1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
24 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
26 PP/02 Make -n do something, by making it not do something.
27 When combined with -bP, the name of an option is not output.
29 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
32 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
33 $sender_host_name and config options to manage this, and basic check
36 PP/05 DSCP support for outbound connections and control modifier for inbound.
38 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
39 (Only plugin which currently uses this is kerberos4, which nobody should
40 be using, but we should make it available and other future plugins might
41 conceivably use it, even though it would break NAT; stuff *should* be
42 using channel bindings instead).
44 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
45 name; added for Sendmail compatibility; requires admin caller.
46 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
47 just ignore it); requires trusted caller.
48 Also parse but ignore: -Ac -Am -X<logfile>
51 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
53 TL/02 Add +smtp_confirmation as a default logging option.
55 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
56 Patch by Magnus Holmgren from 2007-02-20.
58 TL/04 Bugzilla 1281 - Spec typo.
59 Bugzilla 1283 - Spec typo.
60 Bugzilla 1290 - Spec grammar fixes.
62 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
64 TL/06 Add Experimental DMARC support using libopendmarc libraries.
66 TL/07 Fix an out of order global option causing a segfault. Reported to dev
67 mailing list by by Dmitry Isaikin.
69 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
71 JH/02 Support "G" suffix to numbers in ${if comparisons.
73 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
75 NM/01 Bugzilla 1197 - Spec typo
76 Bugzilla 1196 - Spec examples corrections
78 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
80 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
81 gnutls_enable_pkcs11, but renamed to more accurately indicate its
84 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
85 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
87 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
88 "acl {{name}{arg}...}", and optional args on acl condition
91 JH/05 Permit multiple router/transport headers_add/remove lines.
93 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
95 JH/07 Avoid using a waiting database for a single-message-only transport.
96 Performance patch from Paul Fisher. Bugzilla 1262.
98 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
101 JH/09 Add $headers_added variable, with content from use of ACL modifier
102 add_header (but not yet added to the message). Bugzilla 199.
104 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
105 Pulled from Bugzilla 817 by Wolfgang Breyha.
107 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
109 (nb: this is the same fix as in Exim 4.80.1)
111 JH/11 Add A= logging on delivery lines, and a client_set_id option on
114 JH/12 Add optional authenticated_sender logging to A= and a log_selector
117 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
119 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
120 advertise SMTP AUTH mechanism to us, instead of a generic
121 protocol violation error. Also, make Exim more robust to bad
122 data from the Dovecot auth socket.
124 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
126 When a queue runner is handling a message, Exim first routes the
127 recipient addresses, during which it prunes them based on the retry
128 hints database. After that it attempts to deliver the message to
129 any remaining recipients. It then updates the hints database using
132 So if a recipient address works intermittently, it can get repeatedly
133 deferred at routing time. The retry hints record remains fresh so the
134 address never reaches the final cutoff time.
136 This is a fairly common occurrence when a user is bumping up against
137 their storage quota. Exim had some logic in its local delivery code
138 to deal with this. However it did not apply to per-recipient defers
139 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
141 This change adds a proper retry rule check during routing so that the
142 final cutoff time is checked against the message's age. We only do
143 this check if there is an address retry record and there is not a
144 domain retry record; this implies that previous attempts to handle
145 the address had the retry_use_local_parts option turned on. We use
146 this as an approximation for the destination being like a local
147 delivery, as in LMTP.
149 I suspect this new check makes the old local delivery cutoff check
150 redundant, but I have not verified this so I left the code in place.
152 TF/02 Correct gecos expansion when From: is a prefix of the username.
154 Test 0254 submits a message to Exim with the header
158 When I ran the test suite under the user fanf2, Exim expanded
159 the header to contain my full name, whereas it should have added
160 a Resent-Sender: header. It erroneously treats any prefix of the
161 username as equal to the username.
163 This change corrects that bug.
165 GF/01 DCC debug and logging tidyup
166 Error conditions log to paniclog rather than rejectlog.
167 Debug lines prefixed by "DCC: " to remove any ambiguity.
169 TF/03 Avoid unnecessary rebuilds of lookup-related code.
171 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
172 Bug spotted by Jeremy Harris; was flawed since initial commit.
173 Would have resulted in OCSP responses post-SNI triggering an Exim
174 NULL dereference and crash.
176 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
178 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
179 Bug detection, analysis and fix by Samuel Thibault.
180 Bugzilla 1331, Debian bug #698092.
182 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
184 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
185 Server implementation by Todd Lyons, client by JH.
186 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
187 config variable "prdr_enable" controls whether the server
188 advertises the facility. If the client requests PRDR a new
189 acl_data_smtp_prdr ACL is called once for each recipient, after
190 the body content is received and before the acl_smtp_data ACL.
191 The client is controlled by bolth of: a hosts_try_prdr option
192 on the smtp transport, and the server advertisement.
193 Default client logging of deliveries and rejections involving
194 PRDR are flagged with the string "PRDR".
196 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
197 fclose(). Diagnosis by Todd Lyons.
199 PP/17 Update configure.default to handle IPv6 localhost better.
200 Patch by Alain Williams (plus minor tweaks).
203 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
204 This is now consistent with GnuTLS, and is now documented: the
205 previous undocumented portable approach to treating the option as
206 unset was to force an expansion failure. That still works, and
207 an empty string is now equivalent.
209 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
210 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
211 not performing validation itself.
213 PP/20 Added force_command boolean option to pipe transport.
214 Patch from Nick Koston, of cPanel Inc.
216 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
219 TF/04 Added udpsend ACL modifer and hexquote expansion operator
221 PP/21 Fix eximon continuous updating with timestamped log-files.
222 Broken in a format-string cleanup in 4.80, missed when I repaired the
223 other false fix of the same issue.
224 Report and fix from Heiko Schlichting.
227 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
228 Report from Prashanth Katuri.
230 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
231 It's SecureTransport, so affects any MacOS clients which use the
232 system-integrated TLS libraries, including email clients.
234 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
235 using a MIME ACL for non-SMTP local injection.
236 Report and assistance in diagnosis by Warren Baker.
238 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
240 JH/16 Fix comparisons for 64b. Bugzilla 1385.
242 TL/09 Add expansion variable $authenticated_fail_id to keep track of
243 last id that failed so it may be referenced in subsequent ACL's.
245 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
248 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
249 ldap library initialization, allowing self-signed CA's to be
250 used. Also properly sets require_cert option later in code by
251 using NULL (global ldap config) instead of ldap handle (per
252 session). Bug diagnosis and testing by alxgomz.
254 TL/12 Enhanced documentation in the ratelimit.pl script provided in
255 the src/util/ subdirectory.
257 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
258 renamed to Transport Post Delivery Action by Jeremy Harris, as
261 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
262 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
263 redis_servers = needs to be configured which will be used by the redis
264 lookup. Patch from Warren Baker, of The Packet Hub.
266 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
268 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
269 hostname or reverse DNS when processing a host list. Used suggestions
270 from multiple comments on this bug.
272 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
274 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
275 Missed a few lines, added it to make the runtest require no keyboard
278 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
279 contains upper case chars. Make router use caseful_local_part.
281 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
282 support when GnuTLS has been built with p11-kit.
288 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
290 This, or similar/improved, will also be change PP/11 of 4.82.
296 PP/01 Handle short writes when writing local log-files.
297 In practice, only affects FreeBSD (8 onwards).
298 Bugzilla 1053, with thanks to Dmitry Isaikin.
300 NM/01 Bugzilla 949 - Documentation tweak
302 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
305 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
307 PP/02 Implemented gsasl authenticator.
309 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
311 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
312 `pkg-config foo` for cflags/libs.
314 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
315 with rest of GSASL and with heimdal_gssapi.
317 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
318 `pkg-config foo` for cflags/libs for the TLS implementation.
320 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
321 properties get this fed in as external SSF. A number of robustness
322 and debugging improvements to the cyrus_sasl authenticator.
324 PP/08 cyrus_sasl server now expands the server_realm option.
326 PP/09 Bugzilla 1214 - Log authentication information in reject log.
327 Patch by Jeremy Harris.
329 PP/10 Added dbmjz lookup type.
331 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
333 PP/12 MAIL args handles TAB as well as SP, for better interop with
334 non-compliant senders.
335 Analysis and variant patch by Todd Lyons.
337 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
338 Bug report from Lars Müller <lars@samba.org> (via SUSE),
339 Patch from Dirk Mueller <dmueller@suse.com>
341 PP/13 tls_peerdn now print-escaped for spool files.
342 Observed some $tls_peerdn in wild which contained \n, which resulted
343 in spool file corruption.
345 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
346 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
347 or write after TLS renegotiation, which otherwise led to messages
350 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
351 as a tracking header (ie: a signed header comes before the signature).
352 Patch from Wolfgang Breyha.
354 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
355 comma-sep list; embedded commas doubled.
357 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
359 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
361 Report and patch from Dmitry Banschikov.
363 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
364 Removed SSL_clear() after SSL_new() which led to protocol negotiation
365 failures. We appear to now support TLS1.1+ with Exim.
367 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
368 lets Exim select keys and certificates based upon TLS SNI from client.
369 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
370 before an outbound SMTP session. New log_selector, +tls_sni.
372 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
373 NULL dereference. Report and patch from Alun Jones.
375 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
376 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
377 Not seeing resolver debug output on NetBSD, but suspect this is a
378 resolver implementation change.
380 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
381 Left warnings. Added "eximon gdb" invocation mode.
383 PP/21 Defaulting "accept_8bitmime" to true, not false.
385 PP/22 Added -bw for inetd wait mode support.
387 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
388 locate the relevant includes and libraries. Made this the default.
390 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
391 Bugzilla 1246, report and most of solution from Tomasz Kusy.
393 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
394 This may cause build issues on older platforms.
396 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
397 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
398 gnutls_require_mac & gnutls_require_protocols (no longer supported).
399 Added SNI support via GnuTLS too.
400 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
402 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
404 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
405 Applied second patch from Janne, implementing suggestion to default
406 multiple-strings-in-record handling to match SPF spec.
408 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
410 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
411 read-only, out of scope).
412 Patch from Wolfgang Breyha, report from Stuart Northfield.
414 PP/29 Fix three issues highlighted by clang analyser static analysis.
415 Only crash-plausible issue would require the Cambridge-specific
416 iplookup router and a misconfiguration.
417 Report from Marcin Mirosław.
419 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
421 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
422 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
423 As part of this, removing so much warning spew let me fix some minor
424 real issues in debug logging.
426 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
427 assignment on my part. Fixed.
429 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
430 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
431 Janne Snabb (who went above and beyond: thank you).
433 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
434 string otherwise requires a connection and a bunch more work and it's
435 relatively easy to get wrong. Should also expose TLS library linkage
438 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
439 64-bit ${eval} (JH/03).
441 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
442 GNU libc to support some of the 64-bit stuff, should not lead to
443 conflicts. Defined before os.h is pulled in, so if a given platform
444 needs to override this, it can.
446 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
447 protection layer was required, which is not implemented.
448 Bugzilla 1254, patch from Wolfgang Breyha.
450 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
451 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
452 tls_dhparam take prime identifiers. Also unbreak combination of
453 OpenSSL+DH_params+TLSSNI.
455 PP/39 Disable SSLv2 by default in OpenSSL support.
461 PP/01 Solaris build fix for Oracle's LDAP libraries.
462 Bugzilla 1109, patch from Stephen Usher.
464 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
466 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
469 TF/02 Fix a couple more cases where we did not log the error message
470 when unlink() failed. See also change 4.74-TF/03.
472 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
473 lock up or crash if it happened to be inside a call to libc when it
474 got a SIGUSR1 from exiwhat.
476 The SIGUSR1 handler appends the current process status to the process
477 log which is later printed by exiwhat. It used to use the general
478 purpose logging code to do this, but several functions it calls are
479 not safe for signals.
481 The new output code in the SIGUSR1 handler is specific to the process
482 log, and simple enough that it's easy to inspect for signal safety.
483 Removing some special cases also simplifies the general logging code.
484 Removing the spurious timestamps from the process log simplifies
487 TF/04 Improved ratelimit ACL condition.
489 The /noupdate option has been deprecated in favour of /readonly which
490 has clearer semantics. The /leaky, /strict, and /readonly update modes
491 are mutually exclusive. The update mode is no longer included in the
492 database key; it just determines when the database is updated. (This
493 means that when you upgrde Exim will forget old rate measurements.)
495 Exim now checks that the per_* options are used with an update mode that
496 makes sense for the current ACL. For example, when Exim is processing a
497 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
498 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
499 must specify per_mail/readonly. If you omit the update mode it defaults to
500 /leaky where that makes sense (as before) or /readonly where required.
502 The /noupdate option is now undocumented but still supported for
503 backwards compatibility. It is equivalent to /readonly except that in
504 ACLs where /readonly is required you may specify /leaky/noupdate or
505 /strict/noupdate which are treated the same as /readonly.
507 A useful new feature is the /count= option. This is a generalization
508 of the per_byte option, so that you can measure the throughput of other
509 aggregate values. For example, the per_byte option is now equivalent
510 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
512 The per_rcpt option has been generalized using the /count= mechanism
513 (though it's more complicated than the per_byte equivalence). When it is
514 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
515 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
516 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
517 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
518 Note that using per_rcpt with a non-readonly update mode in more than
519 one ACL will cause the recipients to be double-counted. (The per_mail
520 and per_byte options don't have this problem.)
522 The handling of very low rates has changed slightly. If the computed rate
523 is less than the event's count (usually one) then this event is the first
524 after a long gap. In this case the rate is set to the same as this event's
525 count, so that the first message of a spam run is counted properly.
527 The major new feature is a mechanism for counting the rate of unique
528 events. The new per_addr option counts the number of different
529 recipients that someone has sent messages to in the last time period. It
530 behaves like per_rcpt if all the recipient addresses are different, but
531 duplicate recipient addresses do not increase the measured rate. Like
532 the /count= option this is a general mechanism, so the per_addr option
533 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
534 example, measure the rate that a client uses different sender addresses
535 with the options per_mail/unique=$sender_address. There are further
536 details in the main documentation.
538 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
540 TF/06 Removed a few PCRE remnants.
542 TF/07 Automatically extract Exim's version number from tags in the git
543 repository when doing development or release builds.
545 PP/02 Raise smtp_cmd_buffer_size to 16kB.
546 Bugzilla 879. Patch from Paul Fisher.
548 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
549 Heavily based on revision 40f9a89a from Simon Arlott's tree.
552 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
554 PP/05 Variable $av_failed, true if the AV scanner deferred.
555 Bugzilla 1078. Patch from John Horne.
557 PP/06 Stop make process more reliably on build failure.
558 Bugzilla 1087. Patch from Heiko Schlittermann.
560 PP/07 Make maildir_use_size_file an _expandable_ boolean.
561 Bugzilla 1089. Patch from Heiko Schlittermann.
563 PP/08 Handle ${run} returning more data than OS pipe buffer size.
564 Bugzilla 1131. Patch from Holger Weiß.
566 PP/09 Handle IPv6 addresses with SPF.
567 Bugzilla 860. Patch from Wolfgang Breyha.
569 PP/10 GnuTLS: support TLS 1.2 & 1.1.
571 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
574 PP/11 match_* no longer expand right-hand-side by default.
575 New compile-time build option, EXPAND_LISTMATCH_RHS.
576 New expansion conditions, "inlist", "inlisti".
578 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
580 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
582 PP/14 fix log_write() format string regression from TF/03.
583 Bugzilla 1152. Patch from Dmitry Isaikin.
589 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
591 PP/02 Harmonised TLS library version reporting; only show if debugging.
592 Layout now matches that introduced for other libraries in 4.74 PP/03.
594 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
596 PP/04 New "dns_use_edns0" global option.
598 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
601 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
602 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
604 TK/01 Updated PolarSSL code to 0.14.2.
605 Bugzilla 1097. Patch from Andreas Metzler.
607 PP/07 Catch divide-by-zero in ${eval:...}.
610 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
613 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
614 format-string attack -- SECURITY: remote arbitrary code execution.
616 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
617 time unintentionally subject to list matching rules, letting the header
618 cause arbitrary Exim lookups (of items which can occur in lists, *not*
619 arbitrary string expansion). This allowed for information disclosure.
621 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
622 INT_MIN/-1 -- value coerced to INT_MAX.
628 NM/01 Workround for PCRE version dependency in version reporting
631 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
632 This fixes portability to compilers other than gcc, notably
633 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
635 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
636 makefiles for portability to HP-UX and POSIX correctness.
638 PP/01 Permit LOOKUP_foo enabling on the make command-line.
639 Also via indented variable definition in the Makefile.
640 (Debugging by Oliver Heesakkers).
642 PP/02 Restore caching of spamd results with expanded spamd_address.
643 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
645 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
646 Improves build reliability. Fix from: Frank Elsner
648 NM/02 Fix wide character breakage in the rfc2047 coding
649 Fixes bug 1064. Patch from Andrey N. Oktyabrski
651 NM/03 Allow underscore in dnslist lookups
652 Fixes bug 1026. Patch from Graeme Fowler
654 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
655 Code patches from Adam Ciarcinski of NetBSD.
657 NM/04 Fixed exiqgrep to cope with mailq missing size issue
660 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
661 is logged, to avoid truncation. Patch from John Horne.
663 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
664 Patch from Jakob Hirsch.
666 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
667 SQL string expansion failure details.
668 Patch from Andrey Oktyabrski.
670 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
671 Patch from Simon Arlott.
673 PP/09 New lookups functionality failed to compile on old gcc which rejects
674 extern declarations in function scope.
675 Patch from Oliver Fleischmann
677 PP/10 Use sig_atomic_t for flags set from signal handlers.
678 Check getgroups() return and improve debugging.
679 Fixed developed for diagnosis in bug 927 (which turned out to be
682 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
683 Patch from Mark Zealey.
685 PP/12 Bugzilla 1056: Improved spamd server selection.
686 Patch from Mark Zealey.
688 PP/13 Bugzilla 1086: Deal with maildir quota file races.
689 Based on patch from Heiko Schlittermann.
691 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
692 Patch from Uwe Doering, sign-off by Michael Haardt.
694 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
695 variable declaration deep within a block. Bug and patch from
698 PP/15 lookups-Makefile IRIX compatibilty coercion.
700 PP/16 Make DISABLE_DKIM build knob functional.
702 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
703 Patch by Simon Arlott
705 TF/03 Fix valgrind.h portability to C89 compilers that do not support
706 variable argument macros. Our copy now differs from upstream.
712 TF/01 Failure to get a lock on a hints database can have serious
713 consequences so log it to the panic log.
715 TF/02 Log LMTP confirmation messages in the same way as SMTP,
716 controlled using the smtp_confirmation log selector.
718 TF/03 Include the error message when we fail to unlink a spool file.
720 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
721 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
722 for maintaining out-of-tree patches for some time.
724 PP/01 Bugzilla 139: Documentation and portability issues.
725 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
726 Handle per-OS dynamic-module compilation flags.
728 PP/02 Let /dev/null have normal permissions.
729 The 4.73 fixes were a little too stringent and complained about the
730 permissions on /dev/null. Exempt it from some checks.
731 Reported by Andreas M. Kirchwitz.
733 PP/03 Report version information for many libraries, including
734 Exim version information for dynamically loaded libraries. Created
735 version.h, now support a version extension string for distributors
736 who patch heavily. Dynamic module ABI change.
738 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
739 privilege escalation vulnerability whereby the Exim run-time user
740 can cause root to append content of the attacker's choosing to
743 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
746 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
747 If dropping privileges for untrusted macros, we disabled normal logging
748 on the basis that it would fail; for the Exim run-time user, this is not
749 the case, and it resulted in successful deliveries going unlogged.
750 Fixed. Reported by Andreas Metzler.
756 PP/01 Date: & Message-Id: revert to normally being appended to a message,
757 only prepend for the Resent-* case. Fixes regression introduced in
758 Exim 4.70 by NM/22 for Bugzilla 607.
760 PP/02 Include check_rfc2047_length in configure.default because we're seeing
761 increasing numbers of administrators be bitten by this.
763 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
765 PP/03 Bugzilla 994: added openssl_options main configuration option.
767 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
769 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
771 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
773 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
774 without a peer certificate, leading to a segfault because of an
775 assumption that peers always have certificates. Be a little more
776 paranoid. Problem reported by Martin Tscholak.
778 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
779 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
780 NB: ClamAV planning to remove STREAM in "middle of 2010".
781 CL also introduces -bmalware, various -d+acl logging additions and
782 more caution in buffer sizes.
784 PP/09 Implemented reverse_ip expansion operator.
786 PP/10 Bugzilla 937: provide a "debug" ACL control.
788 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
790 PP/12 Bugzilla 973: Implement --version.
792 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
794 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
796 PP/15 Bugzilla 816: support multiple condition rules on Routers.
798 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
799 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
800 ignore trailing whitespace.
802 JJ/02 prevent non-panic DKIM error from being sent to paniclog
804 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
807 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
808 Notification from Dr Andrew Aitchison.
810 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
811 ExtendedDetectionInfo response format.
812 Notification from John Horne.
814 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
817 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
818 XSL and documented dependency on system catalogs, with examples of how
821 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
824 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
825 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
826 configuration file which is writeable by the Exim user or group.
828 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
829 of configuration files to cover files specified with the -C option if
830 they are going to be used with root privileges, not just the default
833 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
834 option (effectively making it always true).
836 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
837 files to be used while preserving root privileges.
839 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
840 that rogue child processes cannot use them.
842 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
843 run-time user, instead of root.
845 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
846 Exim run-time user without dropping privileges.
848 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
849 result string, instead of calling string_vformat() twice with the same
852 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
853 for other users. Others should always drop root privileges if they use
854 -C on the command line, even for a whitelisted configure file.
856 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
858 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
864 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
865 $data_path, and $header_path variables; fixed documentation bugs and
868 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
869 exipick to access non-standard spools, including the "frozen" queue
872 NM/01 Bugzilla 965: Support mysql stored procedures.
873 Patch from Alain Williams
875 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
877 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
878 Patch from Andreas Metzler
880 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
881 Patch from Kirill Miazine
883 NM/05 Bugzilla 671: Added umask to procmail example.
885 JJ/03 installed exipick 20100323.0, fixing doc bug
887 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
888 directory. Notification and patch from Dan Rosenberg.
890 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
892 TK/02 Improve log output when DKIM signing operation fails.
894 MH/01 Treat the transport option dkim_domain as a colon separated
895 list, not as a single string, and sign the message with each element,
896 omitting multiple occurences of the same signer.
898 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
899 Bugzilla 985, 986. Patch by Simon Arlott
901 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
902 Patch by Simon Arlott
904 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
905 MBX locking. Notification from Dan Rosenberg.
911 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
913 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
915 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
917 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
919 NM/04 Bugzilla 915: Flag broken perl installation during build.
925 TK/01 Added patch by Johannes Berg that expands the main option
926 "spamd_address" if it starts with a dollar sign.
928 TK/02 Write list of recipients to X-Envelope-Sender header when building
929 the mbox-format spool file for content scanning (suggested by Jakob
932 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
933 (http://www.dcc-servers.net/) support via dccifd. Activated by
934 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
936 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
937 by Mark Daniel Reidel <mr@df.eu>.
939 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
940 When building exim an external PCRE library is now needed -
941 PCRE is a system library on the majority of modern systems.
942 See entry on PCRE_LIBS in EDITME file.
944 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
945 conversation. Added nologin parameter to request.
946 Patch contributed by Kirill Miazine.
948 TF/01 Do not log submission mode rewrites if they do not change the address.
950 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
952 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
953 log files in place. Contributed by Roberto Lima.
955 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
957 TF/03 Bugzilla 615: When checking the local_parts router precondition
958 after a local_part_suffix or local_part_prefix option, Exim now
959 does not use the address's named list lookup cache, since this
960 contains cached lookups for the whole local part.
962 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
963 Robert Millan. Documentation is in experimental-spec.txt.
965 TF/04 Bugzilla 668: Fix parallel build (make -j).
967 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
969 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
970 Patch provided by Jan Srzednicki.
972 TF/05 Leading white space used to be stripped from $spam_report which
973 wrecked the formatting. Now it is preserved.
975 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
976 that they are available at delivery time.
978 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
980 TF/08 TLS error reporting now respects the incoming_interface and
981 incoming_port log selectors.
983 TF/09 Produce a more useful error message if an SMTP transport's hosts
984 setting expands to an empty string.
986 NM/06 Bugzilla 744: EXPN did not work under TLS.
987 Patch provided by Phil Pennock.
989 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
990 Patch provided by Richard Godbee.
992 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
993 acl_smtp_notquit, added index entry.
995 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
996 Patch provided by Eugene Bujak.
998 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
999 accept(). Patch provided by Maxim Dounin.
1001 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1002 Patch provided by Phil Pennock.
1004 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1006 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1007 Patch provided by Brad "anomie" Jorsch.
1009 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1010 Patch provided by Dean Brooks.
1012 TK/05 Add native DKIM support (does not depend on external libraries).
1014 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1015 Patch provided by Graeme Fowler.
1017 NM/16 Bugzilla 851: Documentation example syntax fix.
1019 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1021 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1024 NM/19 Bugzilla 745: TLS version reporting.
1025 Patch provided by Phil Pennock.
1027 NM/20 Bugzilla 167: bool: condition support.
1028 Patch provided by Phil Pennock.
1030 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1031 clients. Patch provided by Phil Pennock.
1033 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1034 Patch provided by Brad "anomie" Jorsch.
1036 NM/23 Bugzilla 687: Fix misparses in eximstats.
1037 Patch provided by Heiko Schlittermann.
1039 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1040 Patch provided by Heiko Schlittermann.
1042 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1043 plus update to original patch.
1045 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1047 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1048 Patch provided by David Brownlee.
1050 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1052 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1054 NM/30 Bugzilla 888: TLS documentation bugfixes.
1056 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1058 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1059 Unlike the original bugzilla I have changed all shell scripts in src tree.
1061 NM/33 Bugzilla 898: Transport filter timeout fix.
1062 Patch by Todd Rinaldo.
1064 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1065 Patch by Serge Demonchaux.
1067 NM/35 Bugzilla 39: Base64 decode bug fixes.
1068 Patch by Jakob Hirsch.
1070 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1072 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1074 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1076 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1082 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1083 ALT-N's libdkim that I have put here:
1084 http://duncanthrax.net/exim-experimental/
1086 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1087 were called 'true' and it seems that C99 defines that as a reserved
1088 keyword to be used with 'bool' variable types. That means you could
1089 not include C99-style headers which use bools without triggering
1090 build errors in sieve.c.
1092 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1093 as mailq or other aliases. Changed the --help handling significantly
1094 to do whats expected. exim_usage() emits usage/help information.
1096 SC/01 Added the -bylocaldomain option to eximstats.
1098 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1100 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1102 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1108 PH/01 Another patch from the Sieve maintainer.
1110 PH/02 When an IPv6 address is converted to a string for single-key lookup
1111 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1112 dots are used instead of colons so that keys in lsearch files need not
1113 contain colons. This was done some time before quoting was made available
1114 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1115 (notated using the quote facility) so as to distinguish them from IPv4
1116 keys. This meant that lookups for IP addresses in host lists did not work
1117 for iplsearch lookups.
1119 This has been fixed by arranging for IPv6 addresses to be expressed with
1120 colons if the lookup type is iplsearch. This is not incompatible, because
1121 previously such lookups could never work.
1123 The situation is now rather anomolous, since one *can* have colons in
1124 ordinary lsearch keys. However, making the change in all cases is
1125 incompatible and would probably break a number of configurations.
1127 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1130 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1131 conversion specification without a maximum field width, thereby enabling
1132 a rogue spamd server to cause a buffer overflow. While nobody in their
1133 right mind would setup Exim to query an untrusted spamd server, an
1134 attacker that gains access to a server running spamd could potentially
1135 exploit this vulnerability to run arbitrary code as the Exim user.
1137 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1138 $primary_hostname instead of what libspf2 thinks the hosts name is.
1140 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1141 a directory entry by the name of the lookup key. Previously, if a
1142 symlink pointed to a non-existing file or a file in a directory that
1143 Exim lacked permissions to read, a lookup for a key matching that
1144 symlink would fail. Now it is enough that a matching directory entry
1145 exists, symlink or not. (Bugzilla 503.)
1147 PH/03 The body_linecount and body_zerocount variables are now exported in the
1150 PH/04 Added the $dnslist_matched variable.
1152 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1153 This means they are set thereafter only if the connection becomes
1156 PH/06 Added the client_condition to authenticators so that some can be skipped
1157 by clients under certain conditions.
1159 PH/07 The error message for a badly-placed control=no_multiline_responses left
1160 "_responses" off the end of the name.
1162 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1164 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1165 (without spaces) instead of just copying the configuration text.
1167 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1169 PH/11 Added $max_received_linelength.
1171 PH/12 Added +ignore_defer and +include_defer to host lists.
1173 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1174 way in which PCRE > 7.0 is built.
1176 PH/14 Implemented queue_only_load_latch.
1178 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1179 MAIL command. The effect was to mangle the value on 64-bit systems.
1181 PH/16 Another patch from the Sieve maintainer.
1183 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1185 PH/18 If a system quota error occurred while trying to create the file for
1186 a maildir delivery, the message "Mailbox is full" was not appended to the
1187 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1188 only to a quota excession during the actual writing of the file.
1190 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1191 characters?) which causes problems in log lines. The DN values are now
1192 passed through string_printing() before being added to log lines.
1194 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1195 and InterBase are left for another time.)
1197 PH/21 Added message_body_newlines option.
1199 PH/22 Guard against possible overflow in moan_check_errorcopy().
1201 PH/23 POSIX allows open() to be a macro; guard against that.
1203 PH/24 If the recipient of an error message contained an @ in the local part
1204 (suitably quoted, of course), incorrect values were put in $domain and
1205 $local_part during the evaluation of errors_copy.
1211 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1212 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1215 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1216 issue a MAIL command.
1218 PH/02 In an ACL statement such as
1220 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1222 if a client was not listed at all, or was listed with a value other than
1223 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1224 the condition was not true (as it should be), so access was not denied.
1225 The bug was that the ! inversion was incorrectly passed on to the second
1226 item. This has been fixed.
1228 PH/03 Added additional dnslists conditions == and =& which are different from
1229 = and & when the dns lookup returns more than one IP address.
1231 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1232 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1234 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1235 FSYNC, which compiles an option called disable_fsync that allows for
1236 bypassing fsync(). The documentation is heavily laced with warnings.
1238 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1240 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1241 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1242 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1243 including adding "make clean"; (3) Added -fPIC when compiling the test
1244 dynamically loaded module, to get rid of a warning.
1246 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1247 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1248 = 0s. The bug is otherwise harmless.
1250 PH/07 There was a bug in the dovecot authenticator such that the value of
1251 $auth1 could be overwritten, and so not correctly preserved, after a
1252 successful authentication. This usually meant that the value preserved by
1253 the server_setid option was incorrect.
1255 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1257 PH/09 Installed PCRE release 7.0.
1259 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1260 run for batched SMTP input. It is now run at the start of every message
1261 in the batch. While fixing this I discovered that the process information
1262 (output by running exiwhat) was not always getting set for -bs and -bS
1263 input. This is fixed, and it now also says "batched" for BSMTP.
1265 PH/11 Added control=no_pipelining.
1267 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1268 patch, slightly modified), and move the expansion of helo_data till after
1269 the connection is made in the smtp transport (so it can use these
1272 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1274 PH/14 Added log_selector = +pid.
1276 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1278 PH/16 Add ${if forany and ${if forall.
1280 PH/17 Added dsn_from option to vary the From: line in DSNs.
1282 PH/18 Flush SMTP output before performing a callout, unless control =
1283 no_callout_flush is set.
1285 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1286 was true (the default) a successful delivery failed to delete the retry
1287 item, thus causing premature timeout of the address. The bug is now
1290 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1292 PH/21 Long custom messages for fakedefer and fakereject are now split up
1293 into multiline reponses in the same way that messages for "deny" and
1294 other ACL rejections are.
1296 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1297 with slight modification.
1299 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1300 draft, changing the syntax and factoring some duplicate code".
1302 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1303 for deliveries of the second and subsequent messages over the same SMTP
1306 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1307 ${reduce, with only minor "tidies".
1309 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1311 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1312 expansion side effects.
1314 PH/27 When a message times out after an over-quota error from an Exim-imposed
1315 quota, the bounce message says "mailbox is full". This message was not
1316 being given when it was a system quota that was exceeded. It now should
1319 MH/03 Made $recipients available in local_scan(). local_scan() already has
1320 better access to the recipient list through recipients_list[], but
1321 $recipients can be useful in postmaster-provided expansion strings.
1323 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1324 in the case of a MAIL command with additional options following the
1325 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1326 were accidentally chopped off.
1328 PH/29 SMTP synchronization checks are implemented when a command is read -
1329 there is a check that no more input is waiting when there shouldn't be
1330 any. However, for some commands, a delay in an ACL can mean that it is
1331 some time before the response is written. In this time, more input might
1332 arrive, invalidly. So now there are extra checks after an ACL has run for
1333 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1334 pipelining has not been advertised.
1336 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1338 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1339 returned in the SMTP response when smtp_return_error_details was false.
1340 This has been fixed.
1342 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1343 instead of the C I/O that was originally supplied, because problems were
1344 reported on Solaris.
1346 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1347 Exim which did not show up earlier: it was assuming that a call to
1348 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1349 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1350 was a macro that became an assignment, so it seemed to work. This has
1351 changed to a proper function call with a void return, hence the compile
1352 error. Exim's code has been fixed.
1354 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1357 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1359 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1361 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1362 because it turns out to be too big or there is a timeout) there is no
1363 "Completed" line in the log. When some messages of this type were
1364 selected by exigrep, they were listed as "not completed". Others were
1365 picked up by some special patterns. I have improved the selection
1366 criteria to be more general.
1368 PH/38 The host_find_failed option in the manualroute router can now be set
1369 to "ignore", to completely ignore a host whose IP address cannot be
1370 found. If all hosts are ignored, the behaviour is controlled by the new
1371 host_all_ignored option.
1373 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1374 homing or because of multiple MX records with /mx) generated more than
1375 one IP address, and the following item turned out to be the local host,
1376 all the secondary addresses of the first item were incorrectly removed
1377 from the list, along with the local host and any following hosts (which
1378 is what is supposed to happen).
1380 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1381 whoever called Exim into the -H file. In the case of the daemon it was
1382 behaving confusingly. When first started, it used values for whoever
1383 started the daemon, but after a SIGHUP it used the Exim user (because it
1384 calls itself on a restart). I have changed the code so that it now always
1387 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1388 message are rejected with the same error (e.g. no authentication or bad
1389 sender address), and a DATA command is nevertheless sent (as can happen
1390 with PIPELINING or a stupid MUA), the error message that was given to the
1391 RCPT commands is included in the rejection of the DATA command. This is
1392 intended to be helpful for MUAs that show only the final error to their
1395 PH/42 Another patch from the Sieve maintainer.
1397 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1398 Eximstats - Fixed some broken HTML links and added missing column headers
1400 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1401 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1403 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1405 PH/43 Yet another patch from the Sieve maintainer.
1407 PH/44 I found a way to check for a TCP/IP connection going away before sending
1408 the response to the final '.' that terminates a message, but only in the
1409 case where the client has not sent further data following the '.'
1410 (unfortunately, this is allowed). However, in many cases there won't be
1411 any further data because there won't be any more messages to send. A call
1412 to select() can be used: if it shows that the input is "ready", there is
1413 either input waiting, or the socket has been closed. An attempt to read
1414 the next input character can distinguish the two cases. Previously, Exim
1415 would have sent an OK response which the client would never have see.
1416 This could lead to message repetition. This fix should cure that, at
1417 least in a lot of common cases.
1419 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1420 advertised in response to EHLO.
1426 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1427 fixed by 4.65/MH/01 (is this a record?) are fixed:
1429 (i) An empty string was always treated as zero by the numeric comparison
1430 operators. This behaviour has been restored.
1432 (ii) It is documented that the numeric comparison operators always treat
1433 their arguments as decimal numbers. This was broken in that numbers
1434 starting with 0 were being interpreted as octal.
1436 While fixing these problems I realized that there was another issue that
1437 hadn't been noticed. Values of message_size_limit (both the global option
1438 and the transport option) were treated as octal if they started with 0.
1439 The documentation was vague. These values are now always treated as
1440 decimal, and I will make that clear in the documentation.
1446 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1447 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1450 MH/01 Don't check that the operands of numeric comparison operators are
1451 integers when their expansion is in "skipping" mode (fixes bug
1452 introduced by 4.64-PH/07).
1454 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1455 child addresses, Exim now panics and dies. Previously, because the count
1456 is held in a short int, deliveries were likely to be lost. As such a
1457 large number of recipients for a single message is ridiculous
1458 (performance will be very, very poor), I have chosen to impose a limit
1459 rather than extend the field.
1465 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1466 leftover -K file (the existence of which was triggered by #402).
1467 While we were at it, introduced process PID as part of the -K
1468 filename. This should rule out race conditions when creating
1471 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1472 processing considerably. Previous code took too long for large mails,
1473 triggering a timeout which in turn triggers #401.
1475 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1476 in the DK code in transports.c. sendfile() is not really portable,
1477 hence the _LINUX specificness.
1479 TF/01 In the add_headers option to the mail command in an Exim filter,
1480 there was a bug that Exim would claim a syntax error in any
1481 header after the first one which had an odd number of characters
1484 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1485 callout verification, Exim cached a "reject" for the entire domain. This
1486 is correct for most verifications, but it is not correct for a recipient
1487 verification with use_sender or use_postmaster set, because in that case
1488 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1489 case of MAIL FROM:<> rejection from other early rejections (e.g.
1490 rejection of HELO). When verifying a recipient using a non-null MAIL
1491 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1492 Whatever the result of the callout, the value of the domain cache is
1493 left unchanged (for any other kind of callout, getting as far as trying
1494 RCPT means that the domain itself is ok).
1496 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1499 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1500 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1501 session, the daemon ignores ECONNECT errors and logs others; it now
1502 ignores EPIPE as well.
1504 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1505 (quoted-printable decoding).
1507 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1508 later the small subsequent patch to fix an introduced bug.
1510 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1512 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1514 PH/08 An error is now given if message_size_limit is specified negative.
1516 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1517 to be given (somewhat) arbitrary names.
1519 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1522 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1523 miscellaneous code fixes
1525 PH/10 Added the log_reject_target ACL modifier to specify where to log
1528 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1529 hostname. This is wrong, because it relates to the incoming message (and
1530 probably the interface on which it is arriving) and not to the outgoing
1531 callout (which could be using a different interface). This has been
1532 changed to use the value of the helo_data option from the smtp transport
1533 instead - this is what is used when a message is actually being sent. If
1534 there is no remote transport (possible with a router that sets up host
1535 addresses), $smtp_active_hostname is used.
1537 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1538 tweaks were necessary in order to get it to work (see also 21 below):
1539 (a) The code assumed that strncpy() returns a negative number on buffer
1540 overflow, which isn't the case. Replaced with Exim's string_format()
1542 (b) There were several signed/unsigned issues. I just did the minimum
1543 hacking in of casts. There is scope for a larger refactoring.
1544 (c) The code used strcasecmp() which is not a standard C function.
1545 Replaced with Exim's strcmpic() function.
1546 (d) The code set only $1; it now sets $auth1 as well.
1547 (e) A simple test gave the error "authentication client didn't specify
1548 service in request". It would seem that Dovecot has changed its
1549 interface. Fortunately there's a specification; I followed it and
1550 changed what the client sends and it appears to be working now.
1552 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1555 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1556 address A is aliased to B and C, where B exists and C does not. Without
1557 -v the output is "A verified" because verification stops after a
1558 successful redirection if more than one address is generated. However,
1559 with -v the child addresses are also verified. Exim was outputting "A
1560 failed to verify" and then showing the successful verification for C,
1561 with its parentage. It now outputs "B failed to verify", showing B's
1562 parentage before showing the successful verification of C.
1564 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1565 look up a TXT record in a specific list after matching in a combined
1568 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1569 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1570 they consult the DNS. I had assumed they would set it the way they
1571 wanted; and indeed my experiments on Linux seem to show that in some
1572 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1573 To be on the safe side, however, I have now made the interface to
1574 host_find_byname() similar to host_find_bydns(), with an argument
1575 containing the DNS resolver options. The host_find_byname() function now
1576 sets these options at its start, just as host_find_bydns() does. The smtp
1577 transport options dns_qualify_single and dns_search_parents are passed to
1578 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1579 of host_find_byname() use the default settings of RES_DEFNAMES
1580 (qualify_single) but not RES_DNSRCH (search_parents).
1582 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1583 spool_read_header() do less string testing, by means of a preliminary
1584 switch on the second character of optional "-foo" lines. (This is
1585 overdue, caused by the large number of possibilities that now exist.
1586 Originally there were few.) While I was there, I also converted the
1587 str(n)cmp tests so they don't re-test the leading "-" and the first
1588 character, in the hope this might squeeze out yet more improvement.
1590 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1591 flag allowing group syntax was set by the header_syntax check but not
1592 turned off, possible causing trouble later; (2) The flag was not being
1593 set at all for the header_verify test, causing "group"-style headers to
1594 be rejected. I have now set it in this case, and also caused header_
1595 verify to ignore an empty address taken from a group. While doing this, I
1596 came across some other cases where the code for allowing group syntax
1597 while scanning a header line wasn't quite right (mostly, not resetting
1598 the flag correctly in the right place). These bugs could have caused
1599 trouble for malformed header lines. I hope it is now all correct.
1601 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1602 with the "reply" argument non-NULL. The code, however (which originally
1603 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1604 but it didn't always do it. This confused somebody who was copying the
1605 code for some other use. I have removed all the tests.
1607 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1608 feature that was used to support insecure browsers during the U.S. crypto
1609 embargo. It requires special client support, and Exim is probably the
1610 only MTA that supported it -- and would never use it because real RSA is
1611 always available. This code has been removed, because it had the bad
1612 effect of slowing Exim down by computing (never used) parameters for the
1613 RSA_EXPORT functionality.
1615 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1616 authenticator to fail if there's a tab character in the incoming data
1617 (there should never be unless someone is messing about, as it's supposed
1618 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1619 if the connection is using TLS or if the remote IP is the same as the
1620 local IP, and the "valid-client-cert option" if a client certificate has
1623 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1624 authenticators. This can be used for authorization after authentication
1625 succeeds. (In the case of plaintext, it servers for both authentication
1628 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1629 if any retry times were supplied.
1631 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1632 connection, where there is no client IP address to check. In this
1633 situation, the verify now always succeeds.
1635 PH/25 Applied John Jetmore's -Mset patch.
1637 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1639 PH/27 In a string expansion for a processed (not raw) header when multiple
1640 headers of the same name were present, leading whitespace was being
1641 removed from all of them, but trailing whitespace was being removed only
1642 from the last one. Now trailing whitespace is removed from each header
1643 before concatenation. Completely empty headers in a concatenation (as
1644 before) are ignored.
1646 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1647 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1649 PH/29 [Removed. This was a change that I later backed out, and forgot to
1650 correct the ChangeLog entry (that I had efficiently created) before
1651 committing the later change.]
1653 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1654 address errors (4xx response to RCPT) over the same connection as other
1655 messages routed to the same hosts. Such deliveries are always "forced",
1656 so retry times are not inspected. This resulted in far too many retries
1657 for the affected addresses. The effect occurred only when there were more
1658 hosts than the hosts_max_try setting in the smtp transport when it had
1659 the 4xx errors. Those hosts that it had tried were not added to the list
1660 of hosts for which the message was waiting, so if all were tried, there
1661 was no problem. Two fixes have been applied:
1663 (i) If there are any address or message errors in an SMTP delivery, none
1664 of the hosts (tried or untried) are now added to the list of hosts
1665 for which the message is waiting, so the message should not be a
1666 candidate for sending over the same connection that was used for a
1667 successful delivery of some other message. This seems entirely
1668 reasonable: after all the message is NOT "waiting for some host".
1669 This is so "obvious" that I'm not sure why it wasn't done
1670 previously. Hope I haven't missed anything, but it can't do any
1671 harm, as the worst effect is to miss an optimization.
1673 (ii) If, despite (i), such a delivery is accidentally attempted, the
1674 routing retry time is respected, so at least it doesn't keep
1675 hammering the server.
1677 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1678 in ${readsocket because some servers need this prod.
1680 PH/32 Added some extra debug output when updating a wait-xxx database.
1682 PH/33 The hint "could be header name not terminated by colon", which has been
1683 given for certain expansion errors for a long time, was not being given
1684 for the ${if def:h_colon_omitted{... case.
1686 PH/34 The spec says: "With one important exception, whenever a domain list is
1687 being scanned, $domain contains the subject domain." There was at least
1688 one case where this was not true.
1690 PH/35 The error "getsockname() failed: connection reset by peer" was being
1691 written to the panic log as well as the main log, but it isn't really
1692 panic-worthy as it just means the connection died rather early on. I have
1693 removed the panic log writing for the ECONNRESET error when getsockname()
1696 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1697 runs only) independently of the message's sender address. This meant
1698 that, if the 4xx error was in fact related to the sender, a different
1699 message to the same recipient with a different sender could confuse
1700 things. In particualar, this can happen when sending to a greylisting
1701 server, but other circumstances could also provoke similar problems.
1702 I have changed the default so that the retry time for these errors is now
1703 based a combination of the sender and recipient addresses. This change
1704 can be overridden by setting address_retry_include_sender=false in the
1707 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1708 remote server are returned as part of bounce messages. This was not
1709 happening for LMTP over a pipe (the lmtp transport), but now it is the
1710 same for both kinds of LMTP.
1712 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1713 in header lines that were in fact CNAMEs. This is no longer the case.
1715 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1716 and queue runs started by the daemon processed all messages. This has
1717 been fixed so that -R and -S can now usefully be given with -q<time>.
1719 PH/40 Import PCRE release 6.7 (fixes some bugs).
1721 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1723 PH/42 Give an error if -q is specified more than once.
1725 PH/43 Renamed the variables $interface_address and $interface_port as
1726 $received_ip_address and $received_port, to make it clear that these
1727 values apply to message reception, and not to the outgoing interface when
1728 a message is delivered. (The old names remain recognized, of course.)
1730 PH/44 There was no timeout on the connect() call when using a Unix domain
1731 socket in the ${readsocket expansion. There now is.
1733 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1734 be meaningful with "accept".
1736 SC/01 Eximstats V1.43
1737 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1739 SC/02 Eximstats V1.44
1740 Use a glob alias rather than an array ref in the generated
1741 parser. This improves both readability and performance.
1743 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1744 Collect SpamAssassin and rejection statistics.
1745 Don't display local sender or destination tables unless
1746 there is data to show.
1747 Added average volumes into the top table text output.
1749 SC/04 Eximstats V1.46
1750 Collect data on the number of addresses (recipients)
1751 as well as the number of messages.
1753 SC/05 Eximstats V1.47
1754 Added 'Message too big' to the list of mail rejection
1755 reasons (thanks to Marco Gaiarin).
1757 SC/06 Eximstats V1.48
1758 Mainlog lines which have GMT offsets and are too short to
1759 have a flag are now skipped.
1761 SC/07 Eximstats V1.49 (Alain Williams)
1762 Added the -emptyok flag.
1764 SC/08 Eximstats V1.50
1765 Fixes for obtaining the IP address from reject messages.
1767 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1768 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1769 whitesspace changes from 4.64-PH/27
1771 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1774 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1775 are found, allow negative numbers in numeric criteria)
1777 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1779 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1780 to match changes made in 4.64-PH/43
1782 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1784 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1785 "domain" is actually an IP address, and force a failure. This locks out
1786 those revolvers/nameservers that support "A-for-A" lookups, in
1787 contravention of the specifications.
1789 PH/48 When a host name was looked up from an IP address, and the subsequent
1790 forward lookup of the name timed out, the host name was left in
1791 $sender_host_name, contrary to the specification.
1793 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1794 restricted to single-key lookups, Exim was not diagnosing an error if
1795 * or *@ was used with a query-style lookup.
1797 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1799 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1800 long ago, but noone interested enough thought of it. Let's just say that
1801 the "1.1" means that there are some new functions that weren't there at
1802 some point in the past.
1804 PH/51 Error processing for expansion failure of helo_data from an smtp
1805 transport during callout processing was broken.
1807 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1808 tested/used via the -bh/-bhc/-bs options.
1810 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1811 bug, fixed in subsequent PCRE releases).
1813 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1814 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1816 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1822 SC/01 Use a glob alias rather than an array ref in eximstats generated
1823 parser. This improves both readability and performance.
1825 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1826 Don't display local sender or destination tables in eximstats unless
1827 there is data to show.
1828 Added average volumes into the eximstats top table text output.
1830 SC/03 Collect data on the number of addresses (recipients) as well
1831 as the number of messages in eximstats.
1833 TF/01 Correct an error in the documentation for the redirect router. Exim
1834 does (usually) call initgroups() when daemonizing.
1836 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1837 with consistent privilege compared to when running as a daemon.
1839 TF/03 Note in the spec that $authenticated_id is not set for local
1840 submissions from trusted users.
1842 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1843 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1845 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1846 by adding some example configuration directives to the default
1847 configuration file. A little bit of work is required to uncomment the
1848 directives and define how usernames and passwords are checked, but
1849 there is now a framework to start from.
1851 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1852 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1853 without this. I don't know how relevant this is to other LDAP libraries.
1855 PH/02 Add the verb name to the "unknown ACL verb" error.
1857 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1859 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1861 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1862 directory not expanded when it should be if an expanded home directory
1863 was set for the address (which is overridden by the transport).
1865 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1868 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1869 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1870 because it is too late at that time, and has no effect.
1872 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1873 security issue with \' (bugzilla #107). I could not use the
1874 PQescapeStringConn() function, because it needs a PGconn value as one of
1877 PH/08 When testing addresses using -bt, indicate those final addresses that
1878 are duplicates that would not cause an additional delivery. At least one
1879 person was confused, thinking that -bt output corresponded to deliveries.
1880 (Suppressing duplicates isn't a good idea as you lose the information
1881 about possibly different redirections that led to the duplicates.)
1883 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1884 systems where poll() doesn't work, in particular OS X.
1886 PH/10 Added more information to debugging output for retry time not reached.
1888 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1889 operations in malware.c.
1891 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1894 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1895 syslog_duplication set false, log lines that would normally be written
1896 both the the main log and to the reject log were not written to syslog at
1899 PH/14 In the default configuration, change the use of "message" in ACL warn
1900 statements to "add_header".
1902 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1903 not followed by a command (e.g. "seen endif").
1905 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1906 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1909 PH/17 Added extra conditions to the default value of delay_warning_condition
1913 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1914 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1915 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1918 The Auto-Submitted: and various List- headers are standardised, whereas I
1919 don't think Precedence: ever was.
1921 PH/18 Refactored debugging code in route_finduser() to show more information,
1922 in particular, the error code if getpwnam() issues one.
1924 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1925 This is apparently needed in addition to the PH/07 change above to avoid
1926 any possible encoding problems.
1928 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1929 but not after initializing Perl.
1931 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1932 output them only if debugging. By default they are written stderr,
1933 apparently, which is not desirable.
1935 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1938 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1941 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1943 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1944 authenticated or an ident call has been made. Suppress the default
1945 values for $authenticated_id and $authenticated_sender (but permit -oMai
1946 and -oMas) when testing with -bh.
1948 PH/24 Re-jigged the order of the tests in the default configuration so that the
1949 tests for valid domains and recipients precede the DNS black list and CSA
1950 tests, on the grounds that those ones are more expensive.
1952 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1953 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1954 This bug exists in every version of Exim that I still have, right back to
1957 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1958 However, an attempt to turn on case-sensitivity in a regex key by
1959 including (?-i) didn't work because the subject string was already
1960 lowercased, and the effects were non-intuitive. It turns out that a
1961 one-line patch can be used to allow (?-i) to work as expected.
1967 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1968 other effects) broke the use of negated acl sub-conditions.
1970 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1973 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1974 "Deny" causes Exim to reject the incoming connection with a 554 error.
1975 Unfortunately, if there is a major crisis, such as a disk failure,
1976 tcp-wrappers gives "deny", whereas what one would like would be some
1977 kind of temporary error. A kludge has been added to help with this.
1978 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1979 554 error is used if errno is still zero or contains ENOENT (which occurs
1980 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1983 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1985 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1986 errors. Otherwise a message that provokes a temporary error (when other
1987 messages do not) can cause a whole host to time out.
1989 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1990 addresses were routed directly to files or pipes from a redirect router.
1991 File deliveries just didn't batch; pipe deliveries might have suffered
1994 PH/06 A failure to get a lock for a hints database would erroneously always say
1995 "Failed to get write lock", even when it was really a read lock.
1997 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1998 of 0600. This has been changed to use the value of the lockfile_mode
1999 option (which defaults to 0600).
2001 PH/08 Applied small patch from the Sieve maintainer.
2003 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2004 folder from quota calculations, a direct delivery into this folder messed
2005 up the contents of the maildirsize file. This was because the regex was
2006 used only to exclude .Trash (or whatever) when the size of the mailbox
2007 was calculated. There was no check that a delivery was happening into an
2008 excluded directory. This bug has been fixed by ignoring all quota
2009 processing for deliveries into excluded directories.
2011 PH/10 Added the maildirfolder_create_regex option to appendfile.
2017 PH/01 The code for finding all the local interface addresses on a FreeBSD
2018 system running IPv6 was broken. This may well have applied to all BSD
2019 systems, as well as to others that have similar system calls. The broken
2020 code found IPv4 interfaces correctly, but gave incorrect values for the
2021 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2022 that it would not match correctly against @[] and not recognize the IPv6
2025 PH/02 The ipliteral router was not recognizing addresses of the form user@
2026 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2028 PH/03 Added disable_ipv6.
2030 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2031 decoded form, because it is most often used to construct To: headers
2032 lines in autoreplies, and the decoded form may well be syntactically
2033 invalid. However, $reply_address has leading white space removed, and all
2034 newlines turned into spaces so that the autoreply transport does not
2037 PH/05 If group was specified without a user on a router, and no group or user
2038 was specified on a transport, the group from the router was ignored.
2040 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2041 for visible compile-time settings that can be used to change these
2042 numbers, for those that want even more. Backwards compatibility with old
2043 spool files has been maintained. However, going back to a previous Exim
2044 release will lost any variables that are in spool files.
2046 PH/07 Two small changes when running in the test harness: increase delay when
2047 passing a TCP/IP connection to a new process, in case the original
2048 process has to generate a bounce, and remove special handling of
2049 127.0.0.2 (sic), which is no longer necessary.
2051 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2052 be the same on different OS.
2054 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2057 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2058 whether --show-vars was specified or not
2060 JJ/02 exipick: Added support for new ACL variable spool format introduced
2063 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2064 syntactically invalid From: or Reply-to: line, and a filter used this to
2065 generate an autoreply, and therefore failed to obtain an address for the
2066 autoreply, Exim could try to deliver to a non-existent relative file
2067 name, causing unrelated and misleading errors. What now happens is that
2068 it logs this as a hard delivery error, but does not attempt to create a
2071 PH/11 The exinext utility has a -C option for testing purposes, but although
2072 the given file was scanned by exinext itself; it wasn't being passed on
2073 when Exim was called.
2075 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2076 an end-of-file indication when reading a command response.
2078 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2079 compiled. In many other places in Exim, IPv6 addresses are always
2080 recognized, so I have changed this. It also means that IPv4 domain
2081 literals of the form [IPV4:n.n.n.n] are now always recognized.
2083 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2084 used if the router is not running as root, for example, when verifying at
2085 ACL time, or when using -bh. The debugging output from this situation was
2086 non-existent - all you got was a failure to exec. I have made two
2089 (a) Failures to set uid/gid, the current directory, or a process leader
2090 in a subprocess such as that created by queryprogram now generate
2091 suitable debugging ouput when -d is set.
2093 (b) The queryprogram router detects when it is not running as root,
2094 outputs suitable debugging information if -d is set, and then runs
2095 the subprocess without attempting to change uid/gid.
2097 PH/15 Minor change to Makefile for building test_host (undocumented testing
2100 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2101 additional section of a DNS packet that returns MX or SRV records.
2102 Instead, it always explicitly searches for A/AAAA records. This avoids
2103 major problems that occur when a DNS server includes only records of one
2104 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2105 fixed another bug: if SRV records were looked up and the corresponding
2106 address records were *not* found in the additional section, the port
2107 values from the SRV records were lost.
2109 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2110 using the correct key (the original address) when searching the retry
2111 rules in order to find which one to use for generating the retry hint.
2113 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2114 adding the default one. Similarly, if it contains a Reply-To: header, the
2115 errors_reply_to option, if set, is not used.
2117 PH/19 When calculating a retry time, Exim used to measure the "time since
2118 failure" by looking at the "first failed" field in the retry record. Now
2119 it does not use this if it is later than than the arrival time of the
2120 message. Instead it uses the arrival time. This makes for better
2121 behaviour in cases where some deliveries succeed, thus re-setting the
2122 "first failed" field. An example is a quota failure for a huge message
2123 when small messages continue to be delivered. Without this change, the
2124 "time since failure" will always be short, possible causing more frequent
2125 delivery attempts for the huge message than are intended.
2126 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2128 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2129 $1, $2, $3) because the numerical variables can be reset during some
2130 expansion items (e.g. "match"), thereby losing the authentication data.
2132 PH/21 Make -bV show the size of off_t variables so that the test suite can
2133 decide whether to run tests for quotas > 2G.
2135 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2136 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2137 filecount value is greater than 2G or if a quota value is greater than 2G
2138 on a system where the size of off_t is not greater than 4, a panic error
2141 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2142 never match. The debug and -bh output now contains an explicit error
2143 message indicating a malformed IPv4 address or mask.
2145 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2146 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2147 PH/23 above applies.
2149 PH/25 Do not write to syslog when running in the test harness. The only
2150 occasion when this arises is a failure to open the main or panic logs
2151 (for which there is an explicit test).
2153 PH/26 Added the /no_tell option to "control=freeze".
2155 PH/27 If a host name lookup failed very early in a connection, for example, if
2156 the IP address matched host_lookup and the reverse lookup yielded a name
2157 that did not have a forward lookup, an error message of the form "no IP
2158 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2159 could be logged. Now it outputs the IP address instead of "NULL".
2161 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2162 allows the sender and the authenticated sender to be set when
2163 submitting a message from within Exim. Since child_open_exim() is
2164 documented for local_scan(), the new function should be too.
2166 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2167 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2168 results in an empty string is now treated as unset.
2170 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2172 PH/31 Added sender_verify_fail logging option.
2174 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2175 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2176 tidied the source and removed it altogether.
2178 PH/33 When a queue run was abandoned because the load average was too high, a
2179 log line was always written; now it is written only if the queue_run log
2180 selector is set. In addition, the log line for abandonment now contains
2181 information about the queue run such as the pid. This is always present
2182 in "start" and "stop" lines but was omitted from the "abandon" line.
2184 PH/34 Omit spaces between a header name and the colon in the error message that
2185 is given when verify = headers_syntax fails (if there are lots of them,
2186 the message gets confusing).
2188 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2189 names, as there are now some PTR records that contain slashes. This check
2190 is only to protect against broken name servers that fall over on strange
2191 characters, so the fact that it applies to all lookups doesn't matter.
2193 PH/36 Now that the new test suite is complete, we can remove some of the
2194 special code in Exim that was needed for the old test suite. For example,
2195 sorting DNS records because real resolvers return them in an arbitrary
2196 order. The new test suite's fake resolver always returns records in the
2199 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2200 bounces) except when queue_only is set, to avoid logging races between
2201 the different processes.
2203 PH/38 Panic-die if .include specifies a non-absolute path.
2205 PH/39 A tweak to the "H" retry rule from its user.
2207 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2208 a label. They prevented compilation on older perls.
2210 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2211 a warning to be raised on newish perls.
2213 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2214 on queue. Changes to match documented behaviour of showing count of
2215 messages matching specified criteria.
2217 PH/40 Changed the default ident timeout from 30s to 5s.
2219 PH/41 Added support for the use of login_cap features, on those BSD systems
2220 that have them, for controlling the resources used by pipe deliveries.
2222 PH/42 The content-scanning code uses fopen() to create files in which to put
2223 message data. Previously it was not paying any attention to the mode of
2224 the files. Exim runs with umask(0) because the rest of the code creates
2225 files with open(), and sets the required mode explicitly. Thus, these
2226 files were ending up world-writeable. This was not a big issue, because,
2227 being within the spool directory, they were not world-accessible. I have
2228 created a function called modefopen, which takes an additional mode
2229 argument. It sets umask(777), creates the file, chmods it to the required
2230 mode, then resets the umask. All the relevant calls to fopen() in the
2231 content scanning code have been changed to use this function.
2233 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2234 to 24 hours. This avoids potential overflow problems when processing G
2235 and H retry rules. I suspect nobody ever tinkers with this value.
2237 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2239 PH/45 When the plaintext authenticator is running as a client, the server's
2240 challenges are checked to ensure they are valid base64 strings. By
2241 default, the authentication attempt is cancelled if an invalid string is
2242 received. Setting client_ignore_invalid_base64 true ignores these errors.
2243 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2244 they are received. Thus, the responses can be made to depend on the
2245 challenges. If an invalid string is ignored, an empty string is placed in
2248 PH/46 Messages that are created by the autoreply transport now contains a
2249 References: header, in accordance with RFCs 2822 and 3834.
2251 PH/47 Added authenticated_sender_force to the smtp transport.
2253 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2255 PH/49 Installed latest patch from the Sieve maintainer.
2257 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2258 was also set, the appendfile transport was unnecessarily scanning a
2259 directory of message files (e.g. for maildir delivery) to find the count
2260 of files (along with the size), even though it did not need this
2261 information. It now does the scan only if it needs to find either the
2262 size of the count of files.
2264 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2266 PH/52 Two bugs concerned with error handling when the smtp transport is
2269 (i) Exim was not creating retry information for temporary errors given
2270 for individual recipients after the DATA command when the smtp transport
2271 was used in LMTP mode. This meant that they could be retried too
2272 frequently, and not timed out correctly.
2274 (ii) Exim was setting the flag that allows error details to be returned
2275 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2276 recipients that were returned after the DATA command.
2278 PH/53 This is related to PH/52, but is more general: for any failing address,
2279 when detailed error information was permitted to be returned to the
2280 sender, but the error was temporary, then after the final timeout, only
2281 "retry timeout exceeded" was returned. Now it returns the full error as
2282 well as "retry timeout exceeded".
2284 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2285 do this, and (what is worse) MTAs that accept it.
2287 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2288 will now be deprecated.
2290 PH/56 New os.c-cygwin from the Cygwin maintainer.
2292 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2293 formats (previously only available in exim formats via -bpr, -bpru,
2294 and -bpra. Now also available in native and exiqgrep formats)
2296 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2297 with very large, slow to parse queues
2299 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2301 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2303 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2304 responses to authentication challenges, though it was showing the
2305 challenges; (ii) I've removed the CR characters from the debug output for
2308 PH/58 Allow for the insertion of a newline as well as a space when a string
2309 is turned into more than one encoded-word during RFC 2047 encoding. The
2310 Sieve code now uses this.
2312 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2313 data_4xx, lost_connection, tls_required.
2315 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2316 message was being sent as an SMTP response.
2318 PH/61 Add -l and -k options to exicyclog.
2320 PH/62 When verifying, if an address was redirected to one new address, so that
2321 verification continued, and the new address failed or deferred after
2322 having set something in $address_data, the value of $address_data was not
2323 passed back to the ACL. This was different to the case when no
2324 redirection occurred. The value is now passed back in both cases.
2326 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2327 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2328 use login_cap.h, so on its own it isn't the distinguishing feature. The
2329 new name refers directly to the setclassresources() function.
2331 PH/65 Added configuration files for NetBSD3.
2333 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2335 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2336 is preferred over IPv4.
2338 PH/68 The bounce_return_message and bounce_return_body options were not being
2339 honoured for bounces generated during the reception of non-SMTP messages.
2340 In particular, this applied to messages rejected by the ACL. This bug has
2341 been fixed. However, if bounce_return_message is true and bounce_return_
2342 body is false, the headers that are returned for a non-SMTP message
2343 include only those that have been read before the error was detected.
2344 (In the case of an ACL rejection, they have all been read.)
2346 PH/69 The HTML version of the specification is now built in a directory called
2347 spec_html instead of spec.html, because the latter looks like a path with
2348 a MIME-type, and this confuses some software.
2350 PH/70 Catch two compiler warnings in sieve.c.
2352 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2353 function verify_get_ident() calls ip_connect() to connect a socket, but
2354 if the "connect()" function timed out, ip_connect() used to close the
2355 socket. However, verify_get_ident() also closes the socket later, and in
2356 between Exim writes to the log, which may get opened at this point. When
2357 the socket was closed in ip_connect(), the log could get the same file
2358 descriptor number as the socket. This naturally causes chaos. The fix is
2359 not to close the socket in ip_connect(); the socket should be closed by
2360 the function that creates it. There was only one place in the code where
2361 this was missing, in the iplookup router, which I don't think anybody now
2362 uses, but I've fixed it anyway.
2364 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2365 well as to direct DNS lookups. Otherwise the handling of names in host
2366 lists is inconsistent and therefore confusing.
2372 PH/01 Two changes to the default runtime configuration:
2374 (1) Move the checks for relay_from_hosts and authenticated clients from
2375 after to before the (commented out) DNS black list checks.
2377 (2) Add control=submission to the relay_from_hosts and authenticated
2378 clients checks, on the grounds that messages accepted by these
2379 statements are most likely to be submissions.
2381 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2383 (1) Generate an error if the third argument for the ${prvs expansion is
2386 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2389 (3) Reset the variables that are obtained from the first argument of
2390 ${prvscheck and used in the second argument before leaving the code,
2391 because their memory is reclaimed, so using them afterwards may do
2394 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2395 one (it's much easier than Tom thought :-).
2397 (5) Because of (4), we can now allow for the use of $prvscheck_result
2398 inside the third argument.
2400 PH/03 For some reason, the default setting of PATH when running a command from
2401 a pipe transport was just "/usr/bin". I have changed it to
2404 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2405 anything to be listed in the output from -bV.
2407 PH/05 When a filter generated an autoreply, the entire To: header line was
2408 quoted in the delivery log line, like this:
2410 => >A.N.Other <ano@some.domain> <original@ddress> ...
2412 This has been changed so that it extracts the operative address. There
2413 may be more than one such address. If so, they are comma-separated, like
2416 => >ano@some.domain,ona@other.domain <original@ddress> ...
2418 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2419 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2420 not being looked up in the rDNS to get a host name, Exim was showing the
2421 IP address twice in Received: lines, even though the IP addresses were
2422 identical. For example:
2424 Received: from [1.2.3.4] (helo=[1.2.3.4])
2426 However, if the real host name was known, it was omitting the HELO data
2427 if it matched the actual IP address. This has been tidied up so that it
2428 doesn't show the same IP address twice.
2430 PH/07 When both +timestamp and +memory debugging was on, the value given by
2431 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2432 called by the debug printing, thereby overwriting the timestamp buffer.
2433 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2435 PH/08 When the original message was included in an autoreply transport, it
2436 always said "this is a copy of the message, including all the headers",
2437 even if body_only or headers_only was set. It now gives an appropriate
2440 PH/09 Applied a patch from the Sieve maintainer which:
2442 o fixes some comments
2443 o adds the (disabled) notify extension core
2444 o adds some debug output for the result of if/elsif tests
2445 o points to the current vacation draft in the documentation
2446 and documents the missing references header update
2450 o fixes a bug in processing the envelope test (when testing
2451 multiple envelope elements, the last element determinted the
2454 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2455 Electronic Mail") by including:
2457 Auto-submitted: auto-generated
2459 in the messages that it generates (bounce messages and others, such as
2460 warnings). In the case of bounce messages for non-SMTP mesages, there was
2461 also a typo: it was using "Auto_submitted" (underscore instead of
2462 hyphen). Since every message generated by Exim is necessarily in response
2463 to another message, thes have all been changed to:
2465 Auto-Submitted: auto-replied
2467 in accordance with these statements in the RFC:
2469 The auto-replied keyword:
2471 - SHOULD be used on messages sent in direct response to another
2472 message by an automatic process,
2474 - MUST NOT be used on manually-generated messages,
2476 - MAY be used on Delivery Status Notifications (DSNs) and Message
2477 Disposition Notifications (MDNs),
2479 - MUST NOT be used on messages generated by automatic or periodic
2480 processes, except for messages which are automatic responses to
2483 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2484 to the default Received: header definition.
2486 PH/12 Added log selector acl_warn_skipped (default on).
2488 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2489 variables because (a) they are in the wrong storage pool and (b) even if
2490 they were copied, it wouldn't work properly because of the caching.
2492 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2493 checking when decoding. Apparently there are clients that generate
2494 overlong encoded strings. Why am I not surprised?
2496 PH/15 If the first argument of "${if match_address" was not empty, but did not
2497 contain an "@" character, Exim crashed. Now it writes a panic log message
2498 and treats the condition as false.
2500 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2502 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2503 extension "envelope-auth". The code is finished and in agreement with
2504 other implementations, but there is no documentation so far and in fact,
2505 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2506 not changing the active code.
2508 Print executed "if" and "elsif" statements when debugging is used. This
2509 helps a great deal to understand what a filter does.
2511 Document more things not specified clearly in RFC3028. I had all this
2512 sorted out, when out of a sudden new issues came to my mind. Oops."
2514 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2517 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2518 canonical form (as documented). However, after a host name lookup from
2519 the IP address, check_host() was doing a simple string comparison with
2520 addresses acquired from the DNS when checking that the found name did
2521 have the original IP as one of its addresses. Since any found IPv6
2522 addresses are likely to be in abbreviated form, the comparison could
2523 fail. Luckily, there already exists a function for doing the comparison
2524 by converting both addresses to binary, so now that is used instead of
2525 the text comparison.
2527 PH/20 There was another similar case to PH/19, when a complete host name was
2528 given in a host list; looking up its IP address could give an abbreviated
2529 form, whereas the current host's name might or might not be abbreviated.
2530 The same fix has been applied.
2536 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2537 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2540 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2541 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2543 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2545 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2546 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2547 Also, for "Auto-Submitted", treat anything other than "no" as
2548 non-personal, in accordance with RFC 3834. (Previously it treated
2549 anything starting "auto-" as non-personal.)
2551 TF/01 The control=submission/name=... option had a problem with syntax
2552 errors if the name included a slash character. The /name= option
2553 now slurps the rest of the string, so it can include any characters
2554 but it must come last in the list of options (after /sender_retain
2557 PH/05 Some modifications to the interface to the fake nameserver for the new
2565 TK/01 Added the "success_on_redirect" address verification option. See
2566 NewStuff for rationale and an example.
2568 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2570 PH/02 Patch to exigrep to allow it to work on syslog lines.
2572 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2573 fread() to skip over the body file's header line, because in Cygwin the
2574 header line is locked and is inaccessible.
2576 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2577 co-exist for some time) to make it clear that it is the Exim ID that is
2578 referenced, not the Message-ID: header line.
2580 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2581 string_format() function, because snprintf() does not exist on all
2584 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2585 ${stat: expansion item.
2587 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2588 protocol synchronization error", to keep the pedants happy.
2590 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2591 well as for IRIX systems, when gcc is being used. See the host.c source
2594 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2596 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2599 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2600 transport and to the smtp transport in LMTP mode.
2602 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2604 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2605 run a filter in a subprocess. This could lead to confusion in subsequent
2606 lookups in the parent process. There should also be a search_tidyup() at
2607 the end of the subprocess.
2609 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2610 only if the host matched helo_try_verify_hosts, which caused the
2611 verification to occur when the EHLO/HELO command was issued. The ACL just
2612 tested the remembered result. Now, if a previous verification attempt has
2613 not happened, "verify = helo" does it there and then.
2615 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2617 TK/03 Fix log output including CR from clamd.
2619 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2620 exist provoked a memory error which could cause a segfault.
2622 PH/15 Installed PCRE 6.2
2624 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2626 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2627 of the problem. Specifically, suggested +O2 rather than +O1 for the
2630 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2632 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2633 record, the port was not correctly logged when the outgoing_port log
2634 selector was set (it logged the transort's default port).
2636 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2637 fallback_hosts, and "hosts" in the smtp transport.
2639 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2640 host errors such as "Connection refused".
2642 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2643 authentication with radiusclient 0.4.9:
2645 - Error returned from rc_read_config was caught wrongly
2646 - Username/password not passed on to radius server due to wrong length.
2648 The presumption is that some radiusclient API changes for 4.51/PH/17
2649 were not taken care of correctly. The code is still untested by me (my
2650 Linux distribution still has 0.3.2 of radiusclient), but it was
2651 contributed by a Radius user.
2653 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2654 expanding the "port" option of the smtp transport.
2656 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2657 while reading a MIME header. Thanks to Tom Hughes for a patch.
2659 PH/24 Include config.h inside local_scan.h so that configuration settings are
2662 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2663 that in an ACL for RCPT (for example), you can examine exactly what was
2666 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2667 commands, but it was not correctly comparing the address with the actual
2668 client host address. Thus, it would show the EHLO address in Received:
2669 header lines when this was not necessary.
2671 PH/27 Added the % operator to ${eval:}.
2673 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2674 it should be ignoring failures (because with -C, for example, it has lost
2675 privilege). It wasn't ignoring creation failures other than "already
2678 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2681 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2682 because an input line was too long, either on its own, or by virtue of
2683 too many continuations, the temporary file was not being removed, and the
2684 return code was incorrect.
2686 PH/31 Missing "BOOL" in function definition in filtertest.c.
2688 PH/32 Applied Sieve patches from the maintainer.
2690 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2692 PH/33 Added "verify = not_blind".
2694 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2695 Local/Makefile (with some defaults set). These are used in built scripts
2696 such as exicyclog, but they have never been used in the exim_install
2697 script (though there are many overriding facilities there). I have
2698 arranged that the exim_install script now takes note of these two
2701 PH/35 Installed configuration files for Dragonfly.
2703 PH/36 When a locally submitted message by a trusted user did not contain a
2704 From: header, and the sender address was obtained from -f or from an SMTP
2705 MAIL command, and the trusted user did not use -F to supply a sender
2706 name, $originator_name was incorrectly used when constructing a From:
2707 header. Furthermore, $originator_name was used for submission mode
2708 messages from external hosts without From: headers in a similar way,
2709 which is clearly wrong.
2711 PH/37 Added control=suppress_local_fixups.
2713 PH/38 When log_selector = +received_sender was set, and the addition of the
2714 sender made the log line's construction buffer exactly full, or one byte
2715 less than full, an overflow happened when the terminating "\n" was
2718 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2719 when the result of a list match is failure because a DNS lookup failed.
2721 PH/40 RM_COMMAND is now used in the building process.
2723 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2724 the "build-* directories that it finds.
2726 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2727 address in a domain literal was a prefix of an interface address.
2729 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2730 when verifying a sender address, unless rewrite_headers is false.
2732 PH/44 Wrote a long comment about why errors_to addresses are verified as
2733 recipients, not senders.
2735 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2736 the ratelimit ACL was added.
2738 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2740 PH/46 Added extra information about PostgreSQL errors to the error string.
2742 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2743 avoiding the need to install special zones in a real server. This is
2744 backwards compatible; if it can't find the fake resolver, it drops back.
2745 Thus, both old and new test suites can be run.
2747 TF/02 Added util/ratelimit.pl
2749 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2750 clock is set back in time.
2752 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2753 Candler <B.Candler@pobox.com>.
2755 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2756 OK for addresses that are the result of redirections.
2758 PH/48 A number of further additions for the benefit of the new test suite,
2759 including a fake gethostbyname() that interfaces to the fake DNS resolver
2762 TF/06 The fix for widen_domains has also been applied to qualify_single and
2763 search_parents which are the other dnslookup options that can cause
2766 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2769 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2771 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2772 service" if it encountered an MX record with an empty target hostname.
2773 The message is now "an MX or SRV record indicated no SMTP service".
2775 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2776 if the DNS of the sending site is misconfigured. This is quite a
2777 common situation. This change restores the behaviour of treating a
2778 helo verification defer as a failure.
2780 PH/51 If self=fail was set on a router, the bounce message did not include the
2781 actual error message.
2787 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2789 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2790 command itself ended in error, the underlying message about the transport
2791 filter timeout was being overwritten with the pipe command error. Now the
2792 underlying error message should be appended to the second error message.
2794 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2796 PH/02 Reduce the amount of output that "make" produces by default. Full output
2797 can still be requested.
2799 PH/03 The warning log line about a condition test deferring for a "warn" verb
2800 was being output only once per connection, rather than after each
2801 occurrence (because it was using the same function as for successful
2802 "warn" verbs). This seems wrong, so I have changed it.
2804 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2805 it should not have, which might have caused a crash in the right
2806 circumstances, but probably never did.
2808 PH/04 Installed a modified version of Tony Finch's patch to make submission
2809 mode fix the return path as well as the Sender: header line, and to
2810 add a /name= option so that you can make the user's friendly name appear
2813 TF/03 Added the control = fakedefer ACL modifier.
2815 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2816 Mark Lowes for thorough testing.
2818 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2820 TK/03 Merged latest SRS patch from Miles Wilton.
2822 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2823 with the definition in sysexits.h (which is #included earlier).
2824 Fortunately, Exim does not actually use EX_OK. The code used to try to
2825 preserve the sysexits.h value, by assumimg that macro definitions were
2826 scanned for macro replacements. I have been disabused of this notion,
2827 so now the code just undefines EX_OK before #including unistd.h.
2829 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2830 in the smtp transport. When a block could not be written in a single
2831 write() function, the timeout was being re-applied to each part-write.
2832 This seems wrong - if the receiver was accepting one byte at a time it
2833 would take for ever. The timeout is now adjusted when this happens. It
2834 doesn't have to be particularly precise.
2836 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2837 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2839 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2840 without a domain if the check to <postmaster@domain> fails.
2842 SC/01 Eximstats: added -xls and the ability to specify output files
2843 (patch written by Frank Heydlauf).
2845 SC/02 Eximstats: use FileHandles for outputing results.
2847 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2849 SC/04 Eximstats: fixed display of large numbers with -nvr option
2851 SC/05 Eximstats: fixed merging of reports with empty tables.
2853 SC/06 Eximstats: added the -include_original_destination flag
2855 SC/07 Eximstats: removed tabs and trailing whitespace.
2857 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2859 TK/06 MBOX spool code: Add real "From " MBOX separator line
2860 so the .eml file is really in mbox format (even though
2861 most programs do not really care). Patch from Alex Miller.
2863 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2864 The latter is generated from $received_to and is only set if the
2865 message has one envelope recipient. SA can use these headers,
2866 obviously out-of-the-box. Patch from Alex Miller.
2868 PH/08 The ${def test on a variable was returning false if the variable's
2869 value was "0", contrary to what the specification has always said!
2870 The result should be true unless the variable is empty.
2872 PH/09 The syntax error of a character other than { following "${if
2873 def:variable_name" (after optional whitespace) was not being diagnosed.
2874 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2875 accidental colon was present, for example, could give incorrect results.
2877 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2878 result is used (not including appendfile, where other changes are about
2881 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2882 This involved changing a lot of size variables from int to off_t. It
2883 should work with maildirs and everything.
2885 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2886 spamd dying while we are connected to it.
2888 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2891 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2892 function for BDB 4.3.
2894 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2896 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2897 the sources for PCRE have been re-arranged and more files are now
2900 PH/15 The code I had for printing potentially long long variables in PH/11
2901 above was not the best (it lost precision). The length of off_t variables
2902 is now inspected at build time, and an appropriate printing format (%ld
2903 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2904 to be "long long int" or "long int". This is needed for the internal
2905 formatting function string_vformat().
2907 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2908 the configuration file to be ":syslog", then the script "guesses" where
2909 the logs files are, rather than using the compiled in default. In our
2910 case the guess is not the same as the compiled default, so the script
2911 suddenly stopped working when I started to use syslog. The patch checks
2912 to see if log_file_path is "". If so, it attempts to read it from exim
2913 with no configuration file to get the compiled in version, before it
2914 falls back to the previous guessing code."
2916 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2917 implementing BATV in an Exim configuration. See NewStuff for the gory
2920 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2921 Makefile that are specific to HP-UX.
2923 PH/18 If the "use_postmaster" option was set for a recipient callout together
2924 with the "random" option, the postmaster address was used as the MAIL
2925 FROM address for the random test, but not for the subsequent recipient
2926 test. It is now used for both.
2928 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2929 patch removes a few documentation additions to RFC 3028, because the
2930 latest draft now contains them. It adds the new en;ascii-case comparator
2931 and a new error check for 8bit text in MIME parts. Comparator and
2932 require names are now matched exactly. I enabled the subaddress
2933 extension, but it is not well tested yet (read: it works for me)."
2935 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2936 rework some of the code of TK/09 above to avoid the hardwired use of
2937 "%lld" and "long long". Replaced the call to snprintf() with a call to
2940 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2941 records point to non-existent hosts", "retry timeout exceeded", and
2942 "retry time not reached for any host after a long failure period".
2944 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2945 experimental DomainKeys support:
2947 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2948 (2) On an error such as an illegally used "control", the wrong name for
2949 the control was given.
2951 These problems did NOT occur unless DomainKeys support was compiled.
2953 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2955 PH/24 Added ${if match_ip condition.
2957 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2958 databases so that it will be absolutely obvious if a crash occurs in the
2959 DB library. This is a regular occurrence (often caused by mis-matched
2962 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2963 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2964 picked up on a user's system that detects such things. There doesn't seem
2965 to be a gcc warning option for this - only an attribute that has to be
2966 put on the function's prototype. It seems that in Fedora Core 4 they have
2967 set this on a number of new functions. No doubt there will be more in due
2970 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2971 specify a transport. However, if an address that was verified by such a
2972 router was the subject of a callout, Exim crashed because it tried to
2973 read the rcpt_include_affixes from the non-existent transport. Now it
2974 just assumes that the setting of that option is false. This bug was
2975 introduced by 4.51/PH/31.
2977 PH/28 Changed -d+all to exclude +memory, because that information is very
2978 rarely of interest, but it makes the output a lot bigger. People tend to
2979 do -d+all out of habit.
2981 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2982 code in os-type was giving problems when libc.so lives in lib64, like on
2985 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2986 aren't the modern standard, and it seems that some systems' include files
2987 don't always have them. Exim was already checking for some of the newer
2988 ones like T_AAAA, and defining it itself. I've added checks for all the
2989 record types that Exim uses.
2991 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2992 not automatically generating a new one, as it is supposed to. This
2993 prevented TLS from working. If the file did exist, but contained invalid
2994 data, a new version was generated, as expected. It was only the case of a
2995 non-existent file that was broken.
2997 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2998 with a change in libdomainkeys > 0.64.
3000 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3001 from DNS. If the selector record carries the flag, it now has
3002 precedence over the domain-wide flag.
3004 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3006 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3007 the use of a port name that isn't defined in /etc/services) occurred, the
3008 message was deferred as in a normal delivery, and thus remained on the
3009 spool, instead of being failed because of the mua_wrapper setting. This
3010 is now fixed, and I tidied up some of the mua_wrapper messages at the
3013 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3014 the messages in a hash of arrays rather than using individual hashes.
3015 This is a bit cleaner and results in dramatic memory savings, albeit
3016 at a slight CPU cost.
3018 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3019 as requested by Marc Sherman.
3021 SC/10 Eximstats: added histograms for user specified patterns as requested
3024 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3026 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3027 fopen() in the content-scanning modules that did not already have it.
3033 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3034 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3036 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3038 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3040 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3041 to test Sieve filters that use "vacation".
3043 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3044 that changes the way the GnuTLS parameters are stored in the cache file.
3045 The new format can be generated externally. For backward compatibility,
3046 if the data in the cache doesn't make sense, Exim assumes it has read an
3047 old-format file, and it generates new data and writes a new file. This
3048 means that you can't go back to an older release without removing the
3051 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3052 work if there are any delivery delays because "one_time" forces the
3053 parent to be marked "delivered", so its unseen clone is never tried
3054 again. For this reason, Exim now forbids the simultaneous setting of
3057 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3058 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3059 entry if you want to know the details. The fix, however, neglected to
3060 consider the case when local delivery batching is involved. The test for
3061 "previously delivered" was not happening when checking to see if an
3062 address could be batched with a previous (undelivered) one; under
3063 certain circumstances this could lead to multiple deliveries to the same
3066 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3067 in its include files, and this causes problems building Exim.
3069 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3070 header_syntax) but Exim was just ignoring anything given after a slash.
3071 In particular, this caused confusion with an attempt to use "verify =
3072 reverse_host_lookup/defer_ok". An error is now given when options are
3073 supplied for verify items that do not have them. (Maybe reverse_host_
3074 lookup should have a defer_ok option, but that's a different point.)
3076 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3077 defined by RFC 821) to 2048, because there were problems with some AUTH
3078 commands, and RFC 1869 says the size should be increased for extended
3079 SMTP commands that take arguments.
3081 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3084 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3085 "unknown" error; now it says that the functionality isn't in the binary.
3087 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3088 an address' error message when a string expansion fails (syntax or
3089 whatever). Otherwise the password may appear in the log. Following change
3090 PH/42 below, there is no longer a chance of it appearing in a bounce
3093 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3095 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3096 of its addresses was ever tried. (Bugzilla bug #2.)
3098 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3099 the result incorrectly in the debug output. (It correctly added a newline
3100 to what was transported.)
3102 TF/01 Added $received_time.
3104 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3105 commented out examples of how to interface to a virus scanner and to
3106 SpamAssassin. Also added commented examples of av_scanner and
3107 spamd_address settings.
3109 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3110 and controls are allowed in which ACLs. There were a couple of minor
3111 errors. Some of the entries in the conditions table (which is a table of
3112 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3113 negation of where the condition IS allowed.
3115 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3117 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3118 header file does not have a version number, so I've had to invent a new
3119 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3120 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3121 radiusclient), but it was contributed by a Radius user.
3123 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3124 files or directories, for OpenSSL.
3126 PH/19 When an Exim process that is running as root has to create an Exim log
3127 file, it does so in a subprocess that runs as exim:exim so as to get the
3128 ownership right at creation (otherwise, other Exim processes might see
3129 the file with the wrong ownership). There was no test for failure of this
3130 fork() call, which would lead to the process getting stuck as it waited
3131 for a non-existent subprocess. Forks do occasionally fail when resources
3132 run out. I reviewed all the other calls to fork(); they all seem to check
3135 PH/20 When checking for unexpected SMTP input at connect time (before writing
3136 the banner), Exim was not dealing correctly with a non-positive return
3137 from the read() function. If the client had disconnected by this time,
3138 the result was a log entry for a synchronization error with an empty
3139 string after "input=" when read() returned zero. If read() returned -1
3140 (an event I could not check), uninitialized data bytes were printed.
3141 There were reports of junk text (parts of files, etc) appearing after
3144 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3146 PH/22 Added support for macro redefinition, and (re)definition in between
3147 driver and ACL definitions.
3149 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3150 forgetting to use the resulting value; it was using the unexpanded value.
3152 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3153 hadn't been configured. The fix is from Juergen Kreileder, who
3154 understands it better than I do:
3156 "Here's what I see happening with three configured cyrus_sasl
3157 authenticators configured (plain, login, cram-md5):
3159 On startup auth_cyrus_sasl_init() gets called for each of these.
3160 This means three calls to sasl_listmech() without a specified mech_list.
3161 => SASL tests which mechs of all available mechs actually work
3162 => three warnings about OTP not working
3163 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3165 With the patch, sasl_listmech() also gets called three times. But now
3166 SASL's mech_list option is set to the server_mech specified in the the
3167 authenticator. Or in other words, the answer from sasl_listmech()
3168 gets limited to just the mech you're testing for (which is different
3170 => the return list contains just 'plain' or 'login', 'cram-md5' or
3171 nothing depending on the value of ob->server_mech.
3173 I've just tested the patch: Authentication still works fine,
3174 unavailable mechs specified in the exim configuration are still
3175 caught, and the auth.log warnings about OTP are gone."
3177 PH/25 When debugging is enabled, the contents of the command line are added
3178 to the debugging output, even when log_selector=+arguments is not
3181 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3182 answer is "GNU", and only if the return is "GNU/something" is the answer
3185 PH/27 $acl_verify_message is now set immediately after the failure of a
3186 verification in an ACL, and so is available in subsequent modifiers. In
3187 particular, the message can be preserved by coding like this:
3189 warn !verify = sender
3190 set acl_m0 = $acl_verify_message
3192 Previously, $acl_verify_message was set only while expanding "message"
3193 and "log_message" when a very denied access.
3195 PH/28 Modified OS/os.c-Linux with
3197 -#ifndef OS_LOAD_AVERAGE
3198 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3200 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3201 nomenclature these days.)
3203 PH/29 Installed patch from the Sieve maintainer that adds the options
3204 sieve_useraddress and sieve_subaddress to the redirect router.
3206 PH/30 In these circumstances:
3207 . Two addresses routed to the same list of hosts;
3208 . First host does not offer TLS;
3209 . First host accepts first address;
3210 . First host gives temporary error to second address;
3211 . Second host offers TLS and a TLS session is established;
3212 . Second host accepts second address.
3213 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3214 and peerdn, if requested) that were in fact used only for the second
3217 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3218 attention to any local part prefix or suffix that was matched by the
3219 router that accepted the address. It now behaves in the same way as it
3220 does for delivery: the affixes are removed from the local part unless
3221 rcpt_include_affixes is set on the transport.
3223 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3224 timeout during the DATA phase of an incoming message.
3226 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3227 applied a patch sanctioned by the Sieve maintainer.
3229 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3230 the uid or gid is negative. A case of a negative gid caused this to be
3231 noticed. The fix allows for either to be negative.
3233 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3234 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3237 PH/36 The change PH/12 above was broken. Fixed it.
3239 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3240 the grounds that routing the same address twice would always produce the
3241 same answer. This might have been true once, but it is certainly no
3242 longer true now. Routing a child address may depend on the previous
3243 routing that produced that child. Some complicated redirection strategies
3244 went wrong when messages had multiple recipients, and made Exim's
3245 behaviour dependent on the order in which the addresses were given.
3247 I have moved the duplicate checking until after the routing is complete.
3248 Exim scans the addresses that are assigned to local and remote
3249 transports, and removes any duplicates. This means that more work will be
3250 done, as duplicates will always all be routed, but duplicates are
3251 presumably rare, so I don't expect this is of any significance.
3253 For deliveries to pipes, files, and autoreplies, the duplicate checking
3254 still happens during the routing process, since they are not going to be
3257 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3258 It corrects a timeout issue with spamd. This is Ian's comment: "The
3259 background is that sometimes spamd either never reads data from a
3260 connection it has accepted, or it never writes response data. The exiscan
3261 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3262 blindly assumes that writes won't block so it may never time out."
3264 PH/39 Allow G after quota size as well as K and M.
3266 PH/40 The value set for $authenticated_id in an authenticator may not contain
3267 binary zeroes or newlines because the value is written to log lines and
3268 to spool files. There was no check on this. Now the value is run through
3269 the string_printing() function so that such characters are converted to
3270 printable escape sequences.
3272 PH/41 $message_linecount is a new variable that contains the total number of
3273 lines in the message. Compare $body_linecount, which is the count for the
3276 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3277 bounce and delay warning messages, except in certain special cases, which
3280 (a) An SMTP error message from a remote host;
3281 (b) A message specified in a :fail: redirection;
3282 (c) A message specified in a "fail" command in a system filter;
3283 (d) A message specified in a FAIL return from the queryprogram router;
3284 (e) A message specified by the cannot_route_message router option.
3286 In these cases only, Exim does include the error details in bounce and
3287 warning messages. There are also a few cases where bland messages such
3288 as "unrouteable address" or "local delivery error" are given.
3290 PH/43 $value is now also set for the "else" part of a ${run expansion.
3292 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3293 being worked on, but at least Exim now implements the latest version to
3296 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3297 process to complete was treated as a delivery failure, a timeout while
3298 writing the message to the pipe was logged, but erroneously treated as a
3299 successful delivery. Such timeouts include transport filter timeouts. For
3300 consistency with the overall process timeout, these timeouts are now
3301 treated as errors, giving rise to delivery failures by default. However,
3302 there is now a new Boolean option for the pipe transport called
3303 timeout_defer, which, if set TRUE, converts the failures into defers for
3304 both kinds of timeout. A transport filter timeout is now identified in
3307 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3308 systems where "make" and "gmake" are different, calling "gmake" at top
3309 level broke things. I've arranged for the value of $(MAKE) to be passed
3310 from the Makefile to this script so that it can call the same version of
3314 A note about Exim versions 4.44 and 4.50
3315 ----------------------------------------
3317 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3318 changes of various kinds. As a consequence, a big documentation update was
3319 needed. This delayed the release for rather longer than seemed good, especially
3320 in the light of a couple of (minor) security issues. Therefore, the changes
3321 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3322 release. So 4.44 and 4.50 are in effect two different branches that both start
3325 I have left the 4.50 change log unchanged; it contains all the changes since
3326 4.43. The change log for 4.44 is below; many of its items are identical to
3327 those for 4.50. This seems to be the most sensible way to preserve the
3328 historical information.
3334 1. Minor wording change to the doc/README.SIEVE file.
3336 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3337 computation of the current number of files was incorrect.
3339 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3340 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3343 4. Give more explanation in the error message when the command for a transport
3344 filter fails to execute.
3346 5. There are several places where Exim runs a non-Exim command in a
3347 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3348 was being done only for the command run by the queryprogram router. It is
3349 now done for all such subprocesses. The other cases are: ${run, transport
3350 filters, and the commands run by the lmtp and pipe transports.
3352 6. Added CONFIGURE_GROUP build-time option.
3354 7. Some older OS have a limit of 256 on the maximum number of file
3355 descriptors. Exim was using setrlimit() to set 1000 as a large value
3356 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3357 systems. I've change it so that if it can't get 1000, it tries for 256.
3359 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3360 was an oversight, and furthermore, ever since the addition of extra
3361 controls (e.g. 4.43/32), the checks on when to allow different forms of
3362 "control" were broken. There should now be diagnostics for all cases when a
3363 control that does not make sense is encountered.
3365 9. Added the /retain_sender option to "control=submission".
3367 10. $recipients is now available in the predata ACL (oversight).
3369 11. Tidy the search cache before the fork to do a delivery from a message
3370 received from the command line. Otherwise the child will trigger a lookup
3371 failure and thereby defer the delivery if it tries to use (for example) a
3372 cached ldap connection that the parent has called unbind on.
3374 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3375 of $address_data from the recipient verification was clobbered by the
3376 sender verification.
3378 13. The value of address_data from a sender verification is now available in
3379 $sender_address_data in subsequent conditions in the ACL statement.
3381 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3383 15. Added a new option "connect=<time>" to callout options, to set a different
3386 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3387 was its contents. (It was OK if the option was not defined at all.)
3389 17. A "Completed" log line is now written for messages that are removed from
3390 the spool by the -Mrm option.
3392 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3393 information about exactly what failed.
3395 19. Added -dd to debug only the daemon process.
3397 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3398 handles timeouts, both on the server side and network timeouts. Renamed the
3399 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3401 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3402 It is now set to "smtps".
3404 22. $host_address is now set to the target address during the checking of
3405 ignore_target_hosts.
3407 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3408 being passed; this would have caused $sender_host_name to have been used if
3409 matching the list had actually called for a host name (not very likely,
3410 since this list is usually IP addresses). A host name is now passed as
3413 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3414 code that specifies a non-restarting handler (typically sigaction() in
3415 modern systems) in an attempt to fix a rare and obscure crash bug.
3417 25. Narrowed the window for a race in the daemon that could cause it to ignore
3418 SIGCHLD signals. This is not a major problem, because they are used only to
3419 wake it up if nothing else does.
3421 26. A malformed maildirsize file could cause Exim to calculate negative values
3422 for the mailbox size or file count. Odd effects could occur as a result.
3423 The maildirsize information is now recalculated if the size or filecount
3426 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3427 support for a long time. Removed HAVE_SYS_VFS_H.
3429 28. Installed the latest version of exipick from John Jetmore.
3431 29. In an address list, if the pattern was not a regular expression, an empty
3432 subject address (from a bounce message) matched only if the pattern was an
3433 empty string. Non-empty patterns were not even tested. This was the wrong
3434 because it is perfectly reasonable to use an empty address as part of a
3435 database query. An empty address is now tested by patterns that are
3436 lookups. However, all the other forms of pattern expect the subject to
3437 contain a local part and a domain, and therefore, for them, an empty
3438 address still always fails if the pattern is not itself empty.
3440 30. Exim went into a mad DNS loop when attempting to do a callout where the
3441 host was specified on an smtp transport, and looking it up yielded more
3442 than one IP address.
3444 31. Re-factored the code for checking spool and log partition space into a
3445 function that finds that data and another that does the check. The former
3446 is then used to implement four new variables: $spool_space, $log_space,
3447 $spool_inodes, and $log_inodes.
3449 32. The RFC2047 encoding function was originally intended for short strings
3450 such as real names; it was not keeping to the 75-character limit for
3451 encoded words that the RFC imposes. It now respects the limit, and
3452 generates multiple encoded words if necessary. To be on the safe side, I
3453 have increased the buffer size for the ${rfc2047: expansion operator from
3456 33. It is now permitted to omit both strings after an "if" condition; if the
3457 condition is true, the result is "true". As before, when the second string
3458 is omitted, a false condition yields an empty string. This makes it less
3459 cumbersome to write custom ACL and router conditions.
3461 34. Failure to deliver a bounce message always caused it to be frozen, even if
3462 there was an errors_to setting on the router. The errors_to setting is now
3465 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3466 canonical form (fully expanded) before being placed in
3467 $sender_host_address.
3469 36. The table in the code that translates DNS record types into text (T_A to
3470 "A" for instance) was missing entries for NS and CNAME. It is just possible
3471 that this could have caused confusion if both these types were looked up
3472 for the same domain, because the text type is used as part of Exim's
3473 per-process caching. But the chance of anyone hitting this buglet seems
3476 37. The dnsdb lookup has been extended in a number of ways.
3478 (1) There is a new type, "zns", which walks up the domain tree until it
3479 finds some nameserver records. It should be used with care.
3481 (2) There is a new type, "mxh", which is like "mx" except that it returns
3482 just the host names, not the priorities.
3484 (3) It is now possible to give a list of domains (or IP addresses) to be
3485 looked up. The behaviour when one of the lookups defers can be
3486 controlled by a keyword.
3488 (4) It is now possible to specify the separator character for use when
3489 multiple records are returned.
3491 38. The dnslists ACL condition has been extended: it is now possible to supply
3492 a list of IP addresses and/or domains to be looked up in a particular DNS
3495 39. Added log_selector=+queue_time_overall.
3497 40. When running the queue in the test harness, wait just a tad after forking a
3498 delivery process, to get repeatability of debugging output.
3500 41. Include certificate and key file names in error message when GnuTLS fails
3501 to set them up, because the GnuTLS error message doesn't include the name
3502 of the failing file when there is a problem reading it.
3504 42. Allow both -bf and -bF in the same test run.
3506 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3508 44. The "Exiscan patch" is now merged into the mainline Exim source.
3510 45. Sometimes the final signoff response after QUIT could fail to get
3511 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3512 < 0 before doing a fflush(). This bug looks as though it goes back to the
3513 introduction of TLS in release 3.20, but "sometimes" must have been rare
3514 because the tests only now provoked it.
3516 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3517 (this can affect the format of dates).
3519 47. exim_tidydb, when checking for the continued existence of a message for
3520 which it has found a message-specific retry record, was not finding
3521 messages that were in split spool directories. Consequently, it was
3522 deleting retry records that should have stayed in existence.
3524 48. Steve fixed some bugs in eximstats.
3526 49. The SPA authentication driver was not abandoning authentication and moving
3527 on to the next authenticator when an expansion was forced to fail,
3528 contradicting the general specification for all authenticators. Instead it
3529 was generating a temporary error. It now behaves as specified.
3531 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3532 (the order specifies the preference for clients). The order is now AES256,
3533 AES128, 3DES, ARCFOUR128.
3535 51. Small patch to Sieve code - explicitly set From: when generating an
3538 52. Exim crashed if a remote delivery caused a very long error message to be
3539 recorded - for instance if somebody sent an entire SpamAssassin report back
3540 as a large number of 550 error lines. This bug was coincidentally fixed by
3541 increasing the size of one of Exim's internal buffers (big_buffer) that
3542 happened as part of the Exiscan merge. However, to be on the safe side, I
3543 have made the code more robust (and fixed the comments that describe what
3546 53. Now that there can be additional text after "Completed" in log lines (if
3547 the queue_time_overall log selector is set), a one-byte patch to exigrep
3548 was needed to allow it to recognize "Completed" as not the last thing in
3551 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3552 patch that reportedly fixes this has been added. I am not expert enough to
3553 create a test for it. This is what the patch creator wrote:
3555 "I found a little strange behaviour of ldap code when working with
3556 Windows 2003 AD Domain, where users was placed in more than one
3557 Organization Units. When I tried to give exim partial DN, the exit code
3558 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3559 But simultaneously result of request was absolutely normal ldap result,
3560 so I produce this patch..."
3562 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3563 REFERENCE, so I have modified the code to exclude the patch when that macro
3566 55. Some experimental protocols are using DNS PTR records for new purposes. The
3567 keys for these records are domain names, not reversed IP addresses. The
3568 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3569 leaves it alone. Component reversal etc. now happens only for IP addresses.
3572 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3574 57. Double the size of the debug message buffer (to 2048) so that more of very
3575 long debug lines gets shown.
3577 58. The exicyclog utility now does better if the number of log files to keep
3578 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3580 59. Two changes related to the smtp_active_hostname option:
3582 (1) $smtp_active_hostname is now available as a variable.
3583 (2) The default for smtp_banner uses $smtp_active_hostname instead
3584 of $primary_hostname.
3586 60. The host_aton() function is supposed to be passed a string that is known
3587 to be a valid IP address. However, in the case of IPv6 addresses, it was
3588 not checking this. This is a hostage to fortune. Exim now panics and dies
3589 if the condition is not met. A case was found where this could be provoked
3590 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3591 components; fortuitously, this particular loophole had already been fixed
3592 by change 4.50/55 above.
3594 If there are any other similar loopholes, the new check in host_aton()
3595 itself should stop them being exploited. The report I received stated that
3596 data on the command line could provoke the exploit when Exim was running as
3597 exim, but did not say which command line option was involved. All I could
3598 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3599 running as the user.
3602 61. There was a buffer overflow vulnerability in the SPA authentication code
3603 (which came originally from the Samba project). I have added a test to the
3604 spa_base64_to_bits() function which I hope fixes it.
3607 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3608 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3610 63. The daemon start-up calls getloadavg() while still root for those OS that
3611 need the first call to be done as root, but it missed one case: when
3612 deliver_queue_load_max is set with deliver_drop_privilege. This is
3613 necessary for the benefit of the queue runner, because there is no re-exec
3614 when deliver_drop_privilege is set.
3616 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3617 This has been fixed.
3619 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3620 was in use, was not putting the data itself into the right store pool;
3621 consequently, it could be overwritten for a subsequent message in the same
3622 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3625 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3627 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3628 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3629 the function were treating the return as a boolean value, which happened to
3630 work because 0=false and not-0=true, but is not correct code.
3632 68. The host_aton() function was not handling scoped IPv6 addresses (those
3633 with, for example, "%eth0" on the end) correctly.
3635 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3636 negated items (that is, ~something) in unsigned ints. Some compilers
3637 apparently mutter when there is no cast.
3639 70. If an address verification called from an ACL failed, and did not produce a
3640 user-specific message (i.e. there was only a "system" message), nothing was
3641 put in $acl_verify_message. In this situation, it now puts the system
3644 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3645 session; change 4.31/43 added the unwanted input to the log line - except
3646 that it did not do this in the start of session case. It now does.
3648 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3649 This is wrong and can cause the other end to generate a synchronization
3650 error if it is another Exim or anything else that does the synchronization
3651 check. A QUIT command is no longer sent after a timeout.
3653 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3654 during host lookups.
3656 74. The defer_ok option of callout verification was not working if it was used
3657 when verifying addresses in header lines, that is, for this case:
3659 verify = header_sender/callout=defer_ok
3661 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3662 those file descriptors could be used for SMTP connections. If anything
3663 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3664 could be sent to the SMTP client, causing chaos. The daemon now opens
3665 stdin, stdout, and stderr to /dev/null when it puts itself into the
3668 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3669 log by default. The user can override this with suitable Perl magic.
3671 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3672 the log message when discard triggers, was not working for the DATA ACL or
3673 for the non-SMTP ACL.
3675 78. Error message wording change in sieve.c.
3677 79. If smtp_accept_max_per_host was set, the number of connections could be
3678 restricted to fewer than expected, because the daemon was trying to set up
3679 a new connection before checking whether the processes handling previous
3680 connections had finished. The check for completed processes is now done
3681 earlier. On busy systems, this bug wouldn't be noticed because something
3682 else would have woken the daemon, and it would have reaped the completed
3685 80. If a message was submitted locally by a user whose login name contained one
3686 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3687 It caused a spool format error. I have fixed the spool reading code. A
3688 related problem was that the "from" clause in the Received: line became
3689 illegal because of the space(s). It is now covered by ${quote_local_part.
3691 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3694 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3696 83. Patch from Sieve maintainer to fix unterminated string problem in
3697 "vacation" handling.
3699 84. Some minor changes to the Linux configuration files to help with other
3700 OS variants using glibc.
3702 85. One more patch for Sieve to update vacation handling to latest spec.
3705 ----------------------------------------------------
3706 See the note above about the 4.44 and 4.50 releases.
3707 ----------------------------------------------------
3713 1. Change 4.43/35 introduced a bug that caused file counts to be
3714 incorrectly computed when quota_filecount was set in an appendfile
3717 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3718 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3721 3. Give more explanation in the error message when the command for a transport
3722 filter fails to execute.
3724 4. There are several places where Exim runs a non-Exim command in a
3725 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3726 was being done only for the command run by the queryprogram router. It is
3727 now done for all such subprocesses. The other cases are: ${run, transport
3728 filters, and the commands run by the lmtp and pipe transports.
3730 5. Some older OS have a limit of 256 on the maximum number of file
3731 descriptors. Exim was using setrlimit() to set 1000 as a large value
3732 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3733 systems. I've change it so that if it can't get 1000, it tries for 256.
3735 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3736 was an oversight, and furthermore, ever since the addition of extra
3737 controls (e.g. 4.43/32), the checks on when to allow different forms of
3738 "control" were broken. There should now be diagnostics for all cases when a
3739 control that does not make sense is encountered.
3741 7. $recipients is now available in the predata ACL (oversight).
3743 8. Tidy the search cache before the fork to do a delivery from a message
3744 received from the command line. Otherwise the child will trigger a lookup
3745 failure and thereby defer the delivery if it tries to use (for example) a
3746 cached ldap connection that the parent has called unbind on.
3748 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3749 of $address_data from the recipient verification was clobbered by the
3750 sender verification.
3752 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3753 was its contents. (It was OK if the option was not defined at all.)
3755 11. A "Completed" log line is now written for messages that are removed from
3756 the spool by the -Mrm option.
3758 12. $host_address is now set to the target address during the checking of
3759 ignore_target_hosts.
3761 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3762 being passed; this would have caused $sender_host_name to have been used if
3763 matching the list had actually called for a host name (not very likely,
3764 since this list is usually IP addresses). A host name is now passed as
3767 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3768 code that specifies a non-restarting handler (typically sigaction() in
3769 modern systems) in an attempt to fix a rare and obscure crash bug.
3771 15. Narrowed the window for a race in the daemon that could cause it to ignore
3772 SIGCHLD signals. This is not a major problem, because they are used only to
3773 wake it up if nothing else does.
3775 16. A malformed maildirsize file could cause Exim to calculate negative values
3776 for the mailbox size or file count. Odd effects could occur as a result.
3777 The maildirsize information is now recalculated if the size or filecount
3780 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3781 support for a long time. Removed HAVE_SYS_VFS_H.
3783 18. Updated exipick to current release from John Jetmore.
3785 19. Allow an empty sender to be matched against a lookup in an address list.
3786 Previously the only cases considered were a regular expression, or an
3789 20. Exim went into a mad DNS lookup loop when doing a callout where the
3790 host was specified on the transport, if the DNS lookup yielded more than
3793 21. The RFC2047 encoding function was originally intended for short strings
3794 such as real names; it was not keeping to the 75-character limit for
3795 encoded words that the RFC imposes. It now respects the limit, and
3796 generates multiple encoded words if necessary. To be on the safe side, I
3797 have increased the buffer size for the ${rfc2047: expansion operator from
3800 22. Failure to deliver a bounce message always caused it to be frozen, even if
3801 there was an errors_to setting on the router. The errors_to setting is now
3804 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3805 canonical form (fully expanded) before being placed in
3806 $sender_host_address.
3808 24. Updated eximstats to version 1.33
3810 25. Include certificate and key file names in error message when GnuTLS fails
3811 to set them up, because the GnuTLS error message doesn't include the name
3812 of the failing file when there is a problem reading it.
3814 26. Expand error message when OpenSSL has problems setting up cert/key files.
3817 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3818 (this can affect the format of dates).
3820 28. exim_tidydb, when checking for the continued existence of a message for
3821 which it has found a message-specific retry record, was not finding
3822 messages that were in split spool directories. Consequently, it was
3823 deleting retry records that should have stayed in existence.
3825 29. eximstats updated to version 1.35
3826 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3827 1.35 - bugfix such that pie charts by volume are generated correctly
3829 30. The SPA authentication driver was not abandoning authentication and moving
3830 on to the next authenticator when an expansion was forced to fail,
3831 contradicting the general specification for all authenticators. Instead it
3832 was generating a temporary error. It now behaves as specified.
3834 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3835 (the order specifies the preference for clients). The order is now AES256,
3836 AES128, 3DES, ARCFOUR128.
3838 31. Small patch to Sieve code - explicitly set From: when generating an
3841 32. Exim crashed if a remote delivery caused a very long error message to be
3842 recorded - for instance if somebody sent an entire SpamAssassin report back
3843 as a large number of 550 error lines. This bug was coincidentally fixed by
3844 increasing the size of one of Exim's internal buffers (big_buffer) that
3845 happened as part of the Exiscan merge. However, to be on the safe side, I
3846 have made the code more robust (and fixed the comments that describe what
3849 33. Some experimental protocols are using DNS PTR records for new purposes. The
3850 keys for these records are domain names, not reversed IP addresses. The
3851 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3852 leaves it alone. Component reversal etc. now happens only for IP addresses.
3855 34. The host_aton() function is supposed to be passed a string that is known
3856 to be a valid IP address. However, in the case of IPv6 addresses, it was
3857 not checking this. This is a hostage to fortune. Exim now panics and dies
3858 if the condition is not met. A case was found where this could be provoked
3859 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3860 components; fortuitously, this particular loophole had already been fixed
3861 by change 4.50/55 or 4.44/33 above.
3863 If there are any other similar loopholes, the new check in host_aton()
3864 itself should stop them being exploited. The report I received stated that
3865 data on the command line could provoke the exploit when Exim was running as
3866 exim, but did not say which command line option was involved. All I could
3867 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3868 running as the user.
3871 35. There was a buffer overflow vulnerability in the SPA authentication code
3872 (which came originally from the Samba project). I have added a test to the
3873 spa_base64_to_bits() function which I hope fixes it.
3876 36. The daemon start-up calls getloadavg() while still root for those OS that
3877 need the first call to be done as root, but it missed one case: when
3878 deliver_queue_load_max is set with deliver_drop_privilege. This is
3879 necessary for the benefit of the queue runner, because there is no re-exec
3880 when deliver_drop_privilege is set.
3882 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3883 was in use, was not putting the data itself into the right store pool;
3884 consequently, it could be overwritten for a subsequent message in the same
3885 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3888 38. Sometimes the final signoff response after QUIT could fail to get
3889 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3890 < 0 before doing a fflush(). This bug looks as though it goes back to the
3891 introduction of TLS in release 3.20, but "sometimes" must have been rare
3892 because the tests only now provoked it.
3898 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3899 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3900 Now it returns an integer. A number of calls were still expecting a T/F
3901 return. Fortuitously, in all cases, the tests worked in OK situations,
3902 which is the norm. However, things would have gone wrong on any write
3903 failures on the smtp file descriptor. This function is used when sending
3904 messages over SMTP and also when doing verify callouts.
3906 2. When Exim is called to do synchronous delivery of a locally submitted
3907 message (the -odf or -odi options), it no longer closes stderr before doing
3910 3. Implemented the mua_wrapper option.
3912 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3914 5. Implemented the functions header_remove(), header_testname(),
3915 header_add_at_position(), and receive_remove_recipient(), and exported them
3918 6. If an ACL "warn" statement specified the addition of headers, Exim already
3919 inserted X-ACL-Warn: at the start if there was no header name. However, it
3920 was not making this test for the second and subsequent header lines if
3921 there were newlines in the string. This meant that an invalid header could
3922 be inserted if Exim was badly configured.
3924 7. Allow an ACL "warn" statement to add header lines at the start or after all
3925 the Received: headers, as well as at the end.
3927 8. Added the rcpt_4xx retry error code.
3929 9. Added postmaster_mailfrom=xxx to callout verification option.
3931 10. Added mailfrom=xxxx to the callout verification option, for verify=
3934 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3935 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3937 12. Inserted some casts to stop certain compilers warning when using pointer
3938 differences as field lengths or precisions in printf-type calls (mostly
3939 affecting debugging statements).
3941 13. Added optional readline() support for -be (dynamically loaded).
3943 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3944 same clock tick as a message's arrival, so that its received time was the
3945 same as the "first fail" time on the retry record, and that message
3946 remained on the queue past the ultimate address timeout, every queue runner
3947 would try a delivery (because it was past the ultimate address timeout) but
3948 after another failure, the ultimate address timeout, which should have then
3949 bounced the address, did not kick in. This was a "< instead of <=" error;
3950 in most cases the first failure would have been in the next clock tick
3951 after the received time, and all would be well.
3953 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3954 being recognized when the domain list was tested by the match_domain
3955 condition in an expansion string.
3957 16. Added the ${str2b64: operator.
3959 17. Exim was always calling setrlimit() to set a large limit for the number of
3960 processes, without checking whether the existing limit was already
3961 adequate. (It did check for the limit on file descriptors.) Furthermore,
3962 errors from getrlimit() and setrlimit() were being ignored. Now they are
3963 logged to the main and panic logs, but Exim does carry on, to try to do its
3964 job under whatever limits there are.
3966 18. Imported PCRE 5.0.
3968 19. Trivial typo in log message " temporarily refused connection" (the leading
3971 20. If the log selector return_path_on_delivery was set and an address was
3972 redirected to /dev/null, the delivery process crashed because it assumed
3973 that a return path would always be set for a "successful" delivery. In this
3974 case, the whole delivery is bypassed as an optimization, and therefore no
3977 21. Internal re-arrangement: the function for sending a challenge and reading
3978 a response while authentication was assuming a zero-terminated challenge
3979 string. It's now changed to take a pointer and a length, to allow for
3980 binary data in such strings.
3982 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3984 23. Exim was not respecting finduser_retries when seeking the login of the
3985 uid under which it was called; it was always trying 10 times. (The default
3986 setting of finduser_retries is zero.) Also, it was sleeping after the final
3987 failure, which is pointless.
3989 24. Implemented tls_on_connect_ports.
3991 25. Implemented acl_smtp_predata.
3993 26. If the domain in control=submission is set empty, Exim assumes that the
3994 authenticated id is a complete email address when it generates From: or
3995 Sender: header lines.
3997 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3998 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3999 chown and chgrp in /bin and hostname in /usr/bin.
4001 28. Exim was keeping the "process log" file open after each use, just as it
4002 does for the main log. This opens the possibility of it remaining open for
4003 long periods when the USR1 signal hits a daemon. Occasional processlog
4004 errors were reported, that could have been caused by this. Anyway, it seems
4005 much more sensible not to leave this file open at all, so that is what now
4008 29. The long-running daemon process does not normally write to the log once it
4009 has entered its main loop, and it closes the log before doing so. This is
4010 so that log files can straightforwardly be renamed and moved. However,
4011 there are a couple of unusual error situations where the daemon does write
4012 log entries, and I had neglected to close the log afterwards.
4014 30. The text of an SMTP error response that was received during a remote
4015 delivery was being truncated at 512 bytes. This is too short for some of
4016 the long messages that one sometimes sees. I've increased the limit to
4019 31. It is now possible to make retry rules that apply only when a message has a
4020 specific sender, in particular, an empty sender.
4022 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4023 it possible to be selective about when SMTP synchronization is enforced.
4025 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4027 32. Implemented hosts_connection_nolog.
4029 33. Added an ACL for QUIT.
4031 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4034 35. Added mailbox_size and mailbox_filecount to appendfile.
4036 36. Added control = no_multiline_responses to ACLs.
4038 37. There was a bug in the logic of the code that waits for the clock to tick
4039 in the case where the clock went backwards by a substantial amount such
4040 that the microsecond fraction of "now" was more than the microsecond
4041 fraction of "then" (but the whole seconds number was less).
4043 38. Added support for the libradius Radius client library this is found on
4044 FreeBSD (previously only the radiusclient library was supported).
4050 1. When certain lookups returned multiple values in the form name=value, the
4051 quoting of the values was not always being done properly. Specifically:
4052 (a) If the value started with a double quote, but contained no whitespace,
4054 (b) If the value contained whitespace other than a space character (i.e.
4055 tabs or newlines or carriage returns) it was not quoted.
4056 This fix has been applied to the mysql and pgsql lookups by writing a
4057 separate quoting function and calling it from the lookup code. The fix
4058 should probably also be applied to nisplus, ibase and oracle lookups, but
4059 since I cannot test any of those, I have not disturbed their existing code.
4061 2. A hit in the callout cache for a specific address caused a log line with no
4062 reason for rejecting RCPT. Now it says "Previous (cached) callout
4063 verification failure".
4065 3. There was an off-by-one bug in the queryprogram router. An over-long
4066 return line was truncated at 256 instead of 255 characters, thereby
4067 overflowing its buffer with the terminating zero. As well as fixing this, I
4068 have increased the buffer size to 1024 (and made a note to document this).
4070 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4071 when Exim is waiting for an SMTP response from a remote server, Exim
4072 restarts its select() call on the socket, thereby resetting its timeout.
4073 This is not a problem when such interrupts are rare. Somebody set up a cron
4074 job to run exiwhat every 2 minutes, which is less than the normal select()
4075 timeout (5 or 10 minutes). This meant that the select() timeout never
4076 kicked in because it was always reset. I have fixed this by comparing the
4077 time when an interrupt arrives with the time at the start of the first call
4078 to select(). If more time than the timeout has elapsed, the interrupt is
4079 treated as a timeout.
4081 5. Some internal re-factoring in preparation for the addition of Sieve
4082 extensions (by MH). In particular, the "personal" test is moved to a
4083 separate function, and given an option for scanning Cc: and Bcc: (which is
4084 not set for Exim filters).
4086 6. When Exim created an email address using the login of the caller as the
4087 local part (e.g. when creating a From: or Sender: header line), it was not
4088 quoting the local part when it contained special characters such as @.
4090 7. Installed new OpenBSD configuration files.
4092 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4093 try to make them clearer.
4095 9. Callout options, other than the timeout value, were being ignored when
4096 verifying sender addresses in header lines. For example, when using
4098 verify = header_sender/callout=no_cache
4100 the cache was (incorrectly) being used.
4102 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4103 only the Cygwin environment.
4105 11. When return_path_on_delivery was set as a log selector, if different remote
4106 addresses in the same message used different return paths and parallel
4107 remote delivery occurred, the wrong values would sometimes be logged.
4108 (Whenever a remote delivery process finished, the return path value from
4109 the most recently started remote delivery process was logged.)
4111 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4112 lines when AUTH and/or TLS are in use. This is the "received protocol"
4113 field. Exim used to use "asmtp" for authenticated SMTP, without any
4114 indication (in the protocol name) for TLS use. Now it follows the RFC and
4115 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4116 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4117 names appear in log lines as well as in Received: header lines.
4119 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4120 extensions, and comparison tests, and to fix some bugs.
4122 14. Changes to the "personal" filter test:
4124 (1) The test was buggy in that it was just doing the equivalent of
4125 "contains" tests on header lines. For example, if a user's address was
4126 anne@some.where, the "personal" test would incorrectly be true for
4128 To: susanne@some.where
4130 This test is now done by extracting each address from the header in turn,
4131 and checking the entire address. Other tests that are part of "personal"
4132 are now done using regular expressions (for example, to check local parts
4133 of addresses in From: header lines).
4135 (2) The list of non-personal local parts in From: addresses has been
4136 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4137 taken from the Sieve specification recommendations.
4139 (3) If the message contains any header line starting with "List-" it is
4140 treated as non-personal.
4142 (4) The test for "circular" in the Subject: header line has been removed
4143 because it now seems ill-conceived.
4145 15. Minor typos in src/EDITME comments corrected.
4147 16. Installed latest exipick from John Jetmore.
4149 17. If headers_add on a router specified a text string that was too long for
4150 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4151 of string_sprintf() is now avoided.
4153 18. $message_body_size was not set (it was always zero) when running the DATA
4154 ACL and the local_scan() function.
4156 19. For the "mail" command in an Exim filter, no default was being set for
4157 the once_repeat time, causing a random time value to be used if "once" was
4158 specified. (If the value happened to be <= 0, no repeat happened.) The
4159 default is now 0s, meaning "never repeat". The "vacation" command was OK
4160 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4161 (I found it when inspecting the code).
4163 20. There is now an overall timeout for performing a callout verification. It
4164 defaults to 4 times the callout timeout, which applies to individual SMTP
4165 commands during the callout. The overall timeout applies when there is more
4166 than one host that can be tried. The timeout is checked before trying the
4167 next host. This prevents very long delays if there are a large number of
4168 hosts and all are timing out (e.g. when the network connections are timing
4169 out). The value of the overall timeout can be changed by specifying an
4170 additional sub-option for "callout", called "maxwait". For example:
4172 verify = sender/callout=5s,maxwait=20s
4174 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4175 to the end before writing, but this should make it even safer).
4177 22. Exim was forgetting that it had advertised PIPELINING for the second and
4178 subsequent messages on an SMTP connection. It was also not resetting its
4179 memory on STARTTLS and an internal HELO.
4181 23. When Exim logs an SMTP synchronization error within a session, it now
4182 records whether PIPELINING has been advertised or not.
4184 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4185 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4186 rather than long int.
4188 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4190 26. Added the never_mail option to autoreply.
4196 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4197 crash if the getsockname() call failed; this can happen if a connection is
4198 closed very soon after it is established. The problem was simply in the
4199 order in which certain operations were done, causing Exim to try to write
4200 to the SMTP stream before it had set up the file descriptor. The bug has
4201 been fixed by making things happen in the correct order.
4207 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4208 before the connection was closed, thus losing the rejection response.
4210 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4211 some early Solaris releases, but causes trouble in current releases where
4212 socklen_t is defined.
4214 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4217 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4220 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4221 that releasing the top of it at the end releases what was used for sub-
4222 expansions (unless the block got too big). However, discard this block if
4223 the first thing is a variable or header, so that we can use its block when
4224 it is dynamic (useful for very large $message_headers, for example).
4226 6. Lookups now cache *every* query, not just the most recent. A new, separate
4227 store pool is used for this. It can be recovered when all lookup caches are
4228 flushed. Lookups now release memory at the end of their result strings.
4229 This has involved some general refactoring of the lookup sources.
4231 7. Some code has been added to the store_xxx() functions to reduce the amount
4232 of flapping under certain conditions.
4234 8. log_incoming_interface used to affect only the <= reception log lines. Now
4235 it causes the local interface and port to be added to several more SMTP log
4236 lines, for example "SMTP connection from", and rejection lines.
4238 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4240 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4242 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4243 could be overwritten at the end of the current message (or the start of a
4244 new message if it was set in a HELO ACL). The value is now preserved for
4245 the duration of the SMTP connection.
4247 12. If a transport had a headers_rewrite setting, and a matching header line
4248 contained an unqualified address, that address was qualified, even if it
4249 did not match any rewriting rules. The underlying bug was that the values
4250 of the flags that permit the existence of unqualified sender and recipient
4251 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4252 non-local messages, and by -bnq for local messages) were not being
4253 preserved with the message after it was received.
4255 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4256 "next input=" as part of the text comprising the host identity instead of
4257 the correct text. The code was using the same buffer for two different
4258 strings. However, depending on which order the printing function evaluated
4259 its arguments, the bug did not always show up. Under Linux, for example, my
4260 test suite worked just fine.
4262 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4263 doesn't work with some older versions of Perl. It has been changed to "my",
4264 which in any case is probably the better facility to use.
4266 15. A really picky compiler found some instances of statements for creating
4267 error messages that either had too many or two few arguments for the format
4270 16. The size of the buffer for calls to the DNS resolver has been increased
4271 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4272 for addresses that have a lot of PTR records. This alleviates a problem; it
4273 does not fully solve it.
4275 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4276 buffer now truncates the list and logs the incident, which is the same
4277 action as happens when Exim is looking up a host name and its aliases.
4278 Previously in this situation something unpredictable would happen;
4279 sometimes it was "internal error: store_reset failed".
4281 18. If a server dropped the connection unexpectedly when an Exim client was
4282 using GnuTLS and trying to read a response, the client delivery process
4283 crashed while trying to generate an error log message.
4285 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4286 string, for example:
4288 warn message = H1: something\nH2: something
4290 the text was added as a single header line from Exim's point of view
4291 though it ended up OK in the delivered message. However, searching for the
4292 second and subsequent header lines using $h_h2: did not work. This has been
4293 fixed. Similarly, if a system filter added multiple headers in this way,
4294 the routers could not see them.
4296 20. Expanded the error message when iplsearch is called with an invalid key to
4297 suggest using net-iplsearch in a host list.
4299 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4300 ACLs are no longer actually imposed (and a message to that effect is
4303 22. If a "gecos" field in a passwd entry contained escaped characters, in
4304 particular, if it contained a \" sequence, Exim got it wrong when building
4305 a From: or a Sender: header from that name. A second bug also caused
4306 incorrect handling when an unquoted " was present following a character
4307 that needed quoting.
4309 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4310 was not being matched caselessly.
4312 24. Arranged for all hyphens in the exim.8 source to be escaped with
4315 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4316 properly. Recipient callout cache records were still being keyed to include
4317 the sender, even when use_sender was set false. This led to far more
4318 callouts that were necessary. The sender is no longer included in the key
4319 when use_sender is false.
4321 26. Added "control = submission" modifier to ACLs.
4323 27. Added the ${base62d: operator to decode base 62 numbers.
4325 28. dnsdb lookups can now access SRV records.
4327 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4328 the configuration file.
4330 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4331 (-v) mode. This makes the output for a verbose queue run more intelligible.
4333 31. Added a use_postmaster feature to recipient callouts.
4335 32. Added the $body_zerocount variable, containing the number of binary zero
4336 bytes in the message body.
4338 33. The time of last modification of the "new" subdirectory is now used as the
4339 "mailbox time last read" when there is a quota error for a maildir
4342 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4344 35. Added +ignore_unknown as a special item in host lists.
4346 36. Code for decoding IPv6 addresses in host lists is now included, even if
4347 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4348 address was recognized as an IP address, but was then not correctly decoded
4349 into binary, causing unexpected and incorrect effects when compared with
4356 1. Very minor rewording of debugging text in manualroute to say "list of
4357 hosts" instead of "hostlist".
4359 2. If verify=header_syntax was set, and a header line with an unqualified
4360 address (no domain) and a large number of spaces between the end of the
4361 name and the colon was received, the reception process suffered a buffer
4362 overflow, and (when I tested it) crashed. This was caused by some obsolete
4363 code that should have been removed. The fix is to remove it!
4365 3. When running in the test harness, delay a bit after writing a bounce
4366 message to get a bit more predictability in the log output.
4368 4. Added a call to search_tidyup() just before forking a reception process. In
4369 theory, someone could use a lookup in the expansion of smtp_accept_max_
4370 per_host which, without the tidyup, could leave open a database connection.
4372 5. Added the variables $recipient_data and $sender_data which get set from a
4373 lookup success in an ACL "recipients" or "senders" condition, or a router
4374 "senders" option, similar to $domain_data and $local_part_data.
4376 6. Moved the writing of debug_print from before to after the "senders" test
4379 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4380 problems for message scanning, either using a data ACL, or using
4381 local_scan() because the Received: header was not generated till after they
4382 were called (in order to set the time as the time of reception completion).
4383 I have revised the way this works. The header is now generated after the
4384 body is received, but before the ACL or local_scan() are called. After they
4385 are run, the timestamp in the header is updated.
4391 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4392 before starting a queue runner without re-exec. This happened only when
4393 deliver_drop_privilege was set or when the Exim user was set to root. The
4394 effect of the bug was that timeouts during subsequent deliveries caused
4395 crashes instead of being properly handled. The handler is now left at its
4396 default (and expected) setting.
4398 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4399 message, again when deliver_drop_privilege is set or Exim is run as root.
4400 The bug described in (1) was not present in this case, but the tidying up
4401 of the other signals was missing. I have made the two cases consistent.
4403 3. The ignore_target_hosts setting on a manualroute router was being ignored
4404 for hosts that were looked up using the /MX notation.
4406 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4409 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4410 operated on the sender address. After changing the $sender_address to <>
4411 for the sender address verify, Exim was re-instated it as the original
4412 (before rewriting) address, but remembering that it had rewritten it, so it
4413 wasn't rewriting it again. This bug also had the effect of breaking the
4414 sender address verification caching when the sender address was rewritten.
4416 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4417 This has been changed so that if the ip literal address matches
4418 ignore_target_hosts, the router declines.
4420 7. Added expansion conditions match_domain, match_address, and match_local_
4421 part (NOT match_host).
4423 8. The placeholder for the Received: header didn't have a length field set.
4425 9. Added code to Exim itself and to exim_lock to test for a specific race
4426 condition that could lead to file corruption when using MBX delivery. The
4427 issue is with the lockfile that is created in /tmp. If this file is removed
4428 after a process has opened it but before that process has acquired a lock,
4429 there is the potential for a second process to recreate the file and also
4430 acquire a lock. This could lead to two Exim processes writing to the file
4431 at the same time. The added code performs the same test as UW imapd; it
4432 checks after acquiring the lock that its file descriptor still refers to
4433 the same named file.
4435 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4436 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4439 11. Added the smtp_active_hostname option. If used, this will typically be made
4440 to depend on the incoming interface address. Because $interface_address is
4441 not set up until the daemon has forked a reception process, error responses
4442 that can happen earlier (such as "too many connections") no longer contain
4445 12. If an expansion in a condition on a "warn" statement fails because a lookup
4446 defers, the "warn" statement is abandoned, and the next ACL statement is
4447 processed. Previously this caused the whole ACL to be aborted.
4449 13. Added the iplsearch lookup type.
4451 14. Added ident_timeout as a log selector.
4453 15. Added tls_certificate_verified as a log selector.
4455 16. Added a global option tls_require_ciphers (compare the smtp transport
4456 option of the same name). This controls incoming TLS connections.
4458 17. I finally figured out how to make tls_require_ciphers do a similar thing
4459 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4460 before starting the TLS session.
4462 18. Tabs are now shown as \t in -bP output.
4464 19. If the log selector return_path_on_delivery was set, Exim crashed when
4465 bouncing a message because it had too many Received: header lines.
4467 20. If two routers both had headers_remove settings, and the first one included
4468 a superfluous trailing colon, the final name in the first list and the
4469 first name in the second list were incorrectly joined into one item (with a
4470 colon in the middle).
4476 1. Added -C and -D options to the exinext utility, mainly to make it easier
4477 to include in the automated testing, but these could be helpful when
4478 multiple configurations are in use.
4480 2. The exinext utility was not formatting the output nicely when there was
4481 an alternate port involved in the retry record key, nor when there was a
4482 message id as well (for retries that were specific to a specific message
4483 and a specific host). It was also confused by IPv6 addresses, because of
4484 the additional colons they contain. I have fixed the IPv4 problem, and
4485 patched it up to do a reasonable job for IPv6.
4487 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4488 delivery, the log line now contains "pipelined" if PIPELINING was used.
4490 4. An SMTP transport process used to panic and die if the bind() call to set
4491 an explicit outgoing interface failed. This has been changed; it is now
4492 treated in the same way as a connect() failure.
4494 5. A reference to $sender_host_name in the part of a conditional expansion
4495 that was being skipped was still causing a DNS lookup. This no longer
4498 6. The def: expansion condition was not recognizing references to header lines
4499 that used bh_ and bheader_.
4501 7. Added the _cache feature to named lists.
4503 8. The code for checking quota_filecount in the appendfile transport was
4504 allowing one more file than it should have been.
4506 9. For compatibility with Sendmail, the command line option
4514 and sets the incoming protocol and host name (for trusted callers). The
4515 host name and its colon can be omitted when only the protocol is to be set.
4516 Note the Exim already has two private options, -pd and -ps, that refer to
4517 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4518 "s", but I don't think that's a major issue.
4520 10. A number of refactoring changes to the code, none of which should affect
4523 (a) The number of logging options was getting close to filling up the
4524 32-bit word that was used as a bit map. I have split them into two classes:
4525 those that are passed in the argument to log_write(), and those that are
4526 only ever tested independently outside of that function. These are now in
4527 separate 32-bit words, so there is plenty of room for expansion again.
4528 There is no change in the user interface or the logging behaviour.
4530 (b) When building, for example, log lines, the code previously used a
4531 macro that called string_cat() twice, in order to add two strings. This is
4532 not really sufficiently general. Furthermore, there was one instance where
4533 it was actually wrong because one of the argument was used twice, and in
4534 one call a function was used. (As it happened, calling the function twice
4535 did not affect the overall behaviour.) The macro has been replaced by a
4536 function that can join an arbitrary number of extra strings onto a growing
4539 (c) The code for expansion conditions now uses a table and a binary chop
4540 instead of a serial search (which was left over from when there were very
4541 few conditions). Also, it now recognizes conditions like "pam" even when
4542 the relevant support is not compiled in: a suitably worded error message is
4543 given if an attempt is made to use such a condition.
4545 11. Added ${time_interval:xxxxx}.
4547 12. A bug was causing one of the ddress fields not to be passed back correctly
4548 from remote delivery subprocesses. The field in question was not being
4549 subsequently used, so this caused to problems in practice.
4551 13. Added new log selectors queue_time and deliver_time.
4553 14. Might have fixed a bug in maildirsizefile handling that threw up
4554 "unexpected character" debug warnings, and recalculated the data
4555 unnecessarily. In any case, I expanded the warning message to give more
4558 15. Added the message "Restricted characters in address" to the statements in
4559 the default ACL that block characters like @ and % in local parts.
4561 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4562 Three changes have been made:
4564 (a) There was a serious bug; a negative response to MAIL caused the whole
4565 recipient domain to be cached as invalid, thereby blocking all messages
4566 to all local parts at the same domain, from all senders. This bug has
4567 been fixed. The domain is no longer cached after a negative response to
4568 MAIL if the sender used is not empty.
4570 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4573 (c) A new callout option, "use_sender" has been added for people who want
4574 the modified behaviour.
4580 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4583 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4584 indeed breaks things for older releases.
4586 3. Added additional logging to the case where there is a problem reading data
4587 from a filter that is running in a subprocess using a pipe, in order to
4588 try to track down a specific problem.
4590 4. Testing facility fudge: when running in the test harness and attempting
4591 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4592 getting "No route to host". Convert this to a timeout.
4594 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4597 6. Some OS don't have socklen_t but use size_t instead. This affects the
4598 fifth argument of getsockopt() amongst other things. This is now
4599 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4600 can be set for individual OS. I have set it for SunOS5, OSF1, and
4601 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4602 some earlier ones do not.
4604 7. Change 4.30/15 was not doing the test caselessly.
4606 8. The standard form for an IPv6 address literal was being rejected by address
4607 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4608 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4609 this, as well as the form without the "IPv6" on the front (but only when
4610 address literals are enabled, of course).
4612 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4614 10. Exim crashed if a message with an empty sender address specified by -f
4615 encountered a router with an errors_to setting. This could be provoked only
4616 by a command such as
4620 where an empty string was supplied; "<>" did not hit this bug.
4622 11. Installed PCRE release 4.5.
4624 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4625 remained set. It is now erased.
4627 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4628 times from message ids (which are base 36 rather than the normal 62).
4630 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4631 were being counted as actual protocol errors, and logged if the log
4632 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4633 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4634 and DATA following a set of rejected RCPTs do not count as protocol errors.
4635 In other words, Exim assumes they were pipelined, though this may not
4636 actually be the case. Of course, in all cases the client gets an
4637 appropriate error code.
4639 15. If a lookup fails in an ACL condition, a message about the failure may
4640 be available; it is used if testing the ACL cannot continue, because most
4641 such messages specify what the cause of the deferral is. However, some
4642 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4643 that caused an old message to be retained and used if a later statement
4644 caused a defer, replacing the real cause of the deferral.
4646 16. If an IP address had so many PTR records that the DNS lookup buffer
4647 was not large enough to hold them, Exim could crash while trying to process
4648 the truncated data. It now detects and logs this case.
4650 17. Further to 4.21/58, another change has been made: if (and only if) the
4651 first line of a message (the first header line) ends with CRLF, a bare LF
4652 in a subsequent header line has a space inserted after it, so as not to
4653 terminate the header.
4655 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4656 unnecessarily, used atoi() instead of strtol(), and didn't check the
4657 termination when getting file sizes from file names by regex.
4659 19. Completely re-implemented the support for maildirsize files, in the light
4660 of a number of problems with the previous contributed implementation
4661 (4.30/29). In particular:
4663 . If the quota is zero, the maildirsize file is maintained, but no quota is
4666 . If the maildir directory does not exist, it is created before any attempt
4667 to write a maildirsize file.
4669 . The quota value in the file is just a cache; if the quota is changed in
4670 the transport, the new value overrides.
4672 . A regular expression is available for excluding directories from the
4675 20. The autoreply transport checks the characters in options that define the
4676 message's headers; it allows continued headers, but it was checking with
4677 isspace() after an embedded newline instead of explicitly looking for a
4680 21. If all the "regular" hosts to which an address was routed had passed their
4681 expiry times, and had not reached their retry times, the address was
4682 bounced, even if fallback hosts were defined. Now Exim should go on to try
4685 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4686 equivalent code in the SMTP transport. Some hosts send humungous responses
4687 to HELO/EHLO, more than 1024 it seems.
4689 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4690 gives compiler warnings in some environments. I've now done it "properly",
4693 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4694 (because of problems with the built-in one) was declared to return uschar *
4695 instead of char *, causing compiler failure.
4697 25. Fixed a file descriptor leak when processing alias/forward files.
4699 26. Fixed a minor format string issue in dbfn.c.
4701 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4703 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4704 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4707 29. When a sender address is verified, it is cached, to save repeating the test
4708 when there is more than one recipient in a message. However, when the
4709 verification involves a callout, it is possible for different callout
4710 options to be set for different recipients. It is too complicated to keep
4711 track of this in the cache, so now Exim always runs a verification when a
4712 callout is required, relying on the callout cache for the optimization.
4713 The overhead is duplication of the address routing, but this should not be
4716 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4717 to be verified with callout=postmaster, and the main callout worked but the
4718 postmaster check failed, the verification correctly failed. However, if a
4719 subsequent RCPT command asked for sender verification *without* the
4720 postmaster check, incorrect caching caused this verification also to fail,
4723 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4724 it was not caching the DNS options (qualify_single, search_parents) that
4725 were used when the lookup failed. A subsequent lookup with different
4726 options therefore always gave the same answer, though there were cases
4727 where it should not have. (Example: a "domains = !$mx_any" option on a
4728 dnslookup router: the "domains" option is always processed without any
4729 widening, but the router might have qualify_single set.) Now Exim uses the
4730 cached value only when the same options are set.
4732 32. Added John Jetmore's "exipick" utility to the distribution.
4734 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4735 than a timeout (e.g. a certificate is required, and is not provided), an
4736 Exim server now closes the connection immediately. Previously it waited for
4737 the client to close - but if the client is SSL, it seems that they each
4738 wait for each other, leading to a delay before one of them times out.
4740 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4741 maintained 0.8.x compatibility because I don't think many are using it, and
4742 it is clearly obsolete.
4744 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4747 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4748 client certificate was expired. A simple patch fixes this, though I don't
4749 understand the full logic of why the verify callback is called multiple
4752 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4753 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4754 which causes problems with some clients (such as the Certicom SSL Plus
4755 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4756 disables the coutermeasure allowing Eudora to connect."
4758 38. Exim was not checking that a write() to a log file succeeded. This could
4759 lead to Bad Things if a log got too big, in particular if it hit a file
4760 size limit. Exim now panics and dies if it cannot write to a log file, just
4761 as it does if it cannot open a log file.
4763 39. Modified OS/Makefile-Linux so that it now contains
4765 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4767 The two -D definitions ensure that Exim is compiled with large file
4768 support, which makes it possible to handle log files that are bigger than
4771 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4772 instance) a domain was checked against a named list that involved a lookup,
4773 causing $domain_data to be set, then another domain was checked against the
4774 same list, then the first domain was re-checked, the value of $domain_data
4775 after the final check could be wrong. In particular, if the second check
4776 failed, it could be set empty. This bug probably also applied to
4779 41. The strip_trailing_dot option was not being applied to the address given
4780 with the -f command-line option.
4782 42. The code for reading a message's header from the spool was incrementing
4783 $received_count, but never initializing it. This meant that the value was
4784 incorrect (doubled) while delivering a message in the same process in which
4785 it was received. In the most common configuration of Exim, this never
4786 happens - a fresh exec is done - but it can happen when
4787 deliver_drop_privilege is set.
4789 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4790 it now includes up to 150 characters of the unexpected data in the log
4793 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4794 and building data strings. The size of both of these buffers was 10 000
4795 bytes - far larger than anybody would *ever* want, thought I. Needless to
4796 say, somebody hit the limit. I have increased the maximum line length to
4797 20 000 and the maximum data length of concatenated lines to 100 000. I have
4798 also fixed two bugs, because there was no checking on these buffers. Tsk,
4799 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4800 buffer is too small.
4802 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4803 lsearch lookups. Now it does.
4805 46. When parsing a route_list item in a manualroute router, a fixed-length
4806 buffer was used for the list of hosts. I made this 1024 bytes long,
4807 thinking that nobody would ever have a list of hosts that long. Wrong.
4808 Somebody had a whole pile of complicated expansion conditions, and the
4809 string was silently truncated, leading to an expansion error. It turns out
4810 that it is easier to change to an unlimited length (owing to other changes
4811 that have happened since this code was originally written) than to build
4812 structure for giving a limitation error. The length of the item that
4813 expands into the list of hosts is now unlimited.
4815 47. The lsearch lookup could not handle data where the length of text line was
4816 more than 4095 characters. Such lines were truncated, leading to shortened
4817 data being returned. It should now handle lines of any length.
4819 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4820 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4823 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4824 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4825 no longer happens in comment lines. A list of replacements is now placed
4826 at the head of all of the source files, except those whose only change is
4827 to replace PERL_COMMAND in the very first #! line.
4829 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4830 messages on the queue, with a bottom-up merge sort, using code contributed
4831 by Michael Haardt. This should make operations like -bp somewhat faster on
4832 large queues. It won't affect queue runners, except when queue_run_in_order
4835 51. Installed eximstats 1.31 in the distribution.
4837 52. Added support for SRV lookups to the dnslookup router.
4839 53. If an ACL referred to $message_body or $message_body_end, the value was not
4840 reset for any messages that followed in the same SMTP session.
4842 54. The store-handling optimization for building very long strings was not
4843 differentiating between the different store pools. I don't think this
4844 actually made any difference in practice, but I've tidied it.
4846 55. While running the routers to verify a sender address, $sender_address
4847 was still set to the sender address. This is wrong, because when routing to
4848 send a bounce to the sender, it would be empty. Therefore, I have changed
4849 it so that, while verifying a sender address, $sender_address is set to <>.
4850 (There is no change to what happens when verifying a recipient address.)
4852 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4853 target A or AAAA records (if not already returned) without resetting the
4854 qualify_single or search_parents options of the DNS resolver. These are
4855 inappropriate in this case because the targets of MX and SRV records must
4856 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4857 target that, when qualified, matched something in the local domain. These
4858 two options are now turned off when doing these lookups.
4860 57. It seems that at least some releases of Reiserfs (which does not have the
4861 concept of a fixed number of inodes) returns zero and not -1 for the
4862 number of available inodes. This interacted badly with check_spool_inodes,
4863 which assumed that -1 was the "no such thing" setting. What I have done is
4864 to check that the total number of inodes is greater than zero before doing
4865 the test of how many are available.
4867 58. When a "warn" ACL statement has a log_message modifier, the message is
4868 remembered, and not repeated. This is to avoid a lot of repetition when a
4869 message has many recipients that cause the same warning to be written.
4870 Howewer, Exim was preserving the list of already written lines for an
4871 entire SMTP session, which doesn't seem right. The memory is now reset if a
4872 new message is started.
4874 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4875 debugging output unless log_rewrite was also set.
4877 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4878 of (int)(handle) when we know that handle contains (void *)(-1).
4880 61. The Exim daemon panic-logs an error return when it closes the incoming
4881 connection. However "connection reset by peer" seems to be common, and
4882 isn't really an error worthy of noting specially, so that particular error
4885 62. When Exim is trying to find all the local interfaces, it used to panic and
4886 die if the ioctl to get the interface flags failed. However, it seems that
4887 on at least one OS (Solaris 9) it is possible to have an interface that is
4888 included in the list of interfaces, but for which you get a failure error
4889 for this call. This happens when the interface is not "plumbed" into a
4890 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4891 failure of the "get flags" call assumes that the interface is down.
4893 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4894 makes life easier for people who are doing arithmetic on fields extracted
4895 from dates, where you often get leading zeros that should not be
4896 interpreted as octal.
4898 64. Added qualify_domain to the redirect router, to override the global
4901 65. If a pathologically long header line contained very many addresses (the
4902 report of this problem mentioned 10 000) and each of them was rewritten,
4903 Exim could use up a very large amount of memory. (It kept on making new
4904 copies of the header line as it rewrote, and never released the old ones.)
4905 At the expense of a bit more processing, the header rewriting function has
4906 been changed so that it no longer eats memory in this way.
4908 66. The generation of the Received: header has been moved from the time that a
4909 message starts to be received, to the time that it finishes. The timestamp
4910 in the Received: header should now be very close to that of the <= log
4911 line. There are two side-effects of this change:
4913 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4914 logged header lines no longer include the local Received: line, because
4915 it has not yet been created. The same applies to a copy of the message
4916 that is returned to a non-SMTP sender when a message is rejected.
4918 (b) When a filter file is tested using -bf, no additional Received: header
4919 is added to the test message. After some thought, I decided that this
4922 This change does not affect the value of $received_for. It is still set
4923 after address rewriting, but before local_scan() is called.
4925 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4927 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4928 gave an unhelpful panic error message, and a defer error. I have managed to
4929 change this behaviour so that it now rejects any supplied certificate,
4930 which seems right, as the list of acceptable certificates is empty.
4932 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4933 gave an unhelpful defer error. I have not managed to make this reject any
4934 supplied certificates, but the error message it gives is "no certificate
4935 supplied", which is not helpful.
4937 70. exigrep's output now also includes lines that are not associated with any
4938 message, but which match the given pattern. Implemented by a patch from
4939 Martin Sluka, which also tidied up the Perl a bit.
4941 71. Recipient callout verification, like sender verification, was using <> in
4942 the MAIL FROM command. This isn't really the right thing, since the actual
4943 sender may affect whether the remote host accepts the recipient or not. I
4944 have changed it to use the actual sender in the callout; this means that
4945 the cache record is now keyed on a recipient/sender pair, not just the
4946 recipient address. There doesn't seem to be a real danger of callout loops,
4947 since a callout by the remote host to check the sender would use <>.
4948 [SEE ABOVE: changed after hitting problems.]
4950 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4951 temporary errors. However, in the case of such a code being given after
4952 the end of a data transmission (i.e. after ".") Exim was failing to write
4953 a retry record for the message. (Yes, there was some broken host that was
4954 actually sending 8xx at this point.)
4956 73. An unknown lookup type in a host list could cause Exim to panic-die when
4957 the list was checked. (An example that provoked this was putting <; in the
4958 middle of a list instead of at the start.) If this happened during a DATA
4959 ACL check, a -D file could be left lying around. This kind of configuration
4960 error no longer causes Exim to die; instead it causes a defer errror. The
4961 incident is still logged to the main and panic logs.
4963 74. Buglet left over from Exim 3 conversion. The message "too many messages
4964 in one connection" was written to the rejectlog but not the mainlog, except
4965 when address rewriting (yes!) was being logged.
4967 75. Added write_rejectlog option.
4969 76. When a system filter was run not as root (that is, when system_filter_user
4970 was set), the values of the $n variables were not being returned to the
4971 main process; thus, they were not subsequently available in the $sn
4974 77. Added +return_path_on_delivery log selector.
4976 78. A connection timeout was being treated differently from recipients deferred
4977 when testing hosts_max_try with a message that was older than the host's
4978 retry timeout. (The host should not be counted, thus allowing all hosts to
4979 be tried at least once before bouncing.) This may have been the cause of an
4980 occasionally reported bug whereby a message would remain on the queue
4981 longer than the retry timeout, but would be bounced if a delivery was
4982 forced. I say "may" because I never totally pinned down the problem;
4983 setting up timeout/retry tests is difficult. See also the next item.
4985 79. The ultimate address timeout was not being applied to errors that involved
4986 a combination of host plus message (for example, a timeout on a MAIL
4987 command). When an address resolved to a number of possible hosts, and they
4988 were not all tried for each delivery (e.g. because of hosts_max_try), a
4989 message could remain on the queue longer than the retry timeout.
4991 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4994 81. Fixed an obscure SMTP outgoing bug which required at least the following
4995 conditions: (a) there was another message waiting for the same server;
4996 (b) the server returned 5xx to all RCPT commands in the first message so
4997 that the message was not completed; (c) the server dropped the connection
4998 or gave a negative response to the RSET that Exim sends to abort the
4999 transaction. The observed case was a dropped connection after DATA that had
5000 been sent in pipelining mode. That is, the server had advertised PIPELINING
5001 but was not implementing it correctly. The effect of the bug was incorrect
5002 behaviour, such as trying another host, and this could lead to a crash.
5008 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5009 and daemon.c were passed as pointers to ints; they should have been
5010 pointers to socklen_t variables (which are typically unsigned ints).
5012 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5015 3. Fixed a really odd bug that affected only the testing scheme; patching a
5016 certain fixed string in the binary changed the value of another string that
5017 happened to be identical to the end of the original first string.
5019 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5020 name", it returns that address as the IP address. On some operating
5021 systems (e.g. Solaris), it also passes back the IP address string as the
5022 "host name". However, on others (e.g. Linux), it passes back an empty
5023 string. Exim wasn't checking for this, and was changing the host name to an
5024 empty string, assuming it had been canonicized.
5026 5. Although rare, it is permitted to have more than one PTR record for a given
5027 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5028 all the names associated with an address, because they do in Solaris.
5029 However, it seems that they do not in Linux for data that comes from the
5030 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5031 I found this out when I moved to a new Linux workstation and tried to run
5032 the Exim test suite.
5034 To get round this problem I have changed the code so that it now does its
5035 own call to the DNS to look up PTR records when searching for a host name.
5036 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5037 addresses that are only in /etc/hosts are still found.
5039 This behaviour is, however, controlled by an option called host_lookup_
5040 order, which defaults to "bydns:byaddr". If people want to use the other
5041 order, or indeed, just use one or the other means of lookup, they can
5042 specify it in this variable.
5044 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5045 some operating systems, this comes back from gethostbyaddr() as an empty
5046 string, and this is what Exim used to test for. However, it seems that in
5047 other systems, "." is yielded. Exim now tests for this case too.
5049 7. The values of check_spool_space and check_log_space are now held internally
5050 as a number of kilobytes instead of an absolute number of bytes. If a
5051 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5052 kilobyte. This means that much larger values can be stored.
5054 8. Exim monitor: an attempt to get the action menu when not actually pointing
5055 at a message produces an empty menu entitled "No message selected". This
5056 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5057 no entries in it ("Shell widget menu has zero width and/or height"). So I
5058 have added a single, blank menu entry in this case.
5060 9. Added ${quote_local_part.
5062 10. MIME decoding is now applied to the contents of Subject: header lines when
5065 11. Now that a reference to $sender_host_address automatically causes a reverse
5066 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5067 host lookup before query-style lookups in lists that might use this
5068 variable. This has therefore been abolished, and the "net-" prefix is no
5069 longer necessary for query-style lookups.
5071 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5072 have been a typo for LFLAGS, so it has been changed.
5074 13. The install script calls Exim with "-C /dev/null" in order to find the
5075 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5076 to be output. Howeve, since Exim outputs its version number before the
5077 error, it didn't break the script. It just looked ugly. I fixed this by
5078 always allowing "-C /dev/null" if the caller is root.
5080 14. Ignore overlarge ACL variable number when reading spool file - insurance
5081 against a later release with more variables having written the file.
5083 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5084 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5085 this, as well as the form without the "IPv6" on the front.
5087 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5088 OS/Makefile-Darwin file.
5090 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5091 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5093 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5094 variable. However, some OS use ints for inodes. I've added cast to long int
5095 to get rid of the compiler warning.
5097 19. I had forgotten to lock out "/../" in configuration file names when
5098 ALT_CONFIG_PREFIX was set.
5100 20. Routers used for verification do not need to specify transports. However,
5101 if such a router generated a host list, and callout was configured, Exim
5102 crashed, because it could not find a port number from the (non-existent)
5103 transport. It now assumes port 25 in this circumstance.
5105 21. Added the -t option to exigrep.
5107 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5108 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5109 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5110 them is set, LOOKUP_LSEARCH is forced.
5112 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5114 24. Added sender and host information to the "rejected by local_scan()" log
5115 line; previously there was no indication of these.
5117 25. Added .include_if_exists.
5119 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5120 Linux. It turns out that not all file systems support this. Apparently some
5121 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5122 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5123 not supported on the file descriptor, is now ignored when Exim is trying to
5124 sync a directory. This applies only to Linux.
5126 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5128 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5129 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5132 29. Added support for maildirsize files from supplied patch (modified a bit).
5134 30. The use of :fail: followed by an empty string could lead Exim to respond to
5135 sender verification failures with (e.g.):
5137 550 Verification failed for <xxx>
5138 550 Sender verify failed
5140 where the first response line was missing the '-' that indicates it is not
5141 the final line of the response.
5143 31. The loop for finding the name of the user that called Exim had a hardwired
5144 limit of 10; it now uses the value of finduser_retries, which is used for
5145 all other user lookups.
5147 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5150 33. Exim was neglecting to zero errno before one call of strtol() when
5151 expanding a string and expecting an integer value. On some systems this
5152 resulted in spurious "integer overflow" errors. Also, it was casting the
5153 result into an int without checking.
5155 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5156 did not work. The code looks as if it has *never* worked, though it appears
5157 to have been documented since at least releast 1.62. I have made it work.
5159 35. The "timeout_DNS" error in retry rules, also documented since at least
5160 1.62, also never worked. As it isn't clear exactly what this means, and
5161 clearly it isn't a major issue, I have abolished the feature by treating it
5162 as "timeout", and writing a warning to the main and panic logs.
5164 36. The display of retry rules for -brt wasn't always showing the error code
5167 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5168 timeout_connect_A, timeout_connect_MX.
5170 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5171 to the empty sender.
5173 39. The daemon was not analysing the content of -oX till after it had closed
5174 stderr and disconnected from the controlling terminal. This meant that any
5175 syntax errors were only noted on the panic log, and the return code from
5176 the command was 0. By re-arranging the code a little, I've made the
5177 decoding happen first, so such errors now appear on stderr, and the return
5178 code is 1. However, the actual setting up of the sockets still happens in
5179 the disconnected process, so errors there are still only recorded on the
5182 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5183 connections (as happens on some IP stacks) was logged at start up time as
5184 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5185 it from "IPv6 and IPv4", which means that two separate sockets are being
5188 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5189 says whether AF_INET or AF_INET6 was passed as an argument.
5191 42. Exiwhat output was messed up when time zones were included in log
5194 43. Exiwhat now gives more information about the daemon's listening ports,
5195 and whether -tls-on-connect was used.
5197 44. The "port" option of the smtp transport is now expanded.
5199 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5200 silently ignored. Now an error message is written to the main and panic
5203 46. There's a new ACL modifier called "logwrite" which writes to a log file
5204 as soon as it is encountered.
5206 47. Added $local_user_uid and $local_user_gid at routing time.
5208 48. Exim crashed when trying to verify a sender address that was being
5211 49. Exim was recognizing only a space character after ".include". It now also
5212 recognizes a tab character.
5214 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5215 extracting the relevant information from the specification. The man page no
5216 longer contains scrambled data for the -d option, and I've added a section
5217 at the front about calling Exim under different names.
5219 51. Added "extra_headers" argument to the "mail" command in filter files.
5221 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5224 53. Installed eximstats 1.29.
5226 54. Added transport_filter_timeout as a generic transport option.
5228 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5229 Cc: header lines. This was required by RFC 822, but it not required by RFC
5232 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5233 incoming messages that did not have them. Now it does so only if the
5234 message originates locally, that is, if there is no associated remote host
5235 address. When Resent- header lines are present, this applies to the Resent-
5236 lines rather than the non-Resent- lines.
5238 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5239 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5241 58. Messages for configuration errors now include the name of the main
5242 configuration file - useful now that there may be more than one file in a
5243 list (.included file names were always shown).
5245 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5246 for those rare installations that do not start the daemon as root or run it
5247 setuid root. I've cut out the call to initgroups() if the daemon is not
5250 60. The Exim user and group can now be bound into the binary as text strings
5251 that are looked up at the start of Exim's processing.
5253 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5255 62. Added $mailstore_basename variable.
5257 63. Installed patch to sieve.c from Michael Haardt.
5259 64. When Exim failed to open the panic log after failing to open the main log,
5260 the original message it was trying to log was written to stderr and debug
5261 output, but if they were not available (the usual case in production), it
5262 was lost. Now it is written to syslog before the two lines that record the
5263 failures to open the logs.
5265 65. Users' Exim filters run in subprocesses under the user's uid. It is
5266 possible for a "deliver" command or an alias in a "personal" command to
5267 provoke an address rewrite. If logging of address rewriting is configured,
5268 this fails because the process is not running as root or exim. There may be
5269 a better way of dealing with this, but for the moment (because 4.30 needs
5270 to be released), I have disabled address rewrite logging when running a
5271 filter in a non-root, non-exim process.
5277 1. The buildconfig auxiliary program wasn't quoting the value set for
5278 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5279 not defined. This bug was masked in 4.22 by the effect that was fixed in
5282 2. Some messages that were rejected after a message id was allocated were
5283 shown as "incomplete" by exigrep. It no longer does this for messages that
5284 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5286 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5287 have allow_domain_literals set, the ID did not get logged in the <= line.
5288 Domain literals are now always recognized in Message-ID: header lines.
5290 4. The first argument for a ${extract expansion item is the key name or field
5291 number. Leading and trailing spaces in this item were not being ignored,
5292 causing some misleading effects.
5294 5. When deliver_drop_privilege was set, single queue runner processes started
5295 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5296 same command in the process it spins off) were not dropping privilege.
5298 6. When the daemon running as "exim" started a queue runner, it always
5299 re-executed Exim in the spun-off process. This is a waste of effort when
5300 deliver_drop_privilege is set. The new process now just calls the
5301 queue-runner function directly.
5307 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5310 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5311 ignored. Though the use of -oP was forcing the writing of a pid file, it
5312 was always written to the default place.
5314 3. If the message "no IP address found for host xxxx" is generated during
5315 incoming verification, it is now followed by identification of the incoming
5316 connection (so you can more easily find what provoked it).
5318 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5320 5. Added some features to "harden" Exim a bit more against certain attacks:
5322 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5323 be put in Local/Makefile. This is like the never_users runtime option,
5324 but it cannot be overridden. The default setting is "root".
5326 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5327 prefix string with which any file named in a -C command line option
5330 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5331 is retained for -C and -D only if the caller of Exim is root. Without
5332 it, the exim user may also use -C and -D and retain privilege.
5334 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5335 command line option is disabled.
5337 6. Macro names set by the -D option must start with an upper case letter, just
5338 like macro names defined in the configuration file.
5340 7. Added "dereference=" facility to LDAP.
5342 8. Two instances of the typo "uknown" in the source files are fixed.
5344 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5345 the Configure-Makefile script screwed up while processing it.
5347 10. Incorporated PCRE 4.4.
5349 11. The SMTP synchronization check was not operating right at the start of an
5350 SMTP session. For example, it could not catch a HELO sent before the client
5351 waited for the greeting. There is now a check for outstanding input at the
5352 point when the greeting is written. Because of the duplex, asynchronous
5353 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5354 way, but not yet received, when the check is performed.
5356 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5357 on TCP/IP sockets, because this apparently causes some broken clients to
5360 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5361 unchanged) from the Cygwin maintainer.
5363 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5364 was supported instead of testing for mbx. Effectively a typo.
5366 15. The spa authenticator server code was not checking that the input it
5367 received was valid base64.
5369 16. The debug output line for the "set" modifier in ACLs was not showing the
5370 name of the variable that was being set.
5372 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5374 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5375 lookup on its own. Something else was needed to trigger the lookup. For
5376 example, a match in host_lookup or the need for a host name in a host list.
5377 Now, if $sender_host_name is referenced and the host name has not yet been
5378 looked up, a lookup is performed. If the lookup fails, the variable remains
5379 empty, and $host_lookup_failed is set to "1".
5381 19. Added "eqi" as a case-independent comparison operator.
5383 20. The saslauthd authentication condition could segfault if neither service
5384 nor realm was specified.
5386 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5387 error message that was logged was misleading, and incoming SMTP
5388 connections were dropped. The message is now more accurate, and temporary
5389 errors are given to SMTP connections.
5391 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5392 (except RSET) with a 421 error, until QUIT is received. However, it was
5393 failing to send a response to QUIT.
5395 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5396 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5397 ACL is now run after the helo_try_verify_hosts code.
5399 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5400 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5401 case-independent, so other case variants are also recognized). Apparently
5402 some systems use these upper case variants.
5404 25. If more than two messages were waiting for the same host, and a transport
5405 filter was specified for the transport, Exim sent two messages over the
5406 same TCP/IP connection, and then failed with "socket operation on non-
5407 socket" when it tried to send the third.
5409 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5411 27. The extern definition of crypt16() in expand.c was not being excluded when
5412 the OS had its own crypt16() function.
5414 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5415 as a preferred synonym for return_size_limit, both as an option and as an
5418 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5420 30. Changed the default configuration ACL to relax the local part checking rule
5421 for addresses that are not in any local domains. For these addresses,
5422 slashes and pipe symbols are allowed within local parts, but the sequence
5423 /../ is explicitly forbidden.
5425 31. SPA server authentication was not clearing the challenge buffer before
5428 32. log_message in a "warn" ACL statement was writing to the reject log as
5429 well as to the main log, which contradicts the documentation and doesn't
5430 seem right (because no rejection is happening). So I have stopped it.
5432 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5433 However, I am unable to do any testing of this.
5435 34. Fixed an infelicity in the appendfile transport. When checking directories
5436 for a mailbox, to see if any needed to be created, it was accidentally
5437 using path names with one or more superfluous leading slashes; tracing
5438 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5440 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5441 contents are added to the log line that is written for every discarded
5442 recipient. (Previously a log_message setting was ignored.)
5444 36. The ${quote: operator now quotes the string if it is empty.
5446 37. The install script runs exim in order to find its version number. If for
5447 some reason other than non-existence or emptiness, which it checks, it
5448 could not run './exim', it was installing it with an empty version number,
5449 i.e. as "exim-". This error state is now caught, and the installation is
5452 38. An argument was missing from the function that creates an error message
5453 when Exim fails to connect to the socket for saslauthd authentication.
5454 This could cause Exim to crash, or give a corrupted message.
5456 39. Added isip, isip4, and isip6 to ${if conditions.
5458 40. The ACL variables $acl_xx are now saved with the message, and can be
5459 accessed later in routers, transports, and filters.
5461 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5462 strings in the file are not string-expanded.
5464 42. If a MAIL command specified a SIZE value that was too large to fit into an
5465 int variable, the check against message_size_limit failed. Such values are
5466 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5467 day this will have to be increased, but I don't think I want to be around
5468 when emails are that large.
5475 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5476 iconv() is not standard in FreeBSD.
5478 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5479 IPv6 enabled. The observed symptom was a segmentation fault on return from
5480 the function os_common_find_running_interfaces() in src/os.c.
5482 3. In the check_special_case() function in daemon.c I had used "errno" as an
5483 argument name, which causes warnings on some systems. This was basically a
5484 typo, since it was named "eno" in the comments!
5486 4. The code that waits for the clock to tick (at a resolution of some fraction
5487 of a second) so as to ensure message-id uniqueness was always waiting for
5488 at least one whole tick, when it could have waited for less. [This is
5489 almost certainly not relevant at current processor speeds, where it is
5490 unlikely to ever wait at all. But we try to future-proof.]
5492 5. The function that sleeps for a time interval that includes fractions of a
5493 second contained a race. It did not block SIGALRM between setting the
5494 timer, and suspending (a couple of lines later). If the interval was short
5495 and the sigsuspend() was delayed until after it had expired, the suspension
5496 never ended. On busy systems this could lead to processes getting stuck for
5499 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5500 process, before it forks any delivery processes. The open lookup caching
5501 mechanism meant that the open file or database connection was passed into
5502 the delivery process. The problem was that delivery processes always tidy
5503 up cached lookup data. This could cause a problem for the next delivery
5504 process started by the queue runner, because the external queue runner
5505 process does not know about the closure. So the next delivery process
5506 still has data in the lookup cache. In the case of a file lookup, there was
5507 no problem because closing a file descriptor in a subprocess doesn't affect
5508 the parent. However, if the lookup was caching a connection to a database,
5509 the connection was closed, and the second delivery process was likely to
5510 see errors such as "PGSQL: query failed: server closed the connection
5511 unexpectedly". The problem has been fixed by closing all cached lookups
5512 in a queue runner before running a delivery process.
5514 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5515 seem to have the "const" qualifier which it has on other OS. I've
5518 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5519 *of the same type* (client or server) with the same public name that an
5520 error should be diagnosed.
5522 9. When Exim looked up a host name for an IP address, but failed to find the
5523 original IP address when looking up the host name (a safety check), it
5524 output the message "<ip address> does not match any IP for NULL", which was
5525 confusing, to say the least. The bug was that the host name should have
5526 appeared instead of "NULL".
5528 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5529 user that is built into the binary, and the -C or -D options is used, root
5530 privilege is dropped before the configuration file is read. In addition,
5531 logging is switched to stderr instead of the normal log files. If the
5532 configuration then re-defines the Exim user, the unprivileged environment
5533 is probably not what is expected, so Exim logs a panic warning message (but
5536 However, if deliver_drop_privilege is set, the unprivileged state may well
5537 be exactly what is intended, so the warning has been cut out in that case,
5538 and Exim is allowed to try to write to its normal log files.
5544 1. smtp_return_error_details was not giving details for temporary sender
5545 or receiver verification errors.
5547 2. Diagnose a configuration error if two authenticators have the same public
5550 3. Exim used not to create the message log file for a message until the first
5551 delivery attempt. This could be confusing when incoming messages were held
5552 for policy or load reasons. The message log file is now created at the time
5553 the message is received, and an initial "Received" line is written to it.
5555 4. The automatically generated man page for command line options had a minor
5556 bug that caused no ill effects; however, a more serious problem was that
5557 the procedure for building the man page automatically didn't always
5558 operate. Consequently, release 4.20 contains an out-of-date version. This
5559 shouldn't happen again.
5561 5. When building Exim with embedded Perl support, the script that builds the
5562 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5563 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5565 6. The freeze_tell option was not being used for messages that were frozen on
5566 arrival, either by an ACL or by local_scan().
5568 7. Added the smtp_incomplete_transaction log selector.
5570 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5571 was accepting AUTH without a new EHLO.
5573 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5574 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5577 10. Logging of TCP/IP connections (when configured) now happens in the main
5578 daemon process instead of the child process, so that the TCP/IP connection
5579 count is more accurate (but it can never be perfect).
5581 11. The use of "drop" in a nested ACL was not being handled correctly in the
5582 outer ACL. Now, if condition failure induced by the nested "drop" causes
5583 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5584 or "require"), the connection is dropped.
5586 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5587 that yield "discard" can now be used with an "accept" or a "discard" verb,
5588 but an error is generated for any others (because I can't see a useful way
5589 to define what should happen).
5591 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5592 are now processed in the same way as lines in the Exim configuration file.
5593 In particular, continuation lines are supported.
5595 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5597 15. Added -ti meaning -t -i.
5599 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5600 domains, and warn by logging if others are found.
5602 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5603 returned from GIFCONF when Exim is trying to find the list of interfaces on
5604 a host. The code in os.c has been modified to copy each ifreq to an aligned
5605 structure in all cases.
5607 Also, in some cases, the returned ifreq's were being copied to a 'struct
5608 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5609 means the last couple of bytes of an IPv6 address could be chopped if the
5610 ifreq contained only a normal sockaddr (14 bytes storage).
5612 18. Named domain lists were not supported in the hosts_treat_as_local option.
5613 An entry such as +xxxx was not recognized, and was treated as a literal
5616 19. Ensure that header lines added by a DATA ACL are included in the reject log
5617 if the ACL subsequently rejects the message.
5619 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5620 MD5 (which is deprecated).
5622 21. When testing a filter file using -bf, Exim was writing a message when it
5623 took the sender from a "From " line in the message, but it was not doing so
5624 when it took $return_path from a Return-Path: header line. It now does.
5626 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5627 with a valid header line field name (a series of printing characters
5628 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5630 23. Changed "disc" in the source to "disk" to conform to the documentation and
5631 the book and for uniformity.
5633 24. Ignore Sendmail's -Ooption=value command line item.
5635 25. When execve() failed while trying to run a command in a pipe transport,
5636 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5637 could be confused with a return value of 69 from the command itself. This
5638 has been changed to 127, the value the shell returns if it is asked to run
5639 a non-existent command. The wording for the related log line suggests a
5640 non-existent command as the problem.
5642 26. If received_header_text expands to an empty string, do not add a Received:
5643 header line to the message. (Well, it adds a token one on the spool, but
5644 marks it "old" so that it doesn't get used or transmitted.)
5646 27. Installed eximstats 1.28 (addition of -nt option).
5648 28. There was no check for failure on the call to getsockname() in the daemon
5649 code. This can fail if there is a shortage of resources on the system, with
5650 ENOMEM, for example. A temporary error is now given on failure.
5652 29. Contrary to the C standard, it seems that in some environments, the
5653 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5654 program. Exim now does this explicitly; it affects the formatting of
5655 timestamps using strftime().
5657 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5658 complaints. I've now initialized all the variables, to avoid this.
5660 32. Header lines added by a system filter were not being "seen" during
5661 transport-time rewrites.
5663 33. The info_callback() function passed to OpenSSL is set up with type void
5664 (*)(SSL *, int, int), as described somewhere. However, when calling the
5665 function (actually a macro) that sets it up, the type void(*)() is
5666 expected. I've put in a cast to prevent warnings from picky compilers.
5668 34. If a DNS black list lookup found a CNAME record, but there were no A
5669 records associated with the domain it pointed at, Exim crashed.
5671 35. If a DNS black list lookup returned more than one A record, Exim ignored
5672 all but the first. It now scans all returned addresses if a particular IP
5673 value is being sought. In this situation, the contents of the
5674 $dnslist_value variable are a list of all the addresses, separated by a
5677 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5678 to accept a host name and all its aliases if the forward lookup for any of
5679 them yielded the IP address of the incoming connection. Now it accepts only
5680 those names whose forward lookup yields the correct IP address. Any other
5681 names are discarded. This closes a loophole whereby a rogue DNS
5682 administrator could create reverse DNS records to break through a
5683 wildcarded host restriction in an ACL.
5685 37. If a user filter or a system filter that ran in a subprocess used any of
5686 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5687 the wrong values were passed to the pipe command ($thisaddress had the
5688 value of $0, $0 had the value of $1, etc). This bug was introduced by
5689 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5691 38. Improved the line breaking for long SMTP error messages from ACLs.
5692 Previously, if there was no break point between 40 and 75 characters, Exim
5693 left the rest of the message alone. Two changes have been made: (a) I've
5694 reduced the minimum length to 35 characters; (b) if it can't find a break
5695 point between 35 and 75 characters, it looks ahead and uses the first one
5696 that it finds. This may give the occasional overlong line, but at least the
5697 remaining text gets split now.
5699 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5700 file descriptors might be low, and that setting 1000 would always raise it.
5701 It turns out that in some environments, the limit is already over 1000 and
5702 that lowering it causes trouble. So now Exim takes care not to decrease it.
5704 40. When delivering a message, the value of $return_path is set to $sender_
5705 address at the start of routing (routers may change the value). By an
5706 oversight, this default was not being set up when an address was tested by
5707 -bt or -bv, which affected the outcome if any router or filter referred to
5710 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5711 log only when "message" or "log_message" are set. However, if one of the
5712 conditions was an address verification, or a call to a nested ACL, the
5713 messages generated by the underlying test were being passed through. This
5714 no longer happens. The underlying message is available in $acl_verify_
5715 message for both "message" and "log_message" expansions, so it can be
5716 passed through if needed.
5718 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5719 new expansion $bh_ to give the encoded byte string without charset
5720 translation. Translation happens only if iconv() is available; HAVE_ICONV
5721 indicates this at build time. HEADERS_CHARSET gives the charset to
5722 translate to; headers_charset can change it in the configuration, and
5723 "headers charset" can change it in an individual filter file.
5725 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5726 that creates RFC 2047 encoded "words" labels them as that charset instead
5727 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5728 expansion operator; (ii) when Exim creates a From: line for a local
5729 message; (iii) when a header line is rewritten to include a "phrase" part.
5731 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5732 buggy, causing it to skip the first lines of messages whose message ID
5733 ended in 'D'. This would not have bitten before Exim release 4.14, because
5734 message IDs were unlikely to end in 'D' before then. The effect was to have
5735 incorrect size information for certain domains.
5737 45. #include "config.h" was missing at the start of the crypt16.c module. This
5738 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5741 46. If there was a timeout during a "random" callout check, Exim treated it as
5742 a failure of the random address, and carried on sending RSET and the real
5743 address. If the delay was just some slowness somewhere, the response to the
5744 original RCPT would be taken as a response to RSET and so on, causing
5745 mayhem of various kinds.
5747 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5748 when I implemented it. It didn't allow for the fact that some option values
5749 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5750 the right test for positive values.
5752 48. Domain names in DNS records are case-independent. Exim always looks them up
5753 in lower case. Some resolvers return domain names in exactly the case they
5754 appear in the zone file, that is, they may contain uppercase letters. Not
5755 all resolvers do this - some return always lower case. Exim was treating a
5756 change of case by a resolver as a change of domain, similar to a widening
5757 of a domain abbreviation. This triggered its re-routing code and so it was
5758 trying to route what was effectively the same domain again. This normally
5759 caused routing to fail (because the router wouldn't handle the domain
5760 twice). Now Exim checks for this case specially, and just changes the
5761 casing of the domain that it ultimately uses when it transmits the message
5764 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5767 50. If a filter generated a file delivery with a non-absolute name (possible if
5768 no home directory exists for the router), the forbid_file option was not
5771 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5772 the existing equality matching.
5774 52. Exim was using ints instead of ino_t variables in some places where it was
5775 dealing with inode numbers.
5777 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5778 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5779 called TMPDIR, and if it finds it is different, it changes its value.
5781 54. The smtp_printf() function is now made available to local_scan() so
5782 additional output lines can be written before returning. There is also an
5783 smtp_fflush() function to enable the detection of a dropped connection.
5784 The variables smtp_input and smtp_batched_input are exported to
5787 55. Changed the default runtime configuration: the message "Unknown user"
5788 has been removed from the ACL, and instead placed on the localuser router,
5789 using the cannot_route_message feature. This means that any verification
5790 failures that generate their own messages won't get overridden. Similarly,
5791 the "Unrouteable address" message that was in the ACL for unverifiable
5792 relay addresses has also been removed.
5794 56. Added hosts_avoid_esmtp to the smtp transport.
5796 57. The exicyclog script was not checking for the esoteric option
5797 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5798 will work only if exicyclog is run under the appropriate euid.
5800 58. Following a discussion on the list, the rules by which Exim recognises line
5801 endings on incoming messages have been changed. The -dropcr and drop_cr
5802 options are now no-ops, retained only for backwards compatibility. The
5803 following line terminators are recognized: LF CRLF CR. However, special
5804 processing applies to CR:
5806 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5807 nor a local message in the state where . is a terminator.
5809 (ii) If a bare CR is encountered in a header line, an extra space is added
5810 after the line terminator so as not to end the header. The reasoning
5811 behind this is that bare CRs in header lines are most likely either
5812 to be mistakes, or people trying to play silly games.
5814 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5815 was being incorrectly given as 18 bytes larger than it should have been.
5816 This is a VOB (very old bug).
5818 60. This may never have affected anything current, but just in case it has:
5819 When the local host is found other than at the start of a list of hosts,
5820 the local host, those with the same MX, and any that follow, are discarded.
5821 When the list in question was part of a longer list of hosts, the following
5822 hosts (not currently being processed) were also being discarded. This no
5823 longer happens. I'm not sure if this situation could ever has previously
5826 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5829 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5830 header. This is something that is recommended in a new Internet Draft, and
5831 is something that is documented as being done by Sendmail. There are two
5832 possible values. For messages generated by the autoreply transport, Exim
5835 Auto-Submitted: auto-replied
5837 whereas for all other generated messages (e.g. bounces) it adds
5839 Auto-Submitted: auto-generated
5841 63. The "personal" condition in filters now includes a test for the
5842 Auto-Submitted: header. If it contains the string "auto-" the message it
5843 not considered personal.
5845 64. Added rcpt_include_affixes as a generic transport option.
5847 65. Added queue_only_override (default true).
5849 66. Added the syslog_duplication option.
5851 67. If what should have been the first header line of a message consisted of
5852 a space followed by a colon, Exim was mis-interpreting it as a header line.
5853 It isn't of course - it is syntactically invalid and should therefore be
5854 treated as the start of the message body. The misbehaviour could have
5855 caused a number of strange effects, including loss of data in subsequent
5856 header lines, and spool format errors.
5858 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5859 client host had authenticated. This control can now be exercised by an ACL
5860 for more flexibility.
5862 69. By default, callouts do not happen when testing with -bh. There is now a
5863 variant, -bhc, which does actually run the callout code, including
5864 consulting and updating the callout cache.
5866 70. Added support for saslauthd authentication, courtesy of Alexander
5869 71. If statvfs() failed on the spool or log directories while checking their
5870 size for availability, Exim confusingly gave the error "space shortage".
5871 Furthermore, in debugging mode it crashed with a floating point exception.
5872 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5873 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5874 serious problem, Exim now writes to the main and panic logs when this
5875 happens, with details of the failure. It then refuses to accept the
5876 incoming message, giving the message "spool directory problem" or "log
5877 directory problem" with a 421 code for SMTP messages.
5879 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5880 0, 1, and 2 exist, because some OS complain for execs without them (see
5881 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5882 descriptors. However, the code omitted to check that the open succeeded,
5883 causing mysterious errors if for some reason the permissions on /dev/null
5884 got screwed. Now Exim writes a message to the main and panic logs, and
5885 bombs out if it can't open /dev/null.
5887 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5888 interact so that it is all more flexible. It is supposed to remain
5889 backwards compatible. Also added extra_local_interfaces.
5891 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5892 to bomb out with an assertion failure - to the client this appears as a
5893 connection drop. This problem occurs in the part of the code that was taken
5894 from the Samba project. Fortunately, the assertion is in a very simple
5895 function, so I have fixed this by reproducing the function inline in the
5896 one place where it is called, and arranging for authentication to fail
5897 instead of killing the process with assert().
5899 75. The SPA client code was not working when the server requested OEM rather
5900 than Unicode encoding.
5902 76. Added code to make require_files with a specific uid setting more usable in
5903 the case where statting the file as root fails - usually a non-root-mounted
5904 NFS file system. When this happens and the failure is EACCES, Exim now
5905 forks a subprocess and does the per-uid checking as the relevant uid.
5907 77. Added process_log_path.
5909 78. If log_file_path was not explicitly set, a setting of check_log_space or
5910 check_log_inodes was ignored.
5912 79. If a space check for the spool or log partitions fails, the incident is now
5913 logged. Of course, in the latter case the data may get lost...
5915 80. Added the %p formatting code to string_format() so that it can be used to
5916 print addresses in debug_print(). Adjusted all the address printing in the
5917 debugging in store.c to use %p rather than %d.
5919 81. There was a concern that a line of code in smtp_in.c could overflow a
5920 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5921 initially expressed, the concern was not well-founded, because trailing
5922 spaces are removed early. However, if the trailing spaces were followed by
5923 a NULL, they did not get removed, so the overflow was possible. Two fixes
5926 (a) I re-wrote the offending code in a cleaner fashion.
5927 (b) If an incoming SMTP command contains a NULL character, it is rejected
5930 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5931 runs initgroups(), so that if the Exim user is in any additional groups,
5932 they will be used during message reception.
5938 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob