1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
39 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
41 PP/02 Make -n do something, by making it not do something.
42 When combined with -bP, the name of an option is not output.
44 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
47 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
48 $sender_host_name and config options to manage this, and basic check
51 PP/05 DSCP support for outbound connections and control modifier for inbound.
53 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
54 (Only plugin which currently uses this is kerberos4, which nobody should
55 be using, but we should make it available and other future plugins might
56 conceivably use it, even though it would break NAT; stuff *should* be
57 using channel bindings instead).
59 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
60 name; added for Sendmail compatibility; requires admin caller.
61 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
62 just ignore it); requires trusted caller.
63 Also parse but ignore: -Ac -Am -X<logfile>
66 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
68 TL/02 Add +smtp_confirmation as a default logging option.
70 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
71 Patch by Magnus Holmgren from 2007-02-20.
73 TL/04 Bugzilla 1281 - Spec typo.
74 Bugzilla 1283 - Spec typo.
75 Bugzilla 1290 - Spec grammar fixes.
77 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
79 TL/06 Add Experimental DMARC support using libopendmarc libraries.
81 TL/07 Fix an out of order global option causing a segfault. Reported to dev
82 mailing list by by Dmitry Isaikin.
84 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
86 JH/02 Support "G" suffix to numbers in ${if comparisons.
88 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
90 NM/01 Bugzilla 1197 - Spec typo
91 Bugzilla 1196 - Spec examples corrections
93 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
95 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
96 gnutls_enable_pkcs11, but renamed to more accurately indicate its
99 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
100 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
102 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
103 "acl {{name}{arg}...}", and optional args on acl condition
106 JH/05 Permit multiple router/transport headers_add/remove lines.
108 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
110 JH/07 Avoid using a waiting database for a single-message-only transport.
111 Performance patch from Paul Fisher. Bugzilla 1262.
113 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
116 JH/09 Add $headers_added variable, with content from use of ACL modifier
117 add_header (but not yet added to the message). Bugzilla 199.
119 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
120 Pulled from Bugzilla 817 by Wolfgang Breyha.
122 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
124 (nb: this is the same fix as in Exim 4.80.1)
126 JH/11 Add A= logging on delivery lines, and a client_set_id option on
129 JH/12 Add optional authenticated_sender logging to A= and a log_selector
132 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
134 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
135 advertise SMTP AUTH mechanism to us, instead of a generic
136 protocol violation error. Also, make Exim more robust to bad
137 data from the Dovecot auth socket.
139 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
141 When a queue runner is handling a message, Exim first routes the
142 recipient addresses, during which it prunes them based on the retry
143 hints database. After that it attempts to deliver the message to
144 any remaining recipients. It then updates the hints database using
147 So if a recipient address works intermittently, it can get repeatedly
148 deferred at routing time. The retry hints record remains fresh so the
149 address never reaches the final cutoff time.
151 This is a fairly common occurrence when a user is bumping up against
152 their storage quota. Exim had some logic in its local delivery code
153 to deal with this. However it did not apply to per-recipient defers
154 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
156 This change adds a proper retry rule check during routing so that the
157 final cutoff time is checked against the message's age. We only do
158 this check if there is an address retry record and there is not a
159 domain retry record; this implies that previous attempts to handle
160 the address had the retry_use_local_parts option turned on. We use
161 this as an approximation for the destination being like a local
162 delivery, as in LMTP.
164 I suspect this new check makes the old local delivery cutoff check
165 redundant, but I have not verified this so I left the code in place.
167 TF/02 Correct gecos expansion when From: is a prefix of the username.
169 Test 0254 submits a message to Exim with the header
173 When I ran the test suite under the user fanf2, Exim expanded
174 the header to contain my full name, whereas it should have added
175 a Resent-Sender: header. It erroneously treats any prefix of the
176 username as equal to the username.
178 This change corrects that bug.
180 GF/01 DCC debug and logging tidyup
181 Error conditions log to paniclog rather than rejectlog.
182 Debug lines prefixed by "DCC: " to remove any ambiguity.
184 TF/03 Avoid unnecessary rebuilds of lookup-related code.
186 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
187 Bug spotted by Jeremy Harris; was flawed since initial commit.
188 Would have resulted in OCSP responses post-SNI triggering an Exim
189 NULL dereference and crash.
191 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
193 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
194 Bug detection, analysis and fix by Samuel Thibault.
195 Bugzilla 1331, Debian bug #698092.
197 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
199 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
200 Server implementation by Todd Lyons, client by JH.
201 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
202 config variable "prdr_enable" controls whether the server
203 advertises the facility. If the client requests PRDR a new
204 acl_data_smtp_prdr ACL is called once for each recipient, after
205 the body content is received and before the acl_smtp_data ACL.
206 The client is controlled by bolth of: a hosts_try_prdr option
207 on the smtp transport, and the server advertisement.
208 Default client logging of deliveries and rejections involving
209 PRDR are flagged with the string "PRDR".
211 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
212 fclose(). Diagnosis by Todd Lyons.
214 PP/17 Update configure.default to handle IPv6 localhost better.
215 Patch by Alain Williams (plus minor tweaks).
218 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
219 This is now consistent with GnuTLS, and is now documented: the
220 previous undocumented portable approach to treating the option as
221 unset was to force an expansion failure. That still works, and
222 an empty string is now equivalent.
224 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
225 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
226 not performing validation itself.
228 PP/20 Added force_command boolean option to pipe transport.
229 Patch from Nick Koston, of cPanel Inc.
231 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
234 TF/04 Added udpsend ACL modifer and hexquote expansion operator
236 PP/21 Fix eximon continuous updating with timestamped log-files.
237 Broken in a format-string cleanup in 4.80, missed when I repaired the
238 other false fix of the same issue.
239 Report and fix from Heiko Schlichting.
242 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
243 Report from Prashanth Katuri.
245 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
246 It's SecureTransport, so affects any MacOS clients which use the
247 system-integrated TLS libraries, including email clients.
249 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
250 using a MIME ACL for non-SMTP local injection.
251 Report and assistance in diagnosis by Warren Baker.
253 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
255 JH/16 Fix comparisons for 64b. Bugzilla 1385.
257 TL/09 Add expansion variable $authenticated_fail_id to keep track of
258 last id that failed so it may be referenced in subsequent ACL's.
260 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
263 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
264 ldap library initialization, allowing self-signed CA's to be
265 used. Also properly sets require_cert option later in code by
266 using NULL (global ldap config) instead of ldap handle (per
267 session). Bug diagnosis and testing by alxgomz.
269 TL/12 Enhanced documentation in the ratelimit.pl script provided in
270 the src/util/ subdirectory.
272 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
273 renamed to Transport Post Delivery Action by Jeremy Harris, as
276 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
277 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
278 redis_servers = needs to be configured which will be used by the redis
279 lookup. Patch from Warren Baker, of The Packet Hub.
281 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
283 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
284 hostname or reverse DNS when processing a host list. Used suggestions
285 from multiple comments on this bug.
287 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
289 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
290 Missed a few lines, added it to make the runtest require no keyboard
293 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
294 contains upper case chars. Make router use caseful_local_part.
296 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
297 support when GnuTLS has been built with p11-kit.
303 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
305 This, or similar/improved, will also be change PP/11 of 4.82.
311 PP/01 Handle short writes when writing local log-files.
312 In practice, only affects FreeBSD (8 onwards).
313 Bugzilla 1053, with thanks to Dmitry Isaikin.
315 NM/01 Bugzilla 949 - Documentation tweak
317 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
320 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
322 PP/02 Implemented gsasl authenticator.
324 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
326 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
327 `pkg-config foo` for cflags/libs.
329 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
330 with rest of GSASL and with heimdal_gssapi.
332 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
333 `pkg-config foo` for cflags/libs for the TLS implementation.
335 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
336 properties get this fed in as external SSF. A number of robustness
337 and debugging improvements to the cyrus_sasl authenticator.
339 PP/08 cyrus_sasl server now expands the server_realm option.
341 PP/09 Bugzilla 1214 - Log authentication information in reject log.
342 Patch by Jeremy Harris.
344 PP/10 Added dbmjz lookup type.
346 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
348 PP/12 MAIL args handles TAB as well as SP, for better interop with
349 non-compliant senders.
350 Analysis and variant patch by Todd Lyons.
352 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
353 Bug report from Lars Müller <lars@samba.org> (via SUSE),
354 Patch from Dirk Mueller <dmueller@suse.com>
356 PP/13 tls_peerdn now print-escaped for spool files.
357 Observed some $tls_peerdn in wild which contained \n, which resulted
358 in spool file corruption.
360 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
361 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
362 or write after TLS renegotiation, which otherwise led to messages
365 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
366 as a tracking header (ie: a signed header comes before the signature).
367 Patch from Wolfgang Breyha.
369 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
370 comma-sep list; embedded commas doubled.
372 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
374 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
376 Report and patch from Dmitry Banschikov.
378 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
379 Removed SSL_clear() after SSL_new() which led to protocol negotiation
380 failures. We appear to now support TLS1.1+ with Exim.
382 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
383 lets Exim select keys and certificates based upon TLS SNI from client.
384 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
385 before an outbound SMTP session. New log_selector, +tls_sni.
387 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
388 NULL dereference. Report and patch from Alun Jones.
390 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
391 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
392 Not seeing resolver debug output on NetBSD, but suspect this is a
393 resolver implementation change.
395 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
396 Left warnings. Added "eximon gdb" invocation mode.
398 PP/21 Defaulting "accept_8bitmime" to true, not false.
400 PP/22 Added -bw for inetd wait mode support.
402 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
403 locate the relevant includes and libraries. Made this the default.
405 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
406 Bugzilla 1246, report and most of solution from Tomasz Kusy.
408 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
409 This may cause build issues on older platforms.
411 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
412 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
413 gnutls_require_mac & gnutls_require_protocols (no longer supported).
414 Added SNI support via GnuTLS too.
415 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
417 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
419 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
420 Applied second patch from Janne, implementing suggestion to default
421 multiple-strings-in-record handling to match SPF spec.
423 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
425 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
426 read-only, out of scope).
427 Patch from Wolfgang Breyha, report from Stuart Northfield.
429 PP/29 Fix three issues highlighted by clang analyser static analysis.
430 Only crash-plausible issue would require the Cambridge-specific
431 iplookup router and a misconfiguration.
432 Report from Marcin Mirosław.
434 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
436 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
437 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
438 As part of this, removing so much warning spew let me fix some minor
439 real issues in debug logging.
441 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
442 assignment on my part. Fixed.
444 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
445 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
446 Janne Snabb (who went above and beyond: thank you).
448 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
449 string otherwise requires a connection and a bunch more work and it's
450 relatively easy to get wrong. Should also expose TLS library linkage
453 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
454 64-bit ${eval} (JH/03).
456 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
457 GNU libc to support some of the 64-bit stuff, should not lead to
458 conflicts. Defined before os.h is pulled in, so if a given platform
459 needs to override this, it can.
461 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
462 protection layer was required, which is not implemented.
463 Bugzilla 1254, patch from Wolfgang Breyha.
465 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
466 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
467 tls_dhparam take prime identifiers. Also unbreak combination of
468 OpenSSL+DH_params+TLSSNI.
470 PP/39 Disable SSLv2 by default in OpenSSL support.
476 PP/01 Solaris build fix for Oracle's LDAP libraries.
477 Bugzilla 1109, patch from Stephen Usher.
479 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
481 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
484 TF/02 Fix a couple more cases where we did not log the error message
485 when unlink() failed. See also change 4.74-TF/03.
487 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
488 lock up or crash if it happened to be inside a call to libc when it
489 got a SIGUSR1 from exiwhat.
491 The SIGUSR1 handler appends the current process status to the process
492 log which is later printed by exiwhat. It used to use the general
493 purpose logging code to do this, but several functions it calls are
494 not safe for signals.
496 The new output code in the SIGUSR1 handler is specific to the process
497 log, and simple enough that it's easy to inspect for signal safety.
498 Removing some special cases also simplifies the general logging code.
499 Removing the spurious timestamps from the process log simplifies
502 TF/04 Improved ratelimit ACL condition.
504 The /noupdate option has been deprecated in favour of /readonly which
505 has clearer semantics. The /leaky, /strict, and /readonly update modes
506 are mutually exclusive. The update mode is no longer included in the
507 database key; it just determines when the database is updated. (This
508 means that when you upgrde Exim will forget old rate measurements.)
510 Exim now checks that the per_* options are used with an update mode that
511 makes sense for the current ACL. For example, when Exim is processing a
512 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
513 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
514 must specify per_mail/readonly. If you omit the update mode it defaults to
515 /leaky where that makes sense (as before) or /readonly where required.
517 The /noupdate option is now undocumented but still supported for
518 backwards compatibility. It is equivalent to /readonly except that in
519 ACLs where /readonly is required you may specify /leaky/noupdate or
520 /strict/noupdate which are treated the same as /readonly.
522 A useful new feature is the /count= option. This is a generalization
523 of the per_byte option, so that you can measure the throughput of other
524 aggregate values. For example, the per_byte option is now equivalent
525 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
527 The per_rcpt option has been generalized using the /count= mechanism
528 (though it's more complicated than the per_byte equivalence). When it is
529 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
530 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
531 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
532 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
533 Note that using per_rcpt with a non-readonly update mode in more than
534 one ACL will cause the recipients to be double-counted. (The per_mail
535 and per_byte options don't have this problem.)
537 The handling of very low rates has changed slightly. If the computed rate
538 is less than the event's count (usually one) then this event is the first
539 after a long gap. In this case the rate is set to the same as this event's
540 count, so that the first message of a spam run is counted properly.
542 The major new feature is a mechanism for counting the rate of unique
543 events. The new per_addr option counts the number of different
544 recipients that someone has sent messages to in the last time period. It
545 behaves like per_rcpt if all the recipient addresses are different, but
546 duplicate recipient addresses do not increase the measured rate. Like
547 the /count= option this is a general mechanism, so the per_addr option
548 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
549 example, measure the rate that a client uses different sender addresses
550 with the options per_mail/unique=$sender_address. There are further
551 details in the main documentation.
553 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
555 TF/06 Removed a few PCRE remnants.
557 TF/07 Automatically extract Exim's version number from tags in the git
558 repository when doing development or release builds.
560 PP/02 Raise smtp_cmd_buffer_size to 16kB.
561 Bugzilla 879. Patch from Paul Fisher.
563 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
564 Heavily based on revision 40f9a89a from Simon Arlott's tree.
567 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
569 PP/05 Variable $av_failed, true if the AV scanner deferred.
570 Bugzilla 1078. Patch from John Horne.
572 PP/06 Stop make process more reliably on build failure.
573 Bugzilla 1087. Patch from Heiko Schlittermann.
575 PP/07 Make maildir_use_size_file an _expandable_ boolean.
576 Bugzilla 1089. Patch from Heiko Schlittermann.
578 PP/08 Handle ${run} returning more data than OS pipe buffer size.
579 Bugzilla 1131. Patch from Holger Weiß.
581 PP/09 Handle IPv6 addresses with SPF.
582 Bugzilla 860. Patch from Wolfgang Breyha.
584 PP/10 GnuTLS: support TLS 1.2 & 1.1.
586 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
589 PP/11 match_* no longer expand right-hand-side by default.
590 New compile-time build option, EXPAND_LISTMATCH_RHS.
591 New expansion conditions, "inlist", "inlisti".
593 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
595 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
597 PP/14 fix log_write() format string regression from TF/03.
598 Bugzilla 1152. Patch from Dmitry Isaikin.
604 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
606 PP/02 Harmonised TLS library version reporting; only show if debugging.
607 Layout now matches that introduced for other libraries in 4.74 PP/03.
609 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
611 PP/04 New "dns_use_edns0" global option.
613 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
616 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
617 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
619 TK/01 Updated PolarSSL code to 0.14.2.
620 Bugzilla 1097. Patch from Andreas Metzler.
622 PP/07 Catch divide-by-zero in ${eval:...}.
625 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
628 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
629 format-string attack -- SECURITY: remote arbitrary code execution.
631 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
632 time unintentionally subject to list matching rules, letting the header
633 cause arbitrary Exim lookups (of items which can occur in lists, *not*
634 arbitrary string expansion). This allowed for information disclosure.
636 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
637 INT_MIN/-1 -- value coerced to INT_MAX.
643 NM/01 Workround for PCRE version dependency in version reporting
646 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
647 This fixes portability to compilers other than gcc, notably
648 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
650 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
651 makefiles for portability to HP-UX and POSIX correctness.
653 PP/01 Permit LOOKUP_foo enabling on the make command-line.
654 Also via indented variable definition in the Makefile.
655 (Debugging by Oliver Heesakkers).
657 PP/02 Restore caching of spamd results with expanded spamd_address.
658 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
660 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
661 Improves build reliability. Fix from: Frank Elsner
663 NM/02 Fix wide character breakage in the rfc2047 coding
664 Fixes bug 1064. Patch from Andrey N. Oktyabrski
666 NM/03 Allow underscore in dnslist lookups
667 Fixes bug 1026. Patch from Graeme Fowler
669 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
670 Code patches from Adam Ciarcinski of NetBSD.
672 NM/04 Fixed exiqgrep to cope with mailq missing size issue
675 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
676 is logged, to avoid truncation. Patch from John Horne.
678 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
679 Patch from Jakob Hirsch.
681 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
682 SQL string expansion failure details.
683 Patch from Andrey Oktyabrski.
685 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
686 Patch from Simon Arlott.
688 PP/09 New lookups functionality failed to compile on old gcc which rejects
689 extern declarations in function scope.
690 Patch from Oliver Fleischmann
692 PP/10 Use sig_atomic_t for flags set from signal handlers.
693 Check getgroups() return and improve debugging.
694 Fixed developed for diagnosis in bug 927 (which turned out to be
697 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
698 Patch from Mark Zealey.
700 PP/12 Bugzilla 1056: Improved spamd server selection.
701 Patch from Mark Zealey.
703 PP/13 Bugzilla 1086: Deal with maildir quota file races.
704 Based on patch from Heiko Schlittermann.
706 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
707 Patch from Uwe Doering, sign-off by Michael Haardt.
709 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
710 variable declaration deep within a block. Bug and patch from
713 PP/15 lookups-Makefile IRIX compatibilty coercion.
715 PP/16 Make DISABLE_DKIM build knob functional.
717 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
718 Patch by Simon Arlott
720 TF/03 Fix valgrind.h portability to C89 compilers that do not support
721 variable argument macros. Our copy now differs from upstream.
727 TF/01 Failure to get a lock on a hints database can have serious
728 consequences so log it to the panic log.
730 TF/02 Log LMTP confirmation messages in the same way as SMTP,
731 controlled using the smtp_confirmation log selector.
733 TF/03 Include the error message when we fail to unlink a spool file.
735 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
736 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
737 for maintaining out-of-tree patches for some time.
739 PP/01 Bugzilla 139: Documentation and portability issues.
740 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
741 Handle per-OS dynamic-module compilation flags.
743 PP/02 Let /dev/null have normal permissions.
744 The 4.73 fixes were a little too stringent and complained about the
745 permissions on /dev/null. Exempt it from some checks.
746 Reported by Andreas M. Kirchwitz.
748 PP/03 Report version information for many libraries, including
749 Exim version information for dynamically loaded libraries. Created
750 version.h, now support a version extension string for distributors
751 who patch heavily. Dynamic module ABI change.
753 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
754 privilege escalation vulnerability whereby the Exim run-time user
755 can cause root to append content of the attacker's choosing to
758 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
761 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
762 If dropping privileges for untrusted macros, we disabled normal logging
763 on the basis that it would fail; for the Exim run-time user, this is not
764 the case, and it resulted in successful deliveries going unlogged.
765 Fixed. Reported by Andreas Metzler.
771 PP/01 Date: & Message-Id: revert to normally being appended to a message,
772 only prepend for the Resent-* case. Fixes regression introduced in
773 Exim 4.70 by NM/22 for Bugzilla 607.
775 PP/02 Include check_rfc2047_length in configure.default because we're seeing
776 increasing numbers of administrators be bitten by this.
778 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
780 PP/03 Bugzilla 994: added openssl_options main configuration option.
782 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
784 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
786 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
788 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
789 without a peer certificate, leading to a segfault because of an
790 assumption that peers always have certificates. Be a little more
791 paranoid. Problem reported by Martin Tscholak.
793 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
794 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
795 NB: ClamAV planning to remove STREAM in "middle of 2010".
796 CL also introduces -bmalware, various -d+acl logging additions and
797 more caution in buffer sizes.
799 PP/09 Implemented reverse_ip expansion operator.
801 PP/10 Bugzilla 937: provide a "debug" ACL control.
803 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
805 PP/12 Bugzilla 973: Implement --version.
807 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
809 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
811 PP/15 Bugzilla 816: support multiple condition rules on Routers.
813 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
814 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
815 ignore trailing whitespace.
817 JJ/02 prevent non-panic DKIM error from being sent to paniclog
819 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
822 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
823 Notification from Dr Andrew Aitchison.
825 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
826 ExtendedDetectionInfo response format.
827 Notification from John Horne.
829 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
832 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
833 XSL and documented dependency on system catalogs, with examples of how
836 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
839 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
840 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
841 configuration file which is writeable by the Exim user or group.
843 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
844 of configuration files to cover files specified with the -C option if
845 they are going to be used with root privileges, not just the default
848 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
849 option (effectively making it always true).
851 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
852 files to be used while preserving root privileges.
854 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
855 that rogue child processes cannot use them.
857 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
858 run-time user, instead of root.
860 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
861 Exim run-time user without dropping privileges.
863 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
864 result string, instead of calling string_vformat() twice with the same
867 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
868 for other users. Others should always drop root privileges if they use
869 -C on the command line, even for a whitelisted configure file.
871 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
873 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
879 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
880 $data_path, and $header_path variables; fixed documentation bugs and
883 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
884 exipick to access non-standard spools, including the "frozen" queue
887 NM/01 Bugzilla 965: Support mysql stored procedures.
888 Patch from Alain Williams
890 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
892 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
893 Patch from Andreas Metzler
895 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
896 Patch from Kirill Miazine
898 NM/05 Bugzilla 671: Added umask to procmail example.
900 JJ/03 installed exipick 20100323.0, fixing doc bug
902 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
903 directory. Notification and patch from Dan Rosenberg.
905 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
907 TK/02 Improve log output when DKIM signing operation fails.
909 MH/01 Treat the transport option dkim_domain as a colon separated
910 list, not as a single string, and sign the message with each element,
911 omitting multiple occurences of the same signer.
913 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
914 Bugzilla 985, 986. Patch by Simon Arlott
916 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
917 Patch by Simon Arlott
919 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
920 MBX locking. Notification from Dan Rosenberg.
926 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
928 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
930 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
932 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
934 NM/04 Bugzilla 915: Flag broken perl installation during build.
940 TK/01 Added patch by Johannes Berg that expands the main option
941 "spamd_address" if it starts with a dollar sign.
943 TK/02 Write list of recipients to X-Envelope-Sender header when building
944 the mbox-format spool file for content scanning (suggested by Jakob
947 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
948 (http://www.dcc-servers.net/) support via dccifd. Activated by
949 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
951 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
952 by Mark Daniel Reidel <mr@df.eu>.
954 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
955 When building exim an external PCRE library is now needed -
956 PCRE is a system library on the majority of modern systems.
957 See entry on PCRE_LIBS in EDITME file.
959 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
960 conversation. Added nologin parameter to request.
961 Patch contributed by Kirill Miazine.
963 TF/01 Do not log submission mode rewrites if they do not change the address.
965 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
967 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
968 log files in place. Contributed by Roberto Lima.
970 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
972 TF/03 Bugzilla 615: When checking the local_parts router precondition
973 after a local_part_suffix or local_part_prefix option, Exim now
974 does not use the address's named list lookup cache, since this
975 contains cached lookups for the whole local part.
977 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
978 Robert Millan. Documentation is in experimental-spec.txt.
980 TF/04 Bugzilla 668: Fix parallel build (make -j).
982 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
984 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
985 Patch provided by Jan Srzednicki.
987 TF/05 Leading white space used to be stripped from $spam_report which
988 wrecked the formatting. Now it is preserved.
990 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
991 that they are available at delivery time.
993 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
995 TF/08 TLS error reporting now respects the incoming_interface and
996 incoming_port log selectors.
998 TF/09 Produce a more useful error message if an SMTP transport's hosts
999 setting expands to an empty string.
1001 NM/06 Bugzilla 744: EXPN did not work under TLS.
1002 Patch provided by Phil Pennock.
1004 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1005 Patch provided by Richard Godbee.
1007 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1008 acl_smtp_notquit, added index entry.
1010 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1011 Patch provided by Eugene Bujak.
1013 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1014 accept(). Patch provided by Maxim Dounin.
1016 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1017 Patch provided by Phil Pennock.
1019 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1021 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1022 Patch provided by Brad "anomie" Jorsch.
1024 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1025 Patch provided by Dean Brooks.
1027 TK/05 Add native DKIM support (does not depend on external libraries).
1029 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1030 Patch provided by Graeme Fowler.
1032 NM/16 Bugzilla 851: Documentation example syntax fix.
1034 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1036 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1039 NM/19 Bugzilla 745: TLS version reporting.
1040 Patch provided by Phil Pennock.
1042 NM/20 Bugzilla 167: bool: condition support.
1043 Patch provided by Phil Pennock.
1045 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1046 clients. Patch provided by Phil Pennock.
1048 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1049 Patch provided by Brad "anomie" Jorsch.
1051 NM/23 Bugzilla 687: Fix misparses in eximstats.
1052 Patch provided by Heiko Schlittermann.
1054 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1055 Patch provided by Heiko Schlittermann.
1057 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1058 plus update to original patch.
1060 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1062 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1063 Patch provided by David Brownlee.
1065 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1067 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1069 NM/30 Bugzilla 888: TLS documentation bugfixes.
1071 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1073 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1074 Unlike the original bugzilla I have changed all shell scripts in src tree.
1076 NM/33 Bugzilla 898: Transport filter timeout fix.
1077 Patch by Todd Rinaldo.
1079 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1080 Patch by Serge Demonchaux.
1082 NM/35 Bugzilla 39: Base64 decode bug fixes.
1083 Patch by Jakob Hirsch.
1085 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1087 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1089 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1091 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1097 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1098 ALT-N's libdkim that I have put here:
1099 http://duncanthrax.net/exim-experimental/
1101 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1102 were called 'true' and it seems that C99 defines that as a reserved
1103 keyword to be used with 'bool' variable types. That means you could
1104 not include C99-style headers which use bools without triggering
1105 build errors in sieve.c.
1107 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1108 as mailq or other aliases. Changed the --help handling significantly
1109 to do whats expected. exim_usage() emits usage/help information.
1111 SC/01 Added the -bylocaldomain option to eximstats.
1113 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1115 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1117 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1123 PH/01 Another patch from the Sieve maintainer.
1125 PH/02 When an IPv6 address is converted to a string for single-key lookup
1126 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1127 dots are used instead of colons so that keys in lsearch files need not
1128 contain colons. This was done some time before quoting was made available
1129 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1130 (notated using the quote facility) so as to distinguish them from IPv4
1131 keys. This meant that lookups for IP addresses in host lists did not work
1132 for iplsearch lookups.
1134 This has been fixed by arranging for IPv6 addresses to be expressed with
1135 colons if the lookup type is iplsearch. This is not incompatible, because
1136 previously such lookups could never work.
1138 The situation is now rather anomolous, since one *can* have colons in
1139 ordinary lsearch keys. However, making the change in all cases is
1140 incompatible and would probably break a number of configurations.
1142 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1145 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1146 conversion specification without a maximum field width, thereby enabling
1147 a rogue spamd server to cause a buffer overflow. While nobody in their
1148 right mind would setup Exim to query an untrusted spamd server, an
1149 attacker that gains access to a server running spamd could potentially
1150 exploit this vulnerability to run arbitrary code as the Exim user.
1152 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1153 $primary_hostname instead of what libspf2 thinks the hosts name is.
1155 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1156 a directory entry by the name of the lookup key. Previously, if a
1157 symlink pointed to a non-existing file or a file in a directory that
1158 Exim lacked permissions to read, a lookup for a key matching that
1159 symlink would fail. Now it is enough that a matching directory entry
1160 exists, symlink or not. (Bugzilla 503.)
1162 PH/03 The body_linecount and body_zerocount variables are now exported in the
1165 PH/04 Added the $dnslist_matched variable.
1167 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1168 This means they are set thereafter only if the connection becomes
1171 PH/06 Added the client_condition to authenticators so that some can be skipped
1172 by clients under certain conditions.
1174 PH/07 The error message for a badly-placed control=no_multiline_responses left
1175 "_responses" off the end of the name.
1177 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1179 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1180 (without spaces) instead of just copying the configuration text.
1182 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1184 PH/11 Added $max_received_linelength.
1186 PH/12 Added +ignore_defer and +include_defer to host lists.
1188 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1189 way in which PCRE > 7.0 is built.
1191 PH/14 Implemented queue_only_load_latch.
1193 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1194 MAIL command. The effect was to mangle the value on 64-bit systems.
1196 PH/16 Another patch from the Sieve maintainer.
1198 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1200 PH/18 If a system quota error occurred while trying to create the file for
1201 a maildir delivery, the message "Mailbox is full" was not appended to the
1202 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1203 only to a quota excession during the actual writing of the file.
1205 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1206 characters?) which causes problems in log lines. The DN values are now
1207 passed through string_printing() before being added to log lines.
1209 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1210 and InterBase are left for another time.)
1212 PH/21 Added message_body_newlines option.
1214 PH/22 Guard against possible overflow in moan_check_errorcopy().
1216 PH/23 POSIX allows open() to be a macro; guard against that.
1218 PH/24 If the recipient of an error message contained an @ in the local part
1219 (suitably quoted, of course), incorrect values were put in $domain and
1220 $local_part during the evaluation of errors_copy.
1226 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1227 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1230 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1231 issue a MAIL command.
1233 PH/02 In an ACL statement such as
1235 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1237 if a client was not listed at all, or was listed with a value other than
1238 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1239 the condition was not true (as it should be), so access was not denied.
1240 The bug was that the ! inversion was incorrectly passed on to the second
1241 item. This has been fixed.
1243 PH/03 Added additional dnslists conditions == and =& which are different from
1244 = and & when the dns lookup returns more than one IP address.
1246 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1247 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1249 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1250 FSYNC, which compiles an option called disable_fsync that allows for
1251 bypassing fsync(). The documentation is heavily laced with warnings.
1253 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1255 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1256 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1257 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1258 including adding "make clean"; (3) Added -fPIC when compiling the test
1259 dynamically loaded module, to get rid of a warning.
1261 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1262 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1263 = 0s. The bug is otherwise harmless.
1265 PH/07 There was a bug in the dovecot authenticator such that the value of
1266 $auth1 could be overwritten, and so not correctly preserved, after a
1267 successful authentication. This usually meant that the value preserved by
1268 the server_setid option was incorrect.
1270 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1272 PH/09 Installed PCRE release 7.0.
1274 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1275 run for batched SMTP input. It is now run at the start of every message
1276 in the batch. While fixing this I discovered that the process information
1277 (output by running exiwhat) was not always getting set for -bs and -bS
1278 input. This is fixed, and it now also says "batched" for BSMTP.
1280 PH/11 Added control=no_pipelining.
1282 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1283 patch, slightly modified), and move the expansion of helo_data till after
1284 the connection is made in the smtp transport (so it can use these
1287 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1289 PH/14 Added log_selector = +pid.
1291 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1293 PH/16 Add ${if forany and ${if forall.
1295 PH/17 Added dsn_from option to vary the From: line in DSNs.
1297 PH/18 Flush SMTP output before performing a callout, unless control =
1298 no_callout_flush is set.
1300 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1301 was true (the default) a successful delivery failed to delete the retry
1302 item, thus causing premature timeout of the address. The bug is now
1305 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1307 PH/21 Long custom messages for fakedefer and fakereject are now split up
1308 into multiline reponses in the same way that messages for "deny" and
1309 other ACL rejections are.
1311 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1312 with slight modification.
1314 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1315 draft, changing the syntax and factoring some duplicate code".
1317 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1318 for deliveries of the second and subsequent messages over the same SMTP
1321 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1322 ${reduce, with only minor "tidies".
1324 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1326 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1327 expansion side effects.
1329 PH/27 When a message times out after an over-quota error from an Exim-imposed
1330 quota, the bounce message says "mailbox is full". This message was not
1331 being given when it was a system quota that was exceeded. It now should
1334 MH/03 Made $recipients available in local_scan(). local_scan() already has
1335 better access to the recipient list through recipients_list[], but
1336 $recipients can be useful in postmaster-provided expansion strings.
1338 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1339 in the case of a MAIL command with additional options following the
1340 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1341 were accidentally chopped off.
1343 PH/29 SMTP synchronization checks are implemented when a command is read -
1344 there is a check that no more input is waiting when there shouldn't be
1345 any. However, for some commands, a delay in an ACL can mean that it is
1346 some time before the response is written. In this time, more input might
1347 arrive, invalidly. So now there are extra checks after an ACL has run for
1348 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1349 pipelining has not been advertised.
1351 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1353 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1354 returned in the SMTP response when smtp_return_error_details was false.
1355 This has been fixed.
1357 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1358 instead of the C I/O that was originally supplied, because problems were
1359 reported on Solaris.
1361 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1362 Exim which did not show up earlier: it was assuming that a call to
1363 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1364 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1365 was a macro that became an assignment, so it seemed to work. This has
1366 changed to a proper function call with a void return, hence the compile
1367 error. Exim's code has been fixed.
1369 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1372 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1374 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1376 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1377 because it turns out to be too big or there is a timeout) there is no
1378 "Completed" line in the log. When some messages of this type were
1379 selected by exigrep, they were listed as "not completed". Others were
1380 picked up by some special patterns. I have improved the selection
1381 criteria to be more general.
1383 PH/38 The host_find_failed option in the manualroute router can now be set
1384 to "ignore", to completely ignore a host whose IP address cannot be
1385 found. If all hosts are ignored, the behaviour is controlled by the new
1386 host_all_ignored option.
1388 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1389 homing or because of multiple MX records with /mx) generated more than
1390 one IP address, and the following item turned out to be the local host,
1391 all the secondary addresses of the first item were incorrectly removed
1392 from the list, along with the local host and any following hosts (which
1393 is what is supposed to happen).
1395 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1396 whoever called Exim into the -H file. In the case of the daemon it was
1397 behaving confusingly. When first started, it used values for whoever
1398 started the daemon, but after a SIGHUP it used the Exim user (because it
1399 calls itself on a restart). I have changed the code so that it now always
1402 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1403 message are rejected with the same error (e.g. no authentication or bad
1404 sender address), and a DATA command is nevertheless sent (as can happen
1405 with PIPELINING or a stupid MUA), the error message that was given to the
1406 RCPT commands is included in the rejection of the DATA command. This is
1407 intended to be helpful for MUAs that show only the final error to their
1410 PH/42 Another patch from the Sieve maintainer.
1412 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1413 Eximstats - Fixed some broken HTML links and added missing column headers
1415 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1416 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1418 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1420 PH/43 Yet another patch from the Sieve maintainer.
1422 PH/44 I found a way to check for a TCP/IP connection going away before sending
1423 the response to the final '.' that terminates a message, but only in the
1424 case where the client has not sent further data following the '.'
1425 (unfortunately, this is allowed). However, in many cases there won't be
1426 any further data because there won't be any more messages to send. A call
1427 to select() can be used: if it shows that the input is "ready", there is
1428 either input waiting, or the socket has been closed. An attempt to read
1429 the next input character can distinguish the two cases. Previously, Exim
1430 would have sent an OK response which the client would never have see.
1431 This could lead to message repetition. This fix should cure that, at
1432 least in a lot of common cases.
1434 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1435 advertised in response to EHLO.
1441 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1442 fixed by 4.65/MH/01 (is this a record?) are fixed:
1444 (i) An empty string was always treated as zero by the numeric comparison
1445 operators. This behaviour has been restored.
1447 (ii) It is documented that the numeric comparison operators always treat
1448 their arguments as decimal numbers. This was broken in that numbers
1449 starting with 0 were being interpreted as octal.
1451 While fixing these problems I realized that there was another issue that
1452 hadn't been noticed. Values of message_size_limit (both the global option
1453 and the transport option) were treated as octal if they started with 0.
1454 The documentation was vague. These values are now always treated as
1455 decimal, and I will make that clear in the documentation.
1461 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1462 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1465 MH/01 Don't check that the operands of numeric comparison operators are
1466 integers when their expansion is in "skipping" mode (fixes bug
1467 introduced by 4.64-PH/07).
1469 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1470 child addresses, Exim now panics and dies. Previously, because the count
1471 is held in a short int, deliveries were likely to be lost. As such a
1472 large number of recipients for a single message is ridiculous
1473 (performance will be very, very poor), I have chosen to impose a limit
1474 rather than extend the field.
1480 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1481 leftover -K file (the existence of which was triggered by #402).
1482 While we were at it, introduced process PID as part of the -K
1483 filename. This should rule out race conditions when creating
1486 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1487 processing considerably. Previous code took too long for large mails,
1488 triggering a timeout which in turn triggers #401.
1490 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1491 in the DK code in transports.c. sendfile() is not really portable,
1492 hence the _LINUX specificness.
1494 TF/01 In the add_headers option to the mail command in an Exim filter,
1495 there was a bug that Exim would claim a syntax error in any
1496 header after the first one which had an odd number of characters
1499 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1500 callout verification, Exim cached a "reject" for the entire domain. This
1501 is correct for most verifications, but it is not correct for a recipient
1502 verification with use_sender or use_postmaster set, because in that case
1503 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1504 case of MAIL FROM:<> rejection from other early rejections (e.g.
1505 rejection of HELO). When verifying a recipient using a non-null MAIL
1506 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1507 Whatever the result of the callout, the value of the domain cache is
1508 left unchanged (for any other kind of callout, getting as far as trying
1509 RCPT means that the domain itself is ok).
1511 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1514 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1515 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1516 session, the daemon ignores ECONNECT errors and logs others; it now
1517 ignores EPIPE as well.
1519 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1520 (quoted-printable decoding).
1522 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1523 later the small subsequent patch to fix an introduced bug.
1525 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1527 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1529 PH/08 An error is now given if message_size_limit is specified negative.
1531 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1532 to be given (somewhat) arbitrary names.
1534 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1537 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1538 miscellaneous code fixes
1540 PH/10 Added the log_reject_target ACL modifier to specify where to log
1543 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1544 hostname. This is wrong, because it relates to the incoming message (and
1545 probably the interface on which it is arriving) and not to the outgoing
1546 callout (which could be using a different interface). This has been
1547 changed to use the value of the helo_data option from the smtp transport
1548 instead - this is what is used when a message is actually being sent. If
1549 there is no remote transport (possible with a router that sets up host
1550 addresses), $smtp_active_hostname is used.
1552 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1553 tweaks were necessary in order to get it to work (see also 21 below):
1554 (a) The code assumed that strncpy() returns a negative number on buffer
1555 overflow, which isn't the case. Replaced with Exim's string_format()
1557 (b) There were several signed/unsigned issues. I just did the minimum
1558 hacking in of casts. There is scope for a larger refactoring.
1559 (c) The code used strcasecmp() which is not a standard C function.
1560 Replaced with Exim's strcmpic() function.
1561 (d) The code set only $1; it now sets $auth1 as well.
1562 (e) A simple test gave the error "authentication client didn't specify
1563 service in request". It would seem that Dovecot has changed its
1564 interface. Fortunately there's a specification; I followed it and
1565 changed what the client sends and it appears to be working now.
1567 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1570 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1571 address A is aliased to B and C, where B exists and C does not. Without
1572 -v the output is "A verified" because verification stops after a
1573 successful redirection if more than one address is generated. However,
1574 with -v the child addresses are also verified. Exim was outputting "A
1575 failed to verify" and then showing the successful verification for C,
1576 with its parentage. It now outputs "B failed to verify", showing B's
1577 parentage before showing the successful verification of C.
1579 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1580 look up a TXT record in a specific list after matching in a combined
1583 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1584 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1585 they consult the DNS. I had assumed they would set it the way they
1586 wanted; and indeed my experiments on Linux seem to show that in some
1587 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1588 To be on the safe side, however, I have now made the interface to
1589 host_find_byname() similar to host_find_bydns(), with an argument
1590 containing the DNS resolver options. The host_find_byname() function now
1591 sets these options at its start, just as host_find_bydns() does. The smtp
1592 transport options dns_qualify_single and dns_search_parents are passed to
1593 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1594 of host_find_byname() use the default settings of RES_DEFNAMES
1595 (qualify_single) but not RES_DNSRCH (search_parents).
1597 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1598 spool_read_header() do less string testing, by means of a preliminary
1599 switch on the second character of optional "-foo" lines. (This is
1600 overdue, caused by the large number of possibilities that now exist.
1601 Originally there were few.) While I was there, I also converted the
1602 str(n)cmp tests so they don't re-test the leading "-" and the first
1603 character, in the hope this might squeeze out yet more improvement.
1605 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1606 flag allowing group syntax was set by the header_syntax check but not
1607 turned off, possible causing trouble later; (2) The flag was not being
1608 set at all for the header_verify test, causing "group"-style headers to
1609 be rejected. I have now set it in this case, and also caused header_
1610 verify to ignore an empty address taken from a group. While doing this, I
1611 came across some other cases where the code for allowing group syntax
1612 while scanning a header line wasn't quite right (mostly, not resetting
1613 the flag correctly in the right place). These bugs could have caused
1614 trouble for malformed header lines. I hope it is now all correct.
1616 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1617 with the "reply" argument non-NULL. The code, however (which originally
1618 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1619 but it didn't always do it. This confused somebody who was copying the
1620 code for some other use. I have removed all the tests.
1622 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1623 feature that was used to support insecure browsers during the U.S. crypto
1624 embargo. It requires special client support, and Exim is probably the
1625 only MTA that supported it -- and would never use it because real RSA is
1626 always available. This code has been removed, because it had the bad
1627 effect of slowing Exim down by computing (never used) parameters for the
1628 RSA_EXPORT functionality.
1630 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1631 authenticator to fail if there's a tab character in the incoming data
1632 (there should never be unless someone is messing about, as it's supposed
1633 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1634 if the connection is using TLS or if the remote IP is the same as the
1635 local IP, and the "valid-client-cert option" if a client certificate has
1638 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1639 authenticators. This can be used for authorization after authentication
1640 succeeds. (In the case of plaintext, it servers for both authentication
1643 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1644 if any retry times were supplied.
1646 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1647 connection, where there is no client IP address to check. In this
1648 situation, the verify now always succeeds.
1650 PH/25 Applied John Jetmore's -Mset patch.
1652 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1654 PH/27 In a string expansion for a processed (not raw) header when multiple
1655 headers of the same name were present, leading whitespace was being
1656 removed from all of them, but trailing whitespace was being removed only
1657 from the last one. Now trailing whitespace is removed from each header
1658 before concatenation. Completely empty headers in a concatenation (as
1659 before) are ignored.
1661 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1662 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1664 PH/29 [Removed. This was a change that I later backed out, and forgot to
1665 correct the ChangeLog entry (that I had efficiently created) before
1666 committing the later change.]
1668 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1669 address errors (4xx response to RCPT) over the same connection as other
1670 messages routed to the same hosts. Such deliveries are always "forced",
1671 so retry times are not inspected. This resulted in far too many retries
1672 for the affected addresses. The effect occurred only when there were more
1673 hosts than the hosts_max_try setting in the smtp transport when it had
1674 the 4xx errors. Those hosts that it had tried were not added to the list
1675 of hosts for which the message was waiting, so if all were tried, there
1676 was no problem. Two fixes have been applied:
1678 (i) If there are any address or message errors in an SMTP delivery, none
1679 of the hosts (tried or untried) are now added to the list of hosts
1680 for which the message is waiting, so the message should not be a
1681 candidate for sending over the same connection that was used for a
1682 successful delivery of some other message. This seems entirely
1683 reasonable: after all the message is NOT "waiting for some host".
1684 This is so "obvious" that I'm not sure why it wasn't done
1685 previously. Hope I haven't missed anything, but it can't do any
1686 harm, as the worst effect is to miss an optimization.
1688 (ii) If, despite (i), such a delivery is accidentally attempted, the
1689 routing retry time is respected, so at least it doesn't keep
1690 hammering the server.
1692 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1693 in ${readsocket because some servers need this prod.
1695 PH/32 Added some extra debug output when updating a wait-xxx database.
1697 PH/33 The hint "could be header name not terminated by colon", which has been
1698 given for certain expansion errors for a long time, was not being given
1699 for the ${if def:h_colon_omitted{... case.
1701 PH/34 The spec says: "With one important exception, whenever a domain list is
1702 being scanned, $domain contains the subject domain." There was at least
1703 one case where this was not true.
1705 PH/35 The error "getsockname() failed: connection reset by peer" was being
1706 written to the panic log as well as the main log, but it isn't really
1707 panic-worthy as it just means the connection died rather early on. I have
1708 removed the panic log writing for the ECONNRESET error when getsockname()
1711 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1712 runs only) independently of the message's sender address. This meant
1713 that, if the 4xx error was in fact related to the sender, a different
1714 message to the same recipient with a different sender could confuse
1715 things. In particualar, this can happen when sending to a greylisting
1716 server, but other circumstances could also provoke similar problems.
1717 I have changed the default so that the retry time for these errors is now
1718 based a combination of the sender and recipient addresses. This change
1719 can be overridden by setting address_retry_include_sender=false in the
1722 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1723 remote server are returned as part of bounce messages. This was not
1724 happening for LMTP over a pipe (the lmtp transport), but now it is the
1725 same for both kinds of LMTP.
1727 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1728 in header lines that were in fact CNAMEs. This is no longer the case.
1730 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1731 and queue runs started by the daemon processed all messages. This has
1732 been fixed so that -R and -S can now usefully be given with -q<time>.
1734 PH/40 Import PCRE release 6.7 (fixes some bugs).
1736 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1738 PH/42 Give an error if -q is specified more than once.
1740 PH/43 Renamed the variables $interface_address and $interface_port as
1741 $received_ip_address and $received_port, to make it clear that these
1742 values apply to message reception, and not to the outgoing interface when
1743 a message is delivered. (The old names remain recognized, of course.)
1745 PH/44 There was no timeout on the connect() call when using a Unix domain
1746 socket in the ${readsocket expansion. There now is.
1748 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1749 be meaningful with "accept".
1751 SC/01 Eximstats V1.43
1752 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1754 SC/02 Eximstats V1.44
1755 Use a glob alias rather than an array ref in the generated
1756 parser. This improves both readability and performance.
1758 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1759 Collect SpamAssassin and rejection statistics.
1760 Don't display local sender or destination tables unless
1761 there is data to show.
1762 Added average volumes into the top table text output.
1764 SC/04 Eximstats V1.46
1765 Collect data on the number of addresses (recipients)
1766 as well as the number of messages.
1768 SC/05 Eximstats V1.47
1769 Added 'Message too big' to the list of mail rejection
1770 reasons (thanks to Marco Gaiarin).
1772 SC/06 Eximstats V1.48
1773 Mainlog lines which have GMT offsets and are too short to
1774 have a flag are now skipped.
1776 SC/07 Eximstats V1.49 (Alain Williams)
1777 Added the -emptyok flag.
1779 SC/08 Eximstats V1.50
1780 Fixes for obtaining the IP address from reject messages.
1782 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1783 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1784 whitesspace changes from 4.64-PH/27
1786 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1789 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1790 are found, allow negative numbers in numeric criteria)
1792 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1794 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1795 to match changes made in 4.64-PH/43
1797 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1799 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1800 "domain" is actually an IP address, and force a failure. This locks out
1801 those revolvers/nameservers that support "A-for-A" lookups, in
1802 contravention of the specifications.
1804 PH/48 When a host name was looked up from an IP address, and the subsequent
1805 forward lookup of the name timed out, the host name was left in
1806 $sender_host_name, contrary to the specification.
1808 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1809 restricted to single-key lookups, Exim was not diagnosing an error if
1810 * or *@ was used with a query-style lookup.
1812 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1814 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1815 long ago, but noone interested enough thought of it. Let's just say that
1816 the "1.1" means that there are some new functions that weren't there at
1817 some point in the past.
1819 PH/51 Error processing for expansion failure of helo_data from an smtp
1820 transport during callout processing was broken.
1822 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1823 tested/used via the -bh/-bhc/-bs options.
1825 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1826 bug, fixed in subsequent PCRE releases).
1828 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1829 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1831 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1837 SC/01 Use a glob alias rather than an array ref in eximstats generated
1838 parser. This improves both readability and performance.
1840 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1841 Don't display local sender or destination tables in eximstats unless
1842 there is data to show.
1843 Added average volumes into the eximstats top table text output.
1845 SC/03 Collect data on the number of addresses (recipients) as well
1846 as the number of messages in eximstats.
1848 TF/01 Correct an error in the documentation for the redirect router. Exim
1849 does (usually) call initgroups() when daemonizing.
1851 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1852 with consistent privilege compared to when running as a daemon.
1854 TF/03 Note in the spec that $authenticated_id is not set for local
1855 submissions from trusted users.
1857 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1858 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1860 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1861 by adding some example configuration directives to the default
1862 configuration file. A little bit of work is required to uncomment the
1863 directives and define how usernames and passwords are checked, but
1864 there is now a framework to start from.
1866 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1867 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1868 without this. I don't know how relevant this is to other LDAP libraries.
1870 PH/02 Add the verb name to the "unknown ACL verb" error.
1872 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1874 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1876 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1877 directory not expanded when it should be if an expanded home directory
1878 was set for the address (which is overridden by the transport).
1880 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1883 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1884 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1885 because it is too late at that time, and has no effect.
1887 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1888 security issue with \' (bugzilla #107). I could not use the
1889 PQescapeStringConn() function, because it needs a PGconn value as one of
1892 PH/08 When testing addresses using -bt, indicate those final addresses that
1893 are duplicates that would not cause an additional delivery. At least one
1894 person was confused, thinking that -bt output corresponded to deliveries.
1895 (Suppressing duplicates isn't a good idea as you lose the information
1896 about possibly different redirections that led to the duplicates.)
1898 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1899 systems where poll() doesn't work, in particular OS X.
1901 PH/10 Added more information to debugging output for retry time not reached.
1903 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1904 operations in malware.c.
1906 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1909 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1910 syslog_duplication set false, log lines that would normally be written
1911 both the the main log and to the reject log were not written to syslog at
1914 PH/14 In the default configuration, change the use of "message" in ACL warn
1915 statements to "add_header".
1917 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1918 not followed by a command (e.g. "seen endif").
1920 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1921 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1924 PH/17 Added extra conditions to the default value of delay_warning_condition
1928 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1929 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1930 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1933 The Auto-Submitted: and various List- headers are standardised, whereas I
1934 don't think Precedence: ever was.
1936 PH/18 Refactored debugging code in route_finduser() to show more information,
1937 in particular, the error code if getpwnam() issues one.
1939 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1940 This is apparently needed in addition to the PH/07 change above to avoid
1941 any possible encoding problems.
1943 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1944 but not after initializing Perl.
1946 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1947 output them only if debugging. By default they are written stderr,
1948 apparently, which is not desirable.
1950 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1953 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1956 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1958 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1959 authenticated or an ident call has been made. Suppress the default
1960 values for $authenticated_id and $authenticated_sender (but permit -oMai
1961 and -oMas) when testing with -bh.
1963 PH/24 Re-jigged the order of the tests in the default configuration so that the
1964 tests for valid domains and recipients precede the DNS black list and CSA
1965 tests, on the grounds that those ones are more expensive.
1967 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1968 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1969 This bug exists in every version of Exim that I still have, right back to
1972 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1973 However, an attempt to turn on case-sensitivity in a regex key by
1974 including (?-i) didn't work because the subject string was already
1975 lowercased, and the effects were non-intuitive. It turns out that a
1976 one-line patch can be used to allow (?-i) to work as expected.
1982 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1983 other effects) broke the use of negated acl sub-conditions.
1985 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1988 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1989 "Deny" causes Exim to reject the incoming connection with a 554 error.
1990 Unfortunately, if there is a major crisis, such as a disk failure,
1991 tcp-wrappers gives "deny", whereas what one would like would be some
1992 kind of temporary error. A kludge has been added to help with this.
1993 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1994 554 error is used if errno is still zero or contains ENOENT (which occurs
1995 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1998 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2000 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2001 errors. Otherwise a message that provokes a temporary error (when other
2002 messages do not) can cause a whole host to time out.
2004 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2005 addresses were routed directly to files or pipes from a redirect router.
2006 File deliveries just didn't batch; pipe deliveries might have suffered
2009 PH/06 A failure to get a lock for a hints database would erroneously always say
2010 "Failed to get write lock", even when it was really a read lock.
2012 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2013 of 0600. This has been changed to use the value of the lockfile_mode
2014 option (which defaults to 0600).
2016 PH/08 Applied small patch from the Sieve maintainer.
2018 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2019 folder from quota calculations, a direct delivery into this folder messed
2020 up the contents of the maildirsize file. This was because the regex was
2021 used only to exclude .Trash (or whatever) when the size of the mailbox
2022 was calculated. There was no check that a delivery was happening into an
2023 excluded directory. This bug has been fixed by ignoring all quota
2024 processing for deliveries into excluded directories.
2026 PH/10 Added the maildirfolder_create_regex option to appendfile.
2032 PH/01 The code for finding all the local interface addresses on a FreeBSD
2033 system running IPv6 was broken. This may well have applied to all BSD
2034 systems, as well as to others that have similar system calls. The broken
2035 code found IPv4 interfaces correctly, but gave incorrect values for the
2036 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2037 that it would not match correctly against @[] and not recognize the IPv6
2040 PH/02 The ipliteral router was not recognizing addresses of the form user@
2041 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2043 PH/03 Added disable_ipv6.
2045 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2046 decoded form, because it is most often used to construct To: headers
2047 lines in autoreplies, and the decoded form may well be syntactically
2048 invalid. However, $reply_address has leading white space removed, and all
2049 newlines turned into spaces so that the autoreply transport does not
2052 PH/05 If group was specified without a user on a router, and no group or user
2053 was specified on a transport, the group from the router was ignored.
2055 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2056 for visible compile-time settings that can be used to change these
2057 numbers, for those that want even more. Backwards compatibility with old
2058 spool files has been maintained. However, going back to a previous Exim
2059 release will lost any variables that are in spool files.
2061 PH/07 Two small changes when running in the test harness: increase delay when
2062 passing a TCP/IP connection to a new process, in case the original
2063 process has to generate a bounce, and remove special handling of
2064 127.0.0.2 (sic), which is no longer necessary.
2066 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2067 be the same on different OS.
2069 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2072 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2073 whether --show-vars was specified or not
2075 JJ/02 exipick: Added support for new ACL variable spool format introduced
2078 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2079 syntactically invalid From: or Reply-to: line, and a filter used this to
2080 generate an autoreply, and therefore failed to obtain an address for the
2081 autoreply, Exim could try to deliver to a non-existent relative file
2082 name, causing unrelated and misleading errors. What now happens is that
2083 it logs this as a hard delivery error, but does not attempt to create a
2086 PH/11 The exinext utility has a -C option for testing purposes, but although
2087 the given file was scanned by exinext itself; it wasn't being passed on
2088 when Exim was called.
2090 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2091 an end-of-file indication when reading a command response.
2093 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2094 compiled. In many other places in Exim, IPv6 addresses are always
2095 recognized, so I have changed this. It also means that IPv4 domain
2096 literals of the form [IPV4:n.n.n.n] are now always recognized.
2098 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2099 used if the router is not running as root, for example, when verifying at
2100 ACL time, or when using -bh. The debugging output from this situation was
2101 non-existent - all you got was a failure to exec. I have made two
2104 (a) Failures to set uid/gid, the current directory, or a process leader
2105 in a subprocess such as that created by queryprogram now generate
2106 suitable debugging ouput when -d is set.
2108 (b) The queryprogram router detects when it is not running as root,
2109 outputs suitable debugging information if -d is set, and then runs
2110 the subprocess without attempting to change uid/gid.
2112 PH/15 Minor change to Makefile for building test_host (undocumented testing
2115 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2116 additional section of a DNS packet that returns MX or SRV records.
2117 Instead, it always explicitly searches for A/AAAA records. This avoids
2118 major problems that occur when a DNS server includes only records of one
2119 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2120 fixed another bug: if SRV records were looked up and the corresponding
2121 address records were *not* found in the additional section, the port
2122 values from the SRV records were lost.
2124 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2125 using the correct key (the original address) when searching the retry
2126 rules in order to find which one to use for generating the retry hint.
2128 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2129 adding the default one. Similarly, if it contains a Reply-To: header, the
2130 errors_reply_to option, if set, is not used.
2132 PH/19 When calculating a retry time, Exim used to measure the "time since
2133 failure" by looking at the "first failed" field in the retry record. Now
2134 it does not use this if it is later than than the arrival time of the
2135 message. Instead it uses the arrival time. This makes for better
2136 behaviour in cases where some deliveries succeed, thus re-setting the
2137 "first failed" field. An example is a quota failure for a huge message
2138 when small messages continue to be delivered. Without this change, the
2139 "time since failure" will always be short, possible causing more frequent
2140 delivery attempts for the huge message than are intended.
2141 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2143 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2144 $1, $2, $3) because the numerical variables can be reset during some
2145 expansion items (e.g. "match"), thereby losing the authentication data.
2147 PH/21 Make -bV show the size of off_t variables so that the test suite can
2148 decide whether to run tests for quotas > 2G.
2150 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2151 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2152 filecount value is greater than 2G or if a quota value is greater than 2G
2153 on a system where the size of off_t is not greater than 4, a panic error
2156 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2157 never match. The debug and -bh output now contains an explicit error
2158 message indicating a malformed IPv4 address or mask.
2160 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2161 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2162 PH/23 above applies.
2164 PH/25 Do not write to syslog when running in the test harness. The only
2165 occasion when this arises is a failure to open the main or panic logs
2166 (for which there is an explicit test).
2168 PH/26 Added the /no_tell option to "control=freeze".
2170 PH/27 If a host name lookup failed very early in a connection, for example, if
2171 the IP address matched host_lookup and the reverse lookup yielded a name
2172 that did not have a forward lookup, an error message of the form "no IP
2173 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2174 could be logged. Now it outputs the IP address instead of "NULL".
2176 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2177 allows the sender and the authenticated sender to be set when
2178 submitting a message from within Exim. Since child_open_exim() is
2179 documented for local_scan(), the new function should be too.
2181 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2182 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2183 results in an empty string is now treated as unset.
2185 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2187 PH/31 Added sender_verify_fail logging option.
2189 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2190 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2191 tidied the source and removed it altogether.
2193 PH/33 When a queue run was abandoned because the load average was too high, a
2194 log line was always written; now it is written only if the queue_run log
2195 selector is set. In addition, the log line for abandonment now contains
2196 information about the queue run such as the pid. This is always present
2197 in "start" and "stop" lines but was omitted from the "abandon" line.
2199 PH/34 Omit spaces between a header name and the colon in the error message that
2200 is given when verify = headers_syntax fails (if there are lots of them,
2201 the message gets confusing).
2203 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2204 names, as there are now some PTR records that contain slashes. This check
2205 is only to protect against broken name servers that fall over on strange
2206 characters, so the fact that it applies to all lookups doesn't matter.
2208 PH/36 Now that the new test suite is complete, we can remove some of the
2209 special code in Exim that was needed for the old test suite. For example,
2210 sorting DNS records because real resolvers return them in an arbitrary
2211 order. The new test suite's fake resolver always returns records in the
2214 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2215 bounces) except when queue_only is set, to avoid logging races between
2216 the different processes.
2218 PH/38 Panic-die if .include specifies a non-absolute path.
2220 PH/39 A tweak to the "H" retry rule from its user.
2222 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2223 a label. They prevented compilation on older perls.
2225 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2226 a warning to be raised on newish perls.
2228 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2229 on queue. Changes to match documented behaviour of showing count of
2230 messages matching specified criteria.
2232 PH/40 Changed the default ident timeout from 30s to 5s.
2234 PH/41 Added support for the use of login_cap features, on those BSD systems
2235 that have them, for controlling the resources used by pipe deliveries.
2237 PH/42 The content-scanning code uses fopen() to create files in which to put
2238 message data. Previously it was not paying any attention to the mode of
2239 the files. Exim runs with umask(0) because the rest of the code creates
2240 files with open(), and sets the required mode explicitly. Thus, these
2241 files were ending up world-writeable. This was not a big issue, because,
2242 being within the spool directory, they were not world-accessible. I have
2243 created a function called modefopen, which takes an additional mode
2244 argument. It sets umask(777), creates the file, chmods it to the required
2245 mode, then resets the umask. All the relevant calls to fopen() in the
2246 content scanning code have been changed to use this function.
2248 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2249 to 24 hours. This avoids potential overflow problems when processing G
2250 and H retry rules. I suspect nobody ever tinkers with this value.
2252 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2254 PH/45 When the plaintext authenticator is running as a client, the server's
2255 challenges are checked to ensure they are valid base64 strings. By
2256 default, the authentication attempt is cancelled if an invalid string is
2257 received. Setting client_ignore_invalid_base64 true ignores these errors.
2258 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2259 they are received. Thus, the responses can be made to depend on the
2260 challenges. If an invalid string is ignored, an empty string is placed in
2263 PH/46 Messages that are created by the autoreply transport now contains a
2264 References: header, in accordance with RFCs 2822 and 3834.
2266 PH/47 Added authenticated_sender_force to the smtp transport.
2268 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2270 PH/49 Installed latest patch from the Sieve maintainer.
2272 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2273 was also set, the appendfile transport was unnecessarily scanning a
2274 directory of message files (e.g. for maildir delivery) to find the count
2275 of files (along with the size), even though it did not need this
2276 information. It now does the scan only if it needs to find either the
2277 size of the count of files.
2279 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2281 PH/52 Two bugs concerned with error handling when the smtp transport is
2284 (i) Exim was not creating retry information for temporary errors given
2285 for individual recipients after the DATA command when the smtp transport
2286 was used in LMTP mode. This meant that they could be retried too
2287 frequently, and not timed out correctly.
2289 (ii) Exim was setting the flag that allows error details to be returned
2290 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2291 recipients that were returned after the DATA command.
2293 PH/53 This is related to PH/52, but is more general: for any failing address,
2294 when detailed error information was permitted to be returned to the
2295 sender, but the error was temporary, then after the final timeout, only
2296 "retry timeout exceeded" was returned. Now it returns the full error as
2297 well as "retry timeout exceeded".
2299 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2300 do this, and (what is worse) MTAs that accept it.
2302 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2303 will now be deprecated.
2305 PH/56 New os.c-cygwin from the Cygwin maintainer.
2307 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2308 formats (previously only available in exim formats via -bpr, -bpru,
2309 and -bpra. Now also available in native and exiqgrep formats)
2311 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2312 with very large, slow to parse queues
2314 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2316 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2318 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2319 responses to authentication challenges, though it was showing the
2320 challenges; (ii) I've removed the CR characters from the debug output for
2323 PH/58 Allow for the insertion of a newline as well as a space when a string
2324 is turned into more than one encoded-word during RFC 2047 encoding. The
2325 Sieve code now uses this.
2327 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2328 data_4xx, lost_connection, tls_required.
2330 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2331 message was being sent as an SMTP response.
2333 PH/61 Add -l and -k options to exicyclog.
2335 PH/62 When verifying, if an address was redirected to one new address, so that
2336 verification continued, and the new address failed or deferred after
2337 having set something in $address_data, the value of $address_data was not
2338 passed back to the ACL. This was different to the case when no
2339 redirection occurred. The value is now passed back in both cases.
2341 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2342 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2343 use login_cap.h, so on its own it isn't the distinguishing feature. The
2344 new name refers directly to the setclassresources() function.
2346 PH/65 Added configuration files for NetBSD3.
2348 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2350 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2351 is preferred over IPv4.
2353 PH/68 The bounce_return_message and bounce_return_body options were not being
2354 honoured for bounces generated during the reception of non-SMTP messages.
2355 In particular, this applied to messages rejected by the ACL. This bug has
2356 been fixed. However, if bounce_return_message is true and bounce_return_
2357 body is false, the headers that are returned for a non-SMTP message
2358 include only those that have been read before the error was detected.
2359 (In the case of an ACL rejection, they have all been read.)
2361 PH/69 The HTML version of the specification is now built in a directory called
2362 spec_html instead of spec.html, because the latter looks like a path with
2363 a MIME-type, and this confuses some software.
2365 PH/70 Catch two compiler warnings in sieve.c.
2367 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2368 function verify_get_ident() calls ip_connect() to connect a socket, but
2369 if the "connect()" function timed out, ip_connect() used to close the
2370 socket. However, verify_get_ident() also closes the socket later, and in
2371 between Exim writes to the log, which may get opened at this point. When
2372 the socket was closed in ip_connect(), the log could get the same file
2373 descriptor number as the socket. This naturally causes chaos. The fix is
2374 not to close the socket in ip_connect(); the socket should be closed by
2375 the function that creates it. There was only one place in the code where
2376 this was missing, in the iplookup router, which I don't think anybody now
2377 uses, but I've fixed it anyway.
2379 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2380 well as to direct DNS lookups. Otherwise the handling of names in host
2381 lists is inconsistent and therefore confusing.
2387 PH/01 Two changes to the default runtime configuration:
2389 (1) Move the checks for relay_from_hosts and authenticated clients from
2390 after to before the (commented out) DNS black list checks.
2392 (2) Add control=submission to the relay_from_hosts and authenticated
2393 clients checks, on the grounds that messages accepted by these
2394 statements are most likely to be submissions.
2396 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2398 (1) Generate an error if the third argument for the ${prvs expansion is
2401 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2404 (3) Reset the variables that are obtained from the first argument of
2405 ${prvscheck and used in the second argument before leaving the code,
2406 because their memory is reclaimed, so using them afterwards may do
2409 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2410 one (it's much easier than Tom thought :-).
2412 (5) Because of (4), we can now allow for the use of $prvscheck_result
2413 inside the third argument.
2415 PH/03 For some reason, the default setting of PATH when running a command from
2416 a pipe transport was just "/usr/bin". I have changed it to
2419 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2420 anything to be listed in the output from -bV.
2422 PH/05 When a filter generated an autoreply, the entire To: header line was
2423 quoted in the delivery log line, like this:
2425 => >A.N.Other <ano@some.domain> <original@ddress> ...
2427 This has been changed so that it extracts the operative address. There
2428 may be more than one such address. If so, they are comma-separated, like
2431 => >ano@some.domain,ona@other.domain <original@ddress> ...
2433 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2434 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2435 not being looked up in the rDNS to get a host name, Exim was showing the
2436 IP address twice in Received: lines, even though the IP addresses were
2437 identical. For example:
2439 Received: from [1.2.3.4] (helo=[1.2.3.4])
2441 However, if the real host name was known, it was omitting the HELO data
2442 if it matched the actual IP address. This has been tidied up so that it
2443 doesn't show the same IP address twice.
2445 PH/07 When both +timestamp and +memory debugging was on, the value given by
2446 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2447 called by the debug printing, thereby overwriting the timestamp buffer.
2448 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2450 PH/08 When the original message was included in an autoreply transport, it
2451 always said "this is a copy of the message, including all the headers",
2452 even if body_only or headers_only was set. It now gives an appropriate
2455 PH/09 Applied a patch from the Sieve maintainer which:
2457 o fixes some comments
2458 o adds the (disabled) notify extension core
2459 o adds some debug output for the result of if/elsif tests
2460 o points to the current vacation draft in the documentation
2461 and documents the missing references header update
2465 o fixes a bug in processing the envelope test (when testing
2466 multiple envelope elements, the last element determinted the
2469 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2470 Electronic Mail") by including:
2472 Auto-submitted: auto-generated
2474 in the messages that it generates (bounce messages and others, such as
2475 warnings). In the case of bounce messages for non-SMTP mesages, there was
2476 also a typo: it was using "Auto_submitted" (underscore instead of
2477 hyphen). Since every message generated by Exim is necessarily in response
2478 to another message, thes have all been changed to:
2480 Auto-Submitted: auto-replied
2482 in accordance with these statements in the RFC:
2484 The auto-replied keyword:
2486 - SHOULD be used on messages sent in direct response to another
2487 message by an automatic process,
2489 - MUST NOT be used on manually-generated messages,
2491 - MAY be used on Delivery Status Notifications (DSNs) and Message
2492 Disposition Notifications (MDNs),
2494 - MUST NOT be used on messages generated by automatic or periodic
2495 processes, except for messages which are automatic responses to
2498 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2499 to the default Received: header definition.
2501 PH/12 Added log selector acl_warn_skipped (default on).
2503 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2504 variables because (a) they are in the wrong storage pool and (b) even if
2505 they were copied, it wouldn't work properly because of the caching.
2507 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2508 checking when decoding. Apparently there are clients that generate
2509 overlong encoded strings. Why am I not surprised?
2511 PH/15 If the first argument of "${if match_address" was not empty, but did not
2512 contain an "@" character, Exim crashed. Now it writes a panic log message
2513 and treats the condition as false.
2515 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2517 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2518 extension "envelope-auth". The code is finished and in agreement with
2519 other implementations, but there is no documentation so far and in fact,
2520 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2521 not changing the active code.
2523 Print executed "if" and "elsif" statements when debugging is used. This
2524 helps a great deal to understand what a filter does.
2526 Document more things not specified clearly in RFC3028. I had all this
2527 sorted out, when out of a sudden new issues came to my mind. Oops."
2529 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2532 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2533 canonical form (as documented). However, after a host name lookup from
2534 the IP address, check_host() was doing a simple string comparison with
2535 addresses acquired from the DNS when checking that the found name did
2536 have the original IP as one of its addresses. Since any found IPv6
2537 addresses are likely to be in abbreviated form, the comparison could
2538 fail. Luckily, there already exists a function for doing the comparison
2539 by converting both addresses to binary, so now that is used instead of
2540 the text comparison.
2542 PH/20 There was another similar case to PH/19, when a complete host name was
2543 given in a host list; looking up its IP address could give an abbreviated
2544 form, whereas the current host's name might or might not be abbreviated.
2545 The same fix has been applied.
2551 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2552 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2555 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2556 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2558 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2560 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2561 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2562 Also, for "Auto-Submitted", treat anything other than "no" as
2563 non-personal, in accordance with RFC 3834. (Previously it treated
2564 anything starting "auto-" as non-personal.)
2566 TF/01 The control=submission/name=... option had a problem with syntax
2567 errors if the name included a slash character. The /name= option
2568 now slurps the rest of the string, so it can include any characters
2569 but it must come last in the list of options (after /sender_retain
2572 PH/05 Some modifications to the interface to the fake nameserver for the new
2580 TK/01 Added the "success_on_redirect" address verification option. See
2581 NewStuff for rationale and an example.
2583 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2585 PH/02 Patch to exigrep to allow it to work on syslog lines.
2587 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2588 fread() to skip over the body file's header line, because in Cygwin the
2589 header line is locked and is inaccessible.
2591 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2592 co-exist for some time) to make it clear that it is the Exim ID that is
2593 referenced, not the Message-ID: header line.
2595 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2596 string_format() function, because snprintf() does not exist on all
2599 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2600 ${stat: expansion item.
2602 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2603 protocol synchronization error", to keep the pedants happy.
2605 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2606 well as for IRIX systems, when gcc is being used. See the host.c source
2609 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2611 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2614 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2615 transport and to the smtp transport in LMTP mode.
2617 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2619 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2620 run a filter in a subprocess. This could lead to confusion in subsequent
2621 lookups in the parent process. There should also be a search_tidyup() at
2622 the end of the subprocess.
2624 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2625 only if the host matched helo_try_verify_hosts, which caused the
2626 verification to occur when the EHLO/HELO command was issued. The ACL just
2627 tested the remembered result. Now, if a previous verification attempt has
2628 not happened, "verify = helo" does it there and then.
2630 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2632 TK/03 Fix log output including CR from clamd.
2634 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2635 exist provoked a memory error which could cause a segfault.
2637 PH/15 Installed PCRE 6.2
2639 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2641 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2642 of the problem. Specifically, suggested +O2 rather than +O1 for the
2645 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2647 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2648 record, the port was not correctly logged when the outgoing_port log
2649 selector was set (it logged the transort's default port).
2651 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2652 fallback_hosts, and "hosts" in the smtp transport.
2654 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2655 host errors such as "Connection refused".
2657 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2658 authentication with radiusclient 0.4.9:
2660 - Error returned from rc_read_config was caught wrongly
2661 - Username/password not passed on to radius server due to wrong length.
2663 The presumption is that some radiusclient API changes for 4.51/PH/17
2664 were not taken care of correctly. The code is still untested by me (my
2665 Linux distribution still has 0.3.2 of radiusclient), but it was
2666 contributed by a Radius user.
2668 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2669 expanding the "port" option of the smtp transport.
2671 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2672 while reading a MIME header. Thanks to Tom Hughes for a patch.
2674 PH/24 Include config.h inside local_scan.h so that configuration settings are
2677 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2678 that in an ACL for RCPT (for example), you can examine exactly what was
2681 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2682 commands, but it was not correctly comparing the address with the actual
2683 client host address. Thus, it would show the EHLO address in Received:
2684 header lines when this was not necessary.
2686 PH/27 Added the % operator to ${eval:}.
2688 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2689 it should be ignoring failures (because with -C, for example, it has lost
2690 privilege). It wasn't ignoring creation failures other than "already
2693 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2696 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2697 because an input line was too long, either on its own, or by virtue of
2698 too many continuations, the temporary file was not being removed, and the
2699 return code was incorrect.
2701 PH/31 Missing "BOOL" in function definition in filtertest.c.
2703 PH/32 Applied Sieve patches from the maintainer.
2705 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2707 PH/33 Added "verify = not_blind".
2709 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2710 Local/Makefile (with some defaults set). These are used in built scripts
2711 such as exicyclog, but they have never been used in the exim_install
2712 script (though there are many overriding facilities there). I have
2713 arranged that the exim_install script now takes note of these two
2716 PH/35 Installed configuration files for Dragonfly.
2718 PH/36 When a locally submitted message by a trusted user did not contain a
2719 From: header, and the sender address was obtained from -f or from an SMTP
2720 MAIL command, and the trusted user did not use -F to supply a sender
2721 name, $originator_name was incorrectly used when constructing a From:
2722 header. Furthermore, $originator_name was used for submission mode
2723 messages from external hosts without From: headers in a similar way,
2724 which is clearly wrong.
2726 PH/37 Added control=suppress_local_fixups.
2728 PH/38 When log_selector = +received_sender was set, and the addition of the
2729 sender made the log line's construction buffer exactly full, or one byte
2730 less than full, an overflow happened when the terminating "\n" was
2733 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2734 when the result of a list match is failure because a DNS lookup failed.
2736 PH/40 RM_COMMAND is now used in the building process.
2738 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2739 the "build-* directories that it finds.
2741 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2742 address in a domain literal was a prefix of an interface address.
2744 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2745 when verifying a sender address, unless rewrite_headers is false.
2747 PH/44 Wrote a long comment about why errors_to addresses are verified as
2748 recipients, not senders.
2750 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2751 the ratelimit ACL was added.
2753 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2755 PH/46 Added extra information about PostgreSQL errors to the error string.
2757 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2758 avoiding the need to install special zones in a real server. This is
2759 backwards compatible; if it can't find the fake resolver, it drops back.
2760 Thus, both old and new test suites can be run.
2762 TF/02 Added util/ratelimit.pl
2764 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2765 clock is set back in time.
2767 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2768 Candler <B.Candler@pobox.com>.
2770 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2771 OK for addresses that are the result of redirections.
2773 PH/48 A number of further additions for the benefit of the new test suite,
2774 including a fake gethostbyname() that interfaces to the fake DNS resolver
2777 TF/06 The fix for widen_domains has also been applied to qualify_single and
2778 search_parents which are the other dnslookup options that can cause
2781 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2784 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2786 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2787 service" if it encountered an MX record with an empty target hostname.
2788 The message is now "an MX or SRV record indicated no SMTP service".
2790 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2791 if the DNS of the sending site is misconfigured. This is quite a
2792 common situation. This change restores the behaviour of treating a
2793 helo verification defer as a failure.
2795 PH/51 If self=fail was set on a router, the bounce message did not include the
2796 actual error message.
2802 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2804 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2805 command itself ended in error, the underlying message about the transport
2806 filter timeout was being overwritten with the pipe command error. Now the
2807 underlying error message should be appended to the second error message.
2809 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2811 PH/02 Reduce the amount of output that "make" produces by default. Full output
2812 can still be requested.
2814 PH/03 The warning log line about a condition test deferring for a "warn" verb
2815 was being output only once per connection, rather than after each
2816 occurrence (because it was using the same function as for successful
2817 "warn" verbs). This seems wrong, so I have changed it.
2819 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2820 it should not have, which might have caused a crash in the right
2821 circumstances, but probably never did.
2823 PH/04 Installed a modified version of Tony Finch's patch to make submission
2824 mode fix the return path as well as the Sender: header line, and to
2825 add a /name= option so that you can make the user's friendly name appear
2828 TF/03 Added the control = fakedefer ACL modifier.
2830 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2831 Mark Lowes for thorough testing.
2833 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2835 TK/03 Merged latest SRS patch from Miles Wilton.
2837 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2838 with the definition in sysexits.h (which is #included earlier).
2839 Fortunately, Exim does not actually use EX_OK. The code used to try to
2840 preserve the sysexits.h value, by assumimg that macro definitions were
2841 scanned for macro replacements. I have been disabused of this notion,
2842 so now the code just undefines EX_OK before #including unistd.h.
2844 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2845 in the smtp transport. When a block could not be written in a single
2846 write() function, the timeout was being re-applied to each part-write.
2847 This seems wrong - if the receiver was accepting one byte at a time it
2848 would take for ever. The timeout is now adjusted when this happens. It
2849 doesn't have to be particularly precise.
2851 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2852 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2854 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2855 without a domain if the check to <postmaster@domain> fails.
2857 SC/01 Eximstats: added -xls and the ability to specify output files
2858 (patch written by Frank Heydlauf).
2860 SC/02 Eximstats: use FileHandles for outputing results.
2862 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2864 SC/04 Eximstats: fixed display of large numbers with -nvr option
2866 SC/05 Eximstats: fixed merging of reports with empty tables.
2868 SC/06 Eximstats: added the -include_original_destination flag
2870 SC/07 Eximstats: removed tabs and trailing whitespace.
2872 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2874 TK/06 MBOX spool code: Add real "From " MBOX separator line
2875 so the .eml file is really in mbox format (even though
2876 most programs do not really care). Patch from Alex Miller.
2878 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2879 The latter is generated from $received_to and is only set if the
2880 message has one envelope recipient. SA can use these headers,
2881 obviously out-of-the-box. Patch from Alex Miller.
2883 PH/08 The ${def test on a variable was returning false if the variable's
2884 value was "0", contrary to what the specification has always said!
2885 The result should be true unless the variable is empty.
2887 PH/09 The syntax error of a character other than { following "${if
2888 def:variable_name" (after optional whitespace) was not being diagnosed.
2889 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2890 accidental colon was present, for example, could give incorrect results.
2892 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2893 result is used (not including appendfile, where other changes are about
2896 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2897 This involved changing a lot of size variables from int to off_t. It
2898 should work with maildirs and everything.
2900 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2901 spamd dying while we are connected to it.
2903 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2906 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2907 function for BDB 4.3.
2909 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2911 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2912 the sources for PCRE have been re-arranged and more files are now
2915 PH/15 The code I had for printing potentially long long variables in PH/11
2916 above was not the best (it lost precision). The length of off_t variables
2917 is now inspected at build time, and an appropriate printing format (%ld
2918 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2919 to be "long long int" or "long int". This is needed for the internal
2920 formatting function string_vformat().
2922 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2923 the configuration file to be ":syslog", then the script "guesses" where
2924 the logs files are, rather than using the compiled in default. In our
2925 case the guess is not the same as the compiled default, so the script
2926 suddenly stopped working when I started to use syslog. The patch checks
2927 to see if log_file_path is "". If so, it attempts to read it from exim
2928 with no configuration file to get the compiled in version, before it
2929 falls back to the previous guessing code."
2931 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2932 implementing BATV in an Exim configuration. See NewStuff for the gory
2935 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2936 Makefile that are specific to HP-UX.
2938 PH/18 If the "use_postmaster" option was set for a recipient callout together
2939 with the "random" option, the postmaster address was used as the MAIL
2940 FROM address for the random test, but not for the subsequent recipient
2941 test. It is now used for both.
2943 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2944 patch removes a few documentation additions to RFC 3028, because the
2945 latest draft now contains them. It adds the new en;ascii-case comparator
2946 and a new error check for 8bit text in MIME parts. Comparator and
2947 require names are now matched exactly. I enabled the subaddress
2948 extension, but it is not well tested yet (read: it works for me)."
2950 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2951 rework some of the code of TK/09 above to avoid the hardwired use of
2952 "%lld" and "long long". Replaced the call to snprintf() with a call to
2955 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2956 records point to non-existent hosts", "retry timeout exceeded", and
2957 "retry time not reached for any host after a long failure period".
2959 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2960 experimental DomainKeys support:
2962 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2963 (2) On an error such as an illegally used "control", the wrong name for
2964 the control was given.
2966 These problems did NOT occur unless DomainKeys support was compiled.
2968 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2970 PH/24 Added ${if match_ip condition.
2972 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2973 databases so that it will be absolutely obvious if a crash occurs in the
2974 DB library. This is a regular occurrence (often caused by mis-matched
2977 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2978 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2979 picked up on a user's system that detects such things. There doesn't seem
2980 to be a gcc warning option for this - only an attribute that has to be
2981 put on the function's prototype. It seems that in Fedora Core 4 they have
2982 set this on a number of new functions. No doubt there will be more in due
2985 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2986 specify a transport. However, if an address that was verified by such a
2987 router was the subject of a callout, Exim crashed because it tried to
2988 read the rcpt_include_affixes from the non-existent transport. Now it
2989 just assumes that the setting of that option is false. This bug was
2990 introduced by 4.51/PH/31.
2992 PH/28 Changed -d+all to exclude +memory, because that information is very
2993 rarely of interest, but it makes the output a lot bigger. People tend to
2994 do -d+all out of habit.
2996 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2997 code in os-type was giving problems when libc.so lives in lib64, like on
3000 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3001 aren't the modern standard, and it seems that some systems' include files
3002 don't always have them. Exim was already checking for some of the newer
3003 ones like T_AAAA, and defining it itself. I've added checks for all the
3004 record types that Exim uses.
3006 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3007 not automatically generating a new one, as it is supposed to. This
3008 prevented TLS from working. If the file did exist, but contained invalid
3009 data, a new version was generated, as expected. It was only the case of a
3010 non-existent file that was broken.
3012 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3013 with a change in libdomainkeys > 0.64.
3015 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3016 from DNS. If the selector record carries the flag, it now has
3017 precedence over the domain-wide flag.
3019 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3021 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3022 the use of a port name that isn't defined in /etc/services) occurred, the
3023 message was deferred as in a normal delivery, and thus remained on the
3024 spool, instead of being failed because of the mua_wrapper setting. This
3025 is now fixed, and I tidied up some of the mua_wrapper messages at the
3028 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3029 the messages in a hash of arrays rather than using individual hashes.
3030 This is a bit cleaner and results in dramatic memory savings, albeit
3031 at a slight CPU cost.
3033 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3034 as requested by Marc Sherman.
3036 SC/10 Eximstats: added histograms for user specified patterns as requested
3039 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3041 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3042 fopen() in the content-scanning modules that did not already have it.
3048 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3049 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3051 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3053 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3055 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3056 to test Sieve filters that use "vacation".
3058 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3059 that changes the way the GnuTLS parameters are stored in the cache file.
3060 The new format can be generated externally. For backward compatibility,
3061 if the data in the cache doesn't make sense, Exim assumes it has read an
3062 old-format file, and it generates new data and writes a new file. This
3063 means that you can't go back to an older release without removing the
3066 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3067 work if there are any delivery delays because "one_time" forces the
3068 parent to be marked "delivered", so its unseen clone is never tried
3069 again. For this reason, Exim now forbids the simultaneous setting of
3072 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3073 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3074 entry if you want to know the details. The fix, however, neglected to
3075 consider the case when local delivery batching is involved. The test for
3076 "previously delivered" was not happening when checking to see if an
3077 address could be batched with a previous (undelivered) one; under
3078 certain circumstances this could lead to multiple deliveries to the same
3081 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3082 in its include files, and this causes problems building Exim.
3084 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3085 header_syntax) but Exim was just ignoring anything given after a slash.
3086 In particular, this caused confusion with an attempt to use "verify =
3087 reverse_host_lookup/defer_ok". An error is now given when options are
3088 supplied for verify items that do not have them. (Maybe reverse_host_
3089 lookup should have a defer_ok option, but that's a different point.)
3091 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3092 defined by RFC 821) to 2048, because there were problems with some AUTH
3093 commands, and RFC 1869 says the size should be increased for extended
3094 SMTP commands that take arguments.
3096 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3099 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3100 "unknown" error; now it says that the functionality isn't in the binary.
3102 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3103 an address' error message when a string expansion fails (syntax or
3104 whatever). Otherwise the password may appear in the log. Following change
3105 PH/42 below, there is no longer a chance of it appearing in a bounce
3108 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3110 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3111 of its addresses was ever tried. (Bugzilla bug #2.)
3113 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3114 the result incorrectly in the debug output. (It correctly added a newline
3115 to what was transported.)
3117 TF/01 Added $received_time.
3119 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3120 commented out examples of how to interface to a virus scanner and to
3121 SpamAssassin. Also added commented examples of av_scanner and
3122 spamd_address settings.
3124 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3125 and controls are allowed in which ACLs. There were a couple of minor
3126 errors. Some of the entries in the conditions table (which is a table of
3127 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3128 negation of where the condition IS allowed.
3130 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3132 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3133 header file does not have a version number, so I've had to invent a new
3134 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3135 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3136 radiusclient), but it was contributed by a Radius user.
3138 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3139 files or directories, for OpenSSL.
3141 PH/19 When an Exim process that is running as root has to create an Exim log
3142 file, it does so in a subprocess that runs as exim:exim so as to get the
3143 ownership right at creation (otherwise, other Exim processes might see
3144 the file with the wrong ownership). There was no test for failure of this
3145 fork() call, which would lead to the process getting stuck as it waited
3146 for a non-existent subprocess. Forks do occasionally fail when resources
3147 run out. I reviewed all the other calls to fork(); they all seem to check
3150 PH/20 When checking for unexpected SMTP input at connect time (before writing
3151 the banner), Exim was not dealing correctly with a non-positive return
3152 from the read() function. If the client had disconnected by this time,
3153 the result was a log entry for a synchronization error with an empty
3154 string after "input=" when read() returned zero. If read() returned -1
3155 (an event I could not check), uninitialized data bytes were printed.
3156 There were reports of junk text (parts of files, etc) appearing after
3159 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3161 PH/22 Added support for macro redefinition, and (re)definition in between
3162 driver and ACL definitions.
3164 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3165 forgetting to use the resulting value; it was using the unexpanded value.
3167 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3168 hadn't been configured. The fix is from Juergen Kreileder, who
3169 understands it better than I do:
3171 "Here's what I see happening with three configured cyrus_sasl
3172 authenticators configured (plain, login, cram-md5):
3174 On startup auth_cyrus_sasl_init() gets called for each of these.
3175 This means three calls to sasl_listmech() without a specified mech_list.
3176 => SASL tests which mechs of all available mechs actually work
3177 => three warnings about OTP not working
3178 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3180 With the patch, sasl_listmech() also gets called three times. But now
3181 SASL's mech_list option is set to the server_mech specified in the the
3182 authenticator. Or in other words, the answer from sasl_listmech()
3183 gets limited to just the mech you're testing for (which is different
3185 => the return list contains just 'plain' or 'login', 'cram-md5' or
3186 nothing depending on the value of ob->server_mech.
3188 I've just tested the patch: Authentication still works fine,
3189 unavailable mechs specified in the exim configuration are still
3190 caught, and the auth.log warnings about OTP are gone."
3192 PH/25 When debugging is enabled, the contents of the command line are added
3193 to the debugging output, even when log_selector=+arguments is not
3196 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3197 answer is "GNU", and only if the return is "GNU/something" is the answer
3200 PH/27 $acl_verify_message is now set immediately after the failure of a
3201 verification in an ACL, and so is available in subsequent modifiers. In
3202 particular, the message can be preserved by coding like this:
3204 warn !verify = sender
3205 set acl_m0 = $acl_verify_message
3207 Previously, $acl_verify_message was set only while expanding "message"
3208 and "log_message" when a very denied access.
3210 PH/28 Modified OS/os.c-Linux with
3212 -#ifndef OS_LOAD_AVERAGE
3213 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3215 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3216 nomenclature these days.)
3218 PH/29 Installed patch from the Sieve maintainer that adds the options
3219 sieve_useraddress and sieve_subaddress to the redirect router.
3221 PH/30 In these circumstances:
3222 . Two addresses routed to the same list of hosts;
3223 . First host does not offer TLS;
3224 . First host accepts first address;
3225 . First host gives temporary error to second address;
3226 . Second host offers TLS and a TLS session is established;
3227 . Second host accepts second address.
3228 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3229 and peerdn, if requested) that were in fact used only for the second
3232 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3233 attention to any local part prefix or suffix that was matched by the
3234 router that accepted the address. It now behaves in the same way as it
3235 does for delivery: the affixes are removed from the local part unless
3236 rcpt_include_affixes is set on the transport.
3238 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3239 timeout during the DATA phase of an incoming message.
3241 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3242 applied a patch sanctioned by the Sieve maintainer.
3244 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3245 the uid or gid is negative. A case of a negative gid caused this to be
3246 noticed. The fix allows for either to be negative.
3248 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3249 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3252 PH/36 The change PH/12 above was broken. Fixed it.
3254 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3255 the grounds that routing the same address twice would always produce the
3256 same answer. This might have been true once, but it is certainly no
3257 longer true now. Routing a child address may depend on the previous
3258 routing that produced that child. Some complicated redirection strategies
3259 went wrong when messages had multiple recipients, and made Exim's
3260 behaviour dependent on the order in which the addresses were given.
3262 I have moved the duplicate checking until after the routing is complete.
3263 Exim scans the addresses that are assigned to local and remote
3264 transports, and removes any duplicates. This means that more work will be
3265 done, as duplicates will always all be routed, but duplicates are
3266 presumably rare, so I don't expect this is of any significance.
3268 For deliveries to pipes, files, and autoreplies, the duplicate checking
3269 still happens during the routing process, since they are not going to be
3272 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3273 It corrects a timeout issue with spamd. This is Ian's comment: "The
3274 background is that sometimes spamd either never reads data from a
3275 connection it has accepted, or it never writes response data. The exiscan
3276 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3277 blindly assumes that writes won't block so it may never time out."
3279 PH/39 Allow G after quota size as well as K and M.
3281 PH/40 The value set for $authenticated_id in an authenticator may not contain
3282 binary zeroes or newlines because the value is written to log lines and
3283 to spool files. There was no check on this. Now the value is run through
3284 the string_printing() function so that such characters are converted to
3285 printable escape sequences.
3287 PH/41 $message_linecount is a new variable that contains the total number of
3288 lines in the message. Compare $body_linecount, which is the count for the
3291 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3292 bounce and delay warning messages, except in certain special cases, which
3295 (a) An SMTP error message from a remote host;
3296 (b) A message specified in a :fail: redirection;
3297 (c) A message specified in a "fail" command in a system filter;
3298 (d) A message specified in a FAIL return from the queryprogram router;
3299 (e) A message specified by the cannot_route_message router option.
3301 In these cases only, Exim does include the error details in bounce and
3302 warning messages. There are also a few cases where bland messages such
3303 as "unrouteable address" or "local delivery error" are given.
3305 PH/43 $value is now also set for the "else" part of a ${run expansion.
3307 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3308 being worked on, but at least Exim now implements the latest version to
3311 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3312 process to complete was treated as a delivery failure, a timeout while
3313 writing the message to the pipe was logged, but erroneously treated as a
3314 successful delivery. Such timeouts include transport filter timeouts. For
3315 consistency with the overall process timeout, these timeouts are now
3316 treated as errors, giving rise to delivery failures by default. However,
3317 there is now a new Boolean option for the pipe transport called
3318 timeout_defer, which, if set TRUE, converts the failures into defers for
3319 both kinds of timeout. A transport filter timeout is now identified in
3322 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3323 systems where "make" and "gmake" are different, calling "gmake" at top
3324 level broke things. I've arranged for the value of $(MAKE) to be passed
3325 from the Makefile to this script so that it can call the same version of
3329 A note about Exim versions 4.44 and 4.50
3330 ----------------------------------------
3332 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3333 changes of various kinds. As a consequence, a big documentation update was
3334 needed. This delayed the release for rather longer than seemed good, especially
3335 in the light of a couple of (minor) security issues. Therefore, the changes
3336 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3337 release. So 4.44 and 4.50 are in effect two different branches that both start
3340 I have left the 4.50 change log unchanged; it contains all the changes since
3341 4.43. The change log for 4.44 is below; many of its items are identical to
3342 those for 4.50. This seems to be the most sensible way to preserve the
3343 historical information.
3349 1. Minor wording change to the doc/README.SIEVE file.
3351 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3352 computation of the current number of files was incorrect.
3354 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3355 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3358 4. Give more explanation in the error message when the command for a transport
3359 filter fails to execute.
3361 5. There are several places where Exim runs a non-Exim command in a
3362 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3363 was being done only for the command run by the queryprogram router. It is
3364 now done for all such subprocesses. The other cases are: ${run, transport
3365 filters, and the commands run by the lmtp and pipe transports.
3367 6. Added CONFIGURE_GROUP build-time option.
3369 7. Some older OS have a limit of 256 on the maximum number of file
3370 descriptors. Exim was using setrlimit() to set 1000 as a large value
3371 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3372 systems. I've change it so that if it can't get 1000, it tries for 256.
3374 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3375 was an oversight, and furthermore, ever since the addition of extra
3376 controls (e.g. 4.43/32), the checks on when to allow different forms of
3377 "control" were broken. There should now be diagnostics for all cases when a
3378 control that does not make sense is encountered.
3380 9. Added the /retain_sender option to "control=submission".
3382 10. $recipients is now available in the predata ACL (oversight).
3384 11. Tidy the search cache before the fork to do a delivery from a message
3385 received from the command line. Otherwise the child will trigger a lookup
3386 failure and thereby defer the delivery if it tries to use (for example) a
3387 cached ldap connection that the parent has called unbind on.
3389 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3390 of $address_data from the recipient verification was clobbered by the
3391 sender verification.
3393 13. The value of address_data from a sender verification is now available in
3394 $sender_address_data in subsequent conditions in the ACL statement.
3396 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3398 15. Added a new option "connect=<time>" to callout options, to set a different
3401 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3402 was its contents. (It was OK if the option was not defined at all.)
3404 17. A "Completed" log line is now written for messages that are removed from
3405 the spool by the -Mrm option.
3407 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3408 information about exactly what failed.
3410 19. Added -dd to debug only the daemon process.
3412 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3413 handles timeouts, both on the server side and network timeouts. Renamed the
3414 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3416 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3417 It is now set to "smtps".
3419 22. $host_address is now set to the target address during the checking of
3420 ignore_target_hosts.
3422 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3423 being passed; this would have caused $sender_host_name to have been used if
3424 matching the list had actually called for a host name (not very likely,
3425 since this list is usually IP addresses). A host name is now passed as
3428 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3429 code that specifies a non-restarting handler (typically sigaction() in
3430 modern systems) in an attempt to fix a rare and obscure crash bug.
3432 25. Narrowed the window for a race in the daemon that could cause it to ignore
3433 SIGCHLD signals. This is not a major problem, because they are used only to
3434 wake it up if nothing else does.
3436 26. A malformed maildirsize file could cause Exim to calculate negative values
3437 for the mailbox size or file count. Odd effects could occur as a result.
3438 The maildirsize information is now recalculated if the size or filecount
3441 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3442 support for a long time. Removed HAVE_SYS_VFS_H.
3444 28. Installed the latest version of exipick from John Jetmore.
3446 29. In an address list, if the pattern was not a regular expression, an empty
3447 subject address (from a bounce message) matched only if the pattern was an
3448 empty string. Non-empty patterns were not even tested. This was the wrong
3449 because it is perfectly reasonable to use an empty address as part of a
3450 database query. An empty address is now tested by patterns that are
3451 lookups. However, all the other forms of pattern expect the subject to
3452 contain a local part and a domain, and therefore, for them, an empty
3453 address still always fails if the pattern is not itself empty.
3455 30. Exim went into a mad DNS loop when attempting to do a callout where the
3456 host was specified on an smtp transport, and looking it up yielded more
3457 than one IP address.
3459 31. Re-factored the code for checking spool and log partition space into a
3460 function that finds that data and another that does the check. The former
3461 is then used to implement four new variables: $spool_space, $log_space,
3462 $spool_inodes, and $log_inodes.
3464 32. The RFC2047 encoding function was originally intended for short strings
3465 such as real names; it was not keeping to the 75-character limit for
3466 encoded words that the RFC imposes. It now respects the limit, and
3467 generates multiple encoded words if necessary. To be on the safe side, I
3468 have increased the buffer size for the ${rfc2047: expansion operator from
3471 33. It is now permitted to omit both strings after an "if" condition; if the
3472 condition is true, the result is "true". As before, when the second string
3473 is omitted, a false condition yields an empty string. This makes it less
3474 cumbersome to write custom ACL and router conditions.
3476 34. Failure to deliver a bounce message always caused it to be frozen, even if
3477 there was an errors_to setting on the router. The errors_to setting is now
3480 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3481 canonical form (fully expanded) before being placed in
3482 $sender_host_address.
3484 36. The table in the code that translates DNS record types into text (T_A to
3485 "A" for instance) was missing entries for NS and CNAME. It is just possible
3486 that this could have caused confusion if both these types were looked up
3487 for the same domain, because the text type is used as part of Exim's
3488 per-process caching. But the chance of anyone hitting this buglet seems
3491 37. The dnsdb lookup has been extended in a number of ways.
3493 (1) There is a new type, "zns", which walks up the domain tree until it
3494 finds some nameserver records. It should be used with care.
3496 (2) There is a new type, "mxh", which is like "mx" except that it returns
3497 just the host names, not the priorities.
3499 (3) It is now possible to give a list of domains (or IP addresses) to be
3500 looked up. The behaviour when one of the lookups defers can be
3501 controlled by a keyword.
3503 (4) It is now possible to specify the separator character for use when
3504 multiple records are returned.
3506 38. The dnslists ACL condition has been extended: it is now possible to supply
3507 a list of IP addresses and/or domains to be looked up in a particular DNS
3510 39. Added log_selector=+queue_time_overall.
3512 40. When running the queue in the test harness, wait just a tad after forking a
3513 delivery process, to get repeatability of debugging output.
3515 41. Include certificate and key file names in error message when GnuTLS fails
3516 to set them up, because the GnuTLS error message doesn't include the name
3517 of the failing file when there is a problem reading it.
3519 42. Allow both -bf and -bF in the same test run.
3521 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3523 44. The "Exiscan patch" is now merged into the mainline Exim source.
3525 45. Sometimes the final signoff response after QUIT could fail to get
3526 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3527 < 0 before doing a fflush(). This bug looks as though it goes back to the
3528 introduction of TLS in release 3.20, but "sometimes" must have been rare
3529 because the tests only now provoked it.
3531 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3532 (this can affect the format of dates).
3534 47. exim_tidydb, when checking for the continued existence of a message for
3535 which it has found a message-specific retry record, was not finding
3536 messages that were in split spool directories. Consequently, it was
3537 deleting retry records that should have stayed in existence.
3539 48. Steve fixed some bugs in eximstats.
3541 49. The SPA authentication driver was not abandoning authentication and moving
3542 on to the next authenticator when an expansion was forced to fail,
3543 contradicting the general specification for all authenticators. Instead it
3544 was generating a temporary error. It now behaves as specified.
3546 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3547 (the order specifies the preference for clients). The order is now AES256,
3548 AES128, 3DES, ARCFOUR128.
3550 51. Small patch to Sieve code - explicitly set From: when generating an
3553 52. Exim crashed if a remote delivery caused a very long error message to be
3554 recorded - for instance if somebody sent an entire SpamAssassin report back
3555 as a large number of 550 error lines. This bug was coincidentally fixed by
3556 increasing the size of one of Exim's internal buffers (big_buffer) that
3557 happened as part of the Exiscan merge. However, to be on the safe side, I
3558 have made the code more robust (and fixed the comments that describe what
3561 53. Now that there can be additional text after "Completed" in log lines (if
3562 the queue_time_overall log selector is set), a one-byte patch to exigrep
3563 was needed to allow it to recognize "Completed" as not the last thing in
3566 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3567 patch that reportedly fixes this has been added. I am not expert enough to
3568 create a test for it. This is what the patch creator wrote:
3570 "I found a little strange behaviour of ldap code when working with
3571 Windows 2003 AD Domain, where users was placed in more than one
3572 Organization Units. When I tried to give exim partial DN, the exit code
3573 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3574 But simultaneously result of request was absolutely normal ldap result,
3575 so I produce this patch..."
3577 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3578 REFERENCE, so I have modified the code to exclude the patch when that macro
3581 55. Some experimental protocols are using DNS PTR records for new purposes. The
3582 keys for these records are domain names, not reversed IP addresses. The
3583 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3584 leaves it alone. Component reversal etc. now happens only for IP addresses.
3587 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3589 57. Double the size of the debug message buffer (to 2048) so that more of very
3590 long debug lines gets shown.
3592 58. The exicyclog utility now does better if the number of log files to keep
3593 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3595 59. Two changes related to the smtp_active_hostname option:
3597 (1) $smtp_active_hostname is now available as a variable.
3598 (2) The default for smtp_banner uses $smtp_active_hostname instead
3599 of $primary_hostname.
3601 60. The host_aton() function is supposed to be passed a string that is known
3602 to be a valid IP address. However, in the case of IPv6 addresses, it was
3603 not checking this. This is a hostage to fortune. Exim now panics and dies
3604 if the condition is not met. A case was found where this could be provoked
3605 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3606 components; fortuitously, this particular loophole had already been fixed
3607 by change 4.50/55 above.
3609 If there are any other similar loopholes, the new check in host_aton()
3610 itself should stop them being exploited. The report I received stated that
3611 data on the command line could provoke the exploit when Exim was running as
3612 exim, but did not say which command line option was involved. All I could
3613 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3614 running as the user.
3617 61. There was a buffer overflow vulnerability in the SPA authentication code
3618 (which came originally from the Samba project). I have added a test to the
3619 spa_base64_to_bits() function which I hope fixes it.
3622 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3623 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3625 63. The daemon start-up calls getloadavg() while still root for those OS that
3626 need the first call to be done as root, but it missed one case: when
3627 deliver_queue_load_max is set with deliver_drop_privilege. This is
3628 necessary for the benefit of the queue runner, because there is no re-exec
3629 when deliver_drop_privilege is set.
3631 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3632 This has been fixed.
3634 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3635 was in use, was not putting the data itself into the right store pool;
3636 consequently, it could be overwritten for a subsequent message in the same
3637 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3640 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3642 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3643 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3644 the function were treating the return as a boolean value, which happened to
3645 work because 0=false and not-0=true, but is not correct code.
3647 68. The host_aton() function was not handling scoped IPv6 addresses (those
3648 with, for example, "%eth0" on the end) correctly.
3650 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3651 negated items (that is, ~something) in unsigned ints. Some compilers
3652 apparently mutter when there is no cast.
3654 70. If an address verification called from an ACL failed, and did not produce a
3655 user-specific message (i.e. there was only a "system" message), nothing was
3656 put in $acl_verify_message. In this situation, it now puts the system
3659 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3660 session; change 4.31/43 added the unwanted input to the log line - except
3661 that it did not do this in the start of session case. It now does.
3663 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3664 This is wrong and can cause the other end to generate a synchronization
3665 error if it is another Exim or anything else that does the synchronization
3666 check. A QUIT command is no longer sent after a timeout.
3668 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3669 during host lookups.
3671 74. The defer_ok option of callout verification was not working if it was used
3672 when verifying addresses in header lines, that is, for this case:
3674 verify = header_sender/callout=defer_ok
3676 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3677 those file descriptors could be used for SMTP connections. If anything
3678 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3679 could be sent to the SMTP client, causing chaos. The daemon now opens
3680 stdin, stdout, and stderr to /dev/null when it puts itself into the
3683 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3684 log by default. The user can override this with suitable Perl magic.
3686 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3687 the log message when discard triggers, was not working for the DATA ACL or
3688 for the non-SMTP ACL.
3690 78. Error message wording change in sieve.c.
3692 79. If smtp_accept_max_per_host was set, the number of connections could be
3693 restricted to fewer than expected, because the daemon was trying to set up
3694 a new connection before checking whether the processes handling previous
3695 connections had finished. The check for completed processes is now done
3696 earlier. On busy systems, this bug wouldn't be noticed because something
3697 else would have woken the daemon, and it would have reaped the completed
3700 80. If a message was submitted locally by a user whose login name contained one
3701 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3702 It caused a spool format error. I have fixed the spool reading code. A
3703 related problem was that the "from" clause in the Received: line became
3704 illegal because of the space(s). It is now covered by ${quote_local_part.
3706 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3709 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3711 83. Patch from Sieve maintainer to fix unterminated string problem in
3712 "vacation" handling.
3714 84. Some minor changes to the Linux configuration files to help with other
3715 OS variants using glibc.
3717 85. One more patch for Sieve to update vacation handling to latest spec.
3720 ----------------------------------------------------
3721 See the note above about the 4.44 and 4.50 releases.
3722 ----------------------------------------------------
3728 1. Change 4.43/35 introduced a bug that caused file counts to be
3729 incorrectly computed when quota_filecount was set in an appendfile
3732 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3733 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3736 3. Give more explanation in the error message when the command for a transport
3737 filter fails to execute.
3739 4. There are several places where Exim runs a non-Exim command in a
3740 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3741 was being done only for the command run by the queryprogram router. It is
3742 now done for all such subprocesses. The other cases are: ${run, transport
3743 filters, and the commands run by the lmtp and pipe transports.
3745 5. Some older OS have a limit of 256 on the maximum number of file
3746 descriptors. Exim was using setrlimit() to set 1000 as a large value
3747 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3748 systems. I've change it so that if it can't get 1000, it tries for 256.
3750 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3751 was an oversight, and furthermore, ever since the addition of extra
3752 controls (e.g. 4.43/32), the checks on when to allow different forms of
3753 "control" were broken. There should now be diagnostics for all cases when a
3754 control that does not make sense is encountered.
3756 7. $recipients is now available in the predata ACL (oversight).
3758 8. Tidy the search cache before the fork to do a delivery from a message
3759 received from the command line. Otherwise the child will trigger a lookup
3760 failure and thereby defer the delivery if it tries to use (for example) a
3761 cached ldap connection that the parent has called unbind on.
3763 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3764 of $address_data from the recipient verification was clobbered by the
3765 sender verification.
3767 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3768 was its contents. (It was OK if the option was not defined at all.)
3770 11. A "Completed" log line is now written for messages that are removed from
3771 the spool by the -Mrm option.
3773 12. $host_address is now set to the target address during the checking of
3774 ignore_target_hosts.
3776 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3777 being passed; this would have caused $sender_host_name to have been used if
3778 matching the list had actually called for a host name (not very likely,
3779 since this list is usually IP addresses). A host name is now passed as
3782 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3783 code that specifies a non-restarting handler (typically sigaction() in
3784 modern systems) in an attempt to fix a rare and obscure crash bug.
3786 15. Narrowed the window for a race in the daemon that could cause it to ignore
3787 SIGCHLD signals. This is not a major problem, because they are used only to
3788 wake it up if nothing else does.
3790 16. A malformed maildirsize file could cause Exim to calculate negative values
3791 for the mailbox size or file count. Odd effects could occur as a result.
3792 The maildirsize information is now recalculated if the size or filecount
3795 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3796 support for a long time. Removed HAVE_SYS_VFS_H.
3798 18. Updated exipick to current release from John Jetmore.
3800 19. Allow an empty sender to be matched against a lookup in an address list.
3801 Previously the only cases considered were a regular expression, or an
3804 20. Exim went into a mad DNS lookup loop when doing a callout where the
3805 host was specified on the transport, if the DNS lookup yielded more than
3808 21. The RFC2047 encoding function was originally intended for short strings
3809 such as real names; it was not keeping to the 75-character limit for
3810 encoded words that the RFC imposes. It now respects the limit, and
3811 generates multiple encoded words if necessary. To be on the safe side, I
3812 have increased the buffer size for the ${rfc2047: expansion operator from
3815 22. Failure to deliver a bounce message always caused it to be frozen, even if
3816 there was an errors_to setting on the router. The errors_to setting is now
3819 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3820 canonical form (fully expanded) before being placed in
3821 $sender_host_address.
3823 24. Updated eximstats to version 1.33
3825 25. Include certificate and key file names in error message when GnuTLS fails
3826 to set them up, because the GnuTLS error message doesn't include the name
3827 of the failing file when there is a problem reading it.
3829 26. Expand error message when OpenSSL has problems setting up cert/key files.
3832 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3833 (this can affect the format of dates).
3835 28. exim_tidydb, when checking for the continued existence of a message for
3836 which it has found a message-specific retry record, was not finding
3837 messages that were in split spool directories. Consequently, it was
3838 deleting retry records that should have stayed in existence.
3840 29. eximstats updated to version 1.35
3841 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3842 1.35 - bugfix such that pie charts by volume are generated correctly
3844 30. The SPA authentication driver was not abandoning authentication and moving
3845 on to the next authenticator when an expansion was forced to fail,
3846 contradicting the general specification for all authenticators. Instead it
3847 was generating a temporary error. It now behaves as specified.
3849 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3850 (the order specifies the preference for clients). The order is now AES256,
3851 AES128, 3DES, ARCFOUR128.
3853 31. Small patch to Sieve code - explicitly set From: when generating an
3856 32. Exim crashed if a remote delivery caused a very long error message to be
3857 recorded - for instance if somebody sent an entire SpamAssassin report back
3858 as a large number of 550 error lines. This bug was coincidentally fixed by
3859 increasing the size of one of Exim's internal buffers (big_buffer) that
3860 happened as part of the Exiscan merge. However, to be on the safe side, I
3861 have made the code more robust (and fixed the comments that describe what
3864 33. Some experimental protocols are using DNS PTR records for new purposes. The
3865 keys for these records are domain names, not reversed IP addresses. The
3866 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3867 leaves it alone. Component reversal etc. now happens only for IP addresses.
3870 34. The host_aton() function is supposed to be passed a string that is known
3871 to be a valid IP address. However, in the case of IPv6 addresses, it was
3872 not checking this. This is a hostage to fortune. Exim now panics and dies
3873 if the condition is not met. A case was found where this could be provoked
3874 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3875 components; fortuitously, this particular loophole had already been fixed
3876 by change 4.50/55 or 4.44/33 above.
3878 If there are any other similar loopholes, the new check in host_aton()
3879 itself should stop them being exploited. The report I received stated that
3880 data on the command line could provoke the exploit when Exim was running as
3881 exim, but did not say which command line option was involved. All I could
3882 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3883 running as the user.
3886 35. There was a buffer overflow vulnerability in the SPA authentication code
3887 (which came originally from the Samba project). I have added a test to the
3888 spa_base64_to_bits() function which I hope fixes it.
3891 36. The daemon start-up calls getloadavg() while still root for those OS that
3892 need the first call to be done as root, but it missed one case: when
3893 deliver_queue_load_max is set with deliver_drop_privilege. This is
3894 necessary for the benefit of the queue runner, because there is no re-exec
3895 when deliver_drop_privilege is set.
3897 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3898 was in use, was not putting the data itself into the right store pool;
3899 consequently, it could be overwritten for a subsequent message in the same
3900 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3903 38. Sometimes the final signoff response after QUIT could fail to get
3904 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3905 < 0 before doing a fflush(). This bug looks as though it goes back to the
3906 introduction of TLS in release 3.20, but "sometimes" must have been rare
3907 because the tests only now provoked it.
3913 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3914 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3915 Now it returns an integer. A number of calls were still expecting a T/F
3916 return. Fortuitously, in all cases, the tests worked in OK situations,
3917 which is the norm. However, things would have gone wrong on any write
3918 failures on the smtp file descriptor. This function is used when sending
3919 messages over SMTP and also when doing verify callouts.
3921 2. When Exim is called to do synchronous delivery of a locally submitted
3922 message (the -odf or -odi options), it no longer closes stderr before doing
3925 3. Implemented the mua_wrapper option.
3927 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3929 5. Implemented the functions header_remove(), header_testname(),
3930 header_add_at_position(), and receive_remove_recipient(), and exported them
3933 6. If an ACL "warn" statement specified the addition of headers, Exim already
3934 inserted X-ACL-Warn: at the start if there was no header name. However, it
3935 was not making this test for the second and subsequent header lines if
3936 there were newlines in the string. This meant that an invalid header could
3937 be inserted if Exim was badly configured.
3939 7. Allow an ACL "warn" statement to add header lines at the start or after all
3940 the Received: headers, as well as at the end.
3942 8. Added the rcpt_4xx retry error code.
3944 9. Added postmaster_mailfrom=xxx to callout verification option.
3946 10. Added mailfrom=xxxx to the callout verification option, for verify=
3949 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3950 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3952 12. Inserted some casts to stop certain compilers warning when using pointer
3953 differences as field lengths or precisions in printf-type calls (mostly
3954 affecting debugging statements).
3956 13. Added optional readline() support for -be (dynamically loaded).
3958 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3959 same clock tick as a message's arrival, so that its received time was the
3960 same as the "first fail" time on the retry record, and that message
3961 remained on the queue past the ultimate address timeout, every queue runner
3962 would try a delivery (because it was past the ultimate address timeout) but
3963 after another failure, the ultimate address timeout, which should have then
3964 bounced the address, did not kick in. This was a "< instead of <=" error;
3965 in most cases the first failure would have been in the next clock tick
3966 after the received time, and all would be well.
3968 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3969 being recognized when the domain list was tested by the match_domain
3970 condition in an expansion string.
3972 16. Added the ${str2b64: operator.
3974 17. Exim was always calling setrlimit() to set a large limit for the number of
3975 processes, without checking whether the existing limit was already
3976 adequate. (It did check for the limit on file descriptors.) Furthermore,
3977 errors from getrlimit() and setrlimit() were being ignored. Now they are
3978 logged to the main and panic logs, but Exim does carry on, to try to do its
3979 job under whatever limits there are.
3981 18. Imported PCRE 5.0.
3983 19. Trivial typo in log message " temporarily refused connection" (the leading
3986 20. If the log selector return_path_on_delivery was set and an address was
3987 redirected to /dev/null, the delivery process crashed because it assumed
3988 that a return path would always be set for a "successful" delivery. In this
3989 case, the whole delivery is bypassed as an optimization, and therefore no
3992 21. Internal re-arrangement: the function for sending a challenge and reading
3993 a response while authentication was assuming a zero-terminated challenge
3994 string. It's now changed to take a pointer and a length, to allow for
3995 binary data in such strings.
3997 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3999 23. Exim was not respecting finduser_retries when seeking the login of the
4000 uid under which it was called; it was always trying 10 times. (The default
4001 setting of finduser_retries is zero.) Also, it was sleeping after the final
4002 failure, which is pointless.
4004 24. Implemented tls_on_connect_ports.
4006 25. Implemented acl_smtp_predata.
4008 26. If the domain in control=submission is set empty, Exim assumes that the
4009 authenticated id is a complete email address when it generates From: or
4010 Sender: header lines.
4012 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4013 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4014 chown and chgrp in /bin and hostname in /usr/bin.
4016 28. Exim was keeping the "process log" file open after each use, just as it
4017 does for the main log. This opens the possibility of it remaining open for
4018 long periods when the USR1 signal hits a daemon. Occasional processlog
4019 errors were reported, that could have been caused by this. Anyway, it seems
4020 much more sensible not to leave this file open at all, so that is what now
4023 29. The long-running daemon process does not normally write to the log once it
4024 has entered its main loop, and it closes the log before doing so. This is
4025 so that log files can straightforwardly be renamed and moved. However,
4026 there are a couple of unusual error situations where the daemon does write
4027 log entries, and I had neglected to close the log afterwards.
4029 30. The text of an SMTP error response that was received during a remote
4030 delivery was being truncated at 512 bytes. This is too short for some of
4031 the long messages that one sometimes sees. I've increased the limit to
4034 31. It is now possible to make retry rules that apply only when a message has a
4035 specific sender, in particular, an empty sender.
4037 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4038 it possible to be selective about when SMTP synchronization is enforced.
4040 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4042 32. Implemented hosts_connection_nolog.
4044 33. Added an ACL for QUIT.
4046 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4049 35. Added mailbox_size and mailbox_filecount to appendfile.
4051 36. Added control = no_multiline_responses to ACLs.
4053 37. There was a bug in the logic of the code that waits for the clock to tick
4054 in the case where the clock went backwards by a substantial amount such
4055 that the microsecond fraction of "now" was more than the microsecond
4056 fraction of "then" (but the whole seconds number was less).
4058 38. Added support for the libradius Radius client library this is found on
4059 FreeBSD (previously only the radiusclient library was supported).
4065 1. When certain lookups returned multiple values in the form name=value, the
4066 quoting of the values was not always being done properly. Specifically:
4067 (a) If the value started with a double quote, but contained no whitespace,
4069 (b) If the value contained whitespace other than a space character (i.e.
4070 tabs or newlines or carriage returns) it was not quoted.
4071 This fix has been applied to the mysql and pgsql lookups by writing a
4072 separate quoting function and calling it from the lookup code. The fix
4073 should probably also be applied to nisplus, ibase and oracle lookups, but
4074 since I cannot test any of those, I have not disturbed their existing code.
4076 2. A hit in the callout cache for a specific address caused a log line with no
4077 reason for rejecting RCPT. Now it says "Previous (cached) callout
4078 verification failure".
4080 3. There was an off-by-one bug in the queryprogram router. An over-long
4081 return line was truncated at 256 instead of 255 characters, thereby
4082 overflowing its buffer with the terminating zero. As well as fixing this, I
4083 have increased the buffer size to 1024 (and made a note to document this).
4085 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4086 when Exim is waiting for an SMTP response from a remote server, Exim
4087 restarts its select() call on the socket, thereby resetting its timeout.
4088 This is not a problem when such interrupts are rare. Somebody set up a cron
4089 job to run exiwhat every 2 minutes, which is less than the normal select()
4090 timeout (5 or 10 minutes). This meant that the select() timeout never
4091 kicked in because it was always reset. I have fixed this by comparing the
4092 time when an interrupt arrives with the time at the start of the first call
4093 to select(). If more time than the timeout has elapsed, the interrupt is
4094 treated as a timeout.
4096 5. Some internal re-factoring in preparation for the addition of Sieve
4097 extensions (by MH). In particular, the "personal" test is moved to a
4098 separate function, and given an option for scanning Cc: and Bcc: (which is
4099 not set for Exim filters).
4101 6. When Exim created an email address using the login of the caller as the
4102 local part (e.g. when creating a From: or Sender: header line), it was not
4103 quoting the local part when it contained special characters such as @.
4105 7. Installed new OpenBSD configuration files.
4107 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4108 try to make them clearer.
4110 9. Callout options, other than the timeout value, were being ignored when
4111 verifying sender addresses in header lines. For example, when using
4113 verify = header_sender/callout=no_cache
4115 the cache was (incorrectly) being used.
4117 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4118 only the Cygwin environment.
4120 11. When return_path_on_delivery was set as a log selector, if different remote
4121 addresses in the same message used different return paths and parallel
4122 remote delivery occurred, the wrong values would sometimes be logged.
4123 (Whenever a remote delivery process finished, the return path value from
4124 the most recently started remote delivery process was logged.)
4126 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4127 lines when AUTH and/or TLS are in use. This is the "received protocol"
4128 field. Exim used to use "asmtp" for authenticated SMTP, without any
4129 indication (in the protocol name) for TLS use. Now it follows the RFC and
4130 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4131 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4132 names appear in log lines as well as in Received: header lines.
4134 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4135 extensions, and comparison tests, and to fix some bugs.
4137 14. Changes to the "personal" filter test:
4139 (1) The test was buggy in that it was just doing the equivalent of
4140 "contains" tests on header lines. For example, if a user's address was
4141 anne@some.where, the "personal" test would incorrectly be true for
4143 To: susanne@some.where
4145 This test is now done by extracting each address from the header in turn,
4146 and checking the entire address. Other tests that are part of "personal"
4147 are now done using regular expressions (for example, to check local parts
4148 of addresses in From: header lines).
4150 (2) The list of non-personal local parts in From: addresses has been
4151 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4152 taken from the Sieve specification recommendations.
4154 (3) If the message contains any header line starting with "List-" it is
4155 treated as non-personal.
4157 (4) The test for "circular" in the Subject: header line has been removed
4158 because it now seems ill-conceived.
4160 15. Minor typos in src/EDITME comments corrected.
4162 16. Installed latest exipick from John Jetmore.
4164 17. If headers_add on a router specified a text string that was too long for
4165 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4166 of string_sprintf() is now avoided.
4168 18. $message_body_size was not set (it was always zero) when running the DATA
4169 ACL and the local_scan() function.
4171 19. For the "mail" command in an Exim filter, no default was being set for
4172 the once_repeat time, causing a random time value to be used if "once" was
4173 specified. (If the value happened to be <= 0, no repeat happened.) The
4174 default is now 0s, meaning "never repeat". The "vacation" command was OK
4175 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4176 (I found it when inspecting the code).
4178 20. There is now an overall timeout for performing a callout verification. It
4179 defaults to 4 times the callout timeout, which applies to individual SMTP
4180 commands during the callout. The overall timeout applies when there is more
4181 than one host that can be tried. The timeout is checked before trying the
4182 next host. This prevents very long delays if there are a large number of
4183 hosts and all are timing out (e.g. when the network connections are timing
4184 out). The value of the overall timeout can be changed by specifying an
4185 additional sub-option for "callout", called "maxwait". For example:
4187 verify = sender/callout=5s,maxwait=20s
4189 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4190 to the end before writing, but this should make it even safer).
4192 22. Exim was forgetting that it had advertised PIPELINING for the second and
4193 subsequent messages on an SMTP connection. It was also not resetting its
4194 memory on STARTTLS and an internal HELO.
4196 23. When Exim logs an SMTP synchronization error within a session, it now
4197 records whether PIPELINING has been advertised or not.
4199 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4200 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4201 rather than long int.
4203 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4205 26. Added the never_mail option to autoreply.
4211 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4212 crash if the getsockname() call failed; this can happen if a connection is
4213 closed very soon after it is established. The problem was simply in the
4214 order in which certain operations were done, causing Exim to try to write
4215 to the SMTP stream before it had set up the file descriptor. The bug has
4216 been fixed by making things happen in the correct order.
4222 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4223 before the connection was closed, thus losing the rejection response.
4225 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4226 some early Solaris releases, but causes trouble in current releases where
4227 socklen_t is defined.
4229 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4232 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4235 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4236 that releasing the top of it at the end releases what was used for sub-
4237 expansions (unless the block got too big). However, discard this block if
4238 the first thing is a variable or header, so that we can use its block when
4239 it is dynamic (useful for very large $message_headers, for example).
4241 6. Lookups now cache *every* query, not just the most recent. A new, separate
4242 store pool is used for this. It can be recovered when all lookup caches are
4243 flushed. Lookups now release memory at the end of their result strings.
4244 This has involved some general refactoring of the lookup sources.
4246 7. Some code has been added to the store_xxx() functions to reduce the amount
4247 of flapping under certain conditions.
4249 8. log_incoming_interface used to affect only the <= reception log lines. Now
4250 it causes the local interface and port to be added to several more SMTP log
4251 lines, for example "SMTP connection from", and rejection lines.
4253 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4255 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4257 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4258 could be overwritten at the end of the current message (or the start of a
4259 new message if it was set in a HELO ACL). The value is now preserved for
4260 the duration of the SMTP connection.
4262 12. If a transport had a headers_rewrite setting, and a matching header line
4263 contained an unqualified address, that address was qualified, even if it
4264 did not match any rewriting rules. The underlying bug was that the values
4265 of the flags that permit the existence of unqualified sender and recipient
4266 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4267 non-local messages, and by -bnq for local messages) were not being
4268 preserved with the message after it was received.
4270 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4271 "next input=" as part of the text comprising the host identity instead of
4272 the correct text. The code was using the same buffer for two different
4273 strings. However, depending on which order the printing function evaluated
4274 its arguments, the bug did not always show up. Under Linux, for example, my
4275 test suite worked just fine.
4277 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4278 doesn't work with some older versions of Perl. It has been changed to "my",
4279 which in any case is probably the better facility to use.
4281 15. A really picky compiler found some instances of statements for creating
4282 error messages that either had too many or two few arguments for the format
4285 16. The size of the buffer for calls to the DNS resolver has been increased
4286 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4287 for addresses that have a lot of PTR records. This alleviates a problem; it
4288 does not fully solve it.
4290 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4291 buffer now truncates the list and logs the incident, which is the same
4292 action as happens when Exim is looking up a host name and its aliases.
4293 Previously in this situation something unpredictable would happen;
4294 sometimes it was "internal error: store_reset failed".
4296 18. If a server dropped the connection unexpectedly when an Exim client was
4297 using GnuTLS and trying to read a response, the client delivery process
4298 crashed while trying to generate an error log message.
4300 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4301 string, for example:
4303 warn message = H1: something\nH2: something
4305 the text was added as a single header line from Exim's point of view
4306 though it ended up OK in the delivered message. However, searching for the
4307 second and subsequent header lines using $h_h2: did not work. This has been
4308 fixed. Similarly, if a system filter added multiple headers in this way,
4309 the routers could not see them.
4311 20. Expanded the error message when iplsearch is called with an invalid key to
4312 suggest using net-iplsearch in a host list.
4314 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4315 ACLs are no longer actually imposed (and a message to that effect is
4318 22. If a "gecos" field in a passwd entry contained escaped characters, in
4319 particular, if it contained a \" sequence, Exim got it wrong when building
4320 a From: or a Sender: header from that name. A second bug also caused
4321 incorrect handling when an unquoted " was present following a character
4322 that needed quoting.
4324 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4325 was not being matched caselessly.
4327 24. Arranged for all hyphens in the exim.8 source to be escaped with
4330 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4331 properly. Recipient callout cache records were still being keyed to include
4332 the sender, even when use_sender was set false. This led to far more
4333 callouts that were necessary. The sender is no longer included in the key
4334 when use_sender is false.
4336 26. Added "control = submission" modifier to ACLs.
4338 27. Added the ${base62d: operator to decode base 62 numbers.
4340 28. dnsdb lookups can now access SRV records.
4342 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4343 the configuration file.
4345 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4346 (-v) mode. This makes the output for a verbose queue run more intelligible.
4348 31. Added a use_postmaster feature to recipient callouts.
4350 32. Added the $body_zerocount variable, containing the number of binary zero
4351 bytes in the message body.
4353 33. The time of last modification of the "new" subdirectory is now used as the
4354 "mailbox time last read" when there is a quota error for a maildir
4357 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4359 35. Added +ignore_unknown as a special item in host lists.
4361 36. Code for decoding IPv6 addresses in host lists is now included, even if
4362 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4363 address was recognized as an IP address, but was then not correctly decoded
4364 into binary, causing unexpected and incorrect effects when compared with
4371 1. Very minor rewording of debugging text in manualroute to say "list of
4372 hosts" instead of "hostlist".
4374 2. If verify=header_syntax was set, and a header line with an unqualified
4375 address (no domain) and a large number of spaces between the end of the
4376 name and the colon was received, the reception process suffered a buffer
4377 overflow, and (when I tested it) crashed. This was caused by some obsolete
4378 code that should have been removed. The fix is to remove it!
4380 3. When running in the test harness, delay a bit after writing a bounce
4381 message to get a bit more predictability in the log output.
4383 4. Added a call to search_tidyup() just before forking a reception process. In
4384 theory, someone could use a lookup in the expansion of smtp_accept_max_
4385 per_host which, without the tidyup, could leave open a database connection.
4387 5. Added the variables $recipient_data and $sender_data which get set from a
4388 lookup success in an ACL "recipients" or "senders" condition, or a router
4389 "senders" option, similar to $domain_data and $local_part_data.
4391 6. Moved the writing of debug_print from before to after the "senders" test
4394 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4395 problems for message scanning, either using a data ACL, or using
4396 local_scan() because the Received: header was not generated till after they
4397 were called (in order to set the time as the time of reception completion).
4398 I have revised the way this works. The header is now generated after the
4399 body is received, but before the ACL or local_scan() are called. After they
4400 are run, the timestamp in the header is updated.
4406 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4407 before starting a queue runner without re-exec. This happened only when
4408 deliver_drop_privilege was set or when the Exim user was set to root. The
4409 effect of the bug was that timeouts during subsequent deliveries caused
4410 crashes instead of being properly handled. The handler is now left at its
4411 default (and expected) setting.
4413 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4414 message, again when deliver_drop_privilege is set or Exim is run as root.
4415 The bug described in (1) was not present in this case, but the tidying up
4416 of the other signals was missing. I have made the two cases consistent.
4418 3. The ignore_target_hosts setting on a manualroute router was being ignored
4419 for hosts that were looked up using the /MX notation.
4421 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4424 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4425 operated on the sender address. After changing the $sender_address to <>
4426 for the sender address verify, Exim was re-instated it as the original
4427 (before rewriting) address, but remembering that it had rewritten it, so it
4428 wasn't rewriting it again. This bug also had the effect of breaking the
4429 sender address verification caching when the sender address was rewritten.
4431 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4432 This has been changed so that if the ip literal address matches
4433 ignore_target_hosts, the router declines.
4435 7. Added expansion conditions match_domain, match_address, and match_local_
4436 part (NOT match_host).
4438 8. The placeholder for the Received: header didn't have a length field set.
4440 9. Added code to Exim itself and to exim_lock to test for a specific race
4441 condition that could lead to file corruption when using MBX delivery. The
4442 issue is with the lockfile that is created in /tmp. If this file is removed
4443 after a process has opened it but before that process has acquired a lock,
4444 there is the potential for a second process to recreate the file and also
4445 acquire a lock. This could lead to two Exim processes writing to the file
4446 at the same time. The added code performs the same test as UW imapd; it
4447 checks after acquiring the lock that its file descriptor still refers to
4448 the same named file.
4450 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4451 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4454 11. Added the smtp_active_hostname option. If used, this will typically be made
4455 to depend on the incoming interface address. Because $interface_address is
4456 not set up until the daemon has forked a reception process, error responses
4457 that can happen earlier (such as "too many connections") no longer contain
4460 12. If an expansion in a condition on a "warn" statement fails because a lookup
4461 defers, the "warn" statement is abandoned, and the next ACL statement is
4462 processed. Previously this caused the whole ACL to be aborted.
4464 13. Added the iplsearch lookup type.
4466 14. Added ident_timeout as a log selector.
4468 15. Added tls_certificate_verified as a log selector.
4470 16. Added a global option tls_require_ciphers (compare the smtp transport
4471 option of the same name). This controls incoming TLS connections.
4473 17. I finally figured out how to make tls_require_ciphers do a similar thing
4474 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4475 before starting the TLS session.
4477 18. Tabs are now shown as \t in -bP output.
4479 19. If the log selector return_path_on_delivery was set, Exim crashed when
4480 bouncing a message because it had too many Received: header lines.
4482 20. If two routers both had headers_remove settings, and the first one included
4483 a superfluous trailing colon, the final name in the first list and the
4484 first name in the second list were incorrectly joined into one item (with a
4485 colon in the middle).
4491 1. Added -C and -D options to the exinext utility, mainly to make it easier
4492 to include in the automated testing, but these could be helpful when
4493 multiple configurations are in use.
4495 2. The exinext utility was not formatting the output nicely when there was
4496 an alternate port involved in the retry record key, nor when there was a
4497 message id as well (for retries that were specific to a specific message
4498 and a specific host). It was also confused by IPv6 addresses, because of
4499 the additional colons they contain. I have fixed the IPv4 problem, and
4500 patched it up to do a reasonable job for IPv6.
4502 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4503 delivery, the log line now contains "pipelined" if PIPELINING was used.
4505 4. An SMTP transport process used to panic and die if the bind() call to set
4506 an explicit outgoing interface failed. This has been changed; it is now
4507 treated in the same way as a connect() failure.
4509 5. A reference to $sender_host_name in the part of a conditional expansion
4510 that was being skipped was still causing a DNS lookup. This no longer
4513 6. The def: expansion condition was not recognizing references to header lines
4514 that used bh_ and bheader_.
4516 7. Added the _cache feature to named lists.
4518 8. The code for checking quota_filecount in the appendfile transport was
4519 allowing one more file than it should have been.
4521 9. For compatibility with Sendmail, the command line option
4529 and sets the incoming protocol and host name (for trusted callers). The
4530 host name and its colon can be omitted when only the protocol is to be set.
4531 Note the Exim already has two private options, -pd and -ps, that refer to
4532 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4533 "s", but I don't think that's a major issue.
4535 10. A number of refactoring changes to the code, none of which should affect
4538 (a) The number of logging options was getting close to filling up the
4539 32-bit word that was used as a bit map. I have split them into two classes:
4540 those that are passed in the argument to log_write(), and those that are
4541 only ever tested independently outside of that function. These are now in
4542 separate 32-bit words, so there is plenty of room for expansion again.
4543 There is no change in the user interface or the logging behaviour.
4545 (b) When building, for example, log lines, the code previously used a
4546 macro that called string_cat() twice, in order to add two strings. This is
4547 not really sufficiently general. Furthermore, there was one instance where
4548 it was actually wrong because one of the argument was used twice, and in
4549 one call a function was used. (As it happened, calling the function twice
4550 did not affect the overall behaviour.) The macro has been replaced by a
4551 function that can join an arbitrary number of extra strings onto a growing
4554 (c) The code for expansion conditions now uses a table and a binary chop
4555 instead of a serial search (which was left over from when there were very
4556 few conditions). Also, it now recognizes conditions like "pam" even when
4557 the relevant support is not compiled in: a suitably worded error message is
4558 given if an attempt is made to use such a condition.
4560 11. Added ${time_interval:xxxxx}.
4562 12. A bug was causing one of the ddress fields not to be passed back correctly
4563 from remote delivery subprocesses. The field in question was not being
4564 subsequently used, so this caused to problems in practice.
4566 13. Added new log selectors queue_time and deliver_time.
4568 14. Might have fixed a bug in maildirsizefile handling that threw up
4569 "unexpected character" debug warnings, and recalculated the data
4570 unnecessarily. In any case, I expanded the warning message to give more
4573 15. Added the message "Restricted characters in address" to the statements in
4574 the default ACL that block characters like @ and % in local parts.
4576 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4577 Three changes have been made:
4579 (a) There was a serious bug; a negative response to MAIL caused the whole
4580 recipient domain to be cached as invalid, thereby blocking all messages
4581 to all local parts at the same domain, from all senders. This bug has
4582 been fixed. The domain is no longer cached after a negative response to
4583 MAIL if the sender used is not empty.
4585 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4588 (c) A new callout option, "use_sender" has been added for people who want
4589 the modified behaviour.
4595 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4598 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4599 indeed breaks things for older releases.
4601 3. Added additional logging to the case where there is a problem reading data
4602 from a filter that is running in a subprocess using a pipe, in order to
4603 try to track down a specific problem.
4605 4. Testing facility fudge: when running in the test harness and attempting
4606 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4607 getting "No route to host". Convert this to a timeout.
4609 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4612 6. Some OS don't have socklen_t but use size_t instead. This affects the
4613 fifth argument of getsockopt() amongst other things. This is now
4614 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4615 can be set for individual OS. I have set it for SunOS5, OSF1, and
4616 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4617 some earlier ones do not.
4619 7. Change 4.30/15 was not doing the test caselessly.
4621 8. The standard form for an IPv6 address literal was being rejected by address
4622 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4623 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4624 this, as well as the form without the "IPv6" on the front (but only when
4625 address literals are enabled, of course).
4627 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4629 10. Exim crashed if a message with an empty sender address specified by -f
4630 encountered a router with an errors_to setting. This could be provoked only
4631 by a command such as
4635 where an empty string was supplied; "<>" did not hit this bug.
4637 11. Installed PCRE release 4.5.
4639 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4640 remained set. It is now erased.
4642 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4643 times from message ids (which are base 36 rather than the normal 62).
4645 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4646 were being counted as actual protocol errors, and logged if the log
4647 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4648 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4649 and DATA following a set of rejected RCPTs do not count as protocol errors.
4650 In other words, Exim assumes they were pipelined, though this may not
4651 actually be the case. Of course, in all cases the client gets an
4652 appropriate error code.
4654 15. If a lookup fails in an ACL condition, a message about the failure may
4655 be available; it is used if testing the ACL cannot continue, because most
4656 such messages specify what the cause of the deferral is. However, some
4657 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4658 that caused an old message to be retained and used if a later statement
4659 caused a defer, replacing the real cause of the deferral.
4661 16. If an IP address had so many PTR records that the DNS lookup buffer
4662 was not large enough to hold them, Exim could crash while trying to process
4663 the truncated data. It now detects and logs this case.
4665 17. Further to 4.21/58, another change has been made: if (and only if) the
4666 first line of a message (the first header line) ends with CRLF, a bare LF
4667 in a subsequent header line has a space inserted after it, so as not to
4668 terminate the header.
4670 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4671 unnecessarily, used atoi() instead of strtol(), and didn't check the
4672 termination when getting file sizes from file names by regex.
4674 19. Completely re-implemented the support for maildirsize files, in the light
4675 of a number of problems with the previous contributed implementation
4676 (4.30/29). In particular:
4678 . If the quota is zero, the maildirsize file is maintained, but no quota is
4681 . If the maildir directory does not exist, it is created before any attempt
4682 to write a maildirsize file.
4684 . The quota value in the file is just a cache; if the quota is changed in
4685 the transport, the new value overrides.
4687 . A regular expression is available for excluding directories from the
4690 20. The autoreply transport checks the characters in options that define the
4691 message's headers; it allows continued headers, but it was checking with
4692 isspace() after an embedded newline instead of explicitly looking for a
4695 21. If all the "regular" hosts to which an address was routed had passed their
4696 expiry times, and had not reached their retry times, the address was
4697 bounced, even if fallback hosts were defined. Now Exim should go on to try
4700 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4701 equivalent code in the SMTP transport. Some hosts send humungous responses
4702 to HELO/EHLO, more than 1024 it seems.
4704 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4705 gives compiler warnings in some environments. I've now done it "properly",
4708 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4709 (because of problems with the built-in one) was declared to return uschar *
4710 instead of char *, causing compiler failure.
4712 25. Fixed a file descriptor leak when processing alias/forward files.
4714 26. Fixed a minor format string issue in dbfn.c.
4716 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4718 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4719 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4722 29. When a sender address is verified, it is cached, to save repeating the test
4723 when there is more than one recipient in a message. However, when the
4724 verification involves a callout, it is possible for different callout
4725 options to be set for different recipients. It is too complicated to keep
4726 track of this in the cache, so now Exim always runs a verification when a
4727 callout is required, relying on the callout cache for the optimization.
4728 The overhead is duplication of the address routing, but this should not be
4731 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4732 to be verified with callout=postmaster, and the main callout worked but the
4733 postmaster check failed, the verification correctly failed. However, if a
4734 subsequent RCPT command asked for sender verification *without* the
4735 postmaster check, incorrect caching caused this verification also to fail,
4738 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4739 it was not caching the DNS options (qualify_single, search_parents) that
4740 were used when the lookup failed. A subsequent lookup with different
4741 options therefore always gave the same answer, though there were cases
4742 where it should not have. (Example: a "domains = !$mx_any" option on a
4743 dnslookup router: the "domains" option is always processed without any
4744 widening, but the router might have qualify_single set.) Now Exim uses the
4745 cached value only when the same options are set.
4747 32. Added John Jetmore's "exipick" utility to the distribution.
4749 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4750 than a timeout (e.g. a certificate is required, and is not provided), an
4751 Exim server now closes the connection immediately. Previously it waited for
4752 the client to close - but if the client is SSL, it seems that they each
4753 wait for each other, leading to a delay before one of them times out.
4755 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4756 maintained 0.8.x compatibility because I don't think many are using it, and
4757 it is clearly obsolete.
4759 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4762 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4763 client certificate was expired. A simple patch fixes this, though I don't
4764 understand the full logic of why the verify callback is called multiple
4767 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4768 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4769 which causes problems with some clients (such as the Certicom SSL Plus
4770 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4771 disables the coutermeasure allowing Eudora to connect."
4773 38. Exim was not checking that a write() to a log file succeeded. This could
4774 lead to Bad Things if a log got too big, in particular if it hit a file
4775 size limit. Exim now panics and dies if it cannot write to a log file, just
4776 as it does if it cannot open a log file.
4778 39. Modified OS/Makefile-Linux so that it now contains
4780 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4782 The two -D definitions ensure that Exim is compiled with large file
4783 support, which makes it possible to handle log files that are bigger than
4786 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4787 instance) a domain was checked against a named list that involved a lookup,
4788 causing $domain_data to be set, then another domain was checked against the
4789 same list, then the first domain was re-checked, the value of $domain_data
4790 after the final check could be wrong. In particular, if the second check
4791 failed, it could be set empty. This bug probably also applied to
4794 41. The strip_trailing_dot option was not being applied to the address given
4795 with the -f command-line option.
4797 42. The code for reading a message's header from the spool was incrementing
4798 $received_count, but never initializing it. This meant that the value was
4799 incorrect (doubled) while delivering a message in the same process in which
4800 it was received. In the most common configuration of Exim, this never
4801 happens - a fresh exec is done - but it can happen when
4802 deliver_drop_privilege is set.
4804 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4805 it now includes up to 150 characters of the unexpected data in the log
4808 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4809 and building data strings. The size of both of these buffers was 10 000
4810 bytes - far larger than anybody would *ever* want, thought I. Needless to
4811 say, somebody hit the limit. I have increased the maximum line length to
4812 20 000 and the maximum data length of concatenated lines to 100 000. I have
4813 also fixed two bugs, because there was no checking on these buffers. Tsk,
4814 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4815 buffer is too small.
4817 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4818 lsearch lookups. Now it does.
4820 46. When parsing a route_list item in a manualroute router, a fixed-length
4821 buffer was used for the list of hosts. I made this 1024 bytes long,
4822 thinking that nobody would ever have a list of hosts that long. Wrong.
4823 Somebody had a whole pile of complicated expansion conditions, and the
4824 string was silently truncated, leading to an expansion error. It turns out
4825 that it is easier to change to an unlimited length (owing to other changes
4826 that have happened since this code was originally written) than to build
4827 structure for giving a limitation error. The length of the item that
4828 expands into the list of hosts is now unlimited.
4830 47. The lsearch lookup could not handle data where the length of text line was
4831 more than 4095 characters. Such lines were truncated, leading to shortened
4832 data being returned. It should now handle lines of any length.
4834 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4835 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4838 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4839 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4840 no longer happens in comment lines. A list of replacements is now placed
4841 at the head of all of the source files, except those whose only change is
4842 to replace PERL_COMMAND in the very first #! line.
4844 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4845 messages on the queue, with a bottom-up merge sort, using code contributed
4846 by Michael Haardt. This should make operations like -bp somewhat faster on
4847 large queues. It won't affect queue runners, except when queue_run_in_order
4850 51. Installed eximstats 1.31 in the distribution.
4852 52. Added support for SRV lookups to the dnslookup router.
4854 53. If an ACL referred to $message_body or $message_body_end, the value was not
4855 reset for any messages that followed in the same SMTP session.
4857 54. The store-handling optimization for building very long strings was not
4858 differentiating between the different store pools. I don't think this
4859 actually made any difference in practice, but I've tidied it.
4861 55. While running the routers to verify a sender address, $sender_address
4862 was still set to the sender address. This is wrong, because when routing to
4863 send a bounce to the sender, it would be empty. Therefore, I have changed
4864 it so that, while verifying a sender address, $sender_address is set to <>.
4865 (There is no change to what happens when verifying a recipient address.)
4867 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4868 target A or AAAA records (if not already returned) without resetting the
4869 qualify_single or search_parents options of the DNS resolver. These are
4870 inappropriate in this case because the targets of MX and SRV records must
4871 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4872 target that, when qualified, matched something in the local domain. These
4873 two options are now turned off when doing these lookups.
4875 57. It seems that at least some releases of Reiserfs (which does not have the
4876 concept of a fixed number of inodes) returns zero and not -1 for the
4877 number of available inodes. This interacted badly with check_spool_inodes,
4878 which assumed that -1 was the "no such thing" setting. What I have done is
4879 to check that the total number of inodes is greater than zero before doing
4880 the test of how many are available.
4882 58. When a "warn" ACL statement has a log_message modifier, the message is
4883 remembered, and not repeated. This is to avoid a lot of repetition when a
4884 message has many recipients that cause the same warning to be written.
4885 Howewer, Exim was preserving the list of already written lines for an
4886 entire SMTP session, which doesn't seem right. The memory is now reset if a
4887 new message is started.
4889 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4890 debugging output unless log_rewrite was also set.
4892 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4893 of (int)(handle) when we know that handle contains (void *)(-1).
4895 61. The Exim daemon panic-logs an error return when it closes the incoming
4896 connection. However "connection reset by peer" seems to be common, and
4897 isn't really an error worthy of noting specially, so that particular error
4900 62. When Exim is trying to find all the local interfaces, it used to panic and
4901 die if the ioctl to get the interface flags failed. However, it seems that
4902 on at least one OS (Solaris 9) it is possible to have an interface that is
4903 included in the list of interfaces, but for which you get a failure error
4904 for this call. This happens when the interface is not "plumbed" into a
4905 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4906 failure of the "get flags" call assumes that the interface is down.
4908 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4909 makes life easier for people who are doing arithmetic on fields extracted
4910 from dates, where you often get leading zeros that should not be
4911 interpreted as octal.
4913 64. Added qualify_domain to the redirect router, to override the global
4916 65. If a pathologically long header line contained very many addresses (the
4917 report of this problem mentioned 10 000) and each of them was rewritten,
4918 Exim could use up a very large amount of memory. (It kept on making new
4919 copies of the header line as it rewrote, and never released the old ones.)
4920 At the expense of a bit more processing, the header rewriting function has
4921 been changed so that it no longer eats memory in this way.
4923 66. The generation of the Received: header has been moved from the time that a
4924 message starts to be received, to the time that it finishes. The timestamp
4925 in the Received: header should now be very close to that of the <= log
4926 line. There are two side-effects of this change:
4928 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4929 logged header lines no longer include the local Received: line, because
4930 it has not yet been created. The same applies to a copy of the message
4931 that is returned to a non-SMTP sender when a message is rejected.
4933 (b) When a filter file is tested using -bf, no additional Received: header
4934 is added to the test message. After some thought, I decided that this
4937 This change does not affect the value of $received_for. It is still set
4938 after address rewriting, but before local_scan() is called.
4940 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4942 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4943 gave an unhelpful panic error message, and a defer error. I have managed to
4944 change this behaviour so that it now rejects any supplied certificate,
4945 which seems right, as the list of acceptable certificates is empty.
4947 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4948 gave an unhelpful defer error. I have not managed to make this reject any
4949 supplied certificates, but the error message it gives is "no certificate
4950 supplied", which is not helpful.
4952 70. exigrep's output now also includes lines that are not associated with any
4953 message, but which match the given pattern. Implemented by a patch from
4954 Martin Sluka, which also tidied up the Perl a bit.
4956 71. Recipient callout verification, like sender verification, was using <> in
4957 the MAIL FROM command. This isn't really the right thing, since the actual
4958 sender may affect whether the remote host accepts the recipient or not. I
4959 have changed it to use the actual sender in the callout; this means that
4960 the cache record is now keyed on a recipient/sender pair, not just the
4961 recipient address. There doesn't seem to be a real danger of callout loops,
4962 since a callout by the remote host to check the sender would use <>.
4963 [SEE ABOVE: changed after hitting problems.]
4965 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4966 temporary errors. However, in the case of such a code being given after
4967 the end of a data transmission (i.e. after ".") Exim was failing to write
4968 a retry record for the message. (Yes, there was some broken host that was
4969 actually sending 8xx at this point.)
4971 73. An unknown lookup type in a host list could cause Exim to panic-die when
4972 the list was checked. (An example that provoked this was putting <; in the
4973 middle of a list instead of at the start.) If this happened during a DATA
4974 ACL check, a -D file could be left lying around. This kind of configuration
4975 error no longer causes Exim to die; instead it causes a defer errror. The
4976 incident is still logged to the main and panic logs.
4978 74. Buglet left over from Exim 3 conversion. The message "too many messages
4979 in one connection" was written to the rejectlog but not the mainlog, except
4980 when address rewriting (yes!) was being logged.
4982 75. Added write_rejectlog option.
4984 76. When a system filter was run not as root (that is, when system_filter_user
4985 was set), the values of the $n variables were not being returned to the
4986 main process; thus, they were not subsequently available in the $sn
4989 77. Added +return_path_on_delivery log selector.
4991 78. A connection timeout was being treated differently from recipients deferred
4992 when testing hosts_max_try with a message that was older than the host's
4993 retry timeout. (The host should not be counted, thus allowing all hosts to
4994 be tried at least once before bouncing.) This may have been the cause of an
4995 occasionally reported bug whereby a message would remain on the queue
4996 longer than the retry timeout, but would be bounced if a delivery was
4997 forced. I say "may" because I never totally pinned down the problem;
4998 setting up timeout/retry tests is difficult. See also the next item.
5000 79. The ultimate address timeout was not being applied to errors that involved
5001 a combination of host plus message (for example, a timeout on a MAIL
5002 command). When an address resolved to a number of possible hosts, and they
5003 were not all tried for each delivery (e.g. because of hosts_max_try), a
5004 message could remain on the queue longer than the retry timeout.
5006 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5009 81. Fixed an obscure SMTP outgoing bug which required at least the following
5010 conditions: (a) there was another message waiting for the same server;
5011 (b) the server returned 5xx to all RCPT commands in the first message so
5012 that the message was not completed; (c) the server dropped the connection
5013 or gave a negative response to the RSET that Exim sends to abort the
5014 transaction. The observed case was a dropped connection after DATA that had
5015 been sent in pipelining mode. That is, the server had advertised PIPELINING
5016 but was not implementing it correctly. The effect of the bug was incorrect
5017 behaviour, such as trying another host, and this could lead to a crash.
5023 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5024 and daemon.c were passed as pointers to ints; they should have been
5025 pointers to socklen_t variables (which are typically unsigned ints).
5027 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5030 3. Fixed a really odd bug that affected only the testing scheme; patching a
5031 certain fixed string in the binary changed the value of another string that
5032 happened to be identical to the end of the original first string.
5034 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5035 name", it returns that address as the IP address. On some operating
5036 systems (e.g. Solaris), it also passes back the IP address string as the
5037 "host name". However, on others (e.g. Linux), it passes back an empty
5038 string. Exim wasn't checking for this, and was changing the host name to an
5039 empty string, assuming it had been canonicized.
5041 5. Although rare, it is permitted to have more than one PTR record for a given
5042 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5043 all the names associated with an address, because they do in Solaris.
5044 However, it seems that they do not in Linux for data that comes from the
5045 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5046 I found this out when I moved to a new Linux workstation and tried to run
5047 the Exim test suite.
5049 To get round this problem I have changed the code so that it now does its
5050 own call to the DNS to look up PTR records when searching for a host name.
5051 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5052 addresses that are only in /etc/hosts are still found.
5054 This behaviour is, however, controlled by an option called host_lookup_
5055 order, which defaults to "bydns:byaddr". If people want to use the other
5056 order, or indeed, just use one or the other means of lookup, they can
5057 specify it in this variable.
5059 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5060 some operating systems, this comes back from gethostbyaddr() as an empty
5061 string, and this is what Exim used to test for. However, it seems that in
5062 other systems, "." is yielded. Exim now tests for this case too.
5064 7. The values of check_spool_space and check_log_space are now held internally
5065 as a number of kilobytes instead of an absolute number of bytes. If a
5066 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5067 kilobyte. This means that much larger values can be stored.
5069 8. Exim monitor: an attempt to get the action menu when not actually pointing
5070 at a message produces an empty menu entitled "No message selected". This
5071 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5072 no entries in it ("Shell widget menu has zero width and/or height"). So I
5073 have added a single, blank menu entry in this case.
5075 9. Added ${quote_local_part.
5077 10. MIME decoding is now applied to the contents of Subject: header lines when
5080 11. Now that a reference to $sender_host_address automatically causes a reverse
5081 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5082 host lookup before query-style lookups in lists that might use this
5083 variable. This has therefore been abolished, and the "net-" prefix is no
5084 longer necessary for query-style lookups.
5086 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5087 have been a typo for LFLAGS, so it has been changed.
5089 13. The install script calls Exim with "-C /dev/null" in order to find the
5090 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5091 to be output. Howeve, since Exim outputs its version number before the
5092 error, it didn't break the script. It just looked ugly. I fixed this by
5093 always allowing "-C /dev/null" if the caller is root.
5095 14. Ignore overlarge ACL variable number when reading spool file - insurance
5096 against a later release with more variables having written the file.
5098 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5099 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5100 this, as well as the form without the "IPv6" on the front.
5102 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5103 OS/Makefile-Darwin file.
5105 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5106 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5108 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5109 variable. However, some OS use ints for inodes. I've added cast to long int
5110 to get rid of the compiler warning.
5112 19. I had forgotten to lock out "/../" in configuration file names when
5113 ALT_CONFIG_PREFIX was set.
5115 20. Routers used for verification do not need to specify transports. However,
5116 if such a router generated a host list, and callout was configured, Exim
5117 crashed, because it could not find a port number from the (non-existent)
5118 transport. It now assumes port 25 in this circumstance.
5120 21. Added the -t option to exigrep.
5122 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5123 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5124 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5125 them is set, LOOKUP_LSEARCH is forced.
5127 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5129 24. Added sender and host information to the "rejected by local_scan()" log
5130 line; previously there was no indication of these.
5132 25. Added .include_if_exists.
5134 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5135 Linux. It turns out that not all file systems support this. Apparently some
5136 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5137 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5138 not supported on the file descriptor, is now ignored when Exim is trying to
5139 sync a directory. This applies only to Linux.
5141 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5143 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5144 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5147 29. Added support for maildirsize files from supplied patch (modified a bit).
5149 30. The use of :fail: followed by an empty string could lead Exim to respond to
5150 sender verification failures with (e.g.):
5152 550 Verification failed for <xxx>
5153 550 Sender verify failed
5155 where the first response line was missing the '-' that indicates it is not
5156 the final line of the response.
5158 31. The loop for finding the name of the user that called Exim had a hardwired
5159 limit of 10; it now uses the value of finduser_retries, which is used for
5160 all other user lookups.
5162 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5165 33. Exim was neglecting to zero errno before one call of strtol() when
5166 expanding a string and expecting an integer value. On some systems this
5167 resulted in spurious "integer overflow" errors. Also, it was casting the
5168 result into an int without checking.
5170 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5171 did not work. The code looks as if it has *never* worked, though it appears
5172 to have been documented since at least releast 1.62. I have made it work.
5174 35. The "timeout_DNS" error in retry rules, also documented since at least
5175 1.62, also never worked. As it isn't clear exactly what this means, and
5176 clearly it isn't a major issue, I have abolished the feature by treating it
5177 as "timeout", and writing a warning to the main and panic logs.
5179 36. The display of retry rules for -brt wasn't always showing the error code
5182 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5183 timeout_connect_A, timeout_connect_MX.
5185 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5186 to the empty sender.
5188 39. The daemon was not analysing the content of -oX till after it had closed
5189 stderr and disconnected from the controlling terminal. This meant that any
5190 syntax errors were only noted on the panic log, and the return code from
5191 the command was 0. By re-arranging the code a little, I've made the
5192 decoding happen first, so such errors now appear on stderr, and the return
5193 code is 1. However, the actual setting up of the sockets still happens in
5194 the disconnected process, so errors there are still only recorded on the
5197 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5198 connections (as happens on some IP stacks) was logged at start up time as
5199 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5200 it from "IPv6 and IPv4", which means that two separate sockets are being
5203 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5204 says whether AF_INET or AF_INET6 was passed as an argument.
5206 42. Exiwhat output was messed up when time zones were included in log
5209 43. Exiwhat now gives more information about the daemon's listening ports,
5210 and whether -tls-on-connect was used.
5212 44. The "port" option of the smtp transport is now expanded.
5214 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5215 silently ignored. Now an error message is written to the main and panic
5218 46. There's a new ACL modifier called "logwrite" which writes to a log file
5219 as soon as it is encountered.
5221 47. Added $local_user_uid and $local_user_gid at routing time.
5223 48. Exim crashed when trying to verify a sender address that was being
5226 49. Exim was recognizing only a space character after ".include". It now also
5227 recognizes a tab character.
5229 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5230 extracting the relevant information from the specification. The man page no
5231 longer contains scrambled data for the -d option, and I've added a section
5232 at the front about calling Exim under different names.
5234 51. Added "extra_headers" argument to the "mail" command in filter files.
5236 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5239 53. Installed eximstats 1.29.
5241 54. Added transport_filter_timeout as a generic transport option.
5243 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5244 Cc: header lines. This was required by RFC 822, but it not required by RFC
5247 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5248 incoming messages that did not have them. Now it does so only if the
5249 message originates locally, that is, if there is no associated remote host
5250 address. When Resent- header lines are present, this applies to the Resent-
5251 lines rather than the non-Resent- lines.
5253 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5254 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5256 58. Messages for configuration errors now include the name of the main
5257 configuration file - useful now that there may be more than one file in a
5258 list (.included file names were always shown).
5260 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5261 for those rare installations that do not start the daemon as root or run it
5262 setuid root. I've cut out the call to initgroups() if the daemon is not
5265 60. The Exim user and group can now be bound into the binary as text strings
5266 that are looked up at the start of Exim's processing.
5268 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5270 62. Added $mailstore_basename variable.
5272 63. Installed patch to sieve.c from Michael Haardt.
5274 64. When Exim failed to open the panic log after failing to open the main log,
5275 the original message it was trying to log was written to stderr and debug
5276 output, but if they were not available (the usual case in production), it
5277 was lost. Now it is written to syslog before the two lines that record the
5278 failures to open the logs.
5280 65. Users' Exim filters run in subprocesses under the user's uid. It is
5281 possible for a "deliver" command or an alias in a "personal" command to
5282 provoke an address rewrite. If logging of address rewriting is configured,
5283 this fails because the process is not running as root or exim. There may be
5284 a better way of dealing with this, but for the moment (because 4.30 needs
5285 to be released), I have disabled address rewrite logging when running a
5286 filter in a non-root, non-exim process.
5292 1. The buildconfig auxiliary program wasn't quoting the value set for
5293 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5294 not defined. This bug was masked in 4.22 by the effect that was fixed in
5297 2. Some messages that were rejected after a message id was allocated were
5298 shown as "incomplete" by exigrep. It no longer does this for messages that
5299 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5301 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5302 have allow_domain_literals set, the ID did not get logged in the <= line.
5303 Domain literals are now always recognized in Message-ID: header lines.
5305 4. The first argument for a ${extract expansion item is the key name or field
5306 number. Leading and trailing spaces in this item were not being ignored,
5307 causing some misleading effects.
5309 5. When deliver_drop_privilege was set, single queue runner processes started
5310 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5311 same command in the process it spins off) were not dropping privilege.
5313 6. When the daemon running as "exim" started a queue runner, it always
5314 re-executed Exim in the spun-off process. This is a waste of effort when
5315 deliver_drop_privilege is set. The new process now just calls the
5316 queue-runner function directly.
5322 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5325 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5326 ignored. Though the use of -oP was forcing the writing of a pid file, it
5327 was always written to the default place.
5329 3. If the message "no IP address found for host xxxx" is generated during
5330 incoming verification, it is now followed by identification of the incoming
5331 connection (so you can more easily find what provoked it).
5333 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5335 5. Added some features to "harden" Exim a bit more against certain attacks:
5337 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5338 be put in Local/Makefile. This is like the never_users runtime option,
5339 but it cannot be overridden. The default setting is "root".
5341 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5342 prefix string with which any file named in a -C command line option
5345 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5346 is retained for -C and -D only if the caller of Exim is root. Without
5347 it, the exim user may also use -C and -D and retain privilege.
5349 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5350 command line option is disabled.
5352 6. Macro names set by the -D option must start with an upper case letter, just
5353 like macro names defined in the configuration file.
5355 7. Added "dereference=" facility to LDAP.
5357 8. Two instances of the typo "uknown" in the source files are fixed.
5359 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5360 the Configure-Makefile script screwed up while processing it.
5362 10. Incorporated PCRE 4.4.
5364 11. The SMTP synchronization check was not operating right at the start of an
5365 SMTP session. For example, it could not catch a HELO sent before the client
5366 waited for the greeting. There is now a check for outstanding input at the
5367 point when the greeting is written. Because of the duplex, asynchronous
5368 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5369 way, but not yet received, when the check is performed.
5371 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5372 on TCP/IP sockets, because this apparently causes some broken clients to
5375 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5376 unchanged) from the Cygwin maintainer.
5378 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5379 was supported instead of testing for mbx. Effectively a typo.
5381 15. The spa authenticator server code was not checking that the input it
5382 received was valid base64.
5384 16. The debug output line for the "set" modifier in ACLs was not showing the
5385 name of the variable that was being set.
5387 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5389 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5390 lookup on its own. Something else was needed to trigger the lookup. For
5391 example, a match in host_lookup or the need for a host name in a host list.
5392 Now, if $sender_host_name is referenced and the host name has not yet been
5393 looked up, a lookup is performed. If the lookup fails, the variable remains
5394 empty, and $host_lookup_failed is set to "1".
5396 19. Added "eqi" as a case-independent comparison operator.
5398 20. The saslauthd authentication condition could segfault if neither service
5399 nor realm was specified.
5401 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5402 error message that was logged was misleading, and incoming SMTP
5403 connections were dropped. The message is now more accurate, and temporary
5404 errors are given to SMTP connections.
5406 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5407 (except RSET) with a 421 error, until QUIT is received. However, it was
5408 failing to send a response to QUIT.
5410 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5411 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5412 ACL is now run after the helo_try_verify_hosts code.
5414 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5415 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5416 case-independent, so other case variants are also recognized). Apparently
5417 some systems use these upper case variants.
5419 25. If more than two messages were waiting for the same host, and a transport
5420 filter was specified for the transport, Exim sent two messages over the
5421 same TCP/IP connection, and then failed with "socket operation on non-
5422 socket" when it tried to send the third.
5424 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5426 27. The extern definition of crypt16() in expand.c was not being excluded when
5427 the OS had its own crypt16() function.
5429 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5430 as a preferred synonym for return_size_limit, both as an option and as an
5433 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5435 30. Changed the default configuration ACL to relax the local part checking rule
5436 for addresses that are not in any local domains. For these addresses,
5437 slashes and pipe symbols are allowed within local parts, but the sequence
5438 /../ is explicitly forbidden.
5440 31. SPA server authentication was not clearing the challenge buffer before
5443 32. log_message in a "warn" ACL statement was writing to the reject log as
5444 well as to the main log, which contradicts the documentation and doesn't
5445 seem right (because no rejection is happening). So I have stopped it.
5447 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5448 However, I am unable to do any testing of this.
5450 34. Fixed an infelicity in the appendfile transport. When checking directories
5451 for a mailbox, to see if any needed to be created, it was accidentally
5452 using path names with one or more superfluous leading slashes; tracing
5453 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5455 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5456 contents are added to the log line that is written for every discarded
5457 recipient. (Previously a log_message setting was ignored.)
5459 36. The ${quote: operator now quotes the string if it is empty.
5461 37. The install script runs exim in order to find its version number. If for
5462 some reason other than non-existence or emptiness, which it checks, it
5463 could not run './exim', it was installing it with an empty version number,
5464 i.e. as "exim-". This error state is now caught, and the installation is
5467 38. An argument was missing from the function that creates an error message
5468 when Exim fails to connect to the socket for saslauthd authentication.
5469 This could cause Exim to crash, or give a corrupted message.
5471 39. Added isip, isip4, and isip6 to ${if conditions.
5473 40. The ACL variables $acl_xx are now saved with the message, and can be
5474 accessed later in routers, transports, and filters.
5476 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5477 strings in the file are not string-expanded.
5479 42. If a MAIL command specified a SIZE value that was too large to fit into an
5480 int variable, the check against message_size_limit failed. Such values are
5481 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5482 day this will have to be increased, but I don't think I want to be around
5483 when emails are that large.
5490 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5491 iconv() is not standard in FreeBSD.
5493 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5494 IPv6 enabled. The observed symptom was a segmentation fault on return from
5495 the function os_common_find_running_interfaces() in src/os.c.
5497 3. In the check_special_case() function in daemon.c I had used "errno" as an
5498 argument name, which causes warnings on some systems. This was basically a
5499 typo, since it was named "eno" in the comments!
5501 4. The code that waits for the clock to tick (at a resolution of some fraction
5502 of a second) so as to ensure message-id uniqueness was always waiting for
5503 at least one whole tick, when it could have waited for less. [This is
5504 almost certainly not relevant at current processor speeds, where it is
5505 unlikely to ever wait at all. But we try to future-proof.]
5507 5. The function that sleeps for a time interval that includes fractions of a
5508 second contained a race. It did not block SIGALRM between setting the
5509 timer, and suspending (a couple of lines later). If the interval was short
5510 and the sigsuspend() was delayed until after it had expired, the suspension
5511 never ended. On busy systems this could lead to processes getting stuck for
5514 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5515 process, before it forks any delivery processes. The open lookup caching
5516 mechanism meant that the open file or database connection was passed into
5517 the delivery process. The problem was that delivery processes always tidy
5518 up cached lookup data. This could cause a problem for the next delivery
5519 process started by the queue runner, because the external queue runner
5520 process does not know about the closure. So the next delivery process
5521 still has data in the lookup cache. In the case of a file lookup, there was
5522 no problem because closing a file descriptor in a subprocess doesn't affect
5523 the parent. However, if the lookup was caching a connection to a database,
5524 the connection was closed, and the second delivery process was likely to
5525 see errors such as "PGSQL: query failed: server closed the connection
5526 unexpectedly". The problem has been fixed by closing all cached lookups
5527 in a queue runner before running a delivery process.
5529 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5530 seem to have the "const" qualifier which it has on other OS. I've
5533 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5534 *of the same type* (client or server) with the same public name that an
5535 error should be diagnosed.
5537 9. When Exim looked up a host name for an IP address, but failed to find the
5538 original IP address when looking up the host name (a safety check), it
5539 output the message "<ip address> does not match any IP for NULL", which was
5540 confusing, to say the least. The bug was that the host name should have
5541 appeared instead of "NULL".
5543 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5544 user that is built into the binary, and the -C or -D options is used, root
5545 privilege is dropped before the configuration file is read. In addition,
5546 logging is switched to stderr instead of the normal log files. If the
5547 configuration then re-defines the Exim user, the unprivileged environment
5548 is probably not what is expected, so Exim logs a panic warning message (but
5551 However, if deliver_drop_privilege is set, the unprivileged state may well
5552 be exactly what is intended, so the warning has been cut out in that case,
5553 and Exim is allowed to try to write to its normal log files.
5559 1. smtp_return_error_details was not giving details for temporary sender
5560 or receiver verification errors.
5562 2. Diagnose a configuration error if two authenticators have the same public
5565 3. Exim used not to create the message log file for a message until the first
5566 delivery attempt. This could be confusing when incoming messages were held
5567 for policy or load reasons. The message log file is now created at the time
5568 the message is received, and an initial "Received" line is written to it.
5570 4. The automatically generated man page for command line options had a minor
5571 bug that caused no ill effects; however, a more serious problem was that
5572 the procedure for building the man page automatically didn't always
5573 operate. Consequently, release 4.20 contains an out-of-date version. This
5574 shouldn't happen again.
5576 5. When building Exim with embedded Perl support, the script that builds the
5577 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5578 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5580 6. The freeze_tell option was not being used for messages that were frozen on
5581 arrival, either by an ACL or by local_scan().
5583 7. Added the smtp_incomplete_transaction log selector.
5585 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5586 was accepting AUTH without a new EHLO.
5588 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5589 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5592 10. Logging of TCP/IP connections (when configured) now happens in the main
5593 daemon process instead of the child process, so that the TCP/IP connection
5594 count is more accurate (but it can never be perfect).
5596 11. The use of "drop" in a nested ACL was not being handled correctly in the
5597 outer ACL. Now, if condition failure induced by the nested "drop" causes
5598 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5599 or "require"), the connection is dropped.
5601 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5602 that yield "discard" can now be used with an "accept" or a "discard" verb,
5603 but an error is generated for any others (because I can't see a useful way
5604 to define what should happen).
5606 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5607 are now processed in the same way as lines in the Exim configuration file.
5608 In particular, continuation lines are supported.
5610 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5612 15. Added -ti meaning -t -i.
5614 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5615 domains, and warn by logging if others are found.
5617 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5618 returned from GIFCONF when Exim is trying to find the list of interfaces on
5619 a host. The code in os.c has been modified to copy each ifreq to an aligned
5620 structure in all cases.
5622 Also, in some cases, the returned ifreq's were being copied to a 'struct
5623 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5624 means the last couple of bytes of an IPv6 address could be chopped if the
5625 ifreq contained only a normal sockaddr (14 bytes storage).
5627 18. Named domain lists were not supported in the hosts_treat_as_local option.
5628 An entry such as +xxxx was not recognized, and was treated as a literal
5631 19. Ensure that header lines added by a DATA ACL are included in the reject log
5632 if the ACL subsequently rejects the message.
5634 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5635 MD5 (which is deprecated).
5637 21. When testing a filter file using -bf, Exim was writing a message when it
5638 took the sender from a "From " line in the message, but it was not doing so
5639 when it took $return_path from a Return-Path: header line. It now does.
5641 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5642 with a valid header line field name (a series of printing characters
5643 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5645 23. Changed "disc" in the source to "disk" to conform to the documentation and
5646 the book and for uniformity.
5648 24. Ignore Sendmail's -Ooption=value command line item.
5650 25. When execve() failed while trying to run a command in a pipe transport,
5651 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5652 could be confused with a return value of 69 from the command itself. This
5653 has been changed to 127, the value the shell returns if it is asked to run
5654 a non-existent command. The wording for the related log line suggests a
5655 non-existent command as the problem.
5657 26. If received_header_text expands to an empty string, do not add a Received:
5658 header line to the message. (Well, it adds a token one on the spool, but
5659 marks it "old" so that it doesn't get used or transmitted.)
5661 27. Installed eximstats 1.28 (addition of -nt option).
5663 28. There was no check for failure on the call to getsockname() in the daemon
5664 code. This can fail if there is a shortage of resources on the system, with
5665 ENOMEM, for example. A temporary error is now given on failure.
5667 29. Contrary to the C standard, it seems that in some environments, the
5668 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5669 program. Exim now does this explicitly; it affects the formatting of
5670 timestamps using strftime().
5672 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5673 complaints. I've now initialized all the variables, to avoid this.
5675 32. Header lines added by a system filter were not being "seen" during
5676 transport-time rewrites.
5678 33. The info_callback() function passed to OpenSSL is set up with type void
5679 (*)(SSL *, int, int), as described somewhere. However, when calling the
5680 function (actually a macro) that sets it up, the type void(*)() is
5681 expected. I've put in a cast to prevent warnings from picky compilers.
5683 34. If a DNS black list lookup found a CNAME record, but there were no A
5684 records associated with the domain it pointed at, Exim crashed.
5686 35. If a DNS black list lookup returned more than one A record, Exim ignored
5687 all but the first. It now scans all returned addresses if a particular IP
5688 value is being sought. In this situation, the contents of the
5689 $dnslist_value variable are a list of all the addresses, separated by a
5692 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5693 to accept a host name and all its aliases if the forward lookup for any of
5694 them yielded the IP address of the incoming connection. Now it accepts only
5695 those names whose forward lookup yields the correct IP address. Any other
5696 names are discarded. This closes a loophole whereby a rogue DNS
5697 administrator could create reverse DNS records to break through a
5698 wildcarded host restriction in an ACL.
5700 37. If a user filter or a system filter that ran in a subprocess used any of
5701 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5702 the wrong values were passed to the pipe command ($thisaddress had the
5703 value of $0, $0 had the value of $1, etc). This bug was introduced by
5704 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5706 38. Improved the line breaking for long SMTP error messages from ACLs.
5707 Previously, if there was no break point between 40 and 75 characters, Exim
5708 left the rest of the message alone. Two changes have been made: (a) I've
5709 reduced the minimum length to 35 characters; (b) if it can't find a break
5710 point between 35 and 75 characters, it looks ahead and uses the first one
5711 that it finds. This may give the occasional overlong line, but at least the
5712 remaining text gets split now.
5714 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5715 file descriptors might be low, and that setting 1000 would always raise it.
5716 It turns out that in some environments, the limit is already over 1000 and
5717 that lowering it causes trouble. So now Exim takes care not to decrease it.
5719 40. When delivering a message, the value of $return_path is set to $sender_
5720 address at the start of routing (routers may change the value). By an
5721 oversight, this default was not being set up when an address was tested by
5722 -bt or -bv, which affected the outcome if any router or filter referred to
5725 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5726 log only when "message" or "log_message" are set. However, if one of the
5727 conditions was an address verification, or a call to a nested ACL, the
5728 messages generated by the underlying test were being passed through. This
5729 no longer happens. The underlying message is available in $acl_verify_
5730 message for both "message" and "log_message" expansions, so it can be
5731 passed through if needed.
5733 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5734 new expansion $bh_ to give the encoded byte string without charset
5735 translation. Translation happens only if iconv() is available; HAVE_ICONV
5736 indicates this at build time. HEADERS_CHARSET gives the charset to
5737 translate to; headers_charset can change it in the configuration, and
5738 "headers charset" can change it in an individual filter file.
5740 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5741 that creates RFC 2047 encoded "words" labels them as that charset instead
5742 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5743 expansion operator; (ii) when Exim creates a From: line for a local
5744 message; (iii) when a header line is rewritten to include a "phrase" part.
5746 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5747 buggy, causing it to skip the first lines of messages whose message ID
5748 ended in 'D'. This would not have bitten before Exim release 4.14, because
5749 message IDs were unlikely to end in 'D' before then. The effect was to have
5750 incorrect size information for certain domains.
5752 45. #include "config.h" was missing at the start of the crypt16.c module. This
5753 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5756 46. If there was a timeout during a "random" callout check, Exim treated it as
5757 a failure of the random address, and carried on sending RSET and the real
5758 address. If the delay was just some slowness somewhere, the response to the
5759 original RCPT would be taken as a response to RSET and so on, causing
5760 mayhem of various kinds.
5762 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5763 when I implemented it. It didn't allow for the fact that some option values
5764 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5765 the right test for positive values.
5767 48. Domain names in DNS records are case-independent. Exim always looks them up
5768 in lower case. Some resolvers return domain names in exactly the case they
5769 appear in the zone file, that is, they may contain uppercase letters. Not
5770 all resolvers do this - some return always lower case. Exim was treating a
5771 change of case by a resolver as a change of domain, similar to a widening
5772 of a domain abbreviation. This triggered its re-routing code and so it was
5773 trying to route what was effectively the same domain again. This normally
5774 caused routing to fail (because the router wouldn't handle the domain
5775 twice). Now Exim checks for this case specially, and just changes the
5776 casing of the domain that it ultimately uses when it transmits the message
5779 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5782 50. If a filter generated a file delivery with a non-absolute name (possible if
5783 no home directory exists for the router), the forbid_file option was not
5786 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5787 the existing equality matching.
5789 52. Exim was using ints instead of ino_t variables in some places where it was
5790 dealing with inode numbers.
5792 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5793 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5794 called TMPDIR, and if it finds it is different, it changes its value.
5796 54. The smtp_printf() function is now made available to local_scan() so
5797 additional output lines can be written before returning. There is also an
5798 smtp_fflush() function to enable the detection of a dropped connection.
5799 The variables smtp_input and smtp_batched_input are exported to
5802 55. Changed the default runtime configuration: the message "Unknown user"
5803 has been removed from the ACL, and instead placed on the localuser router,
5804 using the cannot_route_message feature. This means that any verification
5805 failures that generate their own messages won't get overridden. Similarly,
5806 the "Unrouteable address" message that was in the ACL for unverifiable
5807 relay addresses has also been removed.
5809 56. Added hosts_avoid_esmtp to the smtp transport.
5811 57. The exicyclog script was not checking for the esoteric option
5812 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5813 will work only if exicyclog is run under the appropriate euid.
5815 58. Following a discussion on the list, the rules by which Exim recognises line
5816 endings on incoming messages have been changed. The -dropcr and drop_cr
5817 options are now no-ops, retained only for backwards compatibility. The
5818 following line terminators are recognized: LF CRLF CR. However, special
5819 processing applies to CR:
5821 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5822 nor a local message in the state where . is a terminator.
5824 (ii) If a bare CR is encountered in a header line, an extra space is added
5825 after the line terminator so as not to end the header. The reasoning
5826 behind this is that bare CRs in header lines are most likely either
5827 to be mistakes, or people trying to play silly games.
5829 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5830 was being incorrectly given as 18 bytes larger than it should have been.
5831 This is a VOB (very old bug).
5833 60. This may never have affected anything current, but just in case it has:
5834 When the local host is found other than at the start of a list of hosts,
5835 the local host, those with the same MX, and any that follow, are discarded.
5836 When the list in question was part of a longer list of hosts, the following
5837 hosts (not currently being processed) were also being discarded. This no
5838 longer happens. I'm not sure if this situation could ever has previously
5841 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5844 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5845 header. This is something that is recommended in a new Internet Draft, and
5846 is something that is documented as being done by Sendmail. There are two
5847 possible values. For messages generated by the autoreply transport, Exim
5850 Auto-Submitted: auto-replied
5852 whereas for all other generated messages (e.g. bounces) it adds
5854 Auto-Submitted: auto-generated
5856 63. The "personal" condition in filters now includes a test for the
5857 Auto-Submitted: header. If it contains the string "auto-" the message it
5858 not considered personal.
5860 64. Added rcpt_include_affixes as a generic transport option.
5862 65. Added queue_only_override (default true).
5864 66. Added the syslog_duplication option.
5866 67. If what should have been the first header line of a message consisted of
5867 a space followed by a colon, Exim was mis-interpreting it as a header line.
5868 It isn't of course - it is syntactically invalid and should therefore be
5869 treated as the start of the message body. The misbehaviour could have
5870 caused a number of strange effects, including loss of data in subsequent
5871 header lines, and spool format errors.
5873 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5874 client host had authenticated. This control can now be exercised by an ACL
5875 for more flexibility.
5877 69. By default, callouts do not happen when testing with -bh. There is now a
5878 variant, -bhc, which does actually run the callout code, including
5879 consulting and updating the callout cache.
5881 70. Added support for saslauthd authentication, courtesy of Alexander
5884 71. If statvfs() failed on the spool or log directories while checking their
5885 size for availability, Exim confusingly gave the error "space shortage".
5886 Furthermore, in debugging mode it crashed with a floating point exception.
5887 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5888 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5889 serious problem, Exim now writes to the main and panic logs when this
5890 happens, with details of the failure. It then refuses to accept the
5891 incoming message, giving the message "spool directory problem" or "log
5892 directory problem" with a 421 code for SMTP messages.
5894 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5895 0, 1, and 2 exist, because some OS complain for execs without them (see
5896 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5897 descriptors. However, the code omitted to check that the open succeeded,
5898 causing mysterious errors if for some reason the permissions on /dev/null
5899 got screwed. Now Exim writes a message to the main and panic logs, and
5900 bombs out if it can't open /dev/null.
5902 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5903 interact so that it is all more flexible. It is supposed to remain
5904 backwards compatible. Also added extra_local_interfaces.
5906 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5907 to bomb out with an assertion failure - to the client this appears as a
5908 connection drop. This problem occurs in the part of the code that was taken
5909 from the Samba project. Fortunately, the assertion is in a very simple
5910 function, so I have fixed this by reproducing the function inline in the
5911 one place where it is called, and arranging for authentication to fail
5912 instead of killing the process with assert().
5914 75. The SPA client code was not working when the server requested OEM rather
5915 than Unicode encoding.
5917 76. Added code to make require_files with a specific uid setting more usable in
5918 the case where statting the file as root fails - usually a non-root-mounted
5919 NFS file system. When this happens and the failure is EACCES, Exim now
5920 forks a subprocess and does the per-uid checking as the relevant uid.
5922 77. Added process_log_path.
5924 78. If log_file_path was not explicitly set, a setting of check_log_space or
5925 check_log_inodes was ignored.
5927 79. If a space check for the spool or log partitions fails, the incident is now
5928 logged. Of course, in the latter case the data may get lost...
5930 80. Added the %p formatting code to string_format() so that it can be used to
5931 print addresses in debug_print(). Adjusted all the address printing in the
5932 debugging in store.c to use %p rather than %d.
5934 81. There was a concern that a line of code in smtp_in.c could overflow a
5935 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5936 initially expressed, the concern was not well-founded, because trailing
5937 spaces are removed early. However, if the trailing spaces were followed by
5938 a NULL, they did not get removed, so the overflow was possible. Two fixes
5941 (a) I re-wrote the offending code in a cleaner fashion.
5942 (b) If an incoming SMTP command contains a NULL character, it is rejected
5945 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5946 runs initgroups(), so that if the Exim user is in any additional groups,
5947 they will be used during message reception.
5953 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob