numbers, for those that want even more. Backwards compatibility with old
2040 spool files has been maintained. However, going back to a previous Exim
2041 release will lost any variables that are in spool files.
2043 PH/07 Two small changes when running in the test harness: increase delay when
2044 passing a TCP/IP connection to a new process, in case the original
2045 process has to generate a bounce, and remove special handling of
2046 127.0.0.2 (sic), which is no longer necessary.
2048 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2049 be the same on different OS.
2051 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2054 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2055 whether --show-vars was specified or not
2057 JJ/02 exipick: Added support for new ACL variable spool format introduced
2060 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2061 syntactically invalid From: or Reply-to: line, and a filter used this to
2062 generate an autoreply, and therefore failed to obtain an address for the
2063 autoreply, Exim could try to deliver to a non-existent relative file
2064 name, causing unrelated and misleading errors. What now happens is that
2065 it logs this as a hard delivery error, but does not attempt to create a
2068 PH/11 The exinext utility has a -C option for testing purposes, but although
2069 the given file was scanned by exinext itself; it wasn't being passed on
2070 when Exim was called.
2072 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2073 an end-of-file indication when reading a command response.
2075 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2076 compiled. In many other places in Exim, IPv6 addresses are always
2077 recognized, so I have changed this. It also means that IPv4 domain
2078 literals of the form [IPV4:n.n.n.n] are now always recognized.
2080 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2081 used if the router is not running as root, for example, when verifying at
2082 ACL time, or when using -bh. The debugging output from this situation was
2083 non-existent - all you got was a failure to exec. I have made two
2086 (a) Failures to set uid/gid, the current directory, or a process leader
2087 in a subprocess such as that created by queryprogram now generate
2088 suitable debugging ouput when -d is set.
2090 (b) The queryprogram router detects when it is not running as root,
2091 outputs suitable debugging information if -d is set, and then runs
2092 the subprocess without attempting to change uid/gid.
2094 PH/15 Minor change to Makefile for building test_host (undocumented testing
2097 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2098 additional section of a DNS packet that returns MX or SRV records.
2099 Instead, it always explicitly searches for A/AAAA records. This avoids
2100 major problems that occur when a DNS server includes only records of one
2101 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2102 fixed another bug: if SRV records were looked up and the corresponding
2103 address records were *not* found in the additional section, the port
2104 values from the SRV records were lost.
2106 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2107 using the correct key (the original address) when searching the retry
2108 rules in order to find which one to use for generating the retry hint.
2110 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2111 adding the default one. Similarly, if it contains a Reply-To: header, the
2112 errors_reply_to option, if set, is not used.
2114 PH/19 When calculating a retry time, Exim used to measure the "time since
2115 failure" by looking at the "first failed" field in the retry record. Now
2116 it does not use this if it is later than than the arrival time of the
2117 message. Instead it uses the arrival time. This makes for better
2118 behaviour in cases where some deliveries succeed, thus re-setting the
2119 "first failed" field. An example is a quota failure for a huge message
2120 when small messages continue to be delivered. Without this change, the
2121 "time since failure" will always be short, possible causing more frequent
2122 delivery attempts for the huge message than are intended.
2123 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2125 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2126 $1, $2, $3) because the numerical variables can be reset during some
2127 expansion items (e.g. "match"), thereby losing the authentication data.
2129 PH/21 Make -bV show the size of off_t variables so that the test suite can
2130 decide whether to run tests for quotas > 2G.
2132 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2133 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2134 filecount value is greater than 2G or if a quota value is greater than 2G
2135 on a system where the size of off_t is not greater than 4, a panic error
2138 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2139 never match. The debug and -bh output now contains an explicit error
2140 message indicating a malformed IPv4 address or mask.
2142 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2143 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2144 PH/23 above applies.
2146 PH/25 Do not write to syslog when running in the test harness. The only
2147 occasion when this arises is a failure to open the main or panic logs
2148 (for which there is an explicit test).
2150 PH/26 Added the /no_tell option to "control=freeze".
2152 PH/27 If a host name lookup failed very early in a connection, for example, if
2153 the IP address matched host_lookup and the reverse lookup yielded a name
2154 that did not have a forward lookup, an error message of the form "no IP
2155 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2156 could be logged. Now it outputs the IP address instead of "NULL".
2158 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2159 allows the sender and the authenticated sender to be set when
2160 submitting a message from within Exim. Since child_open_exim() is
2161 documented for local_scan(), the new function should be too.
2163 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2164 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2165 results in an empty string is now treated as unset.
2167 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2169 PH/31 Added sender_verify_fail logging option.
2171 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2172 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2173 tidied the source and removed it altogether.
2175 PH/33 When a queue run was abandoned because the load average was too high, a
2176 log line was always written; now it is written only if the queue_run log
2177 selector is set. In addition, the log line for abandonment now contains
2178 information about the queue run such as the pid. This is always present
2179 in "start" and "stop" lines but was omitted from the "abandon" line.
2181 PH/34 Omit spaces between a header name and the colon in the error message that
2182 is given when verify = headers_syntax fails (if there are lots of them,
2183 the message gets confusing).
2185 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2186 names, as there are now some PTR records that contain slashes. This check
2187 is only to protect against broken name servers that fall over on strange
2188 characters, so the fact that it applies to all lookups doesn't matter.
2190 PH/36 Now that the new test suite is complete, we can remove some of the
2191 special code in Exim that was needed for the old test suite. For example,
2192 sorting DNS records because real resolvers return them in an arbitrary
2193 order. The new test suite's fake resolver always returns records in the
2196 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2197 bounces) except when queue_only is set, to avoid logging races between
2198 the different processes.
2200 PH/38 Panic-die if .include specifies a non-absolute path.
2202 PH/39 A tweak to the "H" retry rule from its user.
2204 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2205 a label. They prevented compilation on older perls.
2207 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2208 a warning to be raised on newish perls.
2210 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2211 on queue. Changes to match documented behaviour of showing count of
2212 messages matching specified criteria.
2214 PH/40 Changed the default ident timeout from 30s to 5s.
2216 PH/41 Added support for the use of login_cap features, on those BSD systems
2217 that have them, for controlling the resources used by pipe deliveries.
2219 PH/42 The content-scanning code uses fopen() to create files in which to put
2220 message data. Previously it was not paying any attention to the mode of
2221 the files. Exim runs with umask(0) because the rest of the code creates
2222 files with open(), and sets the required mode explicitly. Thus, these
2223 files were ending up world-writeable. This was not a big issue, because,
2224 being within the spool directory, they were not world-accessible. I have
2225 created a function called modefopen, which takes an additional mode
2226 argument. It sets umask(777), creates the file, chmods it to the required
2227 mode, then resets the umask. All the relevant calls to fopen() in the
2228 content scanning code have been changed to use this function.
2230 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2231 to 24 hours. This avoids potential overflow problems when processing G
2232 and H retry rules. I suspect nobody ever tinkers with this value.
2234 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2236 PH/45 When the plaintext authenticator is running as a client, the server's
2237 challenges are checked to ensure they are valid base64 strings. By
2238 default, the authentication attempt is cancelled if an invalid string is
2239 received. Setting client_ignore_invalid_base64 true ignores these errors.
2240 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2241 they are received. Thus, the responses can be made to depend on the
2242 challenges. If an invalid string is ignored, an empty string is placed in
2245 PH/46 Messages that are created by the autoreply transport now contains a
2246 References: header, in accordance with RFCs 2822 and 3834.
2248 PH/47 Added authenticated_sender_force to the smtp transport.
2250 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2252 PH/49 Installed latest patch from the Sieve maintainer.
2254 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2255 was also set, the appendfile transport was unnecessarily scanning a
2256 directory of message files (e.g. for maildir delivery) to find the count
2257 of files (along with the size), even though it did not need this
2258 information. It now does the scan only if it needs to find either the
2259 size of the count of files.
2261 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2263 PH/52 Two bugs concerned with error handling when the smtp transport is
2266 (i) Exim was not creating retry information for temporary errors given
2267 for individual recipients after the DATA command when the smtp transport
2268 was used in LMTP mode. This meant that they could be retried too
2269 frequently, and not timed out correctly.
2271 (ii) Exim was setting the flag that allows error details to be returned
2272 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2273 recipients that were returned after the DATA command.
2275 PH/53 This is related to PH/52, but is more general: for any failing address,
2276 when detailed error information was permitted to be returned to the
2277 sender, but the error was temporary, then after the final timeout, only
2278 "retry timeout exceeded" was returned. Now it returns the full error as
2279 well as "retry timeout exceeded".
2281 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2282 do this, and (what is worse) MTAs that accept it.
2284 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2285 will now be deprecated.
2287 PH/56 New os.c-cygwin from the Cygwin maintainer.
2289 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2290 formats (previously only available in exim formats via -bpr, -bpru,
2291 and -bpra. Now also available in native and exiqgrep formats)
2293 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2294 with very large, slow to parse queues
2296 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2298 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2300 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2301 responses to authentication challenges, though it was showing the
2302 challenges; (ii) I've removed the CR characters from the debug output for
2305 PH/58 Allow for the insertion of a newline as well as a space when a string
2306 is turned into more than one encoded-word during RFC 2047 encoding. The
2307 Sieve code now uses this.
2309 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2310 data_4xx, lost_connection, tls_required.
2312 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2313 message was being sent as an SMTP response.
2315 PH/61 Add -l and -k options to exicyclog.
2317 PH/62 When verifying, if an address was redirected to one new address, so that
2318 verification continued, and the new address failed or deferred after
2319 having set something in $address_data, the value of $address_data was not
2320 passed back to the ACL. This was different to the case when no
2321 redirection occurred. The value is now passed back in both cases.
2323 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2324 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2325 use login_cap.h, so on its own it isn't the distinguishing feature. The
2326 new name refers directly to the setclassresources() function.
2328 PH/65 Added configuration files for NetBSD3.
2330 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2332 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2333 is preferred over IPv4.
2335 PH/68 The bounce_return_message and bounce_return_body options were not being
2336 honoured for bounces generated during the reception of non-SMTP messages.
2337 In particular, this applied to messages rejected by the ACL. This bug has
2338 been fixed. However, if bounce_return_message is true and bounce_return_
2339 body is false, the headers that are returned for a non-SMTP message
2340 include only those that have been read before the error was detected.
2341 (In the case of an ACL rejection, they have all been read.)
2343 PH/69 The HTML version of the specification is now built in a directory called
2344 spec_html instead of spec.html, because the latter looks like a path with
2345 a MIME-type, and this confuses some software.
2347 PH/70 Catch two compiler warnings in sieve.c.
2349 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2350 function verify_get_ident() calls ip_connect() to connect a socket, but
2351 if the "connect()" function timed out, ip_connect() used to close the
2352 socket. However, verify_get_ident() also closes the socket later, and in
2353 between Exim writes to the log, which may get opened at this point. When
2354 the socket was closed in ip_connect(), the log could get the same file
2355 descriptor number as the socket. This naturally causes chaos. The fix is
2356 not to close the socket in ip_connect(); the socket should be closed by
2357 the function that creates it. There was only one place in the code where
2358 this was missing, in the iplookup router, which I don't think anybody now
2359 uses, but I've fixed it anyway.
2361 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2362 well as to direct DNS lookups. Otherwise the handling of names in host
2363 lists is inconsistent and therefore confusing.
2369 PH/01 Two changes to the default runtime configuration:
2371 (1) Move the checks for relay_from_hosts and authenticated clients from
2372 after to before the (commented out) DNS black list checks.
2374 (2) Add control=submission to the relay_from_hosts and authenticated
2375 clients checks, on the grounds that messages accepted by these
2376 statements are most likely to be submissions.
2378 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2380 (1) Generate an error if the third argument for the ${prvs expansion is
2383 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2386 (3) Reset the variables that are obtained from the first argument of
2387 ${prvscheck and used in the second argument before leaving the code,
2388 because their memory is reclaimed, so using them afterwards may do
2391 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2392 one (it's much easier than Tom thought :-).
2394 (5) Because of (4), we can now allow for the use of $prvscheck_result
2395 inside the third argument.
2397 PH/03 For some reason, the default setting of PATH when running a command from
2398 a pipe transport was just "/usr/bin". I have changed it to
2401 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2402 anything to be listed in the output from -bV.
2404 PH/05 When a filter generated an autoreply, the entire To: header line was
2405 quoted in the delivery log line, like this:
2407 => >A.N.Other <ano@some.domain> <original@ddress> ...
2409 This has been changed so that it extracts the operative address. There
2410 may be more than one such address. If so, they are comma-separated, like
2413 => >ano@some.domain,ona@other.domain <original@ddress> ...
2415 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2416 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2417 not being looked up in the rDNS to get a host name, Exim was showing the
2418 IP address twice in Received: lines, even though the IP addresses were
2419 identical. For example:
2421 Received: from [1.2.3.4] (helo=[1.2.3.4])
2423 However, if the real host name was known, it was omitting the HELO data
2424 if it matched the actual IP address. This has been tidied up so that it
2425 doesn't show the same IP address twice.
2427 PH/07 When both +timestamp and +memory debugging was on, the value given by
2428 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2429 called by the debug printing, thereby overwriting the timestamp buffer.
2430 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2432 PH/08 When the original message was included in an autoreply transport, it
2433 always said "this is a copy of the message, including all the headers",
2434 even if body_only or headers_only was set. It now gives an appropriate
2437 PH/09 Applied a patch from the Sieve maintainer which:
2439 o fixes some comments
2440 o adds the (disabled) notify extension core
2441 o adds some debug output for the result of if/elsif tests
2442 o points to the current vacation draft in the documentation
2443 and documents the missing references header update
2447 o fixes a bug in processing the envelope test (when testing
2448 multiple envelope elements, the last element determinted the
2451 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2452 Electronic Mail") by including:
2454 Auto-submitted: auto-generated
2456 in the messages that it generates (bounce messages and others, such as
2457 warnings). In the case of bounce messages for non-SMTP mesages, there was
2458 also a typo: it was using "Auto_submitted" (underscore instead of
2459 hyphen). Since every message generated by Exim is necessarily in response
2460 to another message, thes have all been changed to:
2462 Auto-Submitted: auto-replied
2464 in accordance with these statements in the RFC:
2466 The auto-replied keyword:
2468 - SHOULD be used on messages sent in direct response to another
2469 message by an automatic process,
2471 - MUST NOT be used on manually-generated messages,
2473 - MAY be used on Delivery Status Notifications (DSNs) and Message
2474 Disposition Notifications (MDNs),
2476 - MUST NOT be used on messages generated by automatic or periodic
2477 processes, except for messages which are automatic responses to
2480 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2481 to the default Received: header definition.
2483 PH/12 Added log selector acl_warn_skipped (default on).
2485 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2486 variables because (a) they are in the wrong storage pool and (b) even if
2487 they were copied, it wouldn't work properly because of the caching.
2489 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2490 checking when decoding. Apparently there are clients that generate
2491 overlong encoded strings. Why am I not surprised?
2493 PH/15 If the first argument of "${if match_address" was not empty, but did not
2494 contain an "@" character, Exim crashed. Now it writes a panic log message
2495 and treats the condition as false.
2497 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2499 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2500 extension "envelope-auth". The code is finished and in agreement with
2501 other implementations, but there is no documentation so far and in fact,
2502 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2503 not changing the active code.
2505 Print executed "if" and "elsif" statements when debugging is used. This
2506 helps a great deal to understand what a filter does.
2508 Document more things not specified clearly in RFC3028. I had all this
2509 sorted out, when out of a sudden new issues came to my mind. Oops."
2511 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2514 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2515 canonical form (as documented). However, after a host name lookup from
2516 the IP address, check_host() was doing a simple string comparison with
2517 addresses acquired from the DNS when checking that the found name did
2518 have the original IP as one of its addresses. Since any found IPv6
2519 addresses are likely to be in abbreviated form, the comparison could
2520 fail. Luckily, there already exists a function for doing the comparison
2521 by converting both addresses to binary, so now that is used instead of
2522 the text comparison.
2524 PH/20 There was another similar case to PH/19, when a complete host name was
2525 given in a host list; looking up its IP address could give an abbreviated
2526 form, whereas the current host's name might or might not be abbreviated.
2527 The same fix has been applied.
2533 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2534 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2537 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2538 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2540 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2542 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2543 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2544 Also, for "Auto-Submitted", treat anything other than "no" as
2545 non-personal, in accordance with RFC 3834. (Previously it treated
2546 anything starting "auto-" as non-personal.)
2548 TF/01 The control=submission/name=... option had a problem with syntax
2549 errors if the name included a slash character. The /name= option
2550 now slurps the rest of the string, so it can include any characters
2551 but it must come last in the list of options (after /sender_retain
2554 PH/05 Some modifications to the interface to the fake nameserver for the new
2562 TK/01 Added the "success_on_redirect" address verification option. See
2563 NewStuff for rationale and an example.
2565 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2567 PH/02 Patch to exigrep to allow it to work on syslog lines.
2569 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2570 fread() to skip over the body file's header line, because in Cygwin the
2571 header line is locked and is inaccessible.
2573 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2574 co-exist for some time) to make it clear that it is the Exim ID that is
2575 referenced, not the Message-ID: header line.
2577 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2578 string_format() function, because snprintf() does not exist on all
2581 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2582 ${stat: expansion item.
2584 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2585 protocol synchronization error", to keep the pedants happy.
2587 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2588 well as for IRIX systems, when gcc is being used. See the host.c source
2591 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2593 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2596 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2597 transport and to the smtp transport in LMTP mode.
2599 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2601 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2602 run a filter in a subprocess. This could lead to confusion in subsequent
2603 lookups in the parent process. There should also be a search_tidyup() at
2604 the end of the subprocess.
2606 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2607 only if the host matched helo_try_verify_hosts, which caused the
2608 verification to occur when the EHLO/HELO command was issued. The ACL just
2609 tested the remembered result. Now, if a previous verification attempt has
2610 not happened, "verify = helo" does it there and then.
2612 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2614 TK/03 Fix log output including CR from clamd.
2616 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2617 exist provoked a memory error which could cause a segfault.
2619 PH/15 Installed PCRE 6.2
2621 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2623 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2624 of the problem. Specifically, suggested +O2 rather than +O1 for the
2627 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2629 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2630 record, the port was not correctly logged when the outgoing_port log
2631 selector was set (it logged the transort's default port).
2633 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2634 fallback_hosts, and "hosts" in the smtp transport.
2636 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2637 host errors such as "Connection refused".
2639 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2640 authentication with radiusclient 0.4.9:
2642 - Error returned from rc_read_config was caught wrongly
2643 - Username/password not passed on to radius server due to wrong length.
2645 The presumption is that some radiusclient API changes for 4.51/PH/17
2646 were not taken care of correctly. The code is still untested by me (my
2647 Linux distribution still has 0.3.2 of radiusclient), but it was
2648 contributed by a Radius user.
2650 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2651 expanding the "port" option of the smtp transport.
2653 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2654 while reading a MIME header. Thanks to Tom Hughes for a patch.
2656 PH/24 Include config.h inside local_scan.h so that configuration settings are
2659 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2660 that in an ACL for RCPT (for example), you can examine exactly what was
2663 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2664 commands, but it was not correctly comparing the address with the actual
2665 client host address. Thus, it would show the EHLO address in Received:
2666 header lines when this was not necessary.
2668 PH/27 Added the % operator to ${eval:}.
2670 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2671 it should be ignoring failures (because with -C, for example, it has lost
2672 privilege). It wasn't ignoring creation failures other than "already
2675 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2678 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2679 because an input line was too long, either on its own, or by virtue of
2680 too many continuations, the temporary file was not being removed, and the
2681 return code was incorrect.
2683 PH/31 Missing "BOOL" in function definition in filtertest.c.
2685 PH/32 Applied Sieve patches from the maintainer.
2687 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2689 PH/33 Added "verify = not_blind".
2691 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2692 Local/Makefile (with some defaults set). These are used in built scripts
2693 such as exicyclog, but they have never been used in the exim_install
2694 script (though there are many overriding facilities there). I have
2695 arranged that the exim_install script now takes note of these two
2698 PH/35 Installed configuration files for Dragonfly.
2700 PH/36 When a locally submitted message by a trusted user did not contain a
2701 From: header, and the sender address was obtained from -f or from an SMTP
2702 MAIL command, and the trusted user did not use -F to supply a sender
2703 name, $originator_name was incorrectly used when constructing a From:
2704 header. Furthermore, $originator_name was used for submission mode
2705 messages from external hosts without From: headers in a similar way,
2706 which is clearly wrong.
2708 PH/37 Added control=suppress_local_fixups.
2710 PH/38 When log_selector = +received_sender was set, and the addition of the
2711 sender made the log line's construction buffer exactly full, or one byte
2712 less than full, an overflow happened when the terminating "\n" was
2715 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2716 when the result of a list match is failure because a DNS lookup failed.
2718 PH/40 RM_COMMAND is now used in the building process.
2720 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2721 the "build-* directories that it finds.
2723 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2724 address in a domain literal was a prefix of an interface address.
2726 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2727 when verifying a sender address, unless rewrite_headers is false.
2729 PH/44 Wrote a long comment about why errors_to addresses are verified as
2730 recipients, not senders.
2732 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2733 the ratelimit ACL was added.
2735 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2737 PH/46 Added extra information about PostgreSQL errors to the error string.
2739 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2740 avoiding the need to install special zones in a real server. This is
2741 backwards compatible; if it can't find the fake resolver, it drops back.
2742 Thus, both old and new test suites can be run.
2744 TF/02 Added util/ratelimit.pl
2746 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2747 clock is set back in time.
2749 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2750 Candler <B.Candler@pobox.com>.
2752 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2753 OK for addresses that are the result of redirections.
2755 PH/48 A number of further additions for the benefit of the new test suite,
2756 including a fake gethostbyname() that interfaces to the fake DNS resolver
2759 TF/06 The fix for widen_domains has also been applied to qualify_single and
2760 search_parents which are the other dnslookup options that can cause
2763 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2766 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2768 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2769 service" if it encountered an MX record with an empty target hostname.
2770 The message is now "an MX or SRV record indicated no SMTP service".
2772 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2773 if the DNS of the sending site is misconfigured. This is quite a
2774 common situation. This change restores the behaviour of treating a
2775 helo verification defer as a failure.
2777 PH/51 If self=fail was set on a router, the bounce message did not include the
2778 actual error message.
2784 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2786 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2787 command itself ended in error, the underlying message about the transport
2788 filter timeout was being overwritten with the pipe command error. Now the
2789 underlying error message should be appended to the second error message.
2791 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2793 PH/02 Reduce the amount of output that "make" produces by default. Full output
2794 can still be requested.
2796 PH/03 The warning log line about a condition test deferring for a "warn" verb
2797 was being output only once per connection, rather than after each
2798 occurrence (because it was using the same function as for successful
2799 "warn" verbs). This seems wrong, so I have changed it.
2801 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2802 it should not have, which might have caused a crash in the right
2803 circumstances, but probably never did.
2805 PH/04 Installed a modified version of Tony Finch's patch to make submission
2806 mode fix the return path as well as the Sender: header line, and to
2807 add a /name= option so that you can make the user's friendly name appear
2810 TF/03 Added the control = fakedefer ACL modifier.
2812 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2813 Mark Lowes for thorough testing.
2815 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2817 TK/03 Merged latest SRS patch from Miles Wilton.
2819 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2820 with the definition in sysexits.h (which is #included earlier).
2821 Fortunately, Exim does not actually use EX_OK. The code used to try to
2822 preserve the sysexits.h value, by assumimg that macro definitions were
2823 scanned for macro replacements. I have been disabused of this notion,
2824 so now the code just undefines EX_OK before #including unistd.h.
2826 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2827 in the smtp transport. When a block could not be written in a single
2828 write() function, the timeout was being re-applied to each part-write.
2829 This seems wrong - if the receiver was accepting one byte at a time it
2830 would take for ever. The timeout is now adjusted when this happens. It
2831 doesn't have to be particularly precise.
2833 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2834 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2836 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2837 without a domain if the check to <postmaster@domain> fails.
2839 SC/01 Eximstats: added -xls and the ability to specify output files
2840 (patch written by Frank Heydlauf).
2842 SC/02 Eximstats: use FileHandles for outputing results.
2844 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2846 SC/04 Eximstats: fixed display of large numbers with -nvr option
2848 SC/05 Eximstats: fixed merging of reports with empty tables.
2850 SC/06 Eximstats: added the -include_original_destination flag
2852 SC/07 Eximstats: removed tabs and trailing whitespace.
2854 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2856 TK/06 MBOX spool code: Add real "From " MBOX separator line
2857 so the .eml file is really in mbox format (even though
2858 most programs do not really care). Patch from Alex Miller.
2860 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2861 The latter is generated from $received_to and is only set if the
2862 message has one envelope recipient. SA can use these headers,
2863 obviously out-of-the-box. Patch from Alex Miller.
2865 PH/08 The ${def test on a variable was returning false if the variable's
2866 value was "0", contrary to what the specification has always said!
2867 The result should be true unless the variable is empty.
2869 PH/09 The syntax error of a character other than { following "${if
2870 def:variable_name" (after optional whitespace) was not being diagnosed.
2871 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2872 accidental colon was present, for example, could give incorrect results.
2874 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2875 result is used (not including appendfile, where other changes are about
2878 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2879 This involved changing a lot of size variables from int to off_t. It
2880 should work with maildirs and everything.
2882 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2883 spamd dying while we are connected to it.
2885 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2888 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2889 function for BDB 4.3.
2891 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2893 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2894 the sources for PCRE have been re-arranged and more files are now
2897 PH/15 The code I had for printing potentially long long variables in PH/11
2898 above was not the best (it lost precision). The length of off_t variables
2899 is now inspected at build time, and an appropriate printing format (%ld
2900 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2901 to be "long long int" or "long int". This is needed for the internal
2902 formatting function string_vformat().
2904 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2905 the configuration file to be ":syslog", then the script "guesses" where
2906 the logs files are, rather than using the compiled in default. In our
2907 case the guess is not the same as the compiled default, so the script
2908 suddenly stopped working when I started to use syslog. The patch checks
2909 to see if log_file_path is "". If so, it attempts to read it from exim
2910 with no configuration file to get the compiled in version, before it
2911 falls back to the previous guessing code."
2913 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2914 implementing BATV in an Exim configuration. See NewStuff for the gory
2917 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2918 Makefile that are specific to HP-UX.
2920 PH/18 If the "use_postmaster" option was set for a recipient callout together
2921 with the "random" option, the postmaster address was used as the MAIL
2922 FROM address for the random test, but not for the subsequent recipient
2923 test. It is now used for both.
2925 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2926 patch removes a few documentation additions to RFC 3028, because the
2927 latest draft now contains them. It adds the new en;ascii-case comparator
2928 and a new error check for 8bit text in MIME parts. Comparator and
2929 require names are now matched exactly. I enabled the subaddress
2930 extension, but it is not well tested yet (read: it works for me)."
2932 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2933 rework some of the code of TK/09 above to avoid the hardwired use of
2934 "%lld" and "long long". Replaced the call to snprintf() with a call to
2937 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2938 records point to non-existent hosts", "retry timeout exceeded", and
2939 "retry time not reached for any host after a long failure period".
2941 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2942 experimental DomainKeys support:
2944 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2945 (2) On an error such as an illegally used "control", the wrong name for
2946 the control was given.
2948 These problems did NOT occur unless DomainKeys support was compiled.
2950 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2952 PH/24 Added ${if match_ip condition.
2954 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2955 databases so that it will be absolutely obvious if a crash occurs in the
2956 DB library. This is a regular occurrence (often caused by mis-matched
2959 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2960 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2961 picked up on a user's system that detects such things. There doesn't seem
2962 to be a gcc warning option for this - only an attribute that has to be
2963 put on the function's prototype. It seems that in Fedora Core 4 they have
2964 set this on a number of new functions. No doubt there will be more in due
2967 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2968 specify a transport. However, if an address that was verified by such a
2969 router was the subject of a callout, Exim crashed because it tried to
2970 read the rcpt_include_affixes from the non-existent transport. Now it
2971 just assumes that the setting of that option is false. This bug was
2972 introduced by 4.51/PH/31.
2974 PH/28 Changed -d+all to exclude +memory, because that information is very
2975 rarely of interest, but it makes the output a lot bigger. People tend to
2976 do -d+all out of habit.
2978 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2979 code in os-type was giving problems when libc.so lives in lib64, like on
2982 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2983 aren't the modern standard, and it seems that some systems' include files
2984 don't always have them. Exim was already checking for some of the newer
2985 ones like T_AAAA, and defining it itself. I've added checks for all the
2986 record types that Exim uses.
2988 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2989 not automatically generating a new one, as it is supposed to. This
2990 prevented TLS from working. If the file did exist, but contained invalid
2991 data, a new version was generated, as expected. It was only the case of a
2992 non-existent file that was broken.
2994 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2995 with a change in libdomainkeys > 0.64.
2997 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2998 from DNS. If the selector record carries the flag, it now has
2999 precedence over the domain-wide flag.
3001 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3003 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3004 the use of a port name that isn't defined in /etc/services) occurred, the
3005 message was deferred as in a normal delivery, and thus remained on the
3006 spool, instead of being failed because of the mua_wrapper setting. This
3007 is now fixed, and I tidied up some of the mua_wrapper messages at the
3010 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3011 the messages in a hash of arrays rather than using individual hashes.
3012 This is a bit cleaner and results in dramatic memory savings, albeit
3013 at a slight CPU cost.
3015 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3016 as requested by Marc Sherman.
3018 SC/10 Eximstats: added histograms for user specified patterns as requested
3021 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3023 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3024 fopen() in the content-scanning modules that did not already have it.
3030 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3031 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3033 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3035 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3037 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3038 to test Sieve filters that use "vacation".
3040 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3041 that changes the way the GnuTLS parameters are stored in the cache file.
3042 The new format can be generated externally. For backward compatibility,
3043 if the data in the cache doesn't make sense, Exim assumes it has read an
3044 old-format file, and it generates new data and writes a new file. This
3045 means that you can't go back to an older release without removing the
3048 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3049 work if there are any delivery delays because "one_time" forces the
3050 parent to be marked "delivered", so its unseen clone is never tried
3051 again. For this reason, Exim now forbids the simultaneous setting of
3054 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3055 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3056 entry if you want to know the details. The fix, however, neglected to
3057 consider the case when local delivery batching is involved. The test for
3058 "previously delivered" was not happening when checking to see if an
3059 address could be batched with a previous (undelivered) one; under
3060 certain circumstances this could lead to multiple deliveries to the same
3063 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3064 in its include files, and this causes problems building Exim.
3066 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3067 header_syntax) but Exim was just ignoring anything given after a slash.
3068 In particular, this caused confusion with an attempt to use "verify =
3069 reverse_host_lookup/defer_ok". An error is now given when options are
3070 supplied for verify items that do not have them. (Maybe reverse_host_
3071 lookup should have a defer_ok option, but that's a different point.)
3073 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3074 defined by RFC 821) to 2048, because there were problems with some AUTH
3075 commands, and RFC 1869 says the size should be increased for extended
3076 SMTP commands that take arguments.
3078 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3081 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3082 "unknown" error; now it says that the functionality isn't in the binary.
3084 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3085 an address' error message when a string expansion fails (syntax or
3086 whatever). Otherwise the password may appear in the log. Following change
3087 PH/42 below, there is no longer a chance of it appearing in a bounce
3090 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3092 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3093 of its addresses was ever tried. (Bugzilla bug #2.)
3095 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3096 the result incorrectly in the debug output. (It correctly added a newline
3097 to what was transported.)
3099 TF/01 Added $received_time.
3101 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3102 commented out examples of how to interface to a virus scanner and to
3103 SpamAssassin. Also added commented examples of av_scanner and
3104 spamd_address settings.
3106 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3107 and controls are allowed in which ACLs. There were a couple of minor
3108 errors. Some of the entries in the conditions table (which is a table of
3109 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3110 negation of where the condition IS allowed.
3112 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3114 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3115 header file does not have a version number, so I've had to invent a new
3116 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3117 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3118 radiusclient), but it was contributed by a Radius user.
3120 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3121 files or directories, for OpenSSL.
3123 PH/19 When an Exim process that is running as root has to create an Exim log
3124 file, it does so in a subprocess that runs as exim:exim so as to get the
3125 ownership right at creation (otherwise, other Exim processes might see
3126 the file with the wrong ownership). There was no test for failure of this
3127 fork() call, which would lead to the process getting stuck as it waited
3128 for a non-existent subprocess. Forks do occasionally fail when resources
3129 run out. I reviewed all the other calls to fork(); they all seem to check
3132 PH/20 When checking for unexpected SMTP input at connect time (before writing
3133 the banner), Exim was not dealing correctly with a non-positive return
3134 from the read() function. If the client had disconnected by this time,
3135 the result was a log entry for a synchronization error with an empty
3136 string after "input=" when read() returned zero. If read() returned -1
3137 (an event I could not check), uninitialized data bytes were printed.
3138 There were reports of junk text (parts of files, etc) appearing after
3141 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3143 PH/22 Added support for macro redefinition, and (re)definition in between
3144 driver and ACL definitions.
3146 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3147 forgetting to use the resulting value; it was using the unexpanded value.
3149 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3150 hadn't been configured. The fix is from Juergen Kreileder, who
3151 understands it better than I do:
3153 "Here's what I see happening with three configured cyrus_sasl
3154 authenticators configured (plain, login, cram-md5):
3156 On startup auth_cyrus_sasl_init() gets called for each of these.
3157 This means three calls to sasl_listmech() without a specified mech_list.
3158 => SASL tests which mechs of all available mechs actually work
3159 => three warnings about OTP not working
3160 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3162 With the patch, sasl_listmech() also gets called three times. But now
3163 SASL's mech_list option is set to the server_mech specified in the the
3164 authenticator. Or in other words, the answer from sasl_listmech()
3165 gets limited to just the mech you're testing for (which is different
3167 => the return list contains just 'plain' or 'login', 'cram-md5' or
3168 nothing depending on the value of ob->server_mech.
3170 I've just tested the patch: Authentication still works fine,
3171 unavailable mechs specified in the exim configuration are still
3172 caught, and the auth.log warnings about OTP are gone."
3174 PH/25 When debugging is enabled, the contents of the command line are added
3175 to the debugging output, even when log_selector=+arguments is not
3178 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3179 answer is "GNU", and only if the return is "GNU/something" is the answer
3182 PH/27 $acl_verify_message is now set immediately after the failure of a
3183 verification in an ACL, and so is available in subsequent modifiers. In
3184 particular, the message can be preserved by coding like this:
3186 warn !verify = sender
3187 set acl_m0 = $acl_verify_message
3189 Previously, $acl_verify_message was set only while expanding "message"
3190 and "log_message" when a very denied access.
3192 PH/28 Modified OS/os.c-Linux with
3194 -#ifndef OS_LOAD_AVERAGE
3195 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3197 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3198 nomenclature these days.)
3200 PH/29 Installed patch from the Sieve maintainer that adds the options
3201 sieve_useraddress and sieve_subaddress to the redirect router.
3203 PH/30 In these circumstances:
3204 . Two addresses routed to the same list of hosts;
3205 . First host does not offer TLS;
3206 . First host accepts first address;
3207 . First host gives temporary error to second address;
3208 . Second host offers TLS and a TLS session is established;
3209 . Second host accepts second address.
3210 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3211 and peerdn, if requested) that were in fact used only for the second
3214 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3215 attention to any local part prefix or suffix that was matched by the
3216 router that accepted the address. It now behaves in the same way as it
3217 does for delivery: the affixes are removed from the local part unless
3218 rcpt_include_affixes is set on the transport.
3220 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3221 timeout during the DATA phase of an incoming message.
3223 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3224 applied a patch sanctioned by the Sieve maintainer.
3226 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3227 the uid or gid is negative. A case of a negative gid caused this to be
3228 noticed. The fix allows for either to be negative.
3230 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3231 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3234 PH/36 The change PH/12 above was broken. Fixed it.
3236 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3237 the grounds that routing the same address twice would always produce the
3238 same answer. This might have been true once, but it is certainly no
3239 longer true now. Routing a child address may depend on the previous
3240 routing that produced that child. Some complicated redirection strategies
3241 went wrong when messages had multiple recipients, and made Exim's
3242 behaviour dependent on the order in which the addresses were given.
3244 I have moved the duplicate checking until after the routing is complete.
3245 Exim scans the addresses that are assigned to local and remote
3246 transports, and removes any duplicates. This means that more work will be
3247 done, as duplicates will always all be routed, but duplicates are
3248 presumably rare, so I don't expect this is of any significance.
3250 For deliveries to pipes, files, and autoreplies, the duplicate checking
3251 still happens during the routing process, since they are not going to be
3254 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3255 It corrects a timeout issue with spamd. This is Ian's comment: "The
3256 background is that sometimes spamd either never reads data from a
3257 connection it has accepted, or it never writes response data. The exiscan
3258 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3259 blindly assumes that writes won't block so it may never time out."
3261 PH/39 Allow G after quota size as well as K and M.
3263 PH/40 The value set for $authenticated_id in an authenticator may not contain
3264 binary zeroes or newlines because the value is written to log lines and
3265 to spool files. There was no check on this. Now the value is run through
3266 the string_printing() function so that such characters are converted to
3267 printable escape sequences.
3269 PH/41 $message_linecount is a new variable that contains the total number of
3270 lines in the message. Compare $body_linecount, which is the count for the
3273 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3274 bounce and delay warning messages, except in certain special cases, which
3277 (a) An SMTP error message from a remote host;
3278 (b) A message specified in a :fail: redirection;
3279 (c) A message specified in a "fail" command in a system filter;
3280 (d) A message specified in a FAIL return from the queryprogram router;
3281 (e) A message specified by the cannot_route_message router option.
3283 In these cases only, Exim does include the error details in bounce and
3284 warning messages. There are also a few cases where bland messages such
3285 as "unrouteable address" or "local delivery error" are given.
3287 PH/43 $value is now also set for the "else" part of a ${run expansion.
3289 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3290 being worked on, but at least Exim now implements the latest version to
3293 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3294 process to complete was treated as a delivery failure, a timeout while
3295 writing the message to the pipe was logged, but erroneously treated as a
3296 successful delivery. Such timeouts include transport filter timeouts. For
3297 consistency with the overall process timeout, these timeouts are now
3298 treated as errors, giving rise to delivery failures by default. However,
3299 there is now a new Boolean option for the pipe transport called
3300 timeout_defer, which, if set TRUE, converts the failures into defers for
3301 both kinds of timeout. A transport filter timeout is now identified in
3304 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3305 systems where "make" and "gmake" are different, calling "gmake" at top
3306 level broke things. I've arranged for the value of $(MAKE) to be passed
3307 from the Makefile to this script so that it can call the same version of
3311 A note about Exim versions 4.44 and 4.50
3312 ----------------------------------------
3314 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3315 changes of various kinds. As a consequence, a big documentation update was
3316 needed. This delayed the release for rather longer than seemed good, especially
3317 in the light of a couple of (minor) security issues. Therefore, the changes
3318 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3319 release. So 4.44 and 4.50 are in effect two different branches that both start
3322 I have left the 4.50 change log unchanged; it contains all the changes since
3323 4.43. The change log for 4.44 is below; many of its items are identical to
3324 those for 4.50. This seems to be the most sensible way to preserve the
3325 historical information.
3331 1. Minor wording change to the doc/README.SIEVE file.
3333 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3334 computation of the current number of files was incorrect.
3336 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3337 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3340 4. Give more explanation in the error message when the command for a transport
3341 filter fails to execute.
3343 5. There are several places where Exim runs a non-Exim command in a
3344 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3345 was being done only for the command run by the queryprogram router. It is
3346 now done for all such subprocesses. The other cases are: ${run, transport
3347 filters, and the commands run by the lmtp and pipe transports.
3349 6. Added CONFIGURE_GROUP build-time option.
3351 7. Some older OS have a limit of 256 on the maximum number of file
3352 descriptors. Exim was using setrlimit() to set 1000 as a large value
3353 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3354 systems. I've change it so that if it can't get 1000, it tries for 256.
3356 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3357 was an oversight, and furthermore, ever since the addition of extra
3358 controls (e.g. 4.43/32), the checks on when to allow different forms of
3359 "control" were broken. There should now be diagnostics for all cases when a
3360 control that does not make sense is encountered.
3362 9. Added the /retain_sender option to "control=submission".
3364 10. $recipients is now available in the predata ACL (oversight).
3366 11. Tidy the search cache before the fork to do a delivery from a message
3367 received from the command line. Otherwise the child will trigger a lookup
3368 failure and thereby defer the delivery if it tries to use (for example) a
3369 cached ldap connection that the parent has called unbind on.
3371 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3372 of $address_data from the recipient verification was clobbered by the
3373 sender verification.
3375 13. The value of address_data from a sender verification is now available in
3376 $sender_address_data in subsequent conditions in the ACL statement.
3378 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3380 15. Added a new option "connect=<time>" to callout options, to set a different
3383 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3384 was its contents. (It was OK if the option was not defined at all.)
3386 17. A "Completed" log line is now written for messages that are removed from
3387 the spool by the -Mrm option.
3389 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3390 information about exactly what failed.
3392 19. Added -dd to debug only the daemon process.
3394 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3395 handles timeouts, both on the server side and network timeouts. Renamed the
3396 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3398 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3399 It is now set to "smtps".
3401 22. $host_address is now set to the target address during the checking of
3402 ignore_target_hosts.
3404 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3405 being passed; this would have caused $sender_host_name to have been used if
3406 matching the list had actually called for a host name (not very likely,
3407 since this list is usually IP addresses). A host name is now passed as
3410 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3411 code that specifies a non-restarting handler (typically sigaction() in
3412 modern systems) in an attempt to fix a rare and obscure crash bug.
3414 25. Narrowed the window for a race in the daemon that could cause it to ignore
3415 SIGCHLD signals. This is not a major problem, because they are used only to
3416 wake it up if nothing else does.
3418 26. A malformed maildirsize file could cause Exim to calculate negative values
3419 for the mailbox size or file count. Odd effects could occur as a result.
3420 The maildirsize information is now recalculated if the size or filecount
3423 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3424 support for a long time. Removed HAVE_SYS_VFS_H.
3426 28. Installed the latest version of exipick from John Jetmore.
3428 29. In an address list, if the pattern was not a regular expression, an empty
3429 subject address (from a bounce message) matched only if the pattern was an
3430 empty string. Non-empty patterns were not even tested. This was the wrong
3431 because it is perfectly reasonable to use an empty address as part of a
3432 database query. An empty address is now tested by patterns that are
3433 lookups. However, all the other forms of pattern expect the subject to
3434 contain a local part and a domain, and therefore, for them, an empty
3435 address still always fails if the pattern is not itself empty.
3437 30. Exim went into a mad DNS loop when attempting to do a callout where the
3438 host was specified on an smtp transport, and looking it up yielded more
3439 than one IP address.
3441 31. Re-factored the code for checking spool and log partition space into a
3442 function that finds that data and another that does the check. The former
3443 is then used to implement four new variables: $spool_space, $log_space,
3444 $spool_inodes, and $log_inodes.
3446 32. The RFC2047 encoding function was originally intended for short strings
3447 such as real names; it was not keeping to the 75-character limit for
3448 encoded words that the RFC imposes. It now respects the limit, and
3449 generates multiple encoded words if necessary. To be on the safe side, I
3450 have increased the buffer size for the ${rfc2047: expansion operator from
3453 33. It is now permitted to omit both strings after an "if" condition; if the
3454 condition is true, the result is "true". As before, when the second string
3455 is omitted, a false condition yields an empty string. This makes it less
3456 cumbersome to write custom ACL and router conditions.
3458 34. Failure to deliver a bounce message always caused it to be frozen, even if
3459 there was an errors_to setting on the router. The errors_to setting is now
3462 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3463 canonical form (fully expanded) before being placed in
3464 $sender_host_address.
3466 36. The table in the code that translates DNS record types into text (T_A to
3467 "A" for instance) was missing entries for NS and CNAME. It is just possible
3468 that this could have caused confusion if both these types were looked up
3469 for the same domain, because the text type is used as part of Exim's
3470 per-process caching. But the chance of anyone hitting this buglet seems
3473 37. The dnsdb lookup has been extended in a number of ways.
3475 (1) There is a new type, "zns", which walks up the domain tree until it
3476 finds some nameserver records. It should be used with care.
3478 (2) There is a new type, "mxh", which is like "mx" except that it returns
3479 just the host names, not the priorities.
3481 (3) It is now possible to give a list of domains (or IP addresses) to be
3482 looked up. The behaviour when one of the lookups defers can be
3483 controlled by a keyword.
3485 (4) It is now possible to specify the separator character for use when
3486 multiple records are returned.
3488 38. The dnslists ACL condition has been extended: it is now possible to supply
3489 a list of IP addresses and/or domains to be looked up in a particular DNS
3492 39. Added log_selector=+queue_time_overall.
3494 40. When running the queue in the test harness, wait just a tad after forking a
3495 delivery process, to get repeatability of debugging output.
3497 41. Include certificate and key file names in error message when GnuTLS fails
3498 to set them up, because the GnuTLS error message doesn't include the name
3499 of the failing file when there is a problem reading it.
3501 42. Allow both -bf and -bF in the same test run.
3503 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3505 44. The "Exiscan patch" is now merged into the mainline Exim source.
3507 45. Sometimes the final signoff response after QUIT could fail to get
3508 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3509 < 0 before doing a fflush(). This bug looks as though it goes back to the
3510 introduction of TLS in release 3.20, but "sometimes" must have been rare
3511 because the tests only now provoked it.
3513 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3514 (this can affect the format of dates).
3516 47. exim_tidydb, when checking for the continued existence of a message for
3517 which it has found a message-specific retry record, was not finding
3518 messages that were in split spool directories. Consequently, it was
3519 deleting retry records that should have stayed in existence.
3521 48. Steve fixed some bugs in eximstats.
3523 49. The SPA authentication driver was not abandoning authentication and moving
3524 on to the next authenticator when an expansion was forced to fail,
3525 contradicting the general specification for all authenticators. Instead it
3526 was generating a temporary error. It now behaves as specified.
3528 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3529 (the order specifies the preference for clients). The order is now AES256,
3530 AES128, 3DES, ARCFOUR128.
3532 51. Small patch to Sieve code - explicitly set From: when generating an
3535 52. Exim crashed if a remote delivery caused a very long error message to be
3536 recorded - for instance if somebody sent an entire SpamAssassin report back
3537 as a large number of 550 error lines. This bug was coincidentally fixed by
3538 increasing the size of one of Exim's internal buffers (big_buffer) that
3539 happened as part of the Exiscan merge. However, to be on the safe side, I
3540 have made the code more robust (and fixed the comments that describe what
3543 53. Now that there can be additional text after "Completed" in log lines (if
3544 the queue_time_overall log selector is set), a one-byte patch to exigrep
3545 was needed to allow it to recognize "Completed" as not the last thing in
3548 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3549 patch that reportedly fixes this has been added. I am not expert enough to
3550 create a test for it. This is what the patch creator wrote:
3552 "I found a little strange behaviour of ldap code when working with
3553 Windows 2003 AD Domain, where users was placed in more than one
3554 Organization Units. When I tried to give exim partial DN, the exit code
3555 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3556 But simultaneously result of request was absolutely normal ldap result,
3557 so I produce this patch..."
3559 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3560 REFERENCE, so I have modified the code to exclude the patch when that macro
3563 55. Some experimental protocols are using DNS PTR records for new purposes. The
3564 keys for these records are domain names, not reversed IP addresses. The
3565 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3566 leaves it alone. Component reversal etc. now happens only for IP addresses.
3569 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3571 57. Double the size of the debug message buffer (to 2048) so that more of very
3572 long debug lines gets shown.
3574 58. The exicyclog utility now does better if the number of log files to keep
3575 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3577 59. Two changes related to the smtp_active_hostname option:
3579 (1) $smtp_active_hostname is now available as a variable.
3580 (2) The default for smtp_banner uses $smtp_active_hostname instead
3581 of $primary_hostname.
3583 60. The host_aton() function is supposed to be passed a string that is known
3584 to be a valid IP address. However, in the case of IPv6 addresses, it was
3585 not checking this. This is a hostage to fortune. Exim now panics and dies
3586 if the condition is not met. A case was found where this could be provoked
3587 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3588 components; fortuitously, this particular loophole had already been fixed
3589 by change 4.50/55 above.
3591 If there are any other similar loopholes, the new check in host_aton()
3592 itself should stop them being exploited. The report I received stated that
3593 data on the command line could provoke the exploit when Exim was running as
3594 exim, but did not say which command line option was involved. All I could
3595 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3596 running as the user.
3599 61. There was a buffer overflow vulnerability in the SPA authentication code
3600 (which came originally from the Samba project). I have added a test to the
3601 spa_base64_to_bits() function which I hope fixes it.
3604 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3605 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3607 63. The daemon start-up calls getloadavg() while still root for those OS that
3608 need the first call to be done as root, but it missed one case: when
3609 deliver_queue_load_max is set with deliver_drop_privilege. This is
3610 necessary for the benefit of the queue runner, because there is no re-exec
3611 when deliver_drop_privilege is set.
3613 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3614 This has been fixed.
3616 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3617 was in use, was not putting the data itself into the right store pool;
3618 consequently, it could be overwritten for a subsequent message in the same
3619 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3622 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3624 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3625 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3626 the function were treating the return as a boolean value, which happened to
3627 work because 0=false and not-0=true, but is not correct code.
3629 68. The host_aton() function was not handling scoped IPv6 addresses (those
3630 with, for example, "%eth0" on the end) correctly.
3632 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3633 negated items (that is, ~something) in unsigned ints. Some compilers
3634 apparently mutter when there is no cast.
3636 70. If an address verification called from an ACL failed, and did not produce a
3637 user-specific message (i.e. there was only a "system" message), nothing was
3638 put in $acl_verify_message. In this situation, it now puts the system
3641 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3642 session; change 4.31/43 added the unwanted input to the log line - except
3643 that it did not do this in the start of session case. It now does.
3645 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3646 This is wrong and can cause the other end to generate a synchronization
3647 error if it is another Exim or anything else that does the synchronization
3648 check. A QUIT command is no longer sent after a timeout.
3650 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3651 during host lookups.
3653 74. The defer_ok option of callout verification was not working if it was used
3654 when verifying addresses in header lines, that is, for this case:
3656 verify = header_sender/callout=defer_ok
3658 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3659 those file descriptors could be used for SMTP connections. If anything
3660 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3661 could be sent to the SMTP client, causing chaos. The daemon now opens
3662 stdin, stdout, and stderr to /dev/null when it puts itself into the
3665 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3666 log by default. The user can override this with suitable Perl magic.
3668 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3669 the log message when discard triggers, was not working for the DATA ACL or
3670 for the non-SMTP ACL.
3672 78. Error message wording change in sieve.c.
3674 79. If smtp_accept_max_per_host was set, the number of connections could be
3675 restricted to fewer than expected, because the daemon was trying to set up
3676 a new connection before checking whether the processes handling previous
3677 connections had finished. The check for completed processes is now done
3678 earlier. On busy systems, this bug wouldn't be noticed because something
3679 else would have woken the daemon, and it would have reaped the completed
3682 80. If a message was submitted locally by a user whose login name contained one
3683 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3684 It caused a spool format error. I have fixed the spool reading code. A
3685 related problem was that the "from" clause in the Received: line became
3686 illegal because of the space(s). It is now covered by ${quote_local_part.
3688 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3691 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3693 83. Patch from Sieve maintainer to fix unterminated string problem in
3694 "vacation" handling.
3696 84. Some minor changes to the Linux configuration files to help with other
3697 OS variants using glibc.
3699 85. One more patch for Sieve to update vacation handling to latest spec.
3702 ----------------------------------------------------
3703 See the note above about the 4.44 and 4.50 releases.
3704 ----------------------------------------------------
3710 1. Change 4.43/35 introduced a bug that caused file counts to be
3711 incorrectly computed when quota_filecount was set in an appendfile
3714 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3715 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3718 3. Give more explanation in the error message when the command for a transport
3719 filter fails to execute.
3721 4. There are several places where Exim runs a non-Exim command in a
3722 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3723 was being done only for the command run by the queryprogram router. It is
3724 now done for all such subprocesses. The other cases are: ${run, transport
3725 filters, and the commands run by the lmtp and pipe transports.
3727 5. Some older OS have a limit of 256 on the maximum number of file
3728 descriptors. Exim was using setrlimit() to set 1000 as a large value
3729 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3730 systems. I've change it so that if it can't get 1000, it tries for 256.
3732 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3733 was an oversight, and furthermore, ever since the addition of extra
3734 controls (e.g. 4.43/32), the checks on when to allow different forms of
3735 "control" were broken. There should now be diagnostics for all cases when a
3736 control that does not make sense is encountered.
3738 7. $recipients is now available in the predata ACL (oversight).
3740 8. Tidy the search cache before the fork to do a delivery from a message
3741 received from the command line. Otherwise the child will trigger a lookup
3742 failure and thereby defer the delivery if it tries to use (for example) a
3743 cached ldap connection that the parent has called unbind on.
3745 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3746 of $address_data from the recipient verification was clobbered by the
3747 sender verification.
3749 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3750 was its contents. (It was OK if the option was not defined at all.)
3752 11. A "Completed" log line is now written for messages that are removed from
3753 the spool by the -Mrm option.
3755 12. $host_address is now set to the target address during the checking of
3756 ignore_target_hosts.
3758 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3759 being passed; this would have caused $sender_host_name to have been used if
3760 matching the list had actually called for a host name (not very likely,
3761 since this list is usually IP addresses). A host name is now passed as
3764 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3765 code that specifies a non-restarting handler (typically sigaction() in
3766 modern systems) in an attempt to fix a rare and obscure crash bug.
3768 15. Narrowed the window for a race in the daemon that could cause it to ignore
3769 SIGCHLD signals. This is not a major problem, because they are used only to
3770 wake it up if nothing else does.
3772 16. A malformed maildirsize file could cause Exim to calculate negative values
3773 for the mailbox size or file count. Odd effects could occur as a result.
3774 The maildirsize information is now recalculated if the size or filecount
3777 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3778 support for a long time. Removed HAVE_SYS_VFS_H.
3780 18. Updated exipick to current release from John Jetmore.
3782 19. Allow an empty sender to be matched against a lookup in an address list.
3783 Previously the only cases considered were a regular expression, or an
3786 20. Exim went into a mad DNS lookup loop when doing a callout where the
3787 host was specified on the transport, if the DNS lookup yielded more than
3790 21. The RFC2047 encoding function was originally intended for short strings
3791 such as real names; it was not keeping to the 75-character limit for
3792 encoded words that the RFC imposes. It now respects the limit, and
3793 generates multiple encoded words if necessary. To be on the safe side, I
3794 have increased the buffer size for the ${rfc2047: expansion operator from
3797 22. Failure to deliver a bounce message always caused it to be frozen, even if
3798 there was an errors_to setting on the router. The errors_to setting is now
3801 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3802 canonical form (fully expanded) before being placed in
3803 $sender_host_address.
3805 24. Updated eximstats to version 1.33
3807 25. Include certificate and key file names in error message when GnuTLS fails
3808 to set them up, because the GnuTLS error message doesn't include the name
3809 of the failing file when there is a problem reading it.
3811 26. Expand error message when OpenSSL has problems setting up cert/key files.
3814 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3815 (this can affect the format of dates).
3817 28. exim_tidydb, when checking for the continued existence of a message for
3818 which it has found a message-specific retry record, was not finding
3819 messages that were in split spool directories. Consequently, it was
3820 deleting retry records that should have stayed in existence.
3822 29. eximstats updated to version 1.35
3823 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3824 1.35 - bugfix such that pie charts by volume are generated correctly
3826 30. The SPA authentication driver was not abandoning authentication and moving
3827 on to the next authenticator when an expansion was forced to fail,
3828 contradicting the general specification for all authenticators. Instead it
3829 was generating a temporary error. It now behaves as specified.
3831 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3832 (the order specifies the preference for clients). The order is now AES256,
3833 AES128, 3DES, ARCFOUR128.
3835 31. Small patch to Sieve code - explicitly set From: when generating an
3838 32. Exim crashed if a remote delivery caused a very long error message to be
3839 recorded - for instance if somebody sent an entire SpamAssassin report back
3840 as a large number of 550 error lines. This bug was coincidentally fixed by
3841 increasing the size of one of Exim's internal buffers (big_buffer) that
3842 happened as part of the Exiscan merge. However, to be on the safe side, I
3843 have made the code more robust (and fixed the comments that describe what
3846 33. Some experimental protocols are using DNS PTR records for new purposes. The
3847 keys for these records are domain names, not reversed IP addresses. The
3848 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3849 leaves it alone. Component reversal etc. now happens only for IP addresses.
3852 34. The host_aton() function is supposed to be passed a string that is known
3853 to be a valid IP address. However, in the case of IPv6 addresses, it was
3854 not checking this. This is a hostage to fortune. Exim now panics and dies
3855 if the condition is not met. A case was found where this could be provoked
3856 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3857 components; fortuitously, this particular loophole had already been fixed
3858 by change 4.50/55 or 4.44/33 above.
3860 If there are any other similar loopholes, the new check in host_aton()
3861 itself should stop them being exploited. The report I received stated that
3862 data on the command line could provoke the exploit when Exim was running as
3863 exim, but did not say which command line option was involved. All I could
3864 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3865 running as the user.
3868 35. There was a buffer overflow vulnerability in the SPA authentication code
3869 (which came originally from the Samba project). I have added a test to the
3870 spa_base64_to_bits() function which I hope fixes it.
3873 36. The daemon start-up calls getloadavg() while still root for those OS that
3874 need the first call to be done as root, but it missed one case: when
3875 deliver_queue_load_max is set with deliver_drop_privilege. This is
3876 necessary for the benefit of the queue runner, because there is no re-exec
3877 when deliver_drop_privilege is set.
3879 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3880 was in use, was not putting the data itself into the right store pool;
3881 consequently, it could be overwritten for a subsequent message in the same
3882 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3885 38. Sometimes the final signoff response after QUIT could fail to get
3886 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3887 < 0 before doing a fflush(). This bug looks as though it goes back to the
3888 introduction of TLS in release 3.20, but "sometimes" must have been rare
3889 because the tests only now provoked it.
3895 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3896 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3897 Now it returns an integer. A number of calls were still expecting a T/F
3898 return. Fortuitously, in all cases, the tests worked in OK situations,
3899 which is the norm. However, things would have gone wrong on any write
3900 failures on the smtp file descriptor. This function is used when sending
3901 messages over SMTP and also when doing verify callouts.
3903 2. When Exim is called to do synchronous delivery of a locally submitted
3904 message (the -odf or -odi options), it no longer closes stderr before doing
3907 3. Implemented the mua_wrapper option.
3909 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3911 5. Implemented the functions header_remove(), header_testname(),
3912 header_add_at_position(), and receive_remove_recipient(), and exported them
3915 6. If an ACL "warn" statement specified the addition of headers, Exim already
3916 inserted X-ACL-Warn: at the start if there was no header name. However, it
3917 was not making this test for the second and subsequent header lines if
3918 there were newlines in the string. This meant that an invalid header could
3919 be inserted if Exim was badly configured.
3921 7. Allow an ACL "warn" statement to add header lines at the start or after all
3922 the Received: headers, as well as at the end.
3924 8. Added the rcpt_4xx retry error code.
3926 9. Added postmaster_mailfrom=xxx to callout verification option.
3928 10. Added mailfrom=xxxx to the callout verification option, for verify=
3931 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3932 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3934 12. Inserted some casts to stop certain compilers warning when using pointer
3935 differences as field lengths or precisions in printf-type calls (mostly
3936 affecting debugging statements).
3938 13. Added optional readline() support for -be (dynamically loaded).
3940 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3941 same clock tick as a message's arrival, so that its received time was the
3942 same as the "first fail" time on the retry record, and that message
3943 remained on the queue past the ultimate address timeout, every queue runner
3944 would try a delivery (because it was past the ultimate address timeout) but
3945 after another failure, the ultimate address timeout, which should have then
3946 bounced the address, did not kick in. This was a "< instead of <=" error;
3947 in most cases the first failure would have been in the next clock tick
3948 after the received time, and all would be well.
3950 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3951 being recognized when the domain list was tested by the match_domain
3952 condition in an expansion string.
3954 16. Added the ${str2b64: operator.
3956 17. Exim was always calling setrlimit() to set a large limit for the number of
3957 processes, without checking whether the existing limit was already
3958 adequate. (It did check for the limit on file descriptors.) Furthermore,
3959 errors from getrlimit() and setrlimit() were being ignored. Now they are
3960 logged to the main and panic logs, but Exim does carry on, to try to do its
3961 job under whatever limits there are.
3963 18. Imported PCRE 5.0.
3965 19. Trivial typo in log message " temporarily refused connection" (the leading
3968 20. If the log selector return_path_on_delivery was set and an address was
3969 redirected to /dev/null, the delivery process crashed because it assumed
3970 that a return path would always be set for a "successful" delivery. In this
3971 case, the whole delivery is bypassed as an optimization, and therefore no
3974 21. Internal re-arrangement: the function for sending a challenge and reading
3975 a response while authentication was assuming a zero-terminated challenge
3976 string. It's now changed to take a pointer and a length, to allow for
3977 binary data in such strings.
3979 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3981 23. Exim was not respecting finduser_retries when seeking the login of the
3982 uid under which it was called; it was always trying 10 times. (The default
3983 setting of finduser_retries is zero.) Also, it was sleeping after the final
3984 failure, which is pointless.
3986 24. Implemented tls_on_connect_ports.
3988 25. Implemented acl_smtp_predata.
3990 26. If the domain in control=submission is set empty, Exim assumes that the
3991 authenticated id is a complete email address when it generates From: or
3992 Sender: header lines.
3994 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3995 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3996 chown and chgrp in /bin and hostname in /usr/bin.
3998 28. Exim was keeping the "process log" file open after each use, just as it
3999 does for the main log. This opens the possibility of it remaining open for
4000 long periods when the USR1 signal hits a daemon. Occasional processlog
4001 errors were reported, that could have been caused by this. Anyway, it seems
4002 much more sensible not to leave this file open at all, so that is what now
4005 29. The long-running daemon process does not normally write to the log once it
4006 has entered its main loop, and it closes the log before doing so. This is
4007 so that log files can straightforwardly be renamed and moved. However,
4008 there are a couple of unusual error situations where the daemon does write
4009 log entries, and I had neglected to close the log afterwards.
4011 30. The text of an SMTP error response that was received during a remote
4012 delivery was being truncated at 512 bytes. This is too short for some of
4013 the long messages that one sometimes sees. I've increased the limit to
4016 31. It is now possible to make retry rules that apply only when a message has a
4017 specific sender, in particular, an empty sender.
4019 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4020 it possible to be selective about when SMTP synchronization is enforced.
4022 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4024 32. Implemented hosts_connection_nolog.
4026 33. Added an ACL for QUIT.
4028 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4031 35. Added mailbox_size and mailbox_filecount to appendfile.
4033 36. Added control = no_multiline_responses to ACLs.
4035 37. There was a bug in the logic of the code that waits for the clock to tick
4036 in the case where the clock went backwards by a substantial amount such
4037 that the microsecond fraction of "now" was more than the microsecond
4038 fraction of "then" (but the whole seconds number was less).
4040 38. Added support for the libradius Radius client library this is found on
4041 FreeBSD (previously only the radiusclient library was supported).
4047 1. When certain lookups returned multiple values in the form name=value, the
4048 quoting of the values was not always being done properly. Specifically:
4049 (a) If the value started with a double quote, but contained no whitespace,
4051 (b) If the value contained whitespace other than a space character (i.e.
4052 tabs or newlines or carriage returns) it was not quoted.
4053 This fix has been applied to the mysql and pgsql lookups by writing a
4054 separate quoting function and calling it from the lookup code. The fix
4055 should probably also be applied to nisplus, ibase and oracle lookups, but
4056 since I cannot test any of those, I have not disturbed their existing code.
4058 2. A hit in the callout cache for a specific address caused a log line with no
4059 reason for rejecting RCPT. Now it says "Previous (cached) callout
4060 verification failure".
4062 3. There was an off-by-one bug in the queryprogram router. An over-long
4063 return line was truncated at 256 instead of 255 characters, thereby
4064 overflowing its buffer with the terminating zero. As well as fixing this, I
4065 have increased the buffer size to 1024 (and made a note to document this).
4067 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4068 when Exim is waiting for an SMTP response from a remote server, Exim
4069 restarts its select() call on the socket, thereby resetting its timeout.
4070 This is not a problem when such interrupts are rare. Somebody set up a cron
4071 job to run exiwhat every 2 minutes, which is less than the normal select()
4072 timeout (5 or 10 minutes). This meant that the select() timeout never
4073 kicked in because it was always reset. I have fixed this by comparing the
4074 time when an interrupt arrives with the time at the start of the first call
4075 to select(). If more time than the timeout has elapsed, the interrupt is
4076 treated as a timeout.
4078 5. Some internal re-factoring in preparation for the addition of Sieve
4079 extensions (by MH). In particular, the "personal" test is moved to a
4080 separate function, and given an option for scanning Cc: and Bcc: (which is
4081 not set for Exim filters).
4083 6. When Exim created an email address using the login of the caller as the
4084 local part (e.g. when creating a From: or Sender: header line), it was not
4085 quoting the local part when it contained special characters such as @.
4087 7. Installed new OpenBSD configuration files.
4089 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4090 try to make them clearer.
4092 9. Callout options, other than the timeout value, were being ignored when
4093 verifying sender addresses in header lines. For example, when using
4095 verify = header_sender/callout=no_cache
4097 the cache was (incorrectly) being used.
4099 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4100 only the Cygwin environment.
4102 11. When return_path_on_delivery was set as a log selector, if different remote
4103 addresses in the same message used different return paths and parallel
4104 remote delivery occurred, the wrong values would sometimes be logged.
4105 (Whenever a remote delivery process finished, the return path value from
4106 the most recently started remote delivery process was logged.)
4108 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4109 lines when AUTH and/or TLS are in use. This is the "received protocol"
4110 field. Exim used to use "asmtp" for authenticated SMTP, without any
4111 indication (in the protocol name) for TLS use. Now it follows the RFC and
4112 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4113 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4114 names appear in log lines as well as in Received: header lines.
4116 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4117 extensions, and comparison tests, and to fix some bugs.
4119 14. Changes to the "personal" filter test:
4121 (1) The test was buggy in that it was just doing the equivalent of
4122 "contains" tests on header lines. For example, if a user's address was
4123 anne@some.where, the "personal" test would incorrectly be true for
4125 To: susanne@some.where
4127 This test is now done by extracting each address from the header in turn,
4128 and checking the entire address. Other tests that are part of "personal"
4129 are now done using regular expressions (for example, to check local parts
4130 of addresses in From: header lines).
4132 (2) The list of non-personal local parts in From: addresses has been
4133 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4134 taken from the Sieve specification recommendations.
4136 (3) If the message contains any header line starting with "List-" it is
4137 treated as non-personal.
4139 (4) The test for "circular" in the Subject: header line has been removed
4140 because it now seems ill-conceived.
4142 15. Minor typos in src/EDITME comments corrected.
4144 16. Installed latest exipick from John Jetmore.
4146 17. If headers_add on a router specified a text string that was too long for
4147 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4148 of string_sprintf() is now avoided.
4150 18. $message_body_size was not set (it was always zero) when running the DATA
4151 ACL and the local_scan() function.
4153 19. For the "mail" command in an Exim filter, no default was being set for
4154 the once_repeat time, causing a random time value to be used if "once" was
4155 specified. (If the value happened to be <= 0, no repeat happened.) The
4156 default is now 0s, meaning "never repeat". The "vacation" command was OK
4157 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4158 (I found it when inspecting the code).
4160 20. There is now an overall timeout for performing a callout verification. It
4161 defaults to 4 times the callout timeout, which applies to individual SMTP
4162 commands during the callout. The overall timeout applies when there is more
4163 than one host that can be tried. The timeout is checked before trying the
4164 next host. This prevents very long delays if there are a large number of
4165 hosts and all are timing out (e.g. when the network connections are timing
4166 out). The value of the overall timeout can be changed by specifying an
4167 additional sub-option for "callout", called "maxwait". For example:
4169 verify = sender/callout=5s,maxwait=20s
4171 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4172 to the end before writing, but this should make it even safer).
4174 22. Exim was forgetting that it had advertised PIPELINING for the second and
4175 subsequent messages on an SMTP connection. It was also not resetting its
4176 memory on STARTTLS and an internal HELO.
4178 23. When Exim logs an SMTP synchronization error within a session, it now
4179 records whether PIPELINING has been advertised or not.
4181 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4182 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4183 rather than long int.
4185 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4187 26. Added the never_mail option to autoreply.
4193 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4194 crash if the getsockname() call failed; this can happen if a connection is
4195 closed very soon after it is established. The problem was simply in the
4196 order in which certain operations were done, causing Exim to try to write
4197 to the SMTP stream before it had set up the file descriptor. The bug has
4198 been fixed by making things happen in the correct order.
4204 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4205 before the connection was closed, thus losing the rejection response.
4207 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4208 some early Solaris releases, but causes trouble in current releases where
4209 socklen_t is defined.
4211 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4214 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4217 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4218 that releasing the top of it at the end releases what was used for sub-
4219 expansions (unless the block got too big). However, discard this block if
4220 the first thing is a variable or header, so that we can use its block when
4221 it is dynamic (useful for very large $message_headers, for example).
4223 6. Lookups now cache *every* query, not just the most recent. A new, separate
4224 store pool is used for this. It can be recovered when all lookup caches are
4225 flushed. Lookups now release memory at the end of their result strings.
4226 This has involved some general refactoring of the lookup sources.
4228 7. Some code has been added to the store_xxx() functions to reduce the amount
4229 of flapping under certain conditions.
4231 8. log_incoming_interface used to affect only the <= reception log lines. Now
4232 it causes the local interface and port to be added to several more SMTP log
4233 lines, for example "SMTP connection from", and rejection lines.
4235 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4237 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4239 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4240 could be overwritten at the end of the current message (or the start of a
4241 new message if it was set in a HELO ACL). The value is now preserved for
4242 the duration of the SMTP connection.
4244 12. If a transport had a headers_rewrite setting, and a matching header line
4245 contained an unqualified address, that address was qualified, even if it
4246 did not match any rewriting rules. The underlying bug was that the values
4247 of the flags that permit the existence of unqualified sender and recipient
4248 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4249 non-local messages, and by -bnq for local messages) were not being
4250 preserved with the message after it was received.
4252 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4253 "next input=" as part of the text comprising the host identity instead of
4254 the correct text. The code was using the same buffer for two different
4255 strings. However, depending on which order the printing function evaluated
4256 its arguments, the bug did not always show up. Under Linux, for example, my
4257 test suite worked just fine.
4259 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4260 doesn't work with some older versions of Perl. It has been changed to "my",
4261 which in any case is probably the better facility to use.
4263 15. A really picky compiler found some instances of statements for creating
4264 error messages that either had too many or two few arguments for the format
4267 16. The size of the buffer for calls to the DNS resolver has been increased
4268 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4269 for addresses that have a lot of PTR records. This alleviates a problem; it
4270 does not fully solve it.
4272 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4273 buffer now truncates the list and logs the incident, which is the same
4274 action as happens when Exim is looking up a host name and its aliases.
4275 Previously in this situation something unpredictable would happen;
4276 sometimes it was "internal error: store_reset failed".
4278 18. If a server dropped the connection unexpectedly when an Exim client was
4279 using GnuTLS and trying to read a response, the client delivery process
4280 crashed while trying to generate an error log message.
4282 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4283 string, for example:
4285 warn message = H1: something\nH2: something
4287 the text was added as a single header line from Exim's point of view
4288 though it ended up OK in the delivered message. However, searching for the
4289 second and subsequent header lines using $h_h2: did not work. This has been
4290 fixed. Similarly, if a system filter added multiple headers in this way,
4291 the routers could not see them.
4293 20. Expanded the error message when iplsearch is called with an invalid key to
4294 suggest using net-iplsearch in a host list.
4296 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4297 ACLs are no longer actually imposed (and a message to that effect is
4300 22. If a "gecos" field in a passwd entry contained escaped characters, in
4301 particular, if it contained a \" sequence, Exim got it wrong when building
4302 a From: or a Sender: header from that name. A second bug also caused
4303 incorrect handling when an unquoted " was present following a character
4304 that needed quoting.
4306 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4307 was not being matched caselessly.
4309 24. Arranged for all hyphens in the exim.8 source to be escaped with
4312 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4313 properly. Recipient callout cache records were still being keyed to include
4314 the sender, even when use_sender was set false. This led to far more
4315 callouts that were necessary. The sender is no longer included in the key
4316 when use_sender is false.
4318 26. Added "control = submission" modifier to ACLs.
4320 27. Added the ${base62d: operator to decode base 62 numbers.
4322 28. dnsdb lookups can now access SRV records.
4324 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4325 the configuration file.
4327 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4328 (-v) mode. This makes the output for a verbose queue run more intelligible.
4330 31. Added a use_postmaster feature to recipient callouts.
4332 32. Added the $body_zerocount variable, containing the number of binary zero
4333 bytes in the message body.
4335 33. The time of last modification of the "new" subdirectory is now used as the
4336 "mailbox time last read" when there is a quota error for a maildir
4339 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4341 35. Added +ignore_unknown as a special item in host lists.
4343 36. Code for decoding IPv6 addresses in host lists is now included, even if
4344 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4345 address was recognized as an IP address, but was then not correctly decoded
4346 into binary, causing unexpected and incorrect effects when compared with
4353 1. Very minor rewording of debugging text in manualroute to say "list of
4354 hosts" instead of "hostlist".
4356 2. If verify=header_syntax was set, and a header line with an unqualified
4357 address (no domain) and a large number of spaces between the end of the
4358 name and the colon was received, the reception process suffered a buffer
4359 overflow, and (when I tested it) crashed. This was caused by some obsolete
4360 code that should have been removed. The fix is to remove it!
4362 3. When running in the test harness, delay a bit after writing a bounce
4363 message to get a bit more predictability in the log output.
4365 4. Added a call to search_tidyup() just before forking a reception process. In
4366 theory, someone could use a lookup in the expansion of smtp_accept_max_
4367 per_host which, without the tidyup, could leave open a database connection.
4369 5. Added the variables $recipient_data and $sender_data which get set from a
4370 lookup success in an ACL "recipients" or "senders" condition, or a router
4371 "senders" option, similar to $domain_data and $local_part_data.
4373 6. Moved the writing of debug_print from before to after the "senders" test
4376 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4377 problems for message scanning, either using a data ACL, or using
4378 local_scan() because the Received: header was not generated till after they
4379 were called (in order to set the time as the time of reception completion).
4380 I have revised the way this works. The header is now generated after the
4381 body is received, but before the ACL or local_scan() are called. After they
4382 are run, the timestamp in the header is updated.
4388 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4389 before starting a queue runner without re-exec. This happened only when
4390 deliver_drop_privilege was set or when the Exim user was set to root. The
4391 effect of the bug was that timeouts during subsequent deliveries caused
4392 crashes instead of being properly handled. The handler is now left at its
4393 default (and expected) setting.
4395 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4396 message, again when deliver_drop_privilege is set or Exim is run as root.
4397 The bug described in (1) was not present in this case, but the tidying up
4398 of the other signals was missing. I have made the two cases consistent.
4400 3. The ignore_target_hosts setting on a manualroute router was being ignored
4401 for hosts that were looked up using the /MX notation.
4403 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4406 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4407 operated on the sender address. After changing the $sender_address to <>
4408 for the sender address verify, Exim was re-instated it as the original
4409 (before rewriting) address, but remembering that it had rewritten it, so it
4410 wasn't rewriting it again. This bug also had the effect of breaking the
4411 sender address verification caching when the sender address was rewritten.
4413 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4414 This has been changed so that if the ip literal address matches
4415 ignore_target_hosts, the router declines.
4417 7. Added expansion conditions match_domain, match_address, and match_local_
4418 part (NOT match_host).
4420 8. The placeholder for the Received: header didn't have a length field set.
4422 9. Added code to Exim itself and to exim_lock to test for a specific race
4423 condition that could lead to file corruption when using MBX delivery. The
4424 issue is with the lockfile that is created in /tmp. If this file is removed
4425 after a process has opened it but before that process has acquired a lock,
4426 there is the potential for a second process to recreate the file and also
4427 acquire a lock. This could lead to two Exim processes writing to the file
4428 at the same time. The added code performs the same test as UW imapd; it
4429 checks after acquiring the lock that its file descriptor still refers to
4430 the same named file.
4432 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4433 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4436 11. Added the smtp_active_hostname option. If used, this will typically be made
4437 to depend on the incoming interface address. Because $interface_address is
4438 not set up until the daemon has forked a reception process, error responses
4439 that can happen earlier (such as "too many connections") no longer contain
4442 12. If an expansion in a condition on a "warn" statement fails because a lookup
4443 defers, the "warn" statement is abandoned, and the next ACL statement is
4444 processed. Previously this caused the whole ACL to be aborted.
4446 13. Added the iplsearch lookup type.
4448 14. Added ident_timeout as a log selector.
4450 15. Added tls_certificate_verified as a log selector.
4452 16. Added a global option tls_require_ciphers (compare the smtp transport
4453 option of the same name). This controls incoming TLS connections.
4455 17. I finally figured out how to make tls_require_ciphers do a similar thing
4456 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4457 before starting the TLS session.
4459 18. Tabs are now shown as \t in -bP output.
4461 19. If the log selector return_path_on_delivery was set, Exim crashed when
4462 bouncing a message because it had too many Received: header lines.
4464 20. If two routers both had headers_remove settings, and the first one included
4465 a superfluous trailing colon, the final name in the first list and the
4466 first name in the second list were incorrectly joined into one item (with a
4467 colon in the middle).
4473 1. Added -C and -D options to the exinext utility, mainly to make it easier
4474 to include in the automated testing, but these could be helpful when
4475 multiple configurations are in use.
4477 2. The exinext utility was not formatting the output nicely when there was
4478 an alternate port involved in the retry record key, nor when there was a
4479 message id as well (for retries that were specific to a specific message
4480 and a specific host). It was also confused by IPv6 addresses, because of
4481 the additional colons they contain. I have fixed the IPv4 problem, and
4482 patched it up to do a reasonable job for IPv6.
4484 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4485 delivery, the log line now contains "pipelined" if PIPELINING was used.
4487 4. An SMTP transport process used to panic and die if the bind() call to set
4488 an explicit outgoing interface failed. This has been changed; it is now
4489 treated in the same way as a connect() failure.
4491 5. A reference to $sender_host_name in the part of a conditional expansion
4492 that was being skipped was still causing a DNS lookup. This no longer
4495 6. The def: expansion condition was not recognizing references to header lines
4496 that used bh_ and bheader_.
4498 7. Added the _cache feature to named lists.
4500 8. The code for checking quota_filecount in the appendfile transport was
4501 allowing one more file than it should have been.
4503 9. For compatibility with Sendmail, the command line option
4511 and sets the incoming protocol and host name (for trusted callers). The
4512 host name and its colon can be omitted when only the protocol is to be set.
4513 Note the Exim already has two private options, -pd and -ps, that refer to
4514 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4515 "s", but I don't think that's a major issue.
4517 10. A number of refactoring changes to the code, none of which should affect
4520 (a) The number of logging options was getting close to filling up the
4521 32-bit word that was used as a bit map. I have split them into two classes:
4522 those that are passed in the argument to log_write(), and those that are
4523 only ever tested independently outside of that function. These are now in
4524 separate 32-bit words, so there is plenty of room for expansion again.
4525 There is no change in the user interface or the logging behaviour.
4527 (b) When building, for example, log lines, the code previously used a
4528 macro that called string_cat() twice, in order to add two strings. This is
4529 not really sufficiently general. Furthermore, there was one instance where
4530 it was actually wrong because one of the argument was used twice, and in
4531 one call a function was used. (As it happened, calling the function twice
4532 did not affect the overall behaviour.) The macro has been replaced by a
4533 function that can join an arbitrary number of extra strings onto a growing
4536 (c) The code for expansion conditions now uses a table and a binary chop
4537 instead of a serial search (which was left over from when there were very
4538 few conditions). Also, it now recognizes conditions like "pam" even when
4539 the relevant support is not compiled in: a suitably worded error message is
4540 given if an attempt is made to use such a condition.
4542 11. Added ${time_interval:xxxxx}.
4544 12. A bug was causing one of the ddress fields not to be passed back correctly
4545 from remote delivery subprocesses. The field in question was not being
4546 subsequently used, so this caused to problems in practice.
4548 13. Added new log selectors queue_time and deliver_time.
4550 14. Might have fixed a bug in maildirsizefile handling that threw up
4551 "unexpected character" debug warnings, and recalculated the data
4552 unnecessarily. In any case, I expanded the warning message to give more
4555 15. Added the message "Restricted characters in address" to the statements in
4556 the default ACL that block characters like @ and % in local parts.
4558 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4559 Three changes have been made:
4561 (a) There was a serious bug; a negative response to MAIL caused the whole
4562 recipient domain to be cached as invalid, thereby blocking all messages
4563 to all local parts at the same domain, from all senders. This bug has
4564 been fixed. The domain is no longer cached after a negative response to
4565 MAIL if the sender used is not empty.
4567 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4570 (c) A new callout option, "use_sender" has been added for people who want
4571 the modified behaviour.
4577 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4580 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4581 indeed breaks things for older releases.
4583 3. Added additional logging to the case where there is a problem reading data
4584 from a filter that is running in a subprocess using a pipe, in order to
4585 try to track down a specific problem.
4587 4. Testing facility fudge: when running in the test harness and attempting
4588 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4589 getting "No route to host". Convert this to a timeout.
4591 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4594 6. Some OS don't have socklen_t but use size_t instead. This affects the
4595 fifth argument of getsockopt() amongst other things. This is now
4596 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4597 can be set for individual OS. I have set it for SunOS5, OSF1, and
4598 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4599 some earlier ones do not.
4601 7. Change 4.30/15 was not doing the test caselessly.
4603 8. The standard form for an IPv6 address literal was being rejected by address
4604 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4605 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4606 this, as well as the form without the "IPv6" on the front (but only when
4607 address literals are enabled, of course).
4609 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4611 10. Exim crashed if a message with an empty sender address specified by -f
4612 encountered a router with an errors_to setting. This could be provoked only
4613 by a command such as
4617 where an empty string was supplied; "<>" did not hit this bug.
4619 11. Installed PCRE release 4.5.
4621 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4622 remained set. It is now erased.
4624 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4625 times from message ids (which are base 36 rather than the normal 62).
4627 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4628 were being counted as actual protocol errors, and logged if the log
4629 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4630 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4631 and DATA following a set of rejected RCPTs do not count as protocol errors.
4632 In other words, Exim assumes they were pipelined, though this may not
4633 actually be the case. Of course, in all cases the client gets an
4634 appropriate error code.
4636 15. If a lookup fails in an ACL condition, a message about the failure may
4637 be available; it is used if testing the ACL cannot continue, because most
4638 such messages specify what the cause of the deferral is. However, some
4639 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4640 that caused an old message to be retained and used if a later statement
4641 caused a defer, replacing the real cause of the deferral.
4643 16. If an IP address had so many PTR records that the DNS lookup buffer
4644 was not large enough to hold them, Exim could crash while trying to process
4645 the truncated data. It now detects and logs this case.
4647 17. Further to 4.21/58, another change has been made: if (and only if) the
4648 first line of a message (the first header line) ends with CRLF, a bare LF
4649 in a subsequent header line has a space inserted after it, so as not to
4650 terminate the header.
4652 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4653 unnecessarily, used atoi() instead of strtol(), and didn't check the
4654 termination when getting file sizes from file names by regex.
4656 19. Completely re-implemented the support for maildirsize files, in the light
4657 of a number of problems with the previous contributed implementation
4658 (4.30/29). In particular:
4660 . If the quota is zero, the maildirsize file is maintained, but no quota is
4663 . If the maildir directory does not exist, it is created before any attempt
4664 to write a maildirsize file.
4666 . The quota value in the file is just a cache; if the quota is changed in
4667 the transport, the new value overrides.
4669 . A regular expression is available for excluding directories from the
4672 20. The autoreply transport checks the characters in options that define the
4673 message's headers; it allows continued headers, but it was checking with
4674 isspace() after an embedded newline instead of explicitly looking for a
4677 21. If all the "regular" hosts to which an address was routed had passed their
4678 expiry times, and had not reached their retry times, the address was
4679 bounced, even if fallback hosts were defined. Now Exim should go on to try
4682 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4683 equivalent code in the SMTP transport. Some hosts send humungous responses
4684 to HELO/EHLO, more than 1024 it seems.
4686 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4687 gives compiler warnings in some environments. I've now done it "properly",
4690 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4691 (because of problems with the built-in one) was declared to return uschar *
4692 instead of char *, causing compiler failure.
4694 25. Fixed a file descriptor leak when processing alias/forward files.
4696 26. Fixed a minor format string issue in dbfn.c.
4698 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4700 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4701 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4704 29. When a sender address is verified, it is cached, to save repeating the test
4705 when there is more than one recipient in a message. However, when the
4706 verification involves a callout, it is possible for different callout
4707 options to be set for different recipients. It is too complicated to keep
4708 track of this in the cache, so now Exim always runs a verification when a
4709 callout is required, relying on the callout cache for the optimization.
4710 The overhead is duplication of the address routing, but this should not be
4713 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4714 to be verified with callout=postmaster, and the main callout worked but the
4715 postmaster check failed, the verification correctly failed. However, if a
4716 subsequent RCPT command asked for sender verification *without* the
4717 postmaster check, incorrect caching caused this verification also to fail,
4720 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4721 it was not caching the DNS options (qualify_single, search_parents) that
4722 were used when the lookup failed. A subsequent lookup with different
4723 options therefore always gave the same answer, though there were cases
4724 where it should not have. (Example: a "domains = !$mx_any" option on a
4725 dnslookup router: the "domains" option is always processed without any
4726 widening, but the router might have qualify_single set.) Now Exim uses the
4727 cached value only when the same options are set.
4729 32. Added John Jetmore's "exipick" utility to the distribution.
4731 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4732 than a timeout (e.g. a certificate is required, and is not provided), an
4733 Exim server now closes the connection immediately. Previously it waited for
4734 the client to close - but if the client is SSL, it seems that they each
4735 wait for each other, leading to a delay before one of them times out.
4737 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4738 maintained 0.8.x compatibility because I don't think many are using it, and
4739 it is clearly obsolete.
4741 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4744 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4745 client certificate was expired. A simple patch fixes this, though I don't
4746 understand the full logic of why the verify callback is called multiple
4749 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4750 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4751 which causes problems with some clients (such as the Certicom SSL Plus
4752 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4753 disables the coutermeasure allowing Eudora to connect."
4755 38. Exim was not checking that a write() to a log file succeeded. This could
4756 lead to Bad Things if a log got too big, in particular if it hit a file
4757 size limit. Exim now panics and dies if it cannot write to a log file, just
4758 as it does if it cannot open a log file.
4760 39. Modified OS/Makefile-Linux so that it now contains
4762 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4764 The two -D definitions ensure that Exim is compiled with large file
4765 support, which makes it possible to handle log files that are bigger than
4768 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4769 instance) a domain was checked against a named list that involved a lookup,
4770 causing $domain_data to be set, then another domain was checked against the
4771 same list, then the first domain was re-checked, the value of $domain_data
4772 after the final check could be wrong. In particular, if the second check
4773 failed, it could be set empty. This bug probably also applied to
4776 41. The strip_trailing_dot option was not being applied to the address given
4777 with the -f command-line option.
4779 42. The code for reading a message's header from the spool was incrementing
4780 $received_count, but never initializing it. This meant that the value was
4781 incorrect (doubled) while delivering a message in the same process in which
4782 it was received. In the most common configuration of Exim, this never
4783 happens - a fresh exec is done - but it can happen when
4784 deliver_drop_privilege is set.
4786 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4787 it now includes up to 150 characters of the unexpected data in the log
4790 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4791 and building data strings. The size of both of these buffers was 10 000
4792 bytes - far larger than anybody would *ever* want, thought I. Needless to
4793 say, somebody hit the limit. I have increased the maximum line length to
4794 20 000 and the maximum data length of concatenated lines to 100 000. I have
4795 also fixed two bugs, because there was no checking on these buffers. Tsk,
4796 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4797 buffer is too small.
4799 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4800 lsearch lookups. Now it does.
4802 46. When parsing a route_list item in a manualroute router, a fixed-length
4803 buffer was used for the list of hosts. I made this 1024 bytes long,
4804 thinking that nobody would ever have a list of hosts that long. Wrong.
4805 Somebody had a whole pile of complicated expansion conditions, and the
4806 string was silently truncated, leading to an expansion error. It turns out
4807 that it is easier to change to an unlimited length (owing to other changes
4808 that have happened since this code was originally written) than to build
4809 structure for giving a limitation error. The length of the item that
4810 expands into the list of hosts is now unlimited.
4812 47. The lsearch lookup could not handle data where the length of text line was
4813 more than 4095 characters. Such lines were truncated, leading to shortened
4814 data being returned. It should now handle lines of any length.
4816 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4817 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4820 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4821 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4822 no longer happens in comment lines. A list of replacements is now placed
4823 at the head of all of the source files, except those whose only change is
4824 to replace PERL_COMMAND in the very first #! line.
4826 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4827 messages on the queue, with a bottom-up merge sort, using code contributed
4828 by Michael Haardt. This should make operations like -bp somewhat faster on
4829 large queues. It won't affect queue runners, except when queue_run_in_order
4832 51. Installed eximstats 1.31 in the distribution.
4834 52. Added support for SRV lookups to the dnslookup router.
4836 53. If an ACL referred to $message_body or $message_body_end, the value was not
4837 reset for any messages that followed in the same SMTP session.
4839 54. The store-handling optimization for building very long strings was not
4840 differentiating between the different store pools. I don't think this
4841 actually made any difference in practice, but I've tidied it.
4843 55. While running the routers to verify a sender address, $sender_address
4844 was still set to the sender address. This is wrong, because when routing to
4845 send a bounce to the sender, it would be empty. Therefore, I have changed
4846 it so that, while verifying a sender address, $sender_address is set to <>.
4847 (There is no change to what happens when verifying a recipient address.)
4849 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4850 target A or AAAA records (if not already returned) without resetting the
4851 qualify_single or search_parents options of the DNS resolver. These are
4852 inappropriate in this case because the targets of MX and SRV records must
4853 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4854 target that, when qualified, matched something in the local domain. These
4855 two options are now turned off when doing these lookups.
4857 57. It seems that at least some releases of Reiserfs (which does not have the
4858 concept of a fixed number of inodes) returns zero and not -1 for the
4859 number of available inodes. This interacted badly with check_spool_inodes,
4860 which assumed that -1 was the "no such thing" setting. What I have done is
4861 to check that the total number of inodes is greater than zero before doing
4862 the test of how many are available.
4864 58. When a "warn" ACL statement has a log_message modifier, the message is
4865 remembered, and not repeated. This is to avoid a lot of repetition when a
4866 message has many recipients that cause the same warning to be written.
4867 Howewer, Exim was preserving the list of already written lines for an
4868 entire SMTP session, which doesn't seem right. The memory is now reset if a
4869 new message is started.
4871 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4872 debugging output unless log_rewrite was also set.
4874 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4875 of (int)(handle) when we know that handle contains (void *)(-1).
4877 61. The Exim daemon panic-logs an error return when it closes the incoming
4878 connection. However "connection reset by peer" seems to be common, and
4879 isn't really an error worthy of noting specially, so that particular error
4882 62. When Exim is trying to find all the local interfaces, it used to panic and
4883 die if the ioctl to get the interface flags failed. However, it seems that
4884 on at least one OS (Solaris 9) it is possible to have an interface that is
4885 included in the list of interfaces, but for which you get a failure error
4886 for this call. This happens when the interface is not "plumbed" into a
4887 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4888 failure of the "get flags" call assumes that the interface is down.
4890 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4891 makes life easier for people who are doing arithmetic on fields extracted
4892 from dates, where you often get leading zeros that should not be
4893 interpreted as octal.
4895 64. Added qualify_domain to the redirect router, to override the global
4898 65. If a pathologically long header line contained very many addresses (the
4899 report of this problem mentioned 10 000) and each of them was rewritten,
4900 Exim could use up a very large amount of memory. (It kept on making new
4901 copies of the header line as it rewrote, and never released the old ones.)
4902 At the expense of a bit more processing, the header rewriting function has
4903 been changed so that it no longer eats memory in this way.
4905 66. The generation of the Received: header has been moved from the time that a
4906 message starts to be received, to the time that it finishes. The timestamp
4907 in the Received: header should now be very close to that of the <= log
4908 line. There are two side-effects of this change:
4910 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4911 logged header lines no longer include the local Received: line, because
4912 it has not yet been created. The same applies to a copy of the message
4913 that is returned to a non-SMTP sender when a message is rejected.
4915 (b) When a filter file is tested using -bf, no additional Received: header
4916 is added to the test message. After some thought, I decided that this
4919 This change does not affect the value of $received_for. It is still set
4920 after address rewriting, but before local_scan() is called.
4922 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4924 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4925 gave an unhelpful panic error message, and a defer error. I have managed to
4926 change this behaviour so that it now rejects any supplied certificate,
4927 which seems right, as the list of acceptable certificates is empty.
4929 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4930 gave an unhelpful defer error. I have not managed to make this reject any
4931 supplied certificates, but the error message it gives is "no certificate
4932 supplied", which is not helpful.
4934 70. exigrep's output now also includes lines that are not associated with any
4935 message, but which match the given pattern. Implemented by a patch from
4936 Martin Sluka, which also tidied up the Perl a bit.
4938 71. Recipient callout verification, like sender verification, was using <> in
4939 the MAIL FROM command. This isn't really the right thing, since the actual
4940 sender may affect whether the remote host accepts the recipient or not. I
4941 have changed it to use the actual sender in the callout; this means that
4942 the cache record is now keyed on a recipient/sender pair, not just the
4943 recipient address. There doesn't seem to be a real danger of callout loops,
4944 since a callout by the remote host to check the sender would use <>.
4945 [SEE ABOVE: changed after hitting problems.]
4947 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4948 temporary errors. However, in the case of such a code being given after
4949 the end of a data transmission (i.e. after ".") Exim was failing to write
4950 a retry record for the message. (Yes, there was some broken host that was
4951 actually sending 8xx at this point.)
4953 73. An unknown lookup type in a host list could cause Exim to panic-die when
4954 the list was checked. (An example that provoked this was putting <; in the
4955 middle of a list instead of at the start.) If this happened during a DATA
4956 ACL check, a -D file could be left lying around. This kind of configuration
4957 error no longer causes Exim to die; instead it causes a defer errror. The
4958 incident is still logged to the main and panic logs.
4960 74. Buglet left over from Exim 3 conversion. The message "too many messages
4961 in one connection" was written to the rejectlog but not the mainlog, except
4962 when address rewriting (yes!) was being logged.
4964 75. Added write_rejectlog option.
4966 76. When a system filter was run not as root (that is, when system_filter_user
4967 was set), the values of the $n variables were not being returned to the
4968 main process; thus, they were not subsequently available in the $sn
4971 77. Added +return_path_on_delivery log selector.
4973 78. A connection timeout was being treated differently from recipients deferred
4974 when testing hosts_max_try with a message that was older than the host's
4975 retry timeout. (The host should not be counted, thus allowing all hosts to
4976 be tried at least once before bouncing.) This may have been the cause of an
4977 occasionally reported bug whereby a message would remain on the queue
4978 longer than the retry timeout, but would be bounced if a delivery was
4979 forced. I say "may" because I never totally pinned down the problem;
4980 setting up timeout/retry tests is difficult. See also the next item.
4982 79. The ultimate address timeout was not being applied to errors that involved
4983 a combination of host plus message (for example, a timeout on a MAIL
4984 command). When an address resolved to a number of possible hosts, and they
4985 were not all tried for each delivery (e.g. because of hosts_max_try), a
4986 message could remain on the queue longer than the retry timeout.
4988 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4991 81. Fixed an obscure SMTP outgoing bug which required at least the following
4992 conditions: (a) there was another message waiting for the same server;
4993 (b) the server returned 5xx to all RCPT commands in the first message so
4994 that the message was not completed; (c) the server dropped the connection
4995 or gave a negative response to the RSET that Exim sends to abort the
4996 transaction. The observed case was a dropped connection after DATA that had
4997 been sent in pipelining mode. That is, the server had advertised PIPELINING
4998 but was not implementing it correctly. The effect of the bug was incorrect
4999 behaviour, such as trying another host, and this could lead to a crash.
5005 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5006 and daemon.c were passed as pointers to ints; they should have been
5007 pointers to socklen_t variables (which are typically unsigned ints).
5009 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5012 3. Fixed a really odd bug that affected only the testing scheme; patching a
5013 certain fixed string in the binary changed the value of another string that
5014 happened to be identical to the end of the original first string.
5016 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5017 name", it returns that address as the IP address. On some operating
5018 systems (e.g. Solaris), it also passes back the IP address string as the
5019 "host name". However, on others (e.g. Linux), it passes back an empty
5020 string. Exim wasn't checking for this, and was changing the host name to an
5021 empty string, assuming it had been canonicized.
5023 5. Although rare, it is permitted to have more than one PTR record for a given
5024 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5025 all the names associated with an address, because they do in Solaris.
5026 However, it seems that they do not in Linux for data that comes from the
5027 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5028 I found this out when I moved to a new Linux workstation and tried to run
5029 the Exim test suite.
5031 To get round this problem I have changed the code so that it now does its
5032 own call to the DNS to look up PTR records when searching for a host name.
5033 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5034 addresses that are only in /etc/hosts are still found.
5036 This behaviour is, however, controlled by an option called host_lookup_
5037 order, which defaults to "bydns:byaddr". If people want to use the other
5038 order, or indeed, just use one or the other means of lookup, they can
5039 specify it in this variable.
5041 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5042 some operating systems, this comes back from gethostbyaddr() as an empty
5043 string, and this is what Exim used to test for. However, it seems that in
5044 other systems, "." is yielded. Exim now tests for this case too.
5046 7. The values of check_spool_space and check_log_space are now held internally
5047 as a number of kilobytes instead of an absolute number of bytes. If a
5048 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5049 kilobyte. This means that much larger values can be stored.
5051 8. Exim monitor: an attempt to get the action menu when not actually pointing
5052 at a message produces an empty menu entitled "No message selected". This
5053 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5054 no entries in it ("Shell widget menu has zero width and/or height"). So I
5055 have added a single, blank menu entry in this case.
5057 9. Added ${quote_local_part.
5059 10. MIME decoding is now applied to the contents of Subject: header lines when
5062 11. Now that a reference to $sender_host_address automatically causes a reverse
5063 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5064 host lookup before query-style lookups in lists that might use this
5065 variable. This has therefore been abolished, and the "net-" prefix is no
5066 longer necessary for query-style lookups.
5068 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5069 have been a typo for LFLAGS, so it has been changed.
5071 13. The install script calls Exim with "-C /dev/null" in order to find the
5072 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5073 to be output. Howeve, since Exim outputs its version number before the
5074 error, it didn't break the script. It just looked ugly. I fixed this by
5075 always allowing "-C /dev/null" if the caller is root.
5077 14. Ignore overlarge ACL variable number when reading spool file - insurance
5078 against a later release with more variables having written the file.
5080 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5081 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5082 this, as well as the form without the "IPv6" on the front.
5084 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5085 OS/Makefile-Darwin file.
5087 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5088 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5090 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5091 variable. However, some OS use ints for inodes. I've added cast to long int
5092 to get rid of the compiler warning.
5094 19. I had forgotten to lock out "/../" in configuration file names when
5095 ALT_CONFIG_PREFIX was set.
5097 20. Routers used for verification do not need to specify transports. However,
5098 if such a router generated a host list, and callout was configured, Exim
5099 crashed, because it could not find a port number from the (non-existent)
5100 transport. It now assumes port 25 in this circumstance.
5102 21. Added the -t option to exigrep.
5104 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5105 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5106 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5107 them is set, LOOKUP_LSEARCH is forced.
5109 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5111 24. Added sender and host information to the "rejected by local_scan()" log
5112 line; previously there was no indication of these.
5114 25. Added .include_if_exists.
5116 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5117 Linux. It turns out that not all file systems support this. Apparently some
5118 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5119 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5120 not supported on the file descriptor, is now ignored when Exim is trying to
5121 sync a directory. This applies only to Linux.
5123 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5125 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5126 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5129 29. Added support for maildirsize files from supplied patch (modified a bit).
5131 30. The use of :fail: followed by an empty string could lead Exim to respond to
5132 sender verification failures with (e.g.):
5134 550 Verification failed for <xxx>
5135 550 Sender verify failed
5137 where the first response line was missing the '-' that indicates it is not
5138 the final line of the response.
5140 31. The loop for finding the name of the user that called Exim had a hardwired
5141 limit of 10; it now uses the value of finduser_retries, which is used for
5142 all other user lookups.
5144 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5147 33. Exim was neglecting to zero errno before one call of strtol() when
5148 expanding a string and expecting an integer value. On some systems this
5149 resulted in spurious "integer overflow" errors. Also, it was casting the
5150 result into an int without checking.
5152 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5153 did not work. The code looks as if it has *never* worked, though it appears
5154 to have been documented since at least releast 1.62. I have made it work.
5156 35. The "timeout_DNS" error in retry rules, also documented since at least
5157 1.62, also never worked. As it isn't clear exactly what this means, and
5158 clearly it isn't a major issue, I have abolished the feature by treating it
5159 as "timeout", and writing a warning to the main and panic logs.
5161 36. The display of retry rules for -brt wasn't always showing the error code
5164 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5165 timeout_connect_A, timeout_connect_MX.
5167 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5168 to the empty sender.
5170 39. The daemon was not analysing the content of -oX till after it had closed
5171 stderr and disconnected from the controlling terminal. This meant that any
5172 syntax errors were only noted on the panic log, and the return code from
5173 the command was 0. By re-arranging the code a little, I've made the
5174 decoding happen first, so such errors now appear on stderr, and the return
5175 code is 1. However, the actual setting up of the sockets still happens in
5176 the disconnected process, so errors there are still only recorded on the
5179 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5180 connections (as happens on some IP stacks) was logged at start up time as
5181 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5182 it from "IPv6 and IPv4", which means that two separate sockets are being
5185 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5186 says whether AF_INET or AF_INET6 was passed as an argument.
5188 42. Exiwhat output was messed up when time zones were included in log
5191 43. Exiwhat now gives more information about the daemon's listening ports,
5192 and whether -tls-on-connect was used.
5194 44. The "port" option of the smtp transport is now expanded.
5196 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5197 silently ignored. Now an error message is written to the main and panic
5200 46. There's a new ACL modifier called "logwrite" which writes to a log file
5201 as soon as it is encountered.
5203 47. Added $local_user_uid and $local_user_gid at routing time.
5205 48. Exim crashed when trying to verify a sender address that was being
5208 49. Exim was recognizing only a space character after ".include". It now also
5209 recognizes a tab character.
5211 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5212 extracting the relevant information from the specification. The man page no
5213 longer contains scrambled data for the -d option, and I've added a section
5214 at the front about calling Exim under different names.
5216 51. Added "extra_headers" argument to the "mail" command in filter files.
5218 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5221 53. Installed eximstats 1.29.
5223 54. Added transport_filter_timeout as a generic transport option.
5225 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5226 Cc: header lines. This was required by RFC 822, but it not required by RFC
5229 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5230 incoming messages that did not have them. Now it does so only if the
5231 message originates locally, that is, if there is no associated remote host
5232 address. When Resent- header lines are present, this applies to the Resent-
5233 lines rather than the non-Resent- lines.
5235 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5236 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5238 58. Messages for configuration errors now include the name of the main
5239 configuration file - useful now that there may be more than one file in a
5240 list (.included file names were always shown).
5242 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5243 for those rare installations that do not start the daemon as root or run it
5244 setuid root. I've cut out the call to initgroups() if the daemon is not
5247 60. The Exim user and group can now be bound into the binary as text strings
5248 that are looked up at the start of Exim's processing.
5250 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5252 62. Added $mailstore_basename variable.
5254 63. Installed patch to sieve.c from Michael Haardt.
5256 64. When Exim failed to open the panic log after failing to open the main log,
5257 the original message it was trying to log was written to stderr and debug
5258 output, but if they were not available (the usual case in production), it
5259 was lost. Now it is written to syslog before the two lines that record the
5260 failures to open the logs.
5262 65. Users' Exim filters run in subprocesses under the user's uid. It is
5263 possible for a "deliver" command or an alias in a "personal" command to
5264 provoke an address rewrite. If logging of address rewriting is configured,
5265 this fails because the process is not running as root or exim. There may be
5266 a better way of dealing with this, but for the moment (because 4.30 needs
5267 to be released), I have disabled address rewrite logging when running a
5268 filter in a non-root, non-exim process.
5274 1. The buildconfig auxiliary program wasn't quoting the value set for
5275 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5276 not defined. This bug was masked in 4.22 by the effect that was fixed in
5279 2. Some messages that were rejected after a message id was allocated were
5280 shown as "incomplete" by exigrep. It no longer does this for messages that
5281 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5283 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5284 have allow_domain_literals set, the ID did not get logged in the <= line.
5285 Domain literals are now always recognized in Message-ID: header lines.
5287 4. The first argument for a ${extract expansion item is the key name or field
5288 number. Leading and trailing spaces in this item were not being ignored,
5289 causing some misleading effects.
5291 5. When deliver_drop_privilege was set, single queue runner processes started
5292 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5293 same command in the process it spins off) were not dropping privilege.
5295 6. When the daemon running as "exim" started a queue runner, it always
5296 re-executed Exim in the spun-off process. This is a waste of effort when
5297 deliver_drop_privilege is set. The new process now just calls the
5298 queue-runner function directly.
5304 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5307 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5308 ignored. Though the use of -oP was forcing the writing of a pid file, it
5309 was always written to the default place.
5311 3. If the message "no IP address found for host xxxx" is generated during
5312 incoming verification, it is now followed by identification of the incoming
5313 connection (so you can more easily find what provoked it).
5315 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5317 5. Added some features to "harden" Exim a bit more against certain attacks:
5319 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5320 be put in Local/Makefile. This is like the never_users runtime option,
5321 but it cannot be overridden. The default setting is "root".
5323 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5324 prefix string with which any file named in a -C command line option
5327 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5328 is retained for -C and -D only if the caller of Exim is root. Without
5329 it, the exim user may also use -C and -D and retain privilege.
5331 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5332 command line option is disabled.
5334 6. Macro names set by the -D option must start with an upper case letter, just
5335 like macro names defined in the configuration file.
5337 7. Added "dereference=" facility to LDAP.
5339 8. Two instances of the typo "uknown" in the source files are fixed.
5341 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5342 the Configure-Makefile script screwed up while processing it.
5344 10. Incorporated PCRE 4.4.
5346 11. The SMTP synchronization check was not operating right at the start of an
5347 SMTP session. For example, it could not catch a HELO sent before the client
5348 waited for the greeting. There is now a check for outstanding input at the
5349 point when the greeting is written. Because of the duplex, asynchronous
5350 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5351 way, but not yet received, when the check is performed.
5353 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5354 on TCP/IP sockets, because this apparently causes some broken clients to
5357 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5358 unchanged) from the Cygwin maintainer.
5360 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5361 was supported instead of testing for mbx. Effectively a typo.
5363 15. The spa authenticator server code was not checking that the input it
5364 received was valid base64.
5366 16. The debug output line for the "set" modifier in ACLs was not showing the
5367 name of the variable that was being set.
5369 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5371 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5372 lookup on its own. Something else was needed to trigger the lookup. For
5373 example, a match in host_lookup or the need for a host name in a host list.
5374 Now, if $sender_host_name is referenced and the host name has not yet been
5375 looked up, a lookup is performed. If the lookup fails, the variable remains
5376 empty, and $host_lookup_failed is set to "1".
5378 19. Added "eqi" as a case-independent comparison operator.
5380 20. The saslauthd authentication condition could segfault if neither service
5381 nor realm was specified.
5383 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5384 error message that was logged was misleading, and incoming SMTP
5385 connections were dropped. The message is now more accurate, and temporary
5386 errors are given to SMTP connections.
5388 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5389 (except RSET) with a 421 error, until QUIT is received. However, it was
5390 failing to send a response to QUIT.
5392 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5393 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5394 ACL is now run after the helo_try_verify_hosts code.
5396 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5397 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5398 case-independent, so other case variants are also recognized). Apparently
5399 some systems use these upper case variants.
5401 25. If more than two messages were waiting for the same host, and a transport
5402 filter was specified for the transport, Exim sent two messages over the
5403 same TCP/IP connection, and then failed with "socket operation on non-
5404 socket" when it tried to send the third.
5406 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5408 27. The extern definition of crypt16() in expand.c was not being excluded when
5409 the OS had its own crypt16() function.
5411 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5412 as a preferred synonym for return_size_limit, both as an option and as an
5415 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5417 30. Changed the default configuration ACL to relax the local part checking rule
5418 for addresses that are not in any local domains. For these addresses,
5419 slashes and pipe symbols are allowed within local parts, but the sequence
5420 /../ is explicitly forbidden.
5422 31. SPA server authentication was not clearing the challenge buffer before
5425 32. log_message in a "warn" ACL statement was writing to the reject log as
5426 well as to the main log, which contradicts the documentation and doesn't
5427 seem right (because no rejection is happening). So I have stopped it.
5429 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5430 However, I am unable to do any testing of this.
5432 34. Fixed an infelicity in the appendfile transport. When checking directories
5433 for a mailbox, to see if any needed to be created, it was accidentally
5434 using path names with one or more superfluous leading slashes; tracing
5435 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5437 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5438 contents are added to the log line that is written for every discarded
5439 recipient. (Previously a log_message setting was ignored.)
5441 36. The ${quote: operator now quotes the string if it is empty.
5443 37. The install script runs exim in order to find its version number. If for
5444 some reason other than non-existence or emptiness, which it checks, it
5445 could not run './exim', it was installing it with an empty version number,
5446 i.e. as "exim-". This error state is now caught, and the installation is
5449 38. An argument was missing from the function that creates an error message
5450 when Exim fails to connect to the socket for saslauthd authentication.
5451 This could cause Exim to crash, or give a corrupted message.
5453 39. Added isip, isip4, and isip6 to ${if conditions.
5455 40. The ACL variables $acl_xx are now saved with the message, and can be
5456 accessed later in routers, transports, and filters.
5458 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5459 strings in the file are not string-expanded.
5461 42. If a MAIL command specified a SIZE value that was too large to fit into an
5462 int variable, the check against message_size_limit failed. Such values are
5463 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5464 day this will have to be increased, but I don't think I want to be around
5465 when emails are that large.
5472 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5473 iconv() is not standard in FreeBSD.
5475 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5476 IPv6 enabled. The observed symptom was a segmentation fault on return from
5477 the function os_common_find_running_interfaces() in src/os.c.
5479 3. In the check_special_case() function in daemon.c I had used "errno" as an
5480 argument name, which causes warnings on some systems. This was basically a
5481 typo, since it was named "eno" in the comments!
5483 4. The code that waits for the clock to tick (at a resolution of some fraction
5484 of a second) so as to ensure message-id uniqueness was always waiting for
5485 at least one whole tick, when it could have waited for less. [This is
5486 almost certainly not relevant at current processor speeds, where it is
5487 unlikely to ever wait at all. But we try to future-proof.]
5489 5. The function that sleeps for a time interval that includes fractions of a
5490 second contained a race. It did not block SIGALRM between setting the
5491 timer, and suspending (a couple of lines later). If the interval was short
5492 and the sigsuspend() was delayed until after it had expired, the suspension
5493 never ended. On busy systems this could lead to processes getting stuck for
5496 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5497 process, before it forks any delivery processes. The open lookup caching
5498 mechanism meant that the open file or database connection was passed into
5499 the delivery process. The problem was that delivery processes always tidy
5500 up cached lookup data. This could cause a problem for the next delivery
5501 process started by the queue runner, because the external queue runner
5502 process does not know about the closure. So the next delivery process
5503 still has data in the lookup cache. In the case of a file lookup, there was
5504 no problem because closing a file descriptor in a subprocess doesn't affect
5505 the parent. However, if the lookup was caching a connection to a database,
5506 the connection was closed, and the second delivery process was likely to
5507 see errors such as "PGSQL: query failed: server closed the connection
5508 unexpectedly". The problem has been fixed by closing all cached lookups
5509 in a queue runner before running a delivery process.
5511 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5512 seem to have the "const" qualifier which it has on other OS. I've
5515 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5516 *of the same type* (client or server) with the same public name that an
5517 error should be diagnosed.
5519 9. When Exim looked up a host name for an IP address, but failed to find the
5520 original IP address when looking up the host name (a safety check), it
5521 output the message "<ip address> does not match any IP for NULL", which was
5522 confusing, to say the least. The bug was that the host name should have
5523 appeared instead of "NULL".
5525 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5526 user that is built into the binary, and the -C or -D options is used, root
5527 privilege is dropped before the configuration file is read. In addition,
5528 logging is switched to stderr instead of the normal log files. If the
5529 configuration then re-defines the Exim user, the unprivileged environment
5530 is probably not what is expected, so Exim logs a panic warning message (but
5533 However, if deliver_drop_privilege is set, the unprivileged state may well
5534 be exactly what is intended, so the warning has been cut out in that case,
5535 and Exim is allowed to try to write to its normal log files.
5541 1. smtp_return_error_details was not giving details for temporary sender
5542 or receiver verification errors.
5544 2. Diagnose a configuration error if two authenticators have the same public
5547 3. Exim used not to create the message log file for a message until the first
5548 delivery attempt. This could be confusing when incoming messages were held
5549 for policy or load reasons. The message log file is now created at the time
5550 the message is received, and an initial "Received" line is written to it.
5552 4. The automatically generated man page for command line options had a minor
5553 bug that caused no ill effects; however, a more serious problem was that
5554 the procedure for building the man page automatically didn't always
5555 operate. Consequently, release 4.20 contains an out-of-date version. This
5556 shouldn't happen again.
5558 5. When building Exim with embedded Perl support, the script that builds the
5559 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5560 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5562 6. The freeze_tell option was not being used for messages that were frozen on
5563 arrival, either by an ACL or by local_scan().
5565 7. Added the smtp_incomplete_transaction log selector.
5567 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5568 was accepting AUTH without a new EHLO.
5570 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5571 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5574 10. Logging of TCP/IP connections (when configured) now happens in the main
5575 daemon process instead of the child process, so that the TCP/IP connection
5576 count is more accurate (but it can never be perfect).
5578 11. The use of "drop" in a nested ACL was not being handled correctly in the
5579 outer ACL. Now, if condition failure induced by the nested "drop" causes
5580 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5581 or "require"), the connection is dropped.
5583 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5584 that yield "discard" can now be used with an "accept" or a "discard" verb,
5585 but an error is generated for any others (because I can't see a useful way
5586 to define what should happen).
5588 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5589 are now processed in the same way as lines in the Exim configuration file.
5590 In particular, continuation lines are supported.
5592 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5594 15. Added -ti meaning -t -i.
5596 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5597 domains, and warn by logging if others are found.
5599 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5600 returned from GIFCONF when Exim is trying to find the list of interfaces on
5601 a host. The code in os.c has been modified to copy each ifreq to an aligned
5602 structure in all cases.
5604 Also, in some cases, the returned ifreq's were being copied to a 'struct
5605 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5606 means the last couple of bytes of an IPv6 address could be chopped if the
5607 ifreq contained only a normal sockaddr (14 bytes storage).
5609 18. Named domain lists were not supported in the hosts_treat_as_local option.
5610 An entry such as +xxxx was not recognized, and was treated as a literal
5613 19. Ensure that header lines added by a DATA ACL are included in the reject log
5614 if the ACL subsequently rejects the message.
5616 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5617 MD5 (which is deprecated).
5619 21. When testing a filter file using -bf, Exim was writing a message when it
5620 took the sender from a "From " line in the message, but it was not doing so
5621 when it took $return_path from a Return-Path: header line. It now does.
5623 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5624 with a valid header line field name (a series of printing characters
5625 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5627 23. Changed "disc" in the source to "disk" to conform to the documentation and
5628 the book and for uniformity.
5630 24. Ignore Sendmail's -Ooption=value command line item.
5632 25. When execve() failed while trying to run a command in a pipe transport,
5633 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5634 could be confused with a return value of 69 from the command itself. This
5635 has been changed to 127, the value the shell returns if it is asked to run
5636 a non-existent command. The wording for the related log line suggests a
5637 non-existent command as the problem.
5639 26. If received_header_text expands to an empty string, do not add a Received:
5640 header line to the message. (Well, it adds a token one on the spool, but
5641 marks it "old" so that it doesn't get used or transmitted.)
5643 27. Installed eximstats 1.28 (addition of -nt option).
5645 28. There was no check for failure on the call to getsockname() in the daemon
5646 code. This can fail if there is a shortage of resources on the system, with
5647 ENOMEM, for example. A temporary error is now given on failure.
5649 29. Contrary to the C standard, it seems that in some environments, the
5650 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5651 program. Exim now does this explicitly; it affects the formatting of
5652 timestamps using strftime().
5654 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5655 complaints. I've now initialized all the variables, to avoid this.
5657 32. Header lines added by a system filter were not being "seen" during
5658 transport-time rewrites.
5660 33. The info_callback() function passed to OpenSSL is set up with type void
5661 (*)(SSL *, int, int), as described somewhere. However, when calling the
5662 function (actually a macro) that sets it up, the type void(*)() is
5663 expected. I've put in a cast to prevent warnings from picky compilers.
5665 34. If a DNS black list lookup found a CNAME record, but there were no A
5666 records associated with the domain it pointed at, Exim crashed.
5668 35. If a DNS black list lookup returned more than one A record, Exim ignored
5669 all but the first. It now scans all returned addresses if a particular IP
5670 value is being sought. In this situation, the contents of the
5671 $dnslist_value variable are a list of all the addresses, separated by a
5674 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5675 to accept a host name and all its aliases if the forward lookup for any of
5676 them yielded the IP address of the incoming connection. Now it accepts only
5677 those names whose forward lookup yields the correct IP address. Any other
5678 names are discarded. This closes a loophole whereby a rogue DNS
5679 administrator could create reverse DNS records to break through a
5680 wildcarded host restriction in an ACL.
5682 37. If a user filter or a system filter that ran in a subprocess used any of
5683 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5684 the wrong values were passed to the pipe command ($thisaddress had the
5685 value of $0, $0 had the value of $1, etc). This bug was introduced by
5686 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5688 38. Improved the line breaking for long SMTP error messages from ACLs.
5689 Previously, if there was no break point between 40 and 75 characters, Exim
5690 left the rest of the message alone. Two changes have been made: (a) I've
5691 reduced the minimum length to 35 characters; (b) if it can't find a break
5692 point between 35 and 75 characters, it looks ahead and uses the first one
5693 that it finds. This may give the occasional overlong line, but at least the
5694 remaining text gets split now.
5696 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5697 file descriptors might be low, and that setting 1000 would always raise it.
5698 It turns out that in some environments, the limit is already over 1000 and
5699 that lowering it causes trouble. So now Exim takes care not to decrease it.
5701 40. When delivering a message, the value of $return_path is set to $sender_
5702 address at the start of routing (routers may change the value). By an
5703 oversight, this default was not being set up when an address was tested by
5704 -bt or -bv, which affected the outcome if any router or filter referred to
5707 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5708 log only when "message" or "log_message" are set. However, if one of the
5709 conditions was an address verification, or a call to a nested ACL, the
5710 messages generated by the underlying test were being passed through. This
5711 no longer happens. The underlying message is available in $acl_verify_
5712 message for both "message" and "log_message" expansions, so it can be
5713 passed through if needed.
5715 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5716 new expansion $bh_ to give the encoded byte string without charset
5717 translation. Translation happens only if iconv() is available; HAVE_ICONV
5718 indicates this at build time. HEADERS_CHARSET gives the charset to
5719 translate to; headers_charset can change it in the configuration, and
5720 "headers charset" can change it in an individual filter file.
5722 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5723 that creates RFC 2047 encoded "words" labels them as that charset instead
5724 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5725 expansion operator; (ii) when Exim creates a From: line for a local
5726 message; (iii) when a header line is rewritten to include a "phrase" part.
5728 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5729 buggy, causing it to skip the first lines of messages whose message ID
5730 ended in 'D'. This would not have bitten before Exim release 4.14, because
5731 message IDs were unlikely to end in 'D' before then. The effect was to have
5732 incorrect size information for certain domains.
5734 45. #include "config.h" was missing at the start of the crypt16.c module. This
5735 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5738 46. If there was a timeout during a "random" callout check, Exim treated it as
5739 a failure of the random address, and carried on sending RSET and the real
5740 address. If the delay was just some slowness somewhere, the response to the
5741 original RCPT would be taken as a response to RSET and so on, causing
5742 mayhem of various kinds.
5744 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5745 when I implemented it. It didn't allow for the fact that some option values
5746 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5747 the right test for positive values.
5749 48. Domain names in DNS records are case-independent. Exim always looks them up
5750 in lower case. Some resolvers return domain names in exactly the case they
5751 appear in the zone file, that is, they may contain uppercase letters. Not
5752 all resolvers do this - some return always lower case. Exim was treating a
5753 change of case by a resolver as a change of domain, similar to a widening
5754 of a domain abbreviation. This triggered its re-routing code and so it was
5755 trying to route what was effectively the same domain again. This normally
5756 caused routing to fail (because the router wouldn't handle the domain
5757 twice). Now Exim checks for this case specially, and just changes the
5758 casing of the domain that it ultimately uses when it transmits the message
5761 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5764 50. If a filter generated a file delivery with a non-absolute name (possible if
5765 no home directory exists for the router), the forbid_file option was not
5768 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5769 the existing equality matching.
5771 52. Exim was using ints instead of ino_t variables in some places where it was
5772 dealing with inode numbers.
5774 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5775 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5776 called TMPDIR, and if it finds it is different, it changes its value.
5778 54. The smtp_printf() function is now made available to local_scan() so
5779 additional output lines can be written before returning. There is also an
5780 smtp_fflush() function to enable the detection of a dropped connection.
5781 The variables smtp_input and smtp_batched_input are exported to
5784 55. Changed the default runtime configuration: the message "Unknown user"
5785 has been removed from the ACL, and instead placed on the localuser router,
5786 using the cannot_route_message feature. This means that any verification
5787 failures that generate their own messages won't get overridden. Similarly,
5788 the "Unrouteable address" message that was in the ACL for unverifiable
5789 relay addresses has also been removed.
5791 56. Added hosts_avoid_esmtp to the smtp transport.
5793 57. The exicyclog script was not checking for the esoteric option
5794 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5795 will work only if exicyclog is run under the appropriate euid.
5797 58. Following a discussion on the list, the rules by which Exim recognises line
5798 endings on incoming messages have been changed. The -dropcr and drop_cr
5799 options are now no-ops, retained only for backwards compatibility. The
5800 following line terminators are recognized: LF CRLF CR. However, special
5801 processing applies to CR:
5803 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5804 nor a local message in the state where . is a terminator.
5806 (ii) If a bare CR is encountered in a header line, an extra space is added
5807 after the line terminator so as not to end the header. The reasoning
5808 behind this is that bare CRs in header lines are most likely either
5809 to be mistakes, or people trying to play silly games.
5811 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5812 was being incorrectly given as 18 bytes larger than it should have been.
5813 This is a VOB (very old bug).
5815 60. This may never have affected anything current, but just in case it has:
5816 When the local host is found other than at the start of a list of hosts,
5817 the local host, those with the same MX, and any that follow, are discarded.
5818 When the list in question was part of a longer list of hosts, the following
5819 hosts (not currently being processed) were also being discarded. This no
5820 longer happens. I'm not sure if this situation could ever has previously
5823 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5826 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5827 header. This is something that is recommended in a new Internet Draft, and
5828 is something that is documented as being done by Sendmail. There are two
5829 possible values. For messages generated by the autoreply transport, Exim
5832 Auto-Submitted: auto-replied
5834 whereas for all other generated messages (e.g. bounces) it adds
5836 Auto-Submitted: auto-generated
5838 63. The "personal" condition in filters now includes a test for the
5839 Auto-Submitted: header. If it contains the string "auto-" the message it
5840 not considered personal.
5842 64. Added rcpt_include_affixes as a generic transport option.
5844 65. Added queue_only_override (default true).
5846 66. Added the syslog_duplication option.
5848 67. If what should have been the first header line of a message consisted of
5849 a space followed by a colon, Exim was mis-interpreting it as a header line.
5850 It isn't of course - it is syntactically invalid and should therefore be
5851 treated as the start of the message body. The misbehaviour could have
5852 caused a number of strange effects, including loss of data in subsequent
5853 header lines, and spool format errors.
5855 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5856 client host had authenticated. This control can now be exercised by an ACL
5857 for more flexibility.
5859 69. By default, callouts do not happen when testing with -bh. There is now a
5860 variant, -bhc, which does actually run the callout code, including
5861 consulting and updating the callout cache.
5863 70. Added support for saslauthd authentication, courtesy of Alexander
5866 71. If statvfs() failed on the spool or log directories while checking their
5867 size for availability, Exim confusingly gave the error "space shortage".
5868 Furthermore, in debugging mode it crashed with a floating point exception.
5869 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5870 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5871 serious problem, Exim now writes to the main and panic logs when this
5872 happens, with details of the failure. It then refuses to accept the
5873 incoming message, giving the message "spool directory problem" or "log
5874 directory problem" with a 421 code for SMTP messages.
5876 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5877 0, 1, and 2 exist, because some OS complain for execs without them (see
5878 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5879 descriptors. However, the code omitted to check that the open succeeded,
5880 causing mysterious errors if for some reason the permissions on /dev/null
5881 got screwed. Now Exim writes a message to the main and panic logs, and
5882 bombs out if it can't open /dev/null.
5884 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5885 interact so that it is all more flexible. It is supposed to remain
5886 backwards compatible. Also added extra_local_interfaces.
5888 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5889 to bomb out with an assertion failure - to the client this appears as a
5890 connection drop. This problem occurs in the part of the code that was taken
5891 from the Samba project. Fortunately, the assertion is in a very simple
5892 function, so I have fixed this by reproducing the function inline in the
5893 one place where it is called, and arranging for authentication to fail
5894 instead of killing the process with assert().
5896 75. The SPA client code was not working when the server requested OEM rather
5897 than Unicode encoding.
5899 76. Added code to make require_files with a specific uid setting more usable in
5900 the case where statting the file as root fails - usually a non-root-mounted
5901 NFS file system. When this happens and the failure is EACCES, Exim now
5902 forks a subprocess and does the per-uid checking as the relevant uid.
5904 77. Added process_log_path.
5906 78. If log_file_path was not explicitly set, a setting of check_log_space or
5907 check_log_inodes was ignored.
5909 79. If a space check for the spool or log partitions fails, the incident is now
5910 logged. Of course, in the latter case the data may get lost...
5912 80. Added the %p formatting code to string_format() so that it can be used to
5913 print addresses in debug_print(). Adjusted all the address printing in the
5914 debugging in store.c to use %p rather than %d.
5916 81. There was a concern that a line of code in smtp_in.c could overflow a
5917 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5918 initially expressed, the concern was not well-founded, because trailing
5919 spaces are removed early. However, if the trailing spaces were followed by
5920 a NULL, they did not get removed, so the overflow was possible. Two fixes
5923 (a) I re-wrote the offending code in a cleaner fashion.
5924 (b) If an incoming SMTP command contains a NULL character, it is rejected
5927 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5928 runs initgroups(), so that if the Exim user is in any additional groups,
5929 they will be used during message reception.
5935 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob