1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
49 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
51 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
52 selectors, in both main and reject logs.
54 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
57 JH/07 Add malware type "sock" for talking to simple daemon.
59 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
61 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
62 routers/transports under cutthrough routing.
64 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
65 numbers. Touch up "bool" conditional to keep the same definition.
67 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
73 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
75 PP/02 Make -n do something, by making it not do something.
76 When combined with -bP, the name of an option is not output.
78 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
81 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
82 $sender_host_name and config options to manage this, and basic check
85 PP/05 DSCP support for outbound connections and control modifier for inbound.
87 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
88 (Only plugin which currently uses this is kerberos4, which nobody should
89 be using, but we should make it available and other future plugins might
90 conceivably use it, even though it would break NAT; stuff *should* be
91 using channel bindings instead).
93 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
94 name; added for Sendmail compatibility; requires admin caller.
95 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
96 just ignore it); requires trusted caller.
97 Also parse but ignore: -Ac -Am -X<logfile>
100 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
102 TL/02 Add +smtp_confirmation as a default logging option.
104 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
105 Patch by Magnus Holmgren from 2007-02-20.
107 TL/04 Bugzilla 1281 - Spec typo.
108 Bugzilla 1283 - Spec typo.
109 Bugzilla 1290 - Spec grammar fixes.
111 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
113 TL/06 Add Experimental DMARC support using libopendmarc libraries.
115 TL/07 Fix an out of order global option causing a segfault. Reported to dev
116 mailing list by by Dmitry Isaikin.
118 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
120 JH/02 Support "G" suffix to numbers in ${if comparisons.
122 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
124 NM/01 Bugzilla 1197 - Spec typo
125 Bugzilla 1196 - Spec examples corrections
127 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
129 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
130 gnutls_enable_pkcs11, but renamed to more accurately indicate its
133 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
134 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
136 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
137 "acl {{name}{arg}...}", and optional args on acl condition
140 JH/05 Permit multiple router/transport headers_add/remove lines.
142 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
144 JH/07 Avoid using a waiting database for a single-message-only transport.
145 Performance patch from Paul Fisher. Bugzilla 1262.
147 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
150 JH/09 Add $headers_added variable, with content from use of ACL modifier
151 add_header (but not yet added to the message). Bugzilla 199.
153 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
154 Pulled from Bugzilla 817 by Wolfgang Breyha.
156 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
158 (nb: this is the same fix as in Exim 4.80.1)
160 JH/11 Add A= logging on delivery lines, and a client_set_id option on
163 JH/12 Add optional authenticated_sender logging to A= and a log_selector
166 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
168 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
169 advertise SMTP AUTH mechanism to us, instead of a generic
170 protocol violation error. Also, make Exim more robust to bad
171 data from the Dovecot auth socket.
173 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
175 When a queue runner is handling a message, Exim first routes the
176 recipient addresses, during which it prunes them based on the retry
177 hints database. After that it attempts to deliver the message to
178 any remaining recipients. It then updates the hints database using
181 So if a recipient address works intermittently, it can get repeatedly
182 deferred at routing time. The retry hints record remains fresh so the
183 address never reaches the final cutoff time.
185 This is a fairly common occurrence when a user is bumping up against
186 their storage quota. Exim had some logic in its local delivery code
187 to deal with this. However it did not apply to per-recipient defers
188 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
190 This change adds a proper retry rule check during routing so that the
191 final cutoff time is checked against the message's age. We only do
192 this check if there is an address retry record and there is not a
193 domain retry record; this implies that previous attempts to handle
194 the address had the retry_use_local_parts option turned on. We use
195 this as an approximation for the destination being like a local
196 delivery, as in LMTP.
198 I suspect this new check makes the old local delivery cutoff check
199 redundant, but I have not verified this so I left the code in place.
201 TF/02 Correct gecos expansion when From: is a prefix of the username.
203 Test 0254 submits a message to Exim with the header
207 When I ran the test suite under the user fanf2, Exim expanded
208 the header to contain my full name, whereas it should have added
209 a Resent-Sender: header. It erroneously treats any prefix of the
210 username as equal to the username.
212 This change corrects that bug.
214 GF/01 DCC debug and logging tidyup
215 Error conditions log to paniclog rather than rejectlog.
216 Debug lines prefixed by "DCC: " to remove any ambiguity.
218 TF/03 Avoid unnecessary rebuilds of lookup-related code.
220 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
221 Bug spotted by Jeremy Harris; was flawed since initial commit.
222 Would have resulted in OCSP responses post-SNI triggering an Exim
223 NULL dereference and crash.
225 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
227 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
228 Bug detection, analysis and fix by Samuel Thibault.
229 Bugzilla 1331, Debian bug #698092.
231 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
233 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
234 Server implementation by Todd Lyons, client by JH.
235 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
236 config variable "prdr_enable" controls whether the server
237 advertises the facility. If the client requests PRDR a new
238 acl_data_smtp_prdr ACL is called once for each recipient, after
239 the body content is received and before the acl_smtp_data ACL.
240 The client is controlled by bolth of: a hosts_try_prdr option
241 on the smtp transport, and the server advertisement.
242 Default client logging of deliveries and rejections involving
243 PRDR are flagged with the string "PRDR".
245 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
246 fclose(). Diagnosis by Todd Lyons.
248 PP/17 Update configure.default to handle IPv6 localhost better.
249 Patch by Alain Williams (plus minor tweaks).
252 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
253 This is now consistent with GnuTLS, and is now documented: the
254 previous undocumented portable approach to treating the option as
255 unset was to force an expansion failure. That still works, and
256 an empty string is now equivalent.
258 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
259 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
260 not performing validation itself.
262 PP/20 Added force_command boolean option to pipe transport.
263 Patch from Nick Koston, of cPanel Inc.
265 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
268 TF/04 Added udpsend ACL modifer and hexquote expansion operator
270 PP/21 Fix eximon continuous updating with timestamped log-files.
271 Broken in a format-string cleanup in 4.80, missed when I repaired the
272 other false fix of the same issue.
273 Report and fix from Heiko Schlichting.
276 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
277 Report from Prashanth Katuri.
279 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
280 It's SecureTransport, so affects any MacOS clients which use the
281 system-integrated TLS libraries, including email clients.
283 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
284 using a MIME ACL for non-SMTP local injection.
285 Report and assistance in diagnosis by Warren Baker.
287 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
289 JH/16 Fix comparisons for 64b. Bugzilla 1385.
291 TL/09 Add expansion variable $authenticated_fail_id to keep track of
292 last id that failed so it may be referenced in subsequent ACL's.
294 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
297 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
298 ldap library initialization, allowing self-signed CA's to be
299 used. Also properly sets require_cert option later in code by
300 using NULL (global ldap config) instead of ldap handle (per
301 session). Bug diagnosis and testing by alxgomz.
303 TL/12 Enhanced documentation in the ratelimit.pl script provided in
304 the src/util/ subdirectory.
306 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
307 renamed to Transport Post Delivery Action by Jeremy Harris, as
310 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
311 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
312 redis_servers = needs to be configured which will be used by the redis
313 lookup. Patch from Warren Baker, of The Packet Hub.
315 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
317 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
318 hostname or reverse DNS when processing a host list. Used suggestions
319 from multiple comments on this bug.
321 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
323 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
324 Missed a few lines, added it to make the runtest require no keyboard
327 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
328 contains upper case chars. Make router use caseful_local_part.
330 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
331 support when GnuTLS has been built with p11-kit.
337 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
339 This, or similar/improved, will also be change PP/11 of 4.82.
345 PP/01 Handle short writes when writing local log-files.
346 In practice, only affects FreeBSD (8 onwards).
347 Bugzilla 1053, with thanks to Dmitry Isaikin.
349 NM/01 Bugzilla 949 - Documentation tweak
351 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
354 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
356 PP/02 Implemented gsasl authenticator.
358 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
360 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
361 `pkg-config foo` for cflags/libs.
363 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
364 with rest of GSASL and with heimdal_gssapi.
366 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
367 `pkg-config foo` for cflags/libs for the TLS implementation.
369 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
370 properties get this fed in as external SSF. A number of robustness
371 and debugging improvements to the cyrus_sasl authenticator.
373 PP/08 cyrus_sasl server now expands the server_realm option.
375 PP/09 Bugzilla 1214 - Log authentication information in reject log.
376 Patch by Jeremy Harris.
378 PP/10 Added dbmjz lookup type.
380 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
382 PP/12 MAIL args handles TAB as well as SP, for better interop with
383 non-compliant senders.
384 Analysis and variant patch by Todd Lyons.
386 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
387 Bug report from Lars Müller <lars@samba.org> (via SUSE),
388 Patch from Dirk Mueller <dmueller@suse.com>
390 PP/13 tls_peerdn now print-escaped for spool files.
391 Observed some $tls_peerdn in wild which contained \n, which resulted
392 in spool file corruption.
394 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
395 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
396 or write after TLS renegotiation, which otherwise led to messages
399 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
400 as a tracking header (ie: a signed header comes before the signature).
401 Patch from Wolfgang Breyha.
403 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
404 comma-sep list; embedded commas doubled.
406 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
408 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
410 Report and patch from Dmitry Banschikov.
412 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
413 Removed SSL_clear() after SSL_new() which led to protocol negotiation
414 failures. We appear to now support TLS1.1+ with Exim.
416 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
417 lets Exim select keys and certificates based upon TLS SNI from client.
418 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
419 before an outbound SMTP session. New log_selector, +tls_sni.
421 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
422 NULL dereference. Report and patch from Alun Jones.
424 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
425 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
426 Not seeing resolver debug output on NetBSD, but suspect this is a
427 resolver implementation change.
429 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
430 Left warnings. Added "eximon gdb" invocation mode.
432 PP/21 Defaulting "accept_8bitmime" to true, not false.
434 PP/22 Added -bw for inetd wait mode support.
436 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
437 locate the relevant includes and libraries. Made this the default.
439 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
440 Bugzilla 1246, report and most of solution from Tomasz Kusy.
442 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
443 This may cause build issues on older platforms.
445 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
446 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
447 gnutls_require_mac & gnutls_require_protocols (no longer supported).
448 Added SNI support via GnuTLS too.
449 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
451 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
453 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
454 Applied second patch from Janne, implementing suggestion to default
455 multiple-strings-in-record handling to match SPF spec.
457 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
459 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
460 read-only, out of scope).
461 Patch from Wolfgang Breyha, report from Stuart Northfield.
463 PP/29 Fix three issues highlighted by clang analyser static analysis.
464 Only crash-plausible issue would require the Cambridge-specific
465 iplookup router and a misconfiguration.
466 Report from Marcin Mirosław.
468 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
470 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
471 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
472 As part of this, removing so much warning spew let me fix some minor
473 real issues in debug logging.
475 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
476 assignment on my part. Fixed.
478 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
479 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
480 Janne Snabb (who went above and beyond: thank you).
482 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
483 string otherwise requires a connection and a bunch more work and it's
484 relatively easy to get wrong. Should also expose TLS library linkage
487 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
488 64-bit ${eval} (JH/03).
490 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
491 GNU libc to support some of the 64-bit stuff, should not lead to
492 conflicts. Defined before os.h is pulled in, so if a given platform
493 needs to override this, it can.
495 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
496 protection layer was required, which is not implemented.
497 Bugzilla 1254, patch from Wolfgang Breyha.
499 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
500 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
501 tls_dhparam take prime identifiers. Also unbreak combination of
502 OpenSSL+DH_params+TLSSNI.
504 PP/39 Disable SSLv2 by default in OpenSSL support.
510 PP/01 Solaris build fix for Oracle's LDAP libraries.
511 Bugzilla 1109, patch from Stephen Usher.
513 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
515 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
518 TF/02 Fix a couple more cases where we did not log the error message
519 when unlink() failed. See also change 4.74-TF/03.
521 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
522 lock up or crash if it happened to be inside a call to libc when it
523 got a SIGUSR1 from exiwhat.
525 The SIGUSR1 handler appends the current process status to the process
526 log which is later printed by exiwhat. It used to use the general
527 purpose logging code to do this, but several functions it calls are
528 not safe for signals.
530 The new output code in the SIGUSR1 handler is specific to the process
531 log, and simple enough that it's easy to inspect for signal safety.
532 Removing some special cases also simplifies the general logging code.
533 Removing the spurious timestamps from the process log simplifies
536 TF/04 Improved ratelimit ACL condition.
538 The /noupdate option has been deprecated in favour of /readonly which
539 has clearer semantics. The /leaky, /strict, and /readonly update modes
540 are mutually exclusive. The update mode is no longer included in the
541 database key; it just determines when the database is updated. (This
542 means that when you upgrde Exim will forget old rate measurements.)
544 Exim now checks that the per_* options are used with an update mode that
545 makes sense for the current ACL. For example, when Exim is processing a
546 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
547 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
548 must specify per_mail/readonly. If you omit the update mode it defaults to
549 /leaky where that makes sense (as before) or /readonly where required.
551 The /noupdate option is now undocumented but still supported for
552 backwards compatibility. It is equivalent to /readonly except that in
553 ACLs where /readonly is required you may specify /leaky/noupdate or
554 /strict/noupdate which are treated the same as /readonly.
556 A useful new feature is the /count= option. This is a generalization
557 of the per_byte option, so that you can measure the throughput of other
558 aggregate values. For example, the per_byte option is now equivalent
559 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
561 The per_rcpt option has been generalized using the /count= mechanism
562 (though it's more complicated than the per_byte equivalence). When it is
563 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
564 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
565 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
566 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
567 Note that using per_rcpt with a non-readonly update mode in more than
568 one ACL will cause the recipients to be double-counted. (The per_mail
569 and per_byte options don't have this problem.)
571 The handling of very low rates has changed slightly. If the computed rate
572 is less than the event's count (usually one) then this event is the first
573 after a long gap. In this case the rate is set to the same as this event's
574 count, so that the first message of a spam run is counted properly.
576 The major new feature is a mechanism for counting the rate of unique
577 events. The new per_addr option counts the number of different
578 recipients that someone has sent messages to in the last time period. It
579 behaves like per_rcpt if all the recipient addresses are different, but
580 duplicate recipient addresses do not increase the measured rate. Like
581 the /count= option this is a general mechanism, so the per_addr option
582 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
583 example, measure the rate that a client uses different sender addresses
584 with the options per_mail/unique=$sender_address. There are further
585 details in the main documentation.
587 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
589 TF/06 Removed a few PCRE remnants.
591 TF/07 Automatically extract Exim's version number from tags in the git
592 repository when doing development or release builds.
594 PP/02 Raise smtp_cmd_buffer_size to 16kB.
595 Bugzilla 879. Patch from Paul Fisher.
597 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
598 Heavily based on revision 40f9a89a from Simon Arlott's tree.
601 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
603 PP/05 Variable $av_failed, true if the AV scanner deferred.
604 Bugzilla 1078. Patch from John Horne.
606 PP/06 Stop make process more reliably on build failure.
607 Bugzilla 1087. Patch from Heiko Schlittermann.
609 PP/07 Make maildir_use_size_file an _expandable_ boolean.
610 Bugzilla 1089. Patch from Heiko Schlittermann.
612 PP/08 Handle ${run} returning more data than OS pipe buffer size.
613 Bugzilla 1131. Patch from Holger Weiß.
615 PP/09 Handle IPv6 addresses with SPF.
616 Bugzilla 860. Patch from Wolfgang Breyha.
618 PP/10 GnuTLS: support TLS 1.2 & 1.1.
620 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
623 PP/11 match_* no longer expand right-hand-side by default.
624 New compile-time build option, EXPAND_LISTMATCH_RHS.
625 New expansion conditions, "inlist", "inlisti".
627 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
629 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
631 PP/14 fix log_write() format string regression from TF/03.
632 Bugzilla 1152. Patch from Dmitry Isaikin.
638 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
640 PP/02 Harmonised TLS library version reporting; only show if debugging.
641 Layout now matches that introduced for other libraries in 4.74 PP/03.
643 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
645 PP/04 New "dns_use_edns0" global option.
647 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
650 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
651 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
653 TK/01 Updated PolarSSL code to 0.14.2.
654 Bugzilla 1097. Patch from Andreas Metzler.
656 PP/07 Catch divide-by-zero in ${eval:...}.
659 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
662 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
663 format-string attack -- SECURITY: remote arbitrary code execution.
665 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
666 time unintentionally subject to list matching rules, letting the header
667 cause arbitrary Exim lookups (of items which can occur in lists, *not*
668 arbitrary string expansion). This allowed for information disclosure.
670 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
671 INT_MIN/-1 -- value coerced to INT_MAX.
677 NM/01 Workround for PCRE version dependency in version reporting
680 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
681 This fixes portability to compilers other than gcc, notably
682 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
684 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
685 makefiles for portability to HP-UX and POSIX correctness.
687 PP/01 Permit LOOKUP_foo enabling on the make command-line.
688 Also via indented variable definition in the Makefile.
689 (Debugging by Oliver Heesakkers).
691 PP/02 Restore caching of spamd results with expanded spamd_address.
692 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
694 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
695 Improves build reliability. Fix from: Frank Elsner
697 NM/02 Fix wide character breakage in the rfc2047 coding
698 Fixes bug 1064. Patch from Andrey N. Oktyabrski
700 NM/03 Allow underscore in dnslist lookups
701 Fixes bug 1026. Patch from Graeme Fowler
703 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
704 Code patches from Adam Ciarcinski of NetBSD.
706 NM/04 Fixed exiqgrep to cope with mailq missing size issue
709 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
710 is logged, to avoid truncation. Patch from John Horne.
712 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
713 Patch from Jakob Hirsch.
715 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
716 SQL string expansion failure details.
717 Patch from Andrey Oktyabrski.
719 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
720 Patch from Simon Arlott.
722 PP/09 New lookups functionality failed to compile on old gcc which rejects
723 extern declarations in function scope.
724 Patch from Oliver Fleischmann
726 PP/10 Use sig_atomic_t for flags set from signal handlers.
727 Check getgroups() return and improve debugging.
728 Fixed developed for diagnosis in bug 927 (which turned out to be
731 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
732 Patch from Mark Zealey.
734 PP/12 Bugzilla 1056: Improved spamd server selection.
735 Patch from Mark Zealey.
737 PP/13 Bugzilla 1086: Deal with maildir quota file races.
738 Based on patch from Heiko Schlittermann.
740 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
741 Patch from Uwe Doering, sign-off by Michael Haardt.
743 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
744 variable declaration deep within a block. Bug and patch from
747 PP/15 lookups-Makefile IRIX compatibilty coercion.
749 PP/16 Make DISABLE_DKIM build knob functional.
751 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
752 Patch by Simon Arlott
754 TF/03 Fix valgrind.h portability to C89 compilers that do not support
755 variable argument macros. Our copy now differs from upstream.
761 TF/01 Failure to get a lock on a hints database can have serious
762 consequences so log it to the panic log.
764 TF/02 Log LMTP confirmation messages in the same way as SMTP,
765 controlled using the smtp_confirmation log selector.
767 TF/03 Include the error message when we fail to unlink a spool file.
769 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
770 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
771 for maintaining out-of-tree patches for some time.
773 PP/01 Bugzilla 139: Documentation and portability issues.
774 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
775 Handle per-OS dynamic-module compilation flags.
777 PP/02 Let /dev/null have normal permissions.
778 The 4.73 fixes were a little too stringent and complained about the
779 permissions on /dev/null. Exempt it from some checks.
780 Reported by Andreas M. Kirchwitz.
782 PP/03 Report version information for many libraries, including
783 Exim version information for dynamically loaded libraries. Created
784 version.h, now support a version extension string for distributors
785 who patch heavily. Dynamic module ABI change.
787 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
788 privilege escalation vulnerability whereby the Exim run-time user
789 can cause root to append content of the attacker's choosing to
792 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
795 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
796 If dropping privileges for untrusted macros, we disabled normal logging
797 on the basis that it would fail; for the Exim run-time user, this is not
798 the case, and it resulted in successful deliveries going unlogged.
799 Fixed. Reported by Andreas Metzler.
805 PP/01 Date: & Message-Id: revert to normally being appended to a message,
806 only prepend for the Resent-* case. Fixes regression introduced in
807 Exim 4.70 by NM/22 for Bugzilla 607.
809 PP/02 Include check_rfc2047_length in configure.default because we're seeing
810 increasing numbers of administrators be bitten by this.
812 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
814 PP/03 Bugzilla 994: added openssl_options main configuration option.
816 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
818 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
820 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
822 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
823 without a peer certificate, leading to a segfault because of an
824 assumption that peers always have certificates. Be a little more
825 paranoid. Problem reported by Martin Tscholak.
827 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
828 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
829 NB: ClamAV planning to remove STREAM in "middle of 2010".
830 CL also introduces -bmalware, various -d+acl logging additions and
831 more caution in buffer sizes.
833 PP/09 Implemented reverse_ip expansion operator.
835 PP/10 Bugzilla 937: provide a "debug" ACL control.
837 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
839 PP/12 Bugzilla 973: Implement --version.
841 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
843 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
845 PP/15 Bugzilla 816: support multiple condition rules on Routers.
847 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
848 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
849 ignore trailing whitespace.
851 JJ/02 prevent non-panic DKIM error from being sent to paniclog
853 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
856 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
857 Notification from Dr Andrew Aitchison.
859 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
860 ExtendedDetectionInfo response format.
861 Notification from John Horne.
863 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
866 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
867 XSL and documented dependency on system catalogs, with examples of how
870 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
873 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
874 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
875 configuration file which is writeable by the Exim user or group.
877 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
878 of configuration files to cover files specified with the -C option if
879 they are going to be used with root privileges, not just the default
882 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
883 option (effectively making it always true).
885 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
886 files to be used while preserving root privileges.
888 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
889 that rogue child processes cannot use them.
891 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
892 run-time user, instead of root.
894 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
895 Exim run-time user without dropping privileges.
897 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
898 result string, instead of calling string_vformat() twice with the same
901 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
902 for other users. Others should always drop root privileges if they use
903 -C on the command line, even for a whitelisted configure file.
905 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
907 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
913 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
914 $data_path, and $header_path variables; fixed documentation bugs and
917 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
918 exipick to access non-standard spools, including the "frozen" queue
921 NM/01 Bugzilla 965: Support mysql stored procedures.
922 Patch from Alain Williams
924 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
926 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
927 Patch from Andreas Metzler
929 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
930 Patch from Kirill Miazine
932 NM/05 Bugzilla 671: Added umask to procmail example.
934 JJ/03 installed exipick 20100323.0, fixing doc bug
936 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
937 directory. Notification and patch from Dan Rosenberg.
939 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
941 TK/02 Improve log output when DKIM signing operation fails.
943 MH/01 Treat the transport option dkim_domain as a colon separated
944 list, not as a single string, and sign the message with each element,
945 omitting multiple occurences of the same signer.
947 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
948 Bugzilla 985, 986. Patch by Simon Arlott
950 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
951 Patch by Simon Arlott
953 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
954 MBX locking. Notification from Dan Rosenberg.
960 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
962 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
964 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
966 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
968 NM/04 Bugzilla 915: Flag broken perl installation during build.
974 TK/01 Added patch by Johannes Berg that expands the main option
975 "spamd_address" if it starts with a dollar sign.
977 TK/02 Write list of recipients to X-Envelope-Sender header when building
978 the mbox-format spool file for content scanning (suggested by Jakob
981 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
982 (http://www.dcc-servers.net/) support via dccifd. Activated by
983 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
985 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
986 by Mark Daniel Reidel <mr@df.eu>.
988 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
989 When building exim an external PCRE library is now needed -
990 PCRE is a system library on the majority of modern systems.
991 See entry on PCRE_LIBS in EDITME file.
993 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
994 conversation. Added nologin parameter to request.
995 Patch contributed by Kirill Miazine.
997 TF/01 Do not log submission mode rewrites if they do not change the address.
999 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1001 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1002 log files in place. Contributed by Roberto Lima.
1004 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1006 TF/03 Bugzilla 615: When checking the local_parts router precondition
1007 after a local_part_suffix or local_part_prefix option, Exim now
1008 does not use the address's named list lookup cache, since this
1009 contains cached lookups for the whole local part.
1011 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1012 Robert Millan. Documentation is in experimental-spec.txt.
1014 TF/04 Bugzilla 668: Fix parallel build (make -j).
1016 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1018 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1019 Patch provided by Jan Srzednicki.
1021 TF/05 Leading white space used to be stripped from $spam_report which
1022 wrecked the formatting. Now it is preserved.
1024 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1025 that they are available at delivery time.
1027 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1029 TF/08 TLS error reporting now respects the incoming_interface and
1030 incoming_port log selectors.
1032 TF/09 Produce a more useful error message if an SMTP transport's hosts
1033 setting expands to an empty string.
1035 NM/06 Bugzilla 744: EXPN did not work under TLS.
1036 Patch provided by Phil Pennock.
1038 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1039 Patch provided by Richard Godbee.
1041 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1042 acl_smtp_notquit, added index entry.
1044 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1045 Patch provided by Eugene Bujak.
1047 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1048 accept(). Patch provided by Maxim Dounin.
1050 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1051 Patch provided by Phil Pennock.
1053 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1055 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1056 Patch provided by Brad "anomie" Jorsch.
1058 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1059 Patch provided by Dean Brooks.
1061 TK/05 Add native DKIM support (does not depend on external libraries).
1063 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1064 Patch provided by Graeme Fowler.
1066 NM/16 Bugzilla 851: Documentation example syntax fix.
1068 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1070 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1073 NM/19 Bugzilla 745: TLS version reporting.
1074 Patch provided by Phil Pennock.
1076 NM/20 Bugzilla 167: bool: condition support.
1077 Patch provided by Phil Pennock.
1079 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1080 clients. Patch provided by Phil Pennock.
1082 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1083 Patch provided by Brad "anomie" Jorsch.
1085 NM/23 Bugzilla 687: Fix misparses in eximstats.
1086 Patch provided by Heiko Schlittermann.
1088 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1089 Patch provided by Heiko Schlittermann.
1091 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1092 plus update to original patch.
1094 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1096 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1097 Patch provided by David Brownlee.
1099 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1101 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1103 NM/30 Bugzilla 888: TLS documentation bugfixes.
1105 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1107 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1108 Unlike the original bugzilla I have changed all shell scripts in src tree.
1110 NM/33 Bugzilla 898: Transport filter timeout fix.
1111 Patch by Todd Rinaldo.
1113 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1114 Patch by Serge Demonchaux.
1116 NM/35 Bugzilla 39: Base64 decode bug fixes.
1117 Patch by Jakob Hirsch.
1119 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1121 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1123 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1125 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1131 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1132 ALT-N's libdkim that I have put here:
1133 http://duncanthrax.net/exim-experimental/
1135 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1136 were called 'true' and it seems that C99 defines that as a reserved
1137 keyword to be used with 'bool' variable types. That means you could
1138 not include C99-style headers which use bools without triggering
1139 build errors in sieve.c.
1141 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1142 as mailq or other aliases. Changed the --help handling significantly
1143 to do whats expected. exim_usage() emits usage/help information.
1145 SC/01 Added the -bylocaldomain option to eximstats.
1147 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1149 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1151 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1157 PH/01 Another patch from the Sieve maintainer.
1159 PH/02 When an IPv6 address is converted to a string for single-key lookup
1160 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1161 dots are used instead of colons so that keys in lsearch files need not
1162 contain colons. This was done some time before quoting was made available
1163 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1164 (notated using the quote facility) so as to distinguish them from IPv4
1165 keys. This meant that lookups for IP addresses in host lists did not work
1166 for iplsearch lookups.
1168 This has been fixed by arranging for IPv6 addresses to be expressed with
1169 colons if the lookup type is iplsearch. This is not incompatible, because
1170 previously such lookups could never work.
1172 The situation is now rather anomolous, since one *can* have colons in
1173 ordinary lsearch keys. However, making the change in all cases is
1174 incompatible and would probably break a number of configurations.
1176 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1179 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1180 conversion specification without a maximum field width, thereby enabling
1181 a rogue spamd server to cause a buffer overflow. While nobody in their
1182 right mind would setup Exim to query an untrusted spamd server, an
1183 attacker that gains access to a server running spamd could potentially
1184 exploit this vulnerability to run arbitrary code as the Exim user.
1186 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1187 $primary_hostname instead of what libspf2 thinks the hosts name is.
1189 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1190 a directory entry by the name of the lookup key. Previously, if a
1191 symlink pointed to a non-existing file or a file in a directory that
1192 Exim lacked permissions to read, a lookup for a key matching that
1193 symlink would fail. Now it is enough that a matching directory entry
1194 exists, symlink or not. (Bugzilla 503.)
1196 PH/03 The body_linecount and body_zerocount variables are now exported in the
1199 PH/04 Added the $dnslist_matched variable.
1201 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1202 This means they are set thereafter only if the connection becomes
1205 PH/06 Added the client_condition to authenticators so that some can be skipped
1206 by clients under certain conditions.
1208 PH/07 The error message for a badly-placed control=no_multiline_responses left
1209 "_responses" off the end of the name.
1211 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1213 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1214 (without spaces) instead of just copying the configuration text.
1216 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1218 PH/11 Added $max_received_linelength.
1220 PH/12 Added +ignore_defer and +include_defer to host lists.
1222 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1223 way in which PCRE > 7.0 is built.
1225 PH/14 Implemented queue_only_load_latch.
1227 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1228 MAIL command. The effect was to mangle the value on 64-bit systems.
1230 PH/16 Another patch from the Sieve maintainer.
1232 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1234 PH/18 If a system quota error occurred while trying to create the file for
1235 a maildir delivery, the message "Mailbox is full" was not appended to the
1236 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1237 only to a quota excession during the actual writing of the file.
1239 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1240 characters?) which causes problems in log lines. The DN values are now
1241 passed through string_printing() before being added to log lines.
1243 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1244 and InterBase are left for another time.)
1246 PH/21 Added message_body_newlines option.
1248 PH/22 Guard against possible overflow in moan_check_errorcopy().
1250 PH/23 POSIX allows open() to be a macro; guard against that.
1252 PH/24 If the recipient of an error message contained an @ in the local part
1253 (suitably quoted, of course), incorrect values were put in $domain and
1254 $local_part during the evaluation of errors_copy.
1260 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1261 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1264 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1265 issue a MAIL command.
1267 PH/02 In an ACL statement such as
1269 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1271 if a client was not listed at all, or was listed with a value other than
1272 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1273 the condition was not true (as it should be), so access was not denied.
1274 The bug was that the ! inversion was incorrectly passed on to the second
1275 item. This has been fixed.
1277 PH/03 Added additional dnslists conditions == and =& which are different from
1278 = and & when the dns lookup returns more than one IP address.
1280 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1281 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1283 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1284 FSYNC, which compiles an option called disable_fsync that allows for
1285 bypassing fsync(). The documentation is heavily laced with warnings.
1287 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1289 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1290 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1291 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1292 including adding "make clean"; (3) Added -fPIC when compiling the test
1293 dynamically loaded module, to get rid of a warning.
1295 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1296 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1297 = 0s. The bug is otherwise harmless.
1299 PH/07 There was a bug in the dovecot authenticator such that the value of
1300 $auth1 could be overwritten, and so not correctly preserved, after a
1301 successful authentication. This usually meant that the value preserved by
1302 the server_setid option was incorrect.
1304 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1306 PH/09 Installed PCRE release 7.0.
1308 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1309 run for batched SMTP input. It is now run at the start of every message
1310 in the batch. While fixing this I discovered that the process information
1311 (output by running exiwhat) was not always getting set for -bs and -bS
1312 input. This is fixed, and it now also says "batched" for BSMTP.
1314 PH/11 Added control=no_pipelining.
1316 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1317 patch, slightly modified), and move the expansion of helo_data till after
1318 the connection is made in the smtp transport (so it can use these
1321 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1323 PH/14 Added log_selector = +pid.
1325 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1327 PH/16 Add ${if forany and ${if forall.
1329 PH/17 Added dsn_from option to vary the From: line in DSNs.
1331 PH/18 Flush SMTP output before performing a callout, unless control =
1332 no_callout_flush is set.
1334 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1335 was true (the default) a successful delivery failed to delete the retry
1336 item, thus causing premature timeout of the address. The bug is now
1339 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1341 PH/21 Long custom messages for fakedefer and fakereject are now split up
1342 into multiline reponses in the same way that messages for "deny" and
1343 other ACL rejections are.
1345 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1346 with slight modification.
1348 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1349 draft, changing the syntax and factoring some duplicate code".
1351 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1352 for deliveries of the second and subsequent messages over the same SMTP
1355 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1356 ${reduce, with only minor "tidies".
1358 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1360 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1361 expansion side effects.
1363 PH/27 When a message times out after an over-quota error from an Exim-imposed
1364 quota, the bounce message says "mailbox is full". This message was not
1365 being given when it was a system quota that was exceeded. It now should
1368 MH/03 Made $recipients available in local_scan(). local_scan() already has
1369 better access to the recipient list through recipients_list[], but
1370 $recipients can be useful in postmaster-provided expansion strings.
1372 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1373 in the case of a MAIL command with additional options following the
1374 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1375 were accidentally chopped off.
1377 PH/29 SMTP synchronization checks are implemented when a command is read -
1378 there is a check that no more input is waiting when there shouldn't be
1379 any. However, for some commands, a delay in an ACL can mean that it is
1380 some time before the response is written. In this time, more input might
1381 arrive, invalidly. So now there are extra checks after an ACL has run for
1382 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1383 pipelining has not been advertised.
1385 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1387 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1388 returned in the SMTP response when smtp_return_error_details was false.
1389 This has been fixed.
1391 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1392 instead of the C I/O that was originally supplied, because problems were
1393 reported on Solaris.
1395 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1396 Exim which did not show up earlier: it was assuming that a call to
1397 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1398 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1399 was a macro that became an assignment, so it seemed to work. This has
1400 changed to a proper function call with a void return, hence the compile
1401 error. Exim's code has been fixed.
1403 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1406 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1408 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1410 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1411 because it turns out to be too big or there is a timeout) there is no
1412 "Completed" line in the log. When some messages of this type were
1413 selected by exigrep, they were listed as "not completed". Others were
1414 picked up by some special patterns. I have improved the selection
1415 criteria to be more general.
1417 PH/38 The host_find_failed option in the manualroute router can now be set
1418 to "ignore", to completely ignore a host whose IP address cannot be
1419 found. If all hosts are ignored, the behaviour is controlled by the new
1420 host_all_ignored option.
1422 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1423 homing or because of multiple MX records with /mx) generated more than
1424 one IP address, and the following item turned out to be the local host,
1425 all the secondary addresses of the first item were incorrectly removed
1426 from the list, along with the local host and any following hosts (which
1427 is what is supposed to happen).
1429 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1430 whoever called Exim into the -H file. In the case of the daemon it was
1431 behaving confusingly. When first started, it used values for whoever
1432 started the daemon, but after a SIGHUP it used the Exim user (because it
1433 calls itself on a restart). I have changed the code so that it now always
1436 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1437 message are rejected with the same error (e.g. no authentication or bad
1438 sender address), and a DATA command is nevertheless sent (as can happen
1439 with PIPELINING or a stupid MUA), the error message that was given to the
1440 RCPT commands is included in the rejection of the DATA command. This is
1441 intended to be helpful for MUAs that show only the final error to their
1444 PH/42 Another patch from the Sieve maintainer.
1446 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1447 Eximstats - Fixed some broken HTML links and added missing column headers
1449 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1450 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1452 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1454 PH/43 Yet another patch from the Sieve maintainer.
1456 PH/44 I found a way to check for a TCP/IP connection going away before sending
1457 the response to the final '.' that terminates a message, but only in the
1458 case where the client has not sent further data following the '.'
1459 (unfortunately, this is allowed). However, in many cases there won't be
1460 any further data because there won't be any more messages to send. A call
1461 to select() can be used: if it shows that the input is "ready", there is
1462 either input waiting, or the socket has been closed. An attempt to read
1463 the next input character can distinguish the two cases. Previously, Exim
1464 would have sent an OK response which the client would never have see.
1465 This could lead to message repetition. This fix should cure that, at
1466 least in a lot of common cases.
1468 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1469 advertised in response to EHLO.
1475 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1476 fixed by 4.65/MH/01 (is this a record?) are fixed:
1478 (i) An empty string was always treated as zero by the numeric comparison
1479 operators. This behaviour has been restored.
1481 (ii) It is documented that the numeric comparison operators always treat
1482 their arguments as decimal numbers. This was broken in that numbers
1483 starting with 0 were being interpreted as octal.
1485 While fixing these problems I realized that there was another issue that
1486 hadn't been noticed. Values of message_size_limit (both the global option
1487 and the transport option) were treated as octal if they started with 0.
1488 The documentation was vague. These values are now always treated as
1489 decimal, and I will make that clear in the documentation.
1495 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1496 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1499 MH/01 Don't check that the operands of numeric comparison operators are
1500 integers when their expansion is in "skipping" mode (fixes bug
1501 introduced by 4.64-PH/07).
1503 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1504 child addresses, Exim now panics and dies. Previously, because the count
1505 is held in a short int, deliveries were likely to be lost. As such a
1506 large number of recipients for a single message is ridiculous
1507 (performance will be very, very poor), I have chosen to impose a limit
1508 rather than extend the field.
1514 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1515 leftover -K file (the existence of which was triggered by #402).
1516 While we were at it, introduced process PID as part of the -K
1517 filename. This should rule out race conditions when creating
1520 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1521 processing considerably. Previous code took too long for large mails,
1522 triggering a timeout which in turn triggers #401.
1524 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1525 in the DK code in transports.c. sendfile() is not really portable,
1526 hence the _LINUX specificness.
1528 TF/01 In the add_headers option to the mail command in an Exim filter,
1529 there was a bug that Exim would claim a syntax error in any
1530 header after the first one which had an odd number of characters
1533 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1534 callout verification, Exim cached a "reject" for the entire domain. This
1535 is correct for most verifications, but it is not correct for a recipient
1536 verification with use_sender or use_postmaster set, because in that case
1537 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1538 case of MAIL FROM:<> rejection from other early rejections (e.g.
1539 rejection of HELO). When verifying a recipient using a non-null MAIL
1540 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1541 Whatever the result of the callout, the value of the domain cache is
1542 left unchanged (for any other kind of callout, getting as far as trying
1543 RCPT means that the domain itself is ok).
1545 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1548 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1549 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1550 session, the daemon ignores ECONNECT errors and logs others; it now
1551 ignores EPIPE as well.
1553 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1554 (quoted-printable decoding).
1556 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1557 later the small subsequent patch to fix an introduced bug.
1559 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1561 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1563 PH/08 An error is now given if message_size_limit is specified negative.
1565 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1566 to be given (somewhat) arbitrary names.
1568 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1571 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1572 miscellaneous code fixes
1574 PH/10 Added the log_reject_target ACL modifier to specify where to log
1577 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1578 hostname. This is wrong, because it relates to the incoming message (and
1579 probably the interface on which it is arriving) and not to the outgoing
1580 callout (which could be using a different interface). This has been
1581 changed to use the value of the helo_data option from the smtp transport
1582 instead - this is what is used when a message is actually being sent. If
1583 there is no remote transport (possible with a router that sets up host
1584 addresses), $smtp_active_hostname is used.
1586 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1587 tweaks were necessary in order to get it to work (see also 21 below):
1588 (a) The code assumed that strncpy() returns a negative number on buffer
1589 overflow, which isn't the case. Replaced with Exim's string_format()
1591 (b) There were several signed/unsigned issues. I just did the minimum
1592 hacking in of casts. There is scope for a larger refactoring.
1593 (c) The code used strcasecmp() which is not a standard C function.
1594 Replaced with Exim's strcmpic() function.
1595 (d) The code set only $1; it now sets $auth1 as well.
1596 (e) A simple test gave the error "authentication client didn't specify
1597 service in request". It would seem that Dovecot has changed its
1598 interface. Fortunately there's a specification; I followed it and
1599 changed what the client sends and it appears to be working now.
1601 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1604 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1605 address A is aliased to B and C, where B exists and C does not. Without
1606 -v the output is "A verified" because verification stops after a
1607 successful redirection if more than one address is generated. However,
1608 with -v the child addresses are also verified. Exim was outputting "A
1609 failed to verify" and then showing the successful verification for C,
1610 with its parentage. It now outputs "B failed to verify", showing B's
1611 parentage before showing the successful verification of C.
1613 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1614 look up a TXT record in a specific list after matching in a combined
1617 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1618 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1619 they consult the DNS. I had assumed they would set it the way they
1620 wanted; and indeed my experiments on Linux seem to show that in some
1621 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1622 To be on the safe side, however, I have now made the interface to
1623 host_find_byname() similar to host_find_bydns(), with an argument
1624 containing the DNS resolver options. The host_find_byname() function now
1625 sets these options at its start, just as host_find_bydns() does. The smtp
1626 transport options dns_qualify_single and dns_search_parents are passed to
1627 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1628 of host_find_byname() use the default settings of RES_DEFNAMES
1629 (qualify_single) but not RES_DNSRCH (search_parents).
1631 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1632 spool_read_header() do less string testing, by means of a preliminary
1633 switch on the second character of optional "-foo" lines. (This is
1634 overdue, caused by the large number of possibilities that now exist.
1635 Originally there were few.) While I was there, I also converted the
1636 str(n)cmp tests so they don't re-test the leading "-" and the first
1637 character, in the hope this might squeeze out yet more improvement.
1639 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1640 flag allowing group syntax was set by the header_syntax check but not
1641 turned off, possible causing trouble later; (2) The flag was not being
1642 set at all for the header_verify test, causing "group"-style headers to
1643 be rejected. I have now set it in this case, and also caused header_
1644 verify to ignore an empty address taken from a group. While doing this, I
1645 came across some other cases where the code for allowing group syntax
1646 while scanning a header line wasn't quite right (mostly, not resetting
1647 the flag correctly in the right place). These bugs could have caused
1648 trouble for malformed header lines. I hope it is now all correct.
1650 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1651 with the "reply" argument non-NULL. The code, however (which originally
1652 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1653 but it didn't always do it. This confused somebody who was copying the
1654 code for some other use. I have removed all the tests.
1656 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1657 feature that was used to support insecure browsers during the U.S. crypto
1658 embargo. It requires special client support, and Exim is probably the
1659 only MTA that supported it -- and would never use it because real RSA is
1660 always available. This code has been removed, because it had the bad
1661 effect of slowing Exim down by computing (never used) parameters for the
1662 RSA_EXPORT functionality.
1664 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1665 authenticator to fail if there's a tab character in the incoming data
1666 (there should never be unless someone is messing about, as it's supposed
1667 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1668 if the connection is using TLS or if the remote IP is the same as the
1669 local IP, and the "valid-client-cert option" if a client certificate has
1672 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1673 authenticators. This can be used for authorization after authentication
1674 succeeds. (In the case of plaintext, it servers for both authentication
1677 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1678 if any retry times were supplied.
1680 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1681 connection, where there is no client IP address to check. In this
1682 situation, the verify now always succeeds.
1684 PH/25 Applied John Jetmore's -Mset patch.
1686 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1688 PH/27 In a string expansion for a processed (not raw) header when multiple
1689 headers of the same name were present, leading whitespace was being
1690 removed from all of them, but trailing whitespace was being removed only
1691 from the last one. Now trailing whitespace is removed from each header
1692 before concatenation. Completely empty headers in a concatenation (as
1693 before) are ignored.
1695 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1696 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1698 PH/29 [Removed. This was a change that I later backed out, and forgot to
1699 correct the ChangeLog entry (that I had efficiently created) before
1700 committing the later change.]
1702 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1703 address errors (4xx response to RCPT) over the same connection as other
1704 messages routed to the same hosts. Such deliveries are always "forced",
1705 so retry times are not inspected. This resulted in far too many retries
1706 for the affected addresses. The effect occurred only when there were more
1707 hosts than the hosts_max_try setting in the smtp transport when it had
1708 the 4xx errors. Those hosts that it had tried were not added to the list
1709 of hosts for which the message was waiting, so if all were tried, there
1710 was no problem. Two fixes have been applied:
1712 (i) If there are any address or message errors in an SMTP delivery, none
1713 of the hosts (tried or untried) are now added to the list of hosts
1714 for which the message is waiting, so the message should not be a
1715 candidate for sending over the same connection that was used for a
1716 successful delivery of some other message. This seems entirely
1717 reasonable: after all the message is NOT "waiting for some host".
1718 This is so "obvious" that I'm not sure why it wasn't done
1719 previously. Hope I haven't missed anything, but it can't do any
1720 harm, as the worst effect is to miss an optimization.
1722 (ii) If, despite (i), such a delivery is accidentally attempted, the
1723 routing retry time is respected, so at least it doesn't keep
1724 hammering the server.
1726 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1727 in ${readsocket because some servers need this prod.
1729 PH/32 Added some extra debug output when updating a wait-xxx database.
1731 PH/33 The hint "could be header name not terminated by colon", which has been
1732 given for certain expansion errors for a long time, was not being given
1733 for the ${if def:h_colon_omitted{... case.
1735 PH/34 The spec says: "With one important exception, whenever a domain list is
1736 being scanned, $domain contains the subject domain." There was at least
1737 one case where this was not true.
1739 PH/35 The error "getsockname() failed: connection reset by peer" was being
1740 written to the panic log as well as the main log, but it isn't really
1741 panic-worthy as it just means the connection died rather early on. I have
1742 removed the panic log writing for the ECONNRESET error when getsockname()
1745 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1746 runs only) independently of the message's sender address. This meant
1747 that, if the 4xx error was in fact related to the sender, a different
1748 message to the same recipient with a different sender could confuse
1749 things. In particualar, this can happen when sending to a greylisting
1750 server, but other circumstances could also provoke similar problems.
1751 I have changed the default so that the retry time for these errors is now
1752 based a combination of the sender and recipient addresses. This change
1753 can be overridden by setting address_retry_include_sender=false in the
1756 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1757 remote server are returned as part of bounce messages. This was not
1758 happening for LMTP over a pipe (the lmtp transport), but now it is the
1759 same for both kinds of LMTP.
1761 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1762 in header lines that were in fact CNAMEs. This is no longer the case.
1764 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1765 and queue runs started by the daemon processed all messages. This has
1766 been fixed so that -R and -S can now usefully be given with -q<time>.
1768 PH/40 Import PCRE release 6.7 (fixes some bugs).
1770 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1772 PH/42 Give an error if -q is specified more than once.
1774 PH/43 Renamed the variables $interface_address and $interface_port as
1775 $received_ip_address and $received_port, to make it clear that these
1776 values apply to message reception, and not to the outgoing interface when
1777 a message is delivered. (The old names remain recognized, of course.)
1779 PH/44 There was no timeout on the connect() call when using a Unix domain
1780 socket in the ${readsocket expansion. There now is.
1782 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1783 be meaningful with "accept".
1785 SC/01 Eximstats V1.43
1786 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1788 SC/02 Eximstats V1.44
1789 Use a glob alias rather than an array ref in the generated
1790 parser. This improves both readability and performance.
1792 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1793 Collect SpamAssassin and rejection statistics.
1794 Don't display local sender or destination tables unless
1795 there is data to show.
1796 Added average volumes into the top table text output.
1798 SC/04 Eximstats V1.46
1799 Collect data on the number of addresses (recipients)
1800 as well as the number of messages.
1802 SC/05 Eximstats V1.47
1803 Added 'Message too big' to the list of mail rejection
1804 reasons (thanks to Marco Gaiarin).
1806 SC/06 Eximstats V1.48
1807 Mainlog lines which have GMT offsets and are too short to
1808 have a flag are now skipped.
1810 SC/07 Eximstats V1.49 (Alain Williams)
1811 Added the -emptyok flag.
1813 SC/08 Eximstats V1.50
1814 Fixes for obtaining the IP address from reject messages.
1816 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1817 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1818 whitesspace changes from 4.64-PH/27
1820 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1823 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1824 are found, allow negative numbers in numeric criteria)
1826 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1828 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1829 to match changes made in 4.64-PH/43
1831 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1833 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1834 "domain" is actually an IP address, and force a failure. This locks out
1835 those revolvers/nameservers that support "A-for-A" lookups, in
1836 contravention of the specifications.
1838 PH/48 When a host name was looked up from an IP address, and the subsequent
1839 forward lookup of the name timed out, the host name was left in
1840 $sender_host_name, contrary to the specification.
1842 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1843 restricted to single-key lookups, Exim was not diagnosing an error if
1844 * or *@ was used with a query-style lookup.
1846 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1848 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1849 long ago, but noone interested enough thought of it. Let's just say that
1850 the "1.1" means that there are some new functions that weren't there at
1851 some point in the past.
1853 PH/51 Error processing for expansion failure of helo_data from an smtp
1854 transport during callout processing was broken.
1856 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1857 tested/used via the -bh/-bhc/-bs options.
1859 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1860 bug, fixed in subsequent PCRE releases).
1862 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1863 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1865 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1871 SC/01 Use a glob alias rather than an array ref in eximstats generated
1872 parser. This improves both readability and performance.
1874 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1875 Don't display local sender or destination tables in eximstats unless
1876 there is data to show.
1877 Added average volumes into the eximstats top table text output.
1879 SC/03 Collect data on the number of addresses (recipients) as well
1880 as the number of messages in eximstats.
1882 TF/01 Correct an error in the documentation for the redirect router. Exim
1883 does (usually) call initgroups() when daemonizing.
1885 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1886 with consistent privilege compared to when running as a daemon.
1888 TF/03 Note in the spec that $authenticated_id is not set for local
1889 submissions from trusted users.
1891 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1892 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1894 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1895 by adding some example configuration directives to the default
1896 configuration file. A little bit of work is required to uncomment the
1897 directives and define how usernames and passwords are checked, but
1898 there is now a framework to start from.
1900 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1901 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1902 without this. I don't know how relevant this is to other LDAP libraries.
1904 PH/02 Add the verb name to the "unknown ACL verb" error.
1906 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1908 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1910 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1911 directory not expanded when it should be if an expanded home directory
1912 was set for the address (which is overridden by the transport).
1914 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1917 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1918 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1919 because it is too late at that time, and has no effect.
1921 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1922 security issue with \' (bugzilla #107). I could not use the
1923 PQescapeStringConn() function, because it needs a PGconn value as one of
1926 PH/08 When testing addresses using -bt, indicate those final addresses that
1927 are duplicates that would not cause an additional delivery. At least one
1928 person was confused, thinking that -bt output corresponded to deliveries.
1929 (Suppressing duplicates isn't a good idea as you lose the information
1930 about possibly different redirections that led to the duplicates.)
1932 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1933 systems where poll() doesn't work, in particular OS X.
1935 PH/10 Added more information to debugging output for retry time not reached.
1937 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1938 operations in malware.c.
1940 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1943 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1944 syslog_duplication set false, log lines that would normally be written
1945 both the the main log and to the reject log were not written to syslog at
1948 PH/14 In the default configuration, change the use of "message" in ACL warn
1949 statements to "add_header".
1951 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1952 not followed by a command (e.g. "seen endif").
1954 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1955 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1958 PH/17 Added extra conditions to the default value of delay_warning_condition
1962 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1963 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1964 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1967 The Auto-Submitted: and various List- headers are standardised, whereas I
1968 don't think Precedence: ever was.
1970 PH/18 Refactored debugging code in route_finduser() to show more information,
1971 in particular, the error code if getpwnam() issues one.
1973 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1974 This is apparently needed in addition to the PH/07 change above to avoid
1975 any possible encoding problems.
1977 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1978 but not after initializing Perl.
1980 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1981 output them only if debugging. By default they are written stderr,
1982 apparently, which is not desirable.
1984 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1987 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1990 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1992 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1993 authenticated or an ident call has been made. Suppress the default
1994 values for $authenticated_id and $authenticated_sender (but permit -oMai
1995 and -oMas) when testing with -bh.
1997 PH/24 Re-jigged the order of the tests in the default configuration so that the
1998 tests for valid domains and recipients precede the DNS black list and CSA
1999 tests, on the grounds that those ones are more expensive.
2001 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2002 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2003 This bug exists in every version of Exim that I still have, right back to
2006 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2007 However, an attempt to turn on case-sensitivity in a regex key by
2008 including (?-i) didn't work because the subject string was already
2009 lowercased, and the effects were non-intuitive. It turns out that a
2010 one-line patch can be used to allow (?-i) to work as expected.
2016 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2017 other effects) broke the use of negated acl sub-conditions.
2019 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2022 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2023 "Deny" causes Exim to reject the incoming connection with a 554 error.
2024 Unfortunately, if there is a major crisis, such as a disk failure,
2025 tcp-wrappers gives "deny", whereas what one would like would be some
2026 kind of temporary error. A kludge has been added to help with this.
2027 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2028 554 error is used if errno is still zero or contains ENOENT (which occurs
2029 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2032 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2034 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2035 errors. Otherwise a message that provokes a temporary error (when other
2036 messages do not) can cause a whole host to time out.
2038 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2039 addresses were routed directly to files or pipes from a redirect router.
2040 File deliveries just didn't batch; pipe deliveries might have suffered
2043 PH/06 A failure to get a lock for a hints database would erroneously always say
2044 "Failed to get write lock", even when it was really a read lock.
2046 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2047 of 0600. This has been changed to use the value of the lockfile_mode
2048 option (which defaults to 0600).
2050 PH/08 Applied small patch from the Sieve maintainer.
2052 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2053 folder from quota calculations, a direct delivery into this folder messed
2054 up the contents of the maildirsize file. This was because the regex was
2055 used only to exclude .Trash (or whatever) when the size of the mailbox
2056 was calculated. There was no check that a delivery was happening into an
2057 excluded directory. This bug has been fixed by ignoring all quota
2058 processing for deliveries into excluded directories.
2060 PH/10 Added the maildirfolder_create_regex option to appendfile.
2066 PH/01 The code for finding all the local interface addresses on a FreeBSD
2067 system running IPv6 was broken. This may well have applied to all BSD
2068 systems, as well as to others that have similar system calls. The broken
2069 code found IPv4 interfaces correctly, but gave incorrect values for the
2070 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2071 that it would not match correctly against @[] and not recognize the IPv6
2074 PH/02 The ipliteral router was not recognizing addresses of the form user@
2075 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2077 PH/03 Added disable_ipv6.
2079 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2080 decoded form, because it is most often used to construct To: headers
2081 lines in autoreplies, and the decoded form may well be syntactically
2082 invalid. However, $reply_address has leading white space removed, and all
2083 newlines turned into spaces so that the autoreply transport does not
2086 PH/05 If group was specified without a user on a router, and no group or user
2087 was specified on a transport, the group from the router was ignored.
2089 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2090 for visible compile-time settings that can be used to change these
2091 numbers, for those that want even more. Backwards compatibility with old
2092 spool files has been maintained. However, going back to a previous Exim
2093 release will lost any variables that are in spool files.
2095 PH/07 Two small changes when running in the test harness: increase delay when
2096 passing a TCP/IP connection to a new process, in case the original
2097 process has to generate a bounce, and remove special handling of
2098 127.0.0.2 (sic), which is no longer necessary.
2100 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2101 be the same on different OS.
2103 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2106 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2107 whether --show-vars was specified or not
2109 JJ/02 exipick: Added support for new ACL variable spool format introduced
2112 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2113 syntactically invalid From: or Reply-to: line, and a filter used this to
2114 generate an autoreply, and therefore failed to obtain an address for the
2115 autoreply, Exim could try to deliver to a non-existent relative file
2116 name, causing unrelated and misleading errors. What now happens is that
2117 it logs this as a hard delivery error, but does not attempt to create a
2120 PH/11 The exinext utility has a -C option for testing purposes, but although
2121 the given file was scanned by exinext itself; it wasn't being passed on
2122 when Exim was called.
2124 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2125 an end-of-file indication when reading a command response.
2127 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2128 compiled. In many other places in Exim, IPv6 addresses are always
2129 recognized, so I have changed this. It also means that IPv4 domain
2130 literals of the form [IPV4:n.n.n.n] are now always recognized.
2132 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2133 used if the router is not running as root, for example, when verifying at
2134 ACL time, or when using -bh. The debugging output from this situation was
2135 non-existent - all you got was a failure to exec. I have made two
2138 (a) Failures to set uid/gid, the current directory, or a process leader
2139 in a subprocess such as that created by queryprogram now generate
2140 suitable debugging ouput when -d is set.
2142 (b) The queryprogram router detects when it is not running as root,
2143 outputs suitable debugging information if -d is set, and then runs
2144 the subprocess without attempting to change uid/gid.
2146 PH/15 Minor change to Makefile for building test_host (undocumented testing
2149 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2150 additional section of a DNS packet that returns MX or SRV records.
2151 Instead, it always explicitly searches for A/AAAA records. This avoids
2152 major problems that occur when a DNS server includes only records of one
2153 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2154 fixed another bug: if SRV records were looked up and the corresponding
2155 address records were *not* found in the additional section, the port
2156 values from the SRV records were lost.
2158 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2159 using the correct key (the original address) when searching the retry
2160 rules in order to find which one to use for generating the retry hint.
2162 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2163 adding the default one. Similarly, if it contains a Reply-To: header, the
2164 errors_reply_to option, if set, is not used.
2166 PH/19 When calculating a retry time, Exim used to measure the "time since
2167 failure" by looking at the "first failed" field in the retry record. Now
2168 it does not use this if it is later than than the arrival time of the
2169 message. Instead it uses the arrival time. This makes for better
2170 behaviour in cases where some deliveries succeed, thus re-setting the
2171 "first failed" field. An example is a quota failure for a huge message
2172 when small messages continue to be delivered. Without this change, the
2173 "time since failure" will always be short, possible causing more frequent
2174 delivery attempts for the huge message than are intended.
2175 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2177 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2178 $1, $2, $3) because the numerical variables can be reset during some
2179 expansion items (e.g. "match"), thereby losing the authentication data.
2181 PH/21 Make -bV show the size of off_t variables so that the test suite can
2182 decide whether to run tests for quotas > 2G.
2184 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2185 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2186 filecount value is greater than 2G or if a quota value is greater than 2G
2187 on a system where the size of off_t is not greater than 4, a panic error
2190 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2191 never match. The debug and -bh output now contains an explicit error
2192 message indicating a malformed IPv4 address or mask.
2194 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2195 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2196 PH/23 above applies.
2198 PH/25 Do not write to syslog when running in the test harness. The only
2199 occasion when this arises is a failure to open the main or panic logs
2200 (for which there is an explicit test).
2202 PH/26 Added the /no_tell option to "control=freeze".
2204 PH/27 If a host name lookup failed very early in a connection, for example, if
2205 the IP address matched host_lookup and the reverse lookup yielded a name
2206 that did not have a forward lookup, an error message of the form "no IP
2207 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2208 could be logged. Now it outputs the IP address instead of "NULL".
2210 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2211 allows the sender and the authenticated sender to be set when
2212 submitting a message from within Exim. Since child_open_exim() is
2213 documented for local_scan(), the new function should be too.
2215 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2216 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2217 results in an empty string is now treated as unset.
2219 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2221 PH/31 Added sender_verify_fail logging option.
2223 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2224 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2225 tidied the source and removed it altogether.
2227 PH/33 When a queue run was abandoned because the load average was too high, a
2228 log line was always written; now it is written only if the queue_run log
2229 selector is set. In addition, the log line for abandonment now contains
2230 information about the queue run such as the pid. This is always present
2231 in "start" and "stop" lines but was omitted from the "abandon" line.
2233 PH/34 Omit spaces between a header name and the colon in the error message that
2234 is given when verify = headers_syntax fails (if there are lots of them,
2235 the message gets confusing).
2237 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2238 names, as there are now some PTR records that contain slashes. This check
2239 is only to protect against broken name servers that fall over on strange
2240 characters, so the fact that it applies to all lookups doesn't matter.
2242 PH/36 Now that the new test suite is complete, we can remove some of the
2243 special code in Exim that was needed for the old test suite. For example,
2244 sorting DNS records because real resolvers return them in an arbitrary
2245 order. The new test suite's fake resolver always returns records in the
2248 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2249 bounces) except when queue_only is set, to avoid logging races between
2250 the different processes.
2252 PH/38 Panic-die if .include specifies a non-absolute path.
2254 PH/39 A tweak to the "H" retry rule from its user.
2256 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2257 a label. They prevented compilation on older perls.
2259 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2260 a warning to be raised on newish perls.
2262 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2263 on queue. Changes to match documented behaviour of showing count of
2264 messages matching specified criteria.
2266 PH/40 Changed the default ident timeout from 30s to 5s.
2268 PH/41 Added support for the use of login_cap features, on those BSD systems
2269 that have them, for controlling the resources used by pipe deliveries.
2271 PH/42 The content-scanning code uses fopen() to create files in which to put
2272 message data. Previously it was not paying any attention to the mode of
2273 the files. Exim runs with umask(0) because the rest of the code creates
2274 files with open(), and sets the required mode explicitly. Thus, these
2275 files were ending up world-writeable. This was not a big issue, because,
2276 being within the spool directory, they were not world-accessible. I have
2277 created a function called modefopen, which takes an additional mode
2278 argument. It sets umask(777), creates the file, chmods it to the required
2279 mode, then resets the umask. All the relevant calls to fopen() in the
2280 content scanning code have been changed to use this function.
2282 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2283 to 24 hours. This avoids potential overflow problems when processing G
2284 and H retry rules. I suspect nobody ever tinkers with this value.
2286 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2288 PH/45 When the plaintext authenticator is running as a client, the server's
2289 challenges are checked to ensure they are valid base64 strings. By
2290 default, the authentication attempt is cancelled if an invalid string is
2291 received. Setting client_ignore_invalid_base64 true ignores these errors.
2292 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2293 they are received. Thus, the responses can be made to depend on the
2294 challenges. If an invalid string is ignored, an empty string is placed in
2297 PH/46 Messages that are created by the autoreply transport now contains a
2298 References: header, in accordance with RFCs 2822 and 3834.
2300 PH/47 Added authenticated_sender_force to the smtp transport.
2302 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2304 PH/49 Installed latest patch from the Sieve maintainer.
2306 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2307 was also set, the appendfile transport was unnecessarily scanning a
2308 directory of message files (e.g. for maildir delivery) to find the count
2309 of files (along with the size), even though it did not need this
2310 information. It now does the scan only if it needs to find either the
2311 size of the count of files.
2313 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2315 PH/52 Two bugs concerned with error handling when the smtp transport is
2318 (i) Exim was not creating retry information for temporary errors given
2319 for individual recipients after the DATA command when the smtp transport
2320 was used in LMTP mode. This meant that they could be retried too
2321 frequently, and not timed out correctly.
2323 (ii) Exim was setting the flag that allows error details to be returned
2324 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2325 recipients that were returned after the DATA command.
2327 PH/53 This is related to PH/52, but is more general: for any failing address,
2328 when detailed error information was permitted to be returned to the
2329 sender, but the error was temporary, then after the final timeout, only
2330 "retry timeout exceeded" was returned. Now it returns the full error as
2331 well as "retry timeout exceeded".
2333 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2334 do this, and (what is worse) MTAs that accept it.
2336 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2337 will now be deprecated.
2339 PH/56 New os.c-cygwin from the Cygwin maintainer.
2341 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2342 formats (previously only available in exim formats via -bpr, -bpru,
2343 and -bpra. Now also available in native and exiqgrep formats)
2345 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2346 with very large, slow to parse queues
2348 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2350 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2352 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2353 responses to authentication challenges, though it was showing the
2354 challenges; (ii) I've removed the CR characters from the debug output for
2357 PH/58 Allow for the insertion of a newline as well as a space when a string
2358 is turned into more than one encoded-word during RFC 2047 encoding. The
2359 Sieve code now uses this.
2361 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2362 data_4xx, lost_connection, tls_required.
2364 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2365 message was being sent as an SMTP response.
2367 PH/61 Add -l and -k options to exicyclog.
2369 PH/62 When verifying, if an address was redirected to one new address, so that
2370 verification continued, and the new address failed or deferred after
2371 having set something in $address_data, the value of $address_data was not
2372 passed back to the ACL. This was different to the case when no
2373 redirection occurred. The value is now passed back in both cases.
2375 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2376 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2377 use login_cap.h, so on its own it isn't the distinguishing feature. The
2378 new name refers directly to the setclassresources() function.
2380 PH/65 Added configuration files for NetBSD3.
2382 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2384 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2385 is preferred over IPv4.
2387 PH/68 The bounce_return_message and bounce_return_body options were not being
2388 honoured for bounces generated during the reception of non-SMTP messages.
2389 In particular, this applied to messages rejected by the ACL. This bug has
2390 been fixed. However, if bounce_return_message is true and bounce_return_
2391 body is false, the headers that are returned for a non-SMTP message
2392 include only those that have been read before the error was detected.
2393 (In the case of an ACL rejection, they have all been read.)
2395 PH/69 The HTML version of the specification is now built in a directory called
2396 spec_html instead of spec.html, because the latter looks like a path with
2397 a MIME-type, and this confuses some software.
2399 PH/70 Catch two compiler warnings in sieve.c.
2401 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2402 function verify_get_ident() calls ip_connect() to connect a socket, but
2403 if the "connect()" function timed out, ip_connect() used to close the
2404 socket. However, verify_get_ident() also closes the socket later, and in
2405 between Exim writes to the log, which may get opened at this point. When
2406 the socket was closed in ip_connect(), the log could get the same file
2407 descriptor number as the socket. This naturally causes chaos. The fix is
2408 not to close the socket in ip_connect(); the socket should be closed by
2409 the function that creates it. There was only one place in the code where
2410 this was missing, in the iplookup router, which I don't think anybody now
2411 uses, but I've fixed it anyway.
2413 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2414 well as to direct DNS lookups. Otherwise the handling of names in host
2415 lists is inconsistent and therefore confusing.
2421 PH/01 Two changes to the default runtime configuration:
2423 (1) Move the checks for relay_from_hosts and authenticated clients from
2424 after to before the (commented out) DNS black list checks.
2426 (2) Add control=submission to the relay_from_hosts and authenticated
2427 clients checks, on the grounds that messages accepted by these
2428 statements are most likely to be submissions.
2430 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2432 (1) Generate an error if the third argument for the ${prvs expansion is
2435 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2438 (3) Reset the variables that are obtained from the first argument of
2439 ${prvscheck and used in the second argument before leaving the code,
2440 because their memory is reclaimed, so using them afterwards may do
2443 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2444 one (it's much easier than Tom thought :-).
2446 (5) Because of (4), we can now allow for the use of $prvscheck_result
2447 inside the third argument.
2449 PH/03 For some reason, the default setting of PATH when running a command from
2450 a pipe transport was just "/usr/bin". I have changed it to
2453 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2454 anything to be listed in the output from -bV.
2456 PH/05 When a filter generated an autoreply, the entire To: header line was
2457 quoted in the delivery log line, like this:
2459 => >A.N.Other <ano@some.domain> <original@ddress> ...
2461 This has been changed so that it extracts the operative address. There
2462 may be more than one such address. If so, they are comma-separated, like
2465 => >ano@some.domain,ona@other.domain <original@ddress> ...
2467 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2468 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2469 not being looked up in the rDNS to get a host name, Exim was showing the
2470 IP address twice in Received: lines, even though the IP addresses were
2471 identical. For example:
2473 Received: from [1.2.3.4] (helo=[1.2.3.4])
2475 However, if the real host name was known, it was omitting the HELO data
2476 if it matched the actual IP address. This has been tidied up so that it
2477 doesn't show the same IP address twice.
2479 PH/07 When both +timestamp and +memory debugging was on, the value given by
2480 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2481 called by the debug printing, thereby overwriting the timestamp buffer.
2482 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2484 PH/08 When the original message was included in an autoreply transport, it
2485 always said "this is a copy of the message, including all the headers",
2486 even if body_only or headers_only was set. It now gives an appropriate
2489 PH/09 Applied a patch from the Sieve maintainer which:
2491 o fixes some comments
2492 o adds the (disabled) notify extension core
2493 o adds some debug output for the result of if/elsif tests
2494 o points to the current vacation draft in the documentation
2495 and documents the missing references header update
2499 o fixes a bug in processing the envelope test (when testing
2500 multiple envelope elements, the last element determinted the
2503 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2504 Electronic Mail") by including:
2506 Auto-submitted: auto-generated
2508 in the messages that it generates (bounce messages and others, such as
2509 warnings). In the case of bounce messages for non-SMTP mesages, there was
2510 also a typo: it was using "Auto_submitted" (underscore instead of
2511 hyphen). Since every message generated by Exim is necessarily in response
2512 to another message, thes have all been changed to:
2514 Auto-Submitted: auto-replied
2516 in accordance with these statements in the RFC:
2518 The auto-replied keyword:
2520 - SHOULD be used on messages sent in direct response to another
2521 message by an automatic process,
2523 - MUST NOT be used on manually-generated messages,
2525 - MAY be used on Delivery Status Notifications (DSNs) and Message
2526 Disposition Notifications (MDNs),
2528 - MUST NOT be used on messages generated by automatic or periodic
2529 processes, except for messages which are automatic responses to
2532 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2533 to the default Received: header definition.
2535 PH/12 Added log selector acl_warn_skipped (default on).
2537 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2538 variables because (a) they are in the wrong storage pool and (b) even if
2539 they were copied, it wouldn't work properly because of the caching.
2541 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2542 checking when decoding. Apparently there are clients that generate
2543 overlong encoded strings. Why am I not surprised?
2545 PH/15 If the first argument of "${if match_address" was not empty, but did not
2546 contain an "@" character, Exim crashed. Now it writes a panic log message
2547 and treats the condition as false.
2549 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2551 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2552 extension "envelope-auth". The code is finished and in agreement with
2553 other implementations, but there is no documentation so far and in fact,
2554 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2555 not changing the active code.
2557 Print executed "if" and "elsif" statements when debugging is used. This
2558 helps a great deal to understand what a filter does.
2560 Document more things not specified clearly in RFC3028. I had all this
2561 sorted out, when out of a sudden new issues came to my mind. Oops."
2563 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2566 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2567 canonical form (as documented). However, after a host name lookup from
2568 the IP address, check_host() was doing a simple string comparison with
2569 addresses acquired from the DNS when checking that the found name did
2570 have the original IP as one of its addresses. Since any found IPv6
2571 addresses are likely to be in abbreviated form, the comparison could
2572 fail. Luckily, there already exists a function for doing the comparison
2573 by converting both addresses to binary, so now that is used instead of
2574 the text comparison.
2576 PH/20 There was another similar case to PH/19, when a complete host name was
2577 given in a host list; looking up its IP address could give an abbreviated
2578 form, whereas the current host's name might or might not be abbreviated.
2579 The same fix has been applied.
2585 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2586 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2589 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2590 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2592 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2594 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2595 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2596 Also, for "Auto-Submitted", treat anything other than "no" as
2597 non-personal, in accordance with RFC 3834. (Previously it treated
2598 anything starting "auto-" as non-personal.)
2600 TF/01 The control=submission/name=... option had a problem with syntax
2601 errors if the name included a slash character. The /name= option
2602 now slurps the rest of the string, so it can include any characters
2603 but it must come last in the list of options (after /sender_retain
2606 PH/05 Some modifications to the interface to the fake nameserver for the new
2614 TK/01 Added the "success_on_redirect" address verification option. See
2615 NewStuff for rationale and an example.
2617 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2619 PH/02 Patch to exigrep to allow it to work on syslog lines.
2621 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2622 fread() to skip over the body file's header line, because in Cygwin the
2623 header line is locked and is inaccessible.
2625 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2626 co-exist for some time) to make it clear that it is the Exim ID that is
2627 referenced, not the Message-ID: header line.
2629 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2630 string_format() function, because snprintf() does not exist on all
2633 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2634 ${stat: expansion item.
2636 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2637 protocol synchronization error", to keep the pedants happy.
2639 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2640 well as for IRIX systems, when gcc is being used. See the host.c source
2643 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2645 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2648 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2649 transport and to the smtp transport in LMTP mode.
2651 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2653 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2654 run a filter in a subprocess. This could lead to confusion in subsequent
2655 lookups in the parent process. There should also be a search_tidyup() at
2656 the end of the subprocess.
2658 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2659 only if the host matched helo_try_verify_hosts, which caused the
2660 verification to occur when the EHLO/HELO command was issued. The ACL just
2661 tested the remembered result. Now, if a previous verification attempt has
2662 not happened, "verify = helo" does it there and then.
2664 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2666 TK/03 Fix log output including CR from clamd.
2668 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2669 exist provoked a memory error which could cause a segfault.
2671 PH/15 Installed PCRE 6.2
2673 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2675 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2676 of the problem. Specifically, suggested +O2 rather than +O1 for the
2679 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2681 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2682 record, the port was not correctly logged when the outgoing_port log
2683 selector was set (it logged the transort's default port).
2685 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2686 fallback_hosts, and "hosts" in the smtp transport.
2688 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2689 host errors such as "Connection refused".
2691 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2692 authentication with radiusclient 0.4.9:
2694 - Error returned from rc_read_config was caught wrongly
2695 - Username/password not passed on to radius server due to wrong length.
2697 The presumption is that some radiusclient API changes for 4.51/PH/17
2698 were not taken care of correctly. The code is still untested by me (my
2699 Linux distribution still has 0.3.2 of radiusclient), but it was
2700 contributed by a Radius user.
2702 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2703 expanding the "port" option of the smtp transport.
2705 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2706 while reading a MIME header. Thanks to Tom Hughes for a patch.
2708 PH/24 Include config.h inside local_scan.h so that configuration settings are
2711 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2712 that in an ACL for RCPT (for example), you can examine exactly what was
2715 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2716 commands, but it was not correctly comparing the address with the actual
2717 client host address. Thus, it would show the EHLO address in Received:
2718 header lines when this was not necessary.
2720 PH/27 Added the % operator to ${eval:}.
2722 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2723 it should be ignoring failures (because with -C, for example, it has lost
2724 privilege). It wasn't ignoring creation failures other than "already
2727 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2730 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2731 because an input line was too long, either on its own, or by virtue of
2732 too many continuations, the temporary file was not being removed, and the
2733 return code was incorrect.
2735 PH/31 Missing "BOOL" in function definition in filtertest.c.
2737 PH/32 Applied Sieve patches from the maintainer.
2739 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2741 PH/33 Added "verify = not_blind".
2743 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2744 Local/Makefile (with some defaults set). These are used in built scripts
2745 such as exicyclog, but they have never been used in the exim_install
2746 script (though there are many overriding facilities there). I have
2747 arranged that the exim_install script now takes note of these two
2750 PH/35 Installed configuration files for Dragonfly.
2752 PH/36 When a locally submitted message by a trusted user did not contain a
2753 From: header, and the sender address was obtained from -f or from an SMTP
2754 MAIL command, and the trusted user did not use -F to supply a sender
2755 name, $originator_name was incorrectly used when constructing a From:
2756 header. Furthermore, $originator_name was used for submission mode
2757 messages from external hosts without From: headers in a similar way,
2758 which is clearly wrong.
2760 PH/37 Added control=suppress_local_fixups.
2762 PH/38 When log_selector = +received_sender was set, and the addition of the
2763 sender made the log line's construction buffer exactly full, or one byte
2764 less than full, an overflow happened when the terminating "\n" was
2767 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2768 when the result of a list match is failure because a DNS lookup failed.
2770 PH/40 RM_COMMAND is now used in the building process.
2772 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2773 the "build-* directories that it finds.
2775 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2776 address in a domain literal was a prefix of an interface address.
2778 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2779 when verifying a sender address, unless rewrite_headers is false.
2781 PH/44 Wrote a long comment about why errors_to addresses are verified as
2782 recipients, not senders.
2784 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2785 the ratelimit ACL was added.
2787 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2789 PH/46 Added extra information about PostgreSQL errors to the error string.
2791 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2792 avoiding the need to install special zones in a real server. This is
2793 backwards compatible; if it can't find the fake resolver, it drops back.
2794 Thus, both old and new test suites can be run.
2796 TF/02 Added util/ratelimit.pl
2798 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2799 clock is set back in time.
2801 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2802 Candler <B.Candler@pobox.com>.
2804 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2805 OK for addresses that are the result of redirections.
2807 PH/48 A number of further additions for the benefit of the new test suite,
2808 including a fake gethostbyname() that interfaces to the fake DNS resolver
2811 TF/06 The fix for widen_domains has also been applied to qualify_single and
2812 search_parents which are the other dnslookup options that can cause
2815 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2818 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2820 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2821 service" if it encountered an MX record with an empty target hostname.
2822 The message is now "an MX or SRV record indicated no SMTP service".
2824 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2825 if the DNS of the sending site is misconfigured. This is quite a
2826 common situation. This change restores the behaviour of treating a
2827 helo verification defer as a failure.
2829 PH/51 If self=fail was set on a router, the bounce message did not include the
2830 actual error message.
2836 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2838 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2839 command itself ended in error, the underlying message about the transport
2840 filter timeout was being overwritten with the pipe command error. Now the
2841 underlying error message should be appended to the second error message.
2843 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2845 PH/02 Reduce the amount of output that "make" produces by default. Full output
2846 can still be requested.
2848 PH/03 The warning log line about a condition test deferring for a "warn" verb
2849 was being output only once per connection, rather than after each
2850 occurrence (because it was using the same function as for successful
2851 "warn" verbs). This seems wrong, so I have changed it.
2853 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2854 it should not have, which might have caused a crash in the right
2855 circumstances, but probably never did.
2857 PH/04 Installed a modified version of Tony Finch's patch to make submission
2858 mode fix the return path as well as the Sender: header line, and to
2859 add a /name= option so that you can make the user's friendly name appear
2862 TF/03 Added the control = fakedefer ACL modifier.
2864 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2865 Mark Lowes for thorough testing.
2867 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2869 TK/03 Merged latest SRS patch from Miles Wilton.
2871 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2872 with the definition in sysexits.h (which is #included earlier).
2873 Fortunately, Exim does not actually use EX_OK. The code used to try to
2874 preserve the sysexits.h value, by assumimg that macro definitions were
2875 scanned for macro replacements. I have been disabused of this notion,
2876 so now the code just undefines EX_OK before #including unistd.h.
2878 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2879 in the smtp transport. When a block could not be written in a single
2880 write() function, the timeout was being re-applied to each part-write.
2881 This seems wrong - if the receiver was accepting one byte at a time it
2882 would take for ever. The timeout is now adjusted when this happens. It
2883 doesn't have to be particularly precise.
2885 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2886 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2888 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2889 without a domain if the check to <postmaster@domain> fails.
2891 SC/01 Eximstats: added -xls and the ability to specify output files
2892 (patch written by Frank Heydlauf).
2894 SC/02 Eximstats: use FileHandles for outputing results.
2896 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2898 SC/04 Eximstats: fixed display of large numbers with -nvr option
2900 SC/05 Eximstats: fixed merging of reports with empty tables.
2902 SC/06 Eximstats: added the -include_original_destination flag
2904 SC/07 Eximstats: removed tabs and trailing whitespace.
2906 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2908 TK/06 MBOX spool code: Add real "From " MBOX separator line
2909 so the .eml file is really in mbox format (even though
2910 most programs do not really care). Patch from Alex Miller.
2912 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2913 The latter is generated from $received_to and is only set if the
2914 message has one envelope recipient. SA can use these headers,
2915 obviously out-of-the-box. Patch from Alex Miller.
2917 PH/08 The ${def test on a variable was returning false if the variable's
2918 value was "0", contrary to what the specification has always said!
2919 The result should be true unless the variable is empty.
2921 PH/09 The syntax error of a character other than { following "${if
2922 def:variable_name" (after optional whitespace) was not being diagnosed.
2923 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2924 accidental colon was present, for example, could give incorrect results.
2926 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2927 result is used (not including appendfile, where other changes are about
2930 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2931 This involved changing a lot of size variables from int to off_t. It
2932 should work with maildirs and everything.
2934 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2935 spamd dying while we are connected to it.
2937 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2940 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2941 function for BDB 4.3.
2943 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2945 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2946 the sources for PCRE have been re-arranged and more files are now
2949 PH/15 The code I had for printing potentially long long variables in PH/11
2950 above was not the best (it lost precision). The length of off_t variables
2951 is now inspected at build time, and an appropriate printing format (%ld
2952 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2953 to be "long long int" or "long int". This is needed for the internal
2954 formatting function string_vformat().
2956 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2957 the configuration file to be ":syslog", then the script "guesses" where
2958 the logs files are, rather than using the compiled in default. In our
2959 case the guess is not the same as the compiled default, so the script
2960 suddenly stopped working when I started to use syslog. The patch checks
2961 to see if log_file_path is "". If so, it attempts to read it from exim
2962 with no configuration file to get the compiled in version, before it
2963 falls back to the previous guessing code."
2965 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2966 implementing BATV in an Exim configuration. See NewStuff for the gory
2969 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2970 Makefile that are specific to HP-UX.
2972 PH/18 If the "use_postmaster" option was set for a recipient callout together
2973 with the "random" option, the postmaster address was used as the MAIL
2974 FROM address for the random test, but not for the subsequent recipient
2975 test. It is now used for both.
2977 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2978 patch removes a few documentation additions to RFC 3028, because the
2979 latest draft now contains them. It adds the new en;ascii-case comparator
2980 and a new error check for 8bit text in MIME parts. Comparator and
2981 require names are now matched exactly. I enabled the subaddress
2982 extension, but it is not well tested yet (read: it works for me)."
2984 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2985 rework some of the code of TK/09 above to avoid the hardwired use of
2986 "%lld" and "long long". Replaced the call to snprintf() with a call to
2989 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2990 records point to non-existent hosts", "retry timeout exceeded", and
2991 "retry time not reached for any host after a long failure period".
2993 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2994 experimental DomainKeys support:
2996 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2997 (2) On an error such as an illegally used "control", the wrong name for
2998 the control was given.
3000 These problems did NOT occur unless DomainKeys support was compiled.
3002 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3004 PH/24 Added ${if match_ip condition.
3006 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3007 databases so that it will be absolutely obvious if a crash occurs in the
3008 DB library. This is a regular occurrence (often caused by mis-matched
3011 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3012 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3013 picked up on a user's system that detects such things. There doesn't seem
3014 to be a gcc warning option for this - only an attribute that has to be
3015 put on the function's prototype. It seems that in Fedora Core 4 they have
3016 set this on a number of new functions. No doubt there will be more in due
3019 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3020 specify a transport. However, if an address that was verified by such a
3021 router was the subject of a callout, Exim crashed because it tried to
3022 read the rcpt_include_affixes from the non-existent transport. Now it
3023 just assumes that the setting of that option is false. This bug was
3024 introduced by 4.51/PH/31.
3026 PH/28 Changed -d+all to exclude +memory, because that information is very
3027 rarely of interest, but it makes the output a lot bigger. People tend to
3028 do -d+all out of habit.
3030 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3031 code in os-type was giving problems when libc.so lives in lib64, like on
3034 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3035 aren't the modern standard, and it seems that some systems' include files
3036 don't always have them. Exim was already checking for some of the newer
3037 ones like T_AAAA, and defining it itself. I've added checks for all the
3038 record types that Exim uses.
3040 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3041 not automatically generating a new one, as it is supposed to. This
3042 prevented TLS from working. If the file did exist, but contained invalid
3043 data, a new version was generated, as expected. It was only the case of a
3044 non-existent file that was broken.
3046 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3047 with a change in libdomainkeys > 0.64.
3049 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3050 from DNS. If the selector record carries the flag, it now has
3051 precedence over the domain-wide flag.
3053 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3055 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3056 the use of a port name that isn't defined in /etc/services) occurred, the
3057 message was deferred as in a normal delivery, and thus remained on the
3058 spool, instead of being failed because of the mua_wrapper setting. This
3059 is now fixed, and I tidied up some of the mua_wrapper messages at the
3062 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3063 the messages in a hash of arrays rather than using individual hashes.
3064 This is a bit cleaner and results in dramatic memory savings, albeit
3065 at a slight CPU cost.
3067 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3068 as requested by Marc Sherman.
3070 SC/10 Eximstats: added histograms for user specified patterns as requested
3073 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3075 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3076 fopen() in the content-scanning modules that did not already have it.
3082 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3083 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3085 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3087 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3089 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3090 to test Sieve filters that use "vacation".
3092 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3093 that changes the way the GnuTLS parameters are stored in the cache file.
3094 The new format can be generated externally. For backward compatibility,
3095 if the data in the cache doesn't make sense, Exim assumes it has read an
3096 old-format file, and it generates new data and writes a new file. This
3097 means that you can't go back to an older release without removing the
3100 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3101 work if there are any delivery delays because "one_time" forces the
3102 parent to be marked "delivered", so its unseen clone is never tried
3103 again. For this reason, Exim now forbids the simultaneous setting of
3106 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3107 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3108 entry if you want to know the details. The fix, however, neglected to
3109 consider the case when local delivery batching is involved. The test for
3110 "previously delivered" was not happening when checking to see if an
3111 address could be batched with a previous (undelivered) one; under
3112 certain circumstances this could lead to multiple deliveries to the same
3115 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3116 in its include files, and this causes problems building Exim.
3118 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3119 header_syntax) but Exim was just ignoring anything given after a slash.
3120 In particular, this caused confusion with an attempt to use "verify =
3121 reverse_host_lookup/defer_ok". An error is now given when options are
3122 supplied for verify items that do not have them. (Maybe reverse_host_
3123 lookup should have a defer_ok option, but that's a different point.)
3125 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3126 defined by RFC 821) to 2048, because there were problems with some AUTH
3127 commands, and RFC 1869 says the size should be increased for extended
3128 SMTP commands that take arguments.
3130 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3133 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3134 "unknown" error; now it says that the functionality isn't in the binary.
3136 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3137 an address' error message when a string expansion fails (syntax or
3138 whatever). Otherwise the password may appear in the log. Following change
3139 PH/42 below, there is no longer a chance of it appearing in a bounce
3142 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3144 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3145 of its addresses was ever tried. (Bugzilla bug #2.)
3147 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3148 the result incorrectly in the debug output. (It correctly added a newline
3149 to what was transported.)
3151 TF/01 Added $received_time.
3153 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3154 commented out examples of how to interface to a virus scanner and to
3155 SpamAssassin. Also added commented examples of av_scanner and
3156 spamd_address settings.
3158 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3159 and controls are allowed in which ACLs. There were a couple of minor
3160 errors. Some of the entries in the conditions table (which is a table of
3161 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3162 negation of where the condition IS allowed.
3164 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3166 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3167 header file does not have a version number, so I've had to invent a new
3168 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3169 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3170 radiusclient), but it was contributed by a Radius user.
3172 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3173 files or directories, for OpenSSL.
3175 PH/19 When an Exim process that is running as root has to create an Exim log
3176 file, it does so in a subprocess that runs as exim:exim so as to get the
3177 ownership right at creation (otherwise, other Exim processes might see
3178 the file with the wrong ownership). There was no test for failure of this
3179 fork() call, which would lead to the process getting stuck as it waited
3180 for a non-existent subprocess. Forks do occasionally fail when resources
3181 run out. I reviewed all the other calls to fork(); they all seem to check
3184 PH/20 When checking for unexpected SMTP input at connect time (before writing
3185 the banner), Exim was not dealing correctly with a non-positive return
3186 from the read() function. If the client had disconnected by this time,
3187 the result was a log entry for a synchronization error with an empty
3188 string after "input=" when read() returned zero. If read() returned -1
3189 (an event I could not check), uninitialized data bytes were printed.
3190 There were reports of junk text (parts of files, etc) appearing after
3193 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3195 PH/22 Added support for macro redefinition, and (re)definition in between
3196 driver and ACL definitions.
3198 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3199 forgetting to use the resulting value; it was using the unexpanded value.
3201 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3202 hadn't been configured. The fix is from Juergen Kreileder, who
3203 understands it better than I do:
3205 "Here's what I see happening with three configured cyrus_sasl
3206 authenticators configured (plain, login, cram-md5):
3208 On startup auth_cyrus_sasl_init() gets called for each of these.
3209 This means three calls to sasl_listmech() without a specified mech_list.
3210 => SASL tests which mechs of all available mechs actually work
3211 => three warnings about OTP not working
3212 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3214 With the patch, sasl_listmech() also gets called three times. But now
3215 SASL's mech_list option is set to the server_mech specified in the the
3216 authenticator. Or in other words, the answer from sasl_listmech()
3217 gets limited to just the mech you're testing for (which is different
3219 => the return list contains just 'plain' or 'login', 'cram-md5' or
3220 nothing depending on the value of ob->server_mech.
3222 I've just tested the patch: Authentication still works fine,
3223 unavailable mechs specified in the exim configuration are still
3224 caught, and the auth.log warnings about OTP are gone."
3226 PH/25 When debugging is enabled, the contents of the command line are added
3227 to the debugging output, even when log_selector=+arguments is not
3230 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3231 answer is "GNU", and only if the return is "GNU/something" is the answer
3234 PH/27 $acl_verify_message is now set immediately after the failure of a
3235 verification in an ACL, and so is available in subsequent modifiers. In
3236 particular, the message can be preserved by coding like this:
3238 warn !verify = sender
3239 set acl_m0 = $acl_verify_message
3241 Previously, $acl_verify_message was set only while expanding "message"
3242 and "log_message" when a very denied access.
3244 PH/28 Modified OS/os.c-Linux with
3246 -#ifndef OS_LOAD_AVERAGE
3247 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3249 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3250 nomenclature these days.)
3252 PH/29 Installed patch from the Sieve maintainer that adds the options
3253 sieve_useraddress and sieve_subaddress to the redirect router.
3255 PH/30 In these circumstances:
3256 . Two addresses routed to the same list of hosts;
3257 . First host does not offer TLS;
3258 . First host accepts first address;
3259 . First host gives temporary error to second address;
3260 . Second host offers TLS and a TLS session is established;
3261 . Second host accepts second address.
3262 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3263 and peerdn, if requested) that were in fact used only for the second
3266 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3267 attention to any local part prefix or suffix that was matched by the
3268 router that accepted the address. It now behaves in the same way as it
3269 does for delivery: the affixes are removed from the local part unless
3270 rcpt_include_affixes is set on the transport.
3272 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3273 timeout during the DATA phase of an incoming message.
3275 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3276 applied a patch sanctioned by the Sieve maintainer.
3278 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3279 the uid or gid is negative. A case of a negative gid caused this to be
3280 noticed. The fix allows for either to be negative.
3282 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3283 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3286 PH/36 The change PH/12 above was broken. Fixed it.
3288 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3289 the grounds that routing the same address twice would always produce the
3290 same answer. This might have been true once, but it is certainly no
3291 longer true now. Routing a child address may depend on the previous
3292 routing that produced that child. Some complicated redirection strategies
3293 went wrong when messages had multiple recipients, and made Exim's
3294 behaviour dependent on the order in which the addresses were given.
3296 I have moved the duplicate checking until after the routing is complete.
3297 Exim scans the addresses that are assigned to local and remote
3298 transports, and removes any duplicates. This means that more work will be
3299 done, as duplicates will always all be routed, but duplicates are
3300 presumably rare, so I don't expect this is of any significance.
3302 For deliveries to pipes, files, and autoreplies, the duplicate checking
3303 still happens during the routing process, since they are not going to be
3306 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3307 It corrects a timeout issue with spamd. This is Ian's comment: "The
3308 background is that sometimes spamd either never reads data from a
3309 connection it has accepted, or it never writes response data. The exiscan
3310 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3311 blindly assumes that writes won't block so it may never time out."
3313 PH/39 Allow G after quota size as well as K and M.
3315 PH/40 The value set for $authenticated_id in an authenticator may not contain
3316 binary zeroes or newlines because the value is written to log lines and
3317 to spool files. There was no check on this. Now the value is run through
3318 the string_printing() function so that such characters are converted to
3319 printable escape sequences.
3321 PH/41 $message_linecount is a new variable that contains the total number of
3322 lines in the message. Compare $body_linecount, which is the count for the
3325 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3326 bounce and delay warning messages, except in certain special cases, which
3329 (a) An SMTP error message from a remote host;
3330 (b) A message specified in a :fail: redirection;
3331 (c) A message specified in a "fail" command in a system filter;
3332 (d) A message specified in a FAIL return from the queryprogram router;
3333 (e) A message specified by the cannot_route_message router option.
3335 In these cases only, Exim does include the error details in bounce and
3336 warning messages. There are also a few cases where bland messages such
3337 as "unrouteable address" or "local delivery error" are given.
3339 PH/43 $value is now also set for the "else" part of a ${run expansion.
3341 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3342 being worked on, but at least Exim now implements the latest version to
3345 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3346 process to complete was treated as a delivery failure, a timeout while
3347 writing the message to the pipe was logged, but erroneously treated as a
3348 successful delivery. Such timeouts include transport filter timeouts. For
3349 consistency with the overall process timeout, these timeouts are now
3350 treated as errors, giving rise to delivery failures by default. However,
3351 there is now a new Boolean option for the pipe transport called
3352 timeout_defer, which, if set TRUE, converts the failures into defers for
3353 both kinds of timeout. A transport filter timeout is now identified in
3356 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3357 systems where "make" and "gmake" are different, calling "gmake" at top
3358 level broke things. I've arranged for the value of $(MAKE) to be passed
3359 from the Makefile to this script so that it can call the same version of
3363 A note about Exim versions 4.44 and 4.50
3364 ----------------------------------------
3366 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3367 changes of various kinds. As a consequence, a big documentation update was
3368 needed. This delayed the release for rather longer than seemed good, especially
3369 in the light of a couple of (minor) security issues. Therefore, the changes
3370 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3371 release. So 4.44 and 4.50 are in effect two different branches that both start
3374 I have left the 4.50 change log unchanged; it contains all the changes since
3375 4.43. The change log for 4.44 is below; many of its items are identical to
3376 those for 4.50. This seems to be the most sensible way to preserve the
3377 historical information.
3383 1. Minor wording change to the doc/README.SIEVE file.
3385 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3386 computation of the current number of files was incorrect.
3388 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3389 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3392 4. Give more explanation in the error message when the command for a transport
3393 filter fails to execute.
3395 5. There are several places where Exim runs a non-Exim command in a
3396 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3397 was being done only for the command run by the queryprogram router. It is
3398 now done for all such subprocesses. The other cases are: ${run, transport
3399 filters, and the commands run by the lmtp and pipe transports.
3401 6. Added CONFIGURE_GROUP build-time option.
3403 7. Some older OS have a limit of 256 on the maximum number of file
3404 descriptors. Exim was using setrlimit() to set 1000 as a large value
3405 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3406 systems. I've change it so that if it can't get 1000, it tries for 256.
3408 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3409 was an oversight, and furthermore, ever since the addition of extra
3410 controls (e.g. 4.43/32), the checks on when to allow different forms of
3411 "control" were broken. There should now be diagnostics for all cases when a
3412 control that does not make sense is encountered.
3414 9. Added the /retain_sender option to "control=submission".
3416 10. $recipients is now available in the predata ACL (oversight).
3418 11. Tidy the search cache before the fork to do a delivery from a message
3419 received from the command line. Otherwise the child will trigger a lookup
3420 failure and thereby defer the delivery if it tries to use (for example) a
3421 cached ldap connection that the parent has called unbind on.
3423 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3424 of $address_data from the recipient verification was clobbered by the
3425 sender verification.
3427 13. The value of address_data from a sender verification is now available in
3428 $sender_address_data in subsequent conditions in the ACL statement.
3430 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3432 15. Added a new option "connect=<time>" to callout options, to set a different
3435 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3436 was its contents. (It was OK if the option was not defined at all.)
3438 17. A "Completed" log line is now written for messages that are removed from
3439 the spool by the -Mrm option.
3441 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3442 information about exactly what failed.
3444 19. Added -dd to debug only the daemon process.
3446 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3447 handles timeouts, both on the server side and network timeouts. Renamed the
3448 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3450 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3451 It is now set to "smtps".
3453 22. $host_address is now set to the target address during the checking of
3454 ignore_target_hosts.
3456 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3457 being passed; this would have caused $sender_host_name to have been used if
3458 matching the list had actually called for a host name (not very likely,
3459 since this list is usually IP addresses). A host name is now passed as
3462 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3463 code that specifies a non-restarting handler (typically sigaction() in
3464 modern systems) in an attempt to fix a rare and obscure crash bug.
3466 25. Narrowed the window for a race in the daemon that could cause it to ignore
3467 SIGCHLD signals. This is not a major problem, because they are used only to
3468 wake it up if nothing else does.
3470 26. A malformed maildirsize file could cause Exim to calculate negative values
3471 for the mailbox size or file count. Odd effects could occur as a result.
3472 The maildirsize information is now recalculated if the size or filecount
3475 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3476 support for a long time. Removed HAVE_SYS_VFS_H.
3478 28. Installed the latest version of exipick from John Jetmore.
3480 29. In an address list, if the pattern was not a regular expression, an empty
3481 subject address (from a bounce message) matched only if the pattern was an
3482 empty string. Non-empty patterns were not even tested. This was the wrong
3483 because it is perfectly reasonable to use an empty address as part of a
3484 database query. An empty address is now tested by patterns that are
3485 lookups. However, all the other forms of pattern expect the subject to
3486 contain a local part and a domain, and therefore, for them, an empty
3487 address still always fails if the pattern is not itself empty.
3489 30. Exim went into a mad DNS loop when attempting to do a callout where the
3490 host was specified on an smtp transport, and looking it up yielded more
3491 than one IP address.
3493 31. Re-factored the code for checking spool and log partition space into a
3494 function that finds that data and another that does the check. The former
3495 is then used to implement four new variables: $spool_space, $log_space,
3496 $spool_inodes, and $log_inodes.
3498 32. The RFC2047 encoding function was originally intended for short strings
3499 such as real names; it was not keeping to the 75-character limit for
3500 encoded words that the RFC imposes. It now respects the limit, and
3501 generates multiple encoded words if necessary. To be on the safe side, I
3502 have increased the buffer size for the ${rfc2047: expansion operator from
3505 33. It is now permitted to omit both strings after an "if" condition; if the
3506 condition is true, the result is "true". As before, when the second string
3507 is omitted, a false condition yields an empty string. This makes it less
3508 cumbersome to write custom ACL and router conditions.
3510 34. Failure to deliver a bounce message always caused it to be frozen, even if
3511 there was an errors_to setting on the router. The errors_to setting is now
3514 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3515 canonical form (fully expanded) before being placed in
3516 $sender_host_address.
3518 36. The table in the code that translates DNS record types into text (T_A to
3519 "A" for instance) was missing entries for NS and CNAME. It is just possible
3520 that this could have caused confusion if both these types were looked up
3521 for the same domain, because the text type is used as part of Exim's
3522 per-process caching. But the chance of anyone hitting this buglet seems
3525 37. The dnsdb lookup has been extended in a number of ways.
3527 (1) There is a new type, "zns", which walks up the domain tree until it
3528 finds some nameserver records. It should be used with care.
3530 (2) There is a new type, "mxh", which is like "mx" except that it returns
3531 just the host names, not the priorities.
3533 (3) It is now possible to give a list of domains (or IP addresses) to be
3534 looked up. The behaviour when one of the lookups defers can be
3535 controlled by a keyword.
3537 (4) It is now possible to specify the separator character for use when
3538 multiple records are returned.
3540 38. The dnslists ACL condition has been extended: it is now possible to supply
3541 a list of IP addresses and/or domains to be looked up in a particular DNS
3544 39. Added log_selector=+queue_time_overall.
3546 40. When running the queue in the test harness, wait just a tad after forking a
3547 delivery process, to get repeatability of debugging output.
3549 41. Include certificate and key file names in error message when GnuTLS fails
3550 to set them up, because the GnuTLS error message doesn't include the name
3551 of the failing file when there is a problem reading it.
3553 42. Allow both -bf and -bF in the same test run.
3555 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3557 44. The "Exiscan patch" is now merged into the mainline Exim source.
3559 45. Sometimes the final signoff response after QUIT could fail to get
3560 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3561 < 0 before doing a fflush(). This bug looks as though it goes back to the
3562 introduction of TLS in release 3.20, but "sometimes" must have been rare
3563 because the tests only now provoked it.
3565 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3566 (this can affect the format of dates).
3568 47. exim_tidydb, when checking for the continued existence of a message for
3569 which it has found a message-specific retry record, was not finding
3570 messages that were in split spool directories. Consequently, it was
3571 deleting retry records that should have stayed in existence.
3573 48. Steve fixed some bugs in eximstats.
3575 49. The SPA authentication driver was not abandoning authentication and moving
3576 on to the next authenticator when an expansion was forced to fail,
3577 contradicting the general specification for all authenticators. Instead it
3578 was generating a temporary error. It now behaves as specified.
3580 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3581 (the order specifies the preference for clients). The order is now AES256,
3582 AES128, 3DES, ARCFOUR128.
3584 51. Small patch to Sieve code - explicitly set From: when generating an
3587 52. Exim crashed if a remote delivery caused a very long error message to be
3588 recorded - for instance if somebody sent an entire SpamAssassin report back
3589 as a large number of 550 error lines. This bug was coincidentally fixed by
3590 increasing the size of one of Exim's internal buffers (big_buffer) that
3591 happened as part of the Exiscan merge. However, to be on the safe side, I
3592 have made the code more robust (and fixed the comments that describe what
3595 53. Now that there can be additional text after "Completed" in log lines (if
3596 the queue_time_overall log selector is set), a one-byte patch to exigrep
3597 was needed to allow it to recognize "Completed" as not the last thing in
3600 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3601 patch that reportedly fixes this has been added. I am not expert enough to
3602 create a test for it. This is what the patch creator wrote:
3604 "I found a little strange behaviour of ldap code when working with
3605 Windows 2003 AD Domain, where users was placed in more than one
3606 Organization Units. When I tried to give exim partial DN, the exit code
3607 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3608 But simultaneously result of request was absolutely normal ldap result,
3609 so I produce this patch..."
3611 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3612 REFERENCE, so I have modified the code to exclude the patch when that macro
3615 55. Some experimental protocols are using DNS PTR records for new purposes. The
3616 keys for these records are domain names, not reversed IP addresses. The
3617 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3618 leaves it alone. Component reversal etc. now happens only for IP addresses.
3621 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3623 57. Double the size of the debug message buffer (to 2048) so that more of very
3624 long debug lines gets shown.
3626 58. The exicyclog utility now does better if the number of log files to keep
3627 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3629 59. Two changes related to the smtp_active_hostname option:
3631 (1) $smtp_active_hostname is now available as a variable.
3632 (2) The default for smtp_banner uses $smtp_active_hostname instead
3633 of $primary_hostname.
3635 60. The host_aton() function is supposed to be passed a string that is known
3636 to be a valid IP address. However, in the case of IPv6 addresses, it was
3637 not checking this. This is a hostage to fortune. Exim now panics and dies
3638 if the condition is not met. A case was found where this could be provoked
3639 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3640 components; fortuitously, this particular loophole had already been fixed
3641 by change 4.50/55 above.
3643 If there are any other similar loopholes, the new check in host_aton()
3644 itself should stop them being exploited. The report I received stated that
3645 data on the command line could provoke the exploit when Exim was running as
3646 exim, but did not say which command line option was involved. All I could
3647 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3648 running as the user.
3651 61. There was a buffer overflow vulnerability in the SPA authentication code
3652 (which came originally from the Samba project). I have added a test to the
3653 spa_base64_to_bits() function which I hope fixes it.
3656 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3657 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3659 63. The daemon start-up calls getloadavg() while still root for those OS that
3660 need the first call to be done as root, but it missed one case: when
3661 deliver_queue_load_max is set with deliver_drop_privilege. This is
3662 necessary for the benefit of the queue runner, because there is no re-exec
3663 when deliver_drop_privilege is set.
3665 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3666 This has been fixed.
3668 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3669 was in use, was not putting the data itself into the right store pool;
3670 consequently, it could be overwritten for a subsequent message in the same
3671 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3674 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3676 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3677 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3678 the function were treating the return as a boolean value, which happened to
3679 work because 0=false and not-0=true, but is not correct code.
3681 68. The host_aton() function was not handling scoped IPv6 addresses (those
3682 with, for example, "%eth0" on the end) correctly.
3684 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3685 negated items (that is, ~something) in unsigned ints. Some compilers
3686 apparently mutter when there is no cast.
3688 70. If an address verification called from an ACL failed, and did not produce a
3689 user-specific message (i.e. there was only a "system" message), nothing was
3690 put in $acl_verify_message. In this situation, it now puts the system
3693 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3694 session; change 4.31/43 added the unwanted input to the log line - except
3695 that it did not do this in the start of session case. It now does.
3697 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3698 This is wrong and can cause the other end to generate a synchronization
3699 error if it is another Exim or anything else that does the synchronization
3700 check. A QUIT command is no longer sent after a timeout.
3702 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3703 during host lookups.
3705 74. The defer_ok option of callout verification was not working if it was used
3706 when verifying addresses in header lines, that is, for this case:
3708 verify = header_sender/callout=defer_ok
3710 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3711 those file descriptors could be used for SMTP connections. If anything
3712 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3713 could be sent to the SMTP client, causing chaos. The daemon now opens
3714 stdin, stdout, and stderr to /dev/null when it puts itself into the
3717 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3718 log by default. The user can override this with suitable Perl magic.
3720 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3721 the log message when discard triggers, was not working for the DATA ACL or
3722 for the non-SMTP ACL.
3724 78. Error message wording change in sieve.c.
3726 79. If smtp_accept_max_per_host was set, the number of connections could be
3727 restricted to fewer than expected, because the daemon was trying to set up
3728 a new connection before checking whether the processes handling previous
3729 connections had finished. The check for completed processes is now done
3730 earlier. On busy systems, this bug wouldn't be noticed because something
3731 else would have woken the daemon, and it would have reaped the completed
3734 80. If a message was submitted locally by a user whose login name contained one
3735 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3736 It caused a spool format error. I have fixed the spool reading code. A
3737 related problem was that the "from" clause in the Received: line became
3738 illegal because of the space(s). It is now covered by ${quote_local_part.
3740 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3743 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3745 83. Patch from Sieve maintainer to fix unterminated string problem in
3746 "vacation" handling.
3748 84. Some minor changes to the Linux configuration files to help with other
3749 OS variants using glibc.
3751 85. One more patch for Sieve to update vacation handling to latest spec.
3754 ----------------------------------------------------
3755 See the note above about the 4.44 and 4.50 releases.
3756 ----------------------------------------------------
3762 1. Change 4.43/35 introduced a bug that caused file counts to be
3763 incorrectly computed when quota_filecount was set in an appendfile
3766 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3767 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3770 3. Give more explanation in the error message when the command for a transport
3771 filter fails to execute.
3773 4. There are several places where Exim runs a non-Exim command in a
3774 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3775 was being done only for the command run by the queryprogram router. It is
3776 now done for all such subprocesses. The other cases are: ${run, transport
3777 filters, and the commands run by the lmtp and pipe transports.
3779 5. Some older OS have a limit of 256 on the maximum number of file
3780 descriptors. Exim was using setrlimit() to set 1000 as a large value
3781 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3782 systems. I've change it so that if it can't get 1000, it tries for 256.
3784 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3785 was an oversight, and furthermore, ever since the addition of extra
3786 controls (e.g. 4.43/32), the checks on when to allow different forms of
3787 "control" were broken. There should now be diagnostics for all cases when a
3788 control that does not make sense is encountered.
3790 7. $recipients is now available in the predata ACL (oversight).
3792 8. Tidy the search cache before the fork to do a delivery from a message
3793 received from the command line. Otherwise the child will trigger a lookup
3794 failure and thereby defer the delivery if it tries to use (for example) a
3795 cached ldap connection that the parent has called unbind on.
3797 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3798 of $address_data from the recipient verification was clobbered by the
3799 sender verification.
3801 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3802 was its contents. (It was OK if the option was not defined at all.)
3804 11. A "Completed" log line is now written for messages that are removed from
3805 the spool by the -Mrm option.
3807 12. $host_address is now set to the target address during the checking of
3808 ignore_target_hosts.
3810 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3811 being passed; this would have caused $sender_host_name to have been used if
3812 matching the list had actually called for a host name (not very likely,
3813 since this list is usually IP addresses). A host name is now passed as
3816 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3817 code that specifies a non-restarting handler (typically sigaction() in
3818 modern systems) in an attempt to fix a rare and obscure crash bug.
3820 15. Narrowed the window for a race in the daemon that could cause it to ignore
3821 SIGCHLD signals. This is not a major problem, because they are used only to
3822 wake it up if nothing else does.
3824 16. A malformed maildirsize file could cause Exim to calculate negative values
3825 for the mailbox size or file count. Odd effects could occur as a result.
3826 The maildirsize information is now recalculated if the size or filecount
3829 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3830 support for a long time. Removed HAVE_SYS_VFS_H.
3832 18. Updated exipick to current release from John Jetmore.
3834 19. Allow an empty sender to be matched against a lookup in an address list.
3835 Previously the only cases considered were a regular expression, or an
3838 20. Exim went into a mad DNS lookup loop when doing a callout where the
3839 host was specified on the transport, if the DNS lookup yielded more than
3842 21. The RFC2047 encoding function was originally intended for short strings
3843 such as real names; it was not keeping to the 75-character limit for
3844 encoded words that the RFC imposes. It now respects the limit, and
3845 generates multiple encoded words if necessary. To be on the safe side, I
3846 have increased the buffer size for the ${rfc2047: expansion operator from
3849 22. Failure to deliver a bounce message always caused it to be frozen, even if
3850 there was an errors_to setting on the router. The errors_to setting is now
3853 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3854 canonical form (fully expanded) before being placed in
3855 $sender_host_address.
3857 24. Updated eximstats to version 1.33
3859 25. Include certificate and key file names in error message when GnuTLS fails
3860 to set them up, because the GnuTLS error message doesn't include the name
3861 of the failing file when there is a problem reading it.
3863 26. Expand error message when OpenSSL has problems setting up cert/key files.
3866 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3867 (this can affect the format of dates).
3869 28. exim_tidydb, when checking for the continued existence of a message for
3870 which it has found a message-specific retry record, was not finding
3871 messages that were in split spool directories. Consequently, it was
3872 deleting retry records that should have stayed in existence.
3874 29. eximstats updated to version 1.35
3875 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3876 1.35 - bugfix such that pie charts by volume are generated correctly
3878 30. The SPA authentication driver was not abandoning authentication and moving
3879 on to the next authenticator when an expansion was forced to fail,
3880 contradicting the general specification for all authenticators. Instead it
3881 was generating a temporary error. It now behaves as specified.
3883 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3884 (the order specifies the preference for clients). The order is now AES256,
3885 AES128, 3DES, ARCFOUR128.
3887 31. Small patch to Sieve code - explicitly set From: when generating an
3890 32. Exim crashed if a remote delivery caused a very long error message to be
3891 recorded - for instance if somebody sent an entire SpamAssassin report back
3892 as a large number of 550 error lines. This bug was coincidentally fixed by
3893 increasing the size of one of Exim's internal buffers (big_buffer) that
3894 happened as part of the Exiscan merge. However, to be on the safe side, I
3895 have made the code more robust (and fixed the comments that describe what
3898 33. Some experimental protocols are using DNS PTR records for new purposes. The
3899 keys for these records are domain names, not reversed IP addresses. The
3900 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3901 leaves it alone. Component reversal etc. now happens only for IP addresses.
3904 34. The host_aton() function is supposed to be passed a string that is known
3905 to be a valid IP address. However, in the case of IPv6 addresses, it was
3906 not checking this. This is a hostage to fortune. Exim now panics and dies
3907 if the condition is not met. A case was found where this could be provoked
3908 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3909 components; fortuitously, this particular loophole had already been fixed
3910 by change 4.50/55 or 4.44/33 above.
3912 If there are any other similar loopholes, the new check in host_aton()
3913 itself should stop them being exploited. The report I received stated that
3914 data on the command line could provoke the exploit when Exim was running as
3915 exim, but did not say which command line option was involved. All I could
3916 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3917 running as the user.
3920 35. There was a buffer overflow vulnerability in the SPA authentication code
3921 (which came originally from the Samba project). I have added a test to the
3922 spa_base64_to_bits() function which I hope fixes it.
3925 36. The daemon start-up calls getloadavg() while still root for those OS that
3926 need the first call to be done as root, but it missed one case: when
3927 deliver_queue_load_max is set with deliver_drop_privilege. This is
3928 necessary for the benefit of the queue runner, because there is no re-exec
3929 when deliver_drop_privilege is set.
3931 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3932 was in use, was not putting the data itself into the right store pool;
3933 consequently, it could be overwritten for a subsequent message in the same
3934 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3937 38. Sometimes the final signoff response after QUIT could fail to get
3938 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3939 < 0 before doing a fflush(). This bug looks as though it goes back to the
3940 introduction of TLS in release 3.20, but "sometimes" must have been rare
3941 because the tests only now provoked it.
3947 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3948 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3949 Now it returns an integer. A number of calls were still expecting a T/F
3950 return. Fortuitously, in all cases, the tests worked in OK situations,
3951 which is the norm. However, things would have gone wrong on any write
3952 failures on the smtp file descriptor. This function is used when sending
3953 messages over SMTP and also when doing verify callouts.
3955 2. When Exim is called to do synchronous delivery of a locally submitted
3956 message (the -odf or -odi options), it no longer closes stderr before doing
3959 3. Implemented the mua_wrapper option.
3961 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3963 5. Implemented the functions header_remove(), header_testname(),
3964 header_add_at_position(), and receive_remove_recipient(), and exported them
3967 6. If an ACL "warn" statement specified the addition of headers, Exim already
3968 inserted X-ACL-Warn: at the start if there was no header name. However, it
3969 was not making this test for the second and subsequent header lines if
3970 there were newlines in the string. This meant that an invalid header could
3971 be inserted if Exim was badly configured.
3973 7. Allow an ACL "warn" statement to add header lines at the start or after all
3974 the Received: headers, as well as at the end.
3976 8. Added the rcpt_4xx retry error code.
3978 9. Added postmaster_mailfrom=xxx to callout verification option.
3980 10. Added mailfrom=xxxx to the callout verification option, for verify=
3983 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3984 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3986 12. Inserted some casts to stop certain compilers warning when using pointer
3987 differences as field lengths or precisions in printf-type calls (mostly
3988 affecting debugging statements).
3990 13. Added optional readline() support for -be (dynamically loaded).
3992 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3993 same clock tick as a message's arrival, so that its received time was the
3994 same as the "first fail" time on the retry record, and that message
3995 remained on the queue past the ultimate address timeout, every queue runner
3996 would try a delivery (because it was past the ultimate address timeout) but
3997 after another failure, the ultimate address timeout, which should have then
3998 bounced the address, did not kick in. This was a "< instead of <=" error;
3999 in most cases the first failure would have been in the next clock tick
4000 after the received time, and all would be well.
4002 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4003 being recognized when the domain list was tested by the match_domain
4004 condition in an expansion string.
4006 16. Added the ${str2b64: operator.
4008 17. Exim was always calling setrlimit() to set a large limit for the number of
4009 processes, without checking whether the existing limit was already
4010 adequate. (It did check for the limit on file descriptors.) Furthermore,
4011 errors from getrlimit() and setrlimit() were being ignored. Now they are
4012 logged to the main and panic logs, but Exim does carry on, to try to do its
4013 job under whatever limits there are.
4015 18. Imported PCRE 5.0.
4017 19. Trivial typo in log message " temporarily refused connection" (the leading
4020 20. If the log selector return_path_on_delivery was set and an address was
4021 redirected to /dev/null, the delivery process crashed because it assumed
4022 that a return path would always be set for a "successful" delivery. In this
4023 case, the whole delivery is bypassed as an optimization, and therefore no
4026 21. Internal re-arrangement: the function for sending a challenge and reading
4027 a response while authentication was assuming a zero-terminated challenge
4028 string. It's now changed to take a pointer and a length, to allow for
4029 binary data in such strings.
4031 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4033 23. Exim was not respecting finduser_retries when seeking the login of the
4034 uid under which it was called; it was always trying 10 times. (The default
4035 setting of finduser_retries is zero.) Also, it was sleeping after the final
4036 failure, which is pointless.
4038 24. Implemented tls_on_connect_ports.
4040 25. Implemented acl_smtp_predata.
4042 26. If the domain in control=submission is set empty, Exim assumes that the
4043 authenticated id is a complete email address when it generates From: or
4044 Sender: header lines.
4046 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4047 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4048 chown and chgrp in /bin and hostname in /usr/bin.
4050 28. Exim was keeping the "process log" file open after each use, just as it
4051 does for the main log. This opens the possibility of it remaining open for
4052 long periods when the USR1 signal hits a daemon. Occasional processlog
4053 errors were reported, that could have been caused by this. Anyway, it seems
4054 much more sensible not to leave this file open at all, so that is what now
4057 29. The long-running daemon process does not normally write to the log once it
4058 has entered its main loop, and it closes the log before doing so. This is
4059 so that log files can straightforwardly be renamed and moved. However,
4060 there are a couple of unusual error situations where the daemon does write
4061 log entries, and I had neglected to close the log afterwards.
4063 30. The text of an SMTP error response that was received during a remote
4064 delivery was being truncated at 512 bytes. This is too short for some of
4065 the long messages that one sometimes sees. I've increased the limit to
4068 31. It is now possible to make retry rules that apply only when a message has a
4069 specific sender, in particular, an empty sender.
4071 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4072 it possible to be selective about when SMTP synchronization is enforced.
4074 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4076 32. Implemented hosts_connection_nolog.
4078 33. Added an ACL for QUIT.
4080 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4083 35. Added mailbox_size and mailbox_filecount to appendfile.
4085 36. Added control = no_multiline_responses to ACLs.
4087 37. There was a bug in the logic of the code that waits for the clock to tick
4088 in the case where the clock went backwards by a substantial amount such
4089 that the microsecond fraction of "now" was more than the microsecond
4090 fraction of "then" (but the whole seconds number was less).
4092 38. Added support for the libradius Radius client library this is found on
4093 FreeBSD (previously only the radiusclient library was supported).
4099 1. When certain lookups returned multiple values in the form name=value, the
4100 quoting of the values was not always being done properly. Specifically:
4101 (a) If the value started with a double quote, but contained no whitespace,
4103 (b) If the value contained whitespace other than a space character (i.e.
4104 tabs or newlines or carriage returns) it was not quoted.
4105 This fix has been applied to the mysql and pgsql lookups by writing a
4106 separate quoting function and calling it from the lookup code. The fix
4107 should probably also be applied to nisplus, ibase and oracle lookups, but
4108 since I cannot test any of those, I have not disturbed their existing code.
4110 2. A hit in the callout cache for a specific address caused a log line with no
4111 reason for rejecting RCPT. Now it says "Previous (cached) callout
4112 verification failure".
4114 3. There was an off-by-one bug in the queryprogram router. An over-long
4115 return line was truncated at 256 instead of 255 characters, thereby
4116 overflowing its buffer with the terminating zero. As well as fixing this, I
4117 have increased the buffer size to 1024 (and made a note to document this).
4119 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4120 when Exim is waiting for an SMTP response from a remote server, Exim
4121 restarts its select() call on the socket, thereby resetting its timeout.
4122 This is not a problem when such interrupts are rare. Somebody set up a cron
4123 job to run exiwhat every 2 minutes, which is less than the normal select()
4124 timeout (5 or 10 minutes). This meant that the select() timeout never
4125 kicked in because it was always reset. I have fixed this by comparing the
4126 time when an interrupt arrives with the time at the start of the first call
4127 to select(). If more time than the timeout has elapsed, the interrupt is
4128 treated as a timeout.
4130 5. Some internal re-factoring in preparation for the addition of Sieve
4131 extensions (by MH). In particular, the "personal" test is moved to a
4132 separate function, and given an option for scanning Cc: and Bcc: (which is
4133 not set for Exim filters).
4135 6. When Exim created an email address using the login of the caller as the
4136 local part (e.g. when creating a From: or Sender: header line), it was not
4137 quoting the local part when it contained special characters such as @.
4139 7. Installed new OpenBSD configuration files.
4141 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4142 try to make them clearer.
4144 9. Callout options, other than the timeout value, were being ignored when
4145 verifying sender addresses in header lines. For example, when using
4147 verify = header_sender/callout=no_cache
4149 the cache was (incorrectly) being used.
4151 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4152 only the Cygwin environment.
4154 11. When return_path_on_delivery was set as a log selector, if different remote
4155 addresses in the same message used different return paths and parallel
4156 remote delivery occurred, the wrong values would sometimes be logged.
4157 (Whenever a remote delivery process finished, the return path value from
4158 the most recently started remote delivery process was logged.)
4160 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4161 lines when AUTH and/or TLS are in use. This is the "received protocol"
4162 field. Exim used to use "asmtp" for authenticated SMTP, without any
4163 indication (in the protocol name) for TLS use. Now it follows the RFC and
4164 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4165 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4166 names appear in log lines as well as in Received: header lines.
4168 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4169 extensions, and comparison tests, and to fix some bugs.
4171 14. Changes to the "personal" filter test:
4173 (1) The test was buggy in that it was just doing the equivalent of
4174 "contains" tests on header lines. For example, if a user's address was
4175 anne@some.where, the "personal" test would incorrectly be true for
4177 To: susanne@some.where
4179 This test is now done by extracting each address from the header in turn,
4180 and checking the entire address. Other tests that are part of "personal"
4181 are now done using regular expressions (for example, to check local parts
4182 of addresses in From: header lines).
4184 (2) The list of non-personal local parts in From: addresses has been
4185 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4186 taken from the Sieve specification recommendations.
4188 (3) If the message contains any header line starting with "List-" it is
4189 treated as non-personal.
4191 (4) The test for "circular" in the Subject: header line has been removed
4192 because it now seems ill-conceived.
4194 15. Minor typos in src/EDITME comments corrected.
4196 16. Installed latest exipick from John Jetmore.
4198 17. If headers_add on a router specified a text string that was too long for
4199 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4200 of string_sprintf() is now avoided.
4202 18. $message_body_size was not set (it was always zero) when running the DATA
4203 ACL and the local_scan() function.
4205 19. For the "mail" command in an Exim filter, no default was being set for
4206 the once_repeat time, causing a random time value to be used if "once" was
4207 specified. (If the value happened to be <= 0, no repeat happened.) The
4208 default is now 0s, meaning "never repeat". The "vacation" command was OK
4209 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4210 (I found it when inspecting the code).
4212 20. There is now an overall timeout for performing a callout verification. It
4213 defaults to 4 times the callout timeout, which applies to individual SMTP
4214 commands during the callout. The overall timeout applies when there is more
4215 than one host that can be tried. The timeout is checked before trying the
4216 next host. This prevents very long delays if there are a large number of
4217 hosts and all are timing out (e.g. when the network connections are timing
4218 out). The value of the overall timeout can be changed by specifying an
4219 additional sub-option for "callout", called "maxwait". For example:
4221 verify = sender/callout=5s,maxwait=20s
4223 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4224 to the end before writing, but this should make it even safer).
4226 22. Exim was forgetting that it had advertised PIPELINING for the second and
4227 subsequent messages on an SMTP connection. It was also not resetting its
4228 memory on STARTTLS and an internal HELO.
4230 23. When Exim logs an SMTP synchronization error within a session, it now
4231 records whether PIPELINING has been advertised or not.
4233 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4234 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4235 rather than long int.
4237 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4239 26. Added the never_mail option to autoreply.
4245 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4246 crash if the getsockname() call failed; this can happen if a connection is
4247 closed very soon after it is established. The problem was simply in the
4248 order in which certain operations were done, causing Exim to try to write
4249 to the SMTP stream before it had set up the file descriptor. The bug has
4250 been fixed by making things happen in the correct order.
4256 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4257 before the connection was closed, thus losing the rejection response.
4259 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4260 some early Solaris releases, but causes trouble in current releases where
4261 socklen_t is defined.
4263 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4266 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4269 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4270 that releasing the top of it at the end releases what was used for sub-
4271 expansions (unless the block got too big). However, discard this block if
4272 the first thing is a variable or header, so that we can use its block when
4273 it is dynamic (useful for very large $message_headers, for example).
4275 6. Lookups now cache *every* query, not just the most recent. A new, separate
4276 store pool is used for this. It can be recovered when all lookup caches are
4277 flushed. Lookups now release memory at the end of their result strings.
4278 This has involved some general refactoring of the lookup sources.
4280 7. Some code has been added to the store_xxx() functions to reduce the amount
4281 of flapping under certain conditions.
4283 8. log_incoming_interface used to affect only the <= reception log lines. Now
4284 it causes the local interface and port to be added to several more SMTP log
4285 lines, for example "SMTP connection from", and rejection lines.
4287 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4289 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4291 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4292 could be overwritten at the end of the current message (or the start of a
4293 new message if it was set in a HELO ACL). The value is now preserved for
4294 the duration of the SMTP connection.
4296 12. If a transport had a headers_rewrite setting, and a matching header line
4297 contained an unqualified address, that address was qualified, even if it
4298 did not match any rewriting rules. The underlying bug was that the values
4299 of the flags that permit the existence of unqualified sender and recipient
4300 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4301 non-local messages, and by -bnq for local messages) were not being
4302 preserved with the message after it was received.
4304 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4305 "next input=" as part of the text comprising the host identity instead of
4306 the correct text. The code was using the same buffer for two different
4307 strings. However, depending on which order the printing function evaluated
4308 its arguments, the bug did not always show up. Under Linux, for example, my
4309 test suite worked just fine.
4311 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4312 doesn't work with some older versions of Perl. It has been changed to "my",
4313 which in any case is probably the better facility to use.
4315 15. A really picky compiler found some instances of statements for creating
4316 error messages that either had too many or two few arguments for the format
4319 16. The size of the buffer for calls to the DNS resolver has been increased
4320 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4321 for addresses that have a lot of PTR records. This alleviates a problem; it
4322 does not fully solve it.
4324 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4325 buffer now truncates the list and logs the incident, which is the same
4326 action as happens when Exim is looking up a host name and its aliases.
4327 Previously in this situation something unpredictable would happen;
4328 sometimes it was "internal error: store_reset failed".
4330 18. If a server dropped the connection unexpectedly when an Exim client was
4331 using GnuTLS and trying to read a response, the client delivery process
4332 crashed while trying to generate an error log message.
4334 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4335 string, for example:
4337 warn message = H1: something\nH2: something
4339 the text was added as a single header line from Exim's point of view
4340 though it ended up OK in the delivered message. However, searching for the
4341 second and subsequent header lines using $h_h2: did not work. This has been
4342 fixed. Similarly, if a system filter added multiple headers in this way,
4343 the routers could not see them.
4345 20. Expanded the error message when iplsearch is called with an invalid key to
4346 suggest using net-iplsearch in a host list.
4348 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4349 ACLs are no longer actually imposed (and a message to that effect is
4352 22. If a "gecos" field in a passwd entry contained escaped characters, in
4353 particular, if it contained a \" sequence, Exim got it wrong when building
4354 a From: or a Sender: header from that name. A second bug also caused
4355 incorrect handling when an unquoted " was present following a character
4356 that needed quoting.
4358 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4359 was not being matched caselessly.
4361 24. Arranged for all hyphens in the exim.8 source to be escaped with
4364 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4365 properly. Recipient callout cache records were still being keyed to include
4366 the sender, even when use_sender was set false. This led to far more
4367 callouts that were necessary. The sender is no longer included in the key
4368 when use_sender is false.
4370 26. Added "control = submission" modifier to ACLs.
4372 27. Added the ${base62d: operator to decode base 62 numbers.
4374 28. dnsdb lookups can now access SRV records.
4376 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4377 the configuration file.
4379 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4380 (-v) mode. This makes the output for a verbose queue run more intelligible.
4382 31. Added a use_postmaster feature to recipient callouts.
4384 32. Added the $body_zerocount variable, containing the number of binary zero
4385 bytes in the message body.
4387 33. The time of last modification of the "new" subdirectory is now used as the
4388 "mailbox time last read" when there is a quota error for a maildir
4391 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4393 35. Added +ignore_unknown as a special item in host lists.
4395 36. Code for decoding IPv6 addresses in host lists is now included, even if
4396 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4397 address was recognized as an IP address, but was then not correctly decoded
4398 into binary, causing unexpected and incorrect effects when compared with
4405 1. Very minor rewording of debugging text in manualroute to say "list of
4406 hosts" instead of "hostlist".
4408 2. If verify=header_syntax was set, and a header line with an unqualified
4409 address (no domain) and a large number of spaces between the end of the
4410 name and the colon was received, the reception process suffered a buffer
4411 overflow, and (when I tested it) crashed. This was caused by some obsolete
4412 code that should have been removed. The fix is to remove it!
4414 3. When running in the test harness, delay a bit after writing a bounce
4415 message to get a bit more predictability in the log output.
4417 4. Added a call to search_tidyup() just before forking a reception process. In
4418 theory, someone could use a lookup in the expansion of smtp_accept_max_
4419 per_host which, without the tidyup, could leave open a database connection.
4421 5. Added the variables $recipient_data and $sender_data which get set from a
4422 lookup success in an ACL "recipients" or "senders" condition, or a router
4423 "senders" option, similar to $domain_data and $local_part_data.
4425 6. Moved the writing of debug_print from before to after the "senders" test
4428 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4429 problems for message scanning, either using a data ACL, or using
4430 local_scan() because the Received: header was not generated till after they
4431 were called (in order to set the time as the time of reception completion).
4432 I have revised the way this works. The header is now generated after the
4433 body is received, but before the ACL or local_scan() are called. After they
4434 are run, the timestamp in the header is updated.
4440 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4441 before starting a queue runner without re-exec. This happened only when
4442 deliver_drop_privilege was set or when the Exim user was set to root. The
4443 effect of the bug was that timeouts during subsequent deliveries caused
4444 crashes instead of being properly handled. The handler is now left at its
4445 default (and expected) setting.
4447 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4448 message, again when deliver_drop_privilege is set or Exim is run as root.
4449 The bug described in (1) was not present in this case, but the tidying up
4450 of the other signals was missing. I have made the two cases consistent.
4452 3. The ignore_target_hosts setting on a manualroute router was being ignored
4453 for hosts that were looked up using the /MX notation.
4455 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4458 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4459 operated on the sender address. After changing the $sender_address to <>
4460 for the sender address verify, Exim was re-instated it as the original
4461 (before rewriting) address, but remembering that it had rewritten it, so it
4462 wasn't rewriting it again. This bug also had the effect of breaking the
4463 sender address verification caching when the sender address was rewritten.
4465 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4466 This has been changed so that if the ip literal address matches
4467 ignore_target_hosts, the router declines.
4469 7. Added expansion conditions match_domain, match_address, and match_local_
4470 part (NOT match_host).
4472 8. The placeholder for the Received: header didn't have a length field set.
4474 9. Added code to Exim itself and to exim_lock to test for a specific race
4475 condition that could lead to file corruption when using MBX delivery. The
4476 issue is with the lockfile that is created in /tmp. If this file is removed
4477 after a process has opened it but before that process has acquired a lock,
4478 there is the potential for a second process to recreate the file and also
4479 acquire a lock. This could lead to two Exim processes writing to the file
4480 at the same time. The added code performs the same test as UW imapd; it
4481 checks after acquiring the lock that its file descriptor still refers to
4482 the same named file.
4484 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4485 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4488 11. Added the smtp_active_hostname option. If used, this will typically be made
4489 to depend on the incoming interface address. Because $interface_address is
4490 not set up until the daemon has forked a reception process, error responses
4491 that can happen earlier (such as "too many connections") no longer contain
4494 12. If an expansion in a condition on a "warn" statement fails because a lookup
4495 defers, the "warn" statement is abandoned, and the next ACL statement is
4496 processed. Previously this caused the whole ACL to be aborted.
4498 13. Added the iplsearch lookup type.
4500 14. Added ident_timeout as a log selector.
4502 15. Added tls_certificate_verified as a log selector.
4504 16. Added a global option tls_require_ciphers (compare the smtp transport
4505 option of the same name). This controls incoming TLS connections.
4507 17. I finally figured out how to make tls_require_ciphers do a similar thing
4508 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4509 before starting the TLS session.
4511 18. Tabs are now shown as \t in -bP output.
4513 19. If the log selector return_path_on_delivery was set, Exim crashed when
4514 bouncing a message because it had too many Received: header lines.
4516 20. If two routers both had headers_remove settings, and the first one included
4517 a superfluous trailing colon, the final name in the first list and the
4518 first name in the second list were incorrectly joined into one item (with a
4519 colon in the middle).
4525 1. Added -C and -D options to the exinext utility, mainly to make it easier
4526 to include in the automated testing, but these could be helpful when
4527 multiple configurations are in use.
4529 2. The exinext utility was not formatting the output nicely when there was
4530 an alternate port involved in the retry record key, nor when there was a
4531 message id as well (for retries that were specific to a specific message
4532 and a specific host). It was also confused by IPv6 addresses, because of
4533 the additional colons they contain. I have fixed the IPv4 problem, and
4534 patched it up to do a reasonable job for IPv6.
4536 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4537 delivery, the log line now contains "pipelined" if PIPELINING was used.
4539 4. An SMTP transport process used to panic and die if the bind() call to set
4540 an explicit outgoing interface failed. This has been changed; it is now
4541 treated in the same way as a connect() failure.
4543 5. A reference to $sender_host_name in the part of a conditional expansion
4544 that was being skipped was still causing a DNS lookup. This no longer
4547 6. The def: expansion condition was not recognizing references to header lines
4548 that used bh_ and bheader_.
4550 7. Added the _cache feature to named lists.
4552 8. The code for checking quota_filecount in the appendfile transport was
4553 allowing one more file than it should have been.
4555 9. For compatibility with Sendmail, the command line option
4563 and sets the incoming protocol and host name (for trusted callers). The
4564 host name and its colon can be omitted when only the protocol is to be set.
4565 Note the Exim already has two private options, -pd and -ps, that refer to
4566 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4567 "s", but I don't think that's a major issue.
4569 10. A number of refactoring changes to the code, none of which should affect
4572 (a) The number of logging options was getting close to filling up the
4573 32-bit word that was used as a bit map. I have split them into two classes:
4574 those that are passed in the argument to log_write(), and those that are
4575 only ever tested independently outside of that function. These are now in
4576 separate 32-bit words, so there is plenty of room for expansion again.
4577 There is no change in the user interface or the logging behaviour.
4579 (b) When building, for example, log lines, the code previously used a
4580 macro that called string_cat() twice, in order to add two strings. This is
4581 not really sufficiently general. Furthermore, there was one instance where
4582 it was actually wrong because one of the argument was used twice, and in
4583 one call a function was used. (As it happened, calling the function twice
4584 did not affect the overall behaviour.) The macro has been replaced by a
4585 function that can join an arbitrary number of extra strings onto a growing
4588 (c) The code for expansion conditions now uses a table and a binary chop
4589 instead of a serial search (which was left over from when there were very
4590 few conditions). Also, it now recognizes conditions like "pam" even when
4591 the relevant support is not compiled in: a suitably worded error message is
4592 given if an attempt is made to use such a condition.
4594 11. Added ${time_interval:xxxxx}.
4596 12. A bug was causing one of the ddress fields not to be passed back correctly
4597 from remote delivery subprocesses. The field in question was not being
4598 subsequently used, so this caused to problems in practice.
4600 13. Added new log selectors queue_time and deliver_time.
4602 14. Might have fixed a bug in maildirsizefile handling that threw up
4603 "unexpected character" debug warnings, and recalculated the data
4604 unnecessarily. In any case, I expanded the warning message to give more
4607 15. Added the message "Restricted characters in address" to the statements in
4608 the default ACL that block characters like @ and % in local parts.
4610 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4611 Three changes have been made:
4613 (a) There was a serious bug; a negative response to MAIL caused the whole
4614 recipient domain to be cached as invalid, thereby blocking all messages
4615 to all local parts at the same domain, from all senders. This bug has
4616 been fixed. The domain is no longer cached after a negative response to
4617 MAIL if the sender used is not empty.
4619 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4622 (c) A new callout option, "use_sender" has been added for people who want
4623 the modified behaviour.
4629 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4632 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4633 indeed breaks things for older releases.
4635 3. Added additional logging to the case where there is a problem reading data
4636 from a filter that is running in a subprocess using a pipe, in order to
4637 try to track down a specific problem.
4639 4. Testing facility fudge: when running in the test harness and attempting
4640 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4641 getting "No route to host". Convert this to a timeout.
4643 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4646 6. Some OS don't have socklen_t but use size_t instead. This affects the
4647 fifth argument of getsockopt() amongst other things. This is now
4648 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4649 can be set for individual OS. I have set it for SunOS5, OSF1, and
4650 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4651 some earlier ones do not.
4653 7. Change 4.30/15 was not doing the test caselessly.
4655 8. The standard form for an IPv6 address literal was being rejected by address
4656 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4657 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4658 this, as well as the form without the "IPv6" on the front (but only when
4659 address literals are enabled, of course).
4661 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4663 10. Exim crashed if a message with an empty sender address specified by -f
4664 encountered a router with an errors_to setting. This could be provoked only
4665 by a command such as
4669 where an empty string was supplied; "<>" did not hit this bug.
4671 11. Installed PCRE release 4.5.
4673 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4674 remained set. It is now erased.
4676 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4677 times from message ids (which are base 36 rather than the normal 62).
4679 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4680 were being counted as actual protocol errors, and logged if the log
4681 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4682 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4683 and DATA following a set of rejected RCPTs do not count as protocol errors.
4684 In other words, Exim assumes they were pipelined, though this may not
4685 actually be the case. Of course, in all cases the client gets an
4686 appropriate error code.
4688 15. If a lookup fails in an ACL condition, a message about the failure may
4689 be available; it is used if testing the ACL cannot continue, because most
4690 such messages specify what the cause of the deferral is. However, some
4691 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4692 that caused an old message to be retained and used if a later statement
4693 caused a defer, replacing the real cause of the deferral.
4695 16. If an IP address had so many PTR records that the DNS lookup buffer
4696 was not large enough to hold them, Exim could crash while trying to process
4697 the truncated data. It now detects and logs this case.
4699 17. Further to 4.21/58, another change has been made: if (and only if) the
4700 first line of a message (the first header line) ends with CRLF, a bare LF
4701 in a subsequent header line has a space inserted after it, so as not to
4702 terminate the header.
4704 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4705 unnecessarily, used atoi() instead of strtol(), and didn't check the
4706 termination when getting file sizes from file names by regex.
4708 19. Completely re-implemented the support for maildirsize files, in the light
4709 of a number of problems with the previous contributed implementation
4710 (4.30/29). In particular:
4712 . If the quota is zero, the maildirsize file is maintained, but no quota is
4715 . If the maildir directory does not exist, it is created before any attempt
4716 to write a maildirsize file.
4718 . The quota value in the file is just a cache; if the quota is changed in
4719 the transport, the new value overrides.
4721 . A regular expression is available for excluding directories from the
4724 20. The autoreply transport checks the characters in options that define the
4725 message's headers; it allows continued headers, but it was checking with
4726 isspace() after an embedded newline instead of explicitly looking for a
4729 21. If all the "regular" hosts to which an address was routed had passed their
4730 expiry times, and had not reached their retry times, the address was
4731 bounced, even if fallback hosts were defined. Now Exim should go on to try
4734 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4735 equivalent code in the SMTP transport. Some hosts send humungous responses
4736 to HELO/EHLO, more than 1024 it seems.
4738 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4739 gives compiler warnings in some environments. I've now done it "properly",
4742 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4743 (because of problems with the built-in one) was declared to return uschar *
4744 instead of char *, causing compiler failure.
4746 25. Fixed a file descriptor leak when processing alias/forward files.
4748 26. Fixed a minor format string issue in dbfn.c.
4750 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4752 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4753 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4756 29. When a sender address is verified, it is cached, to save repeating the test
4757 when there is more than one recipient in a message. However, when the
4758 verification involves a callout, it is possible for different callout
4759 options to be set for different recipients. It is too complicated to keep
4760 track of this in the cache, so now Exim always runs a verification when a
4761 callout is required, relying on the callout cache for the optimization.
4762 The overhead is duplication of the address routing, but this should not be
4765 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4766 to be verified with callout=postmaster, and the main callout worked but the
4767 postmaster check failed, the verification correctly failed. However, if a
4768 subsequent RCPT command asked for sender verification *without* the
4769 postmaster check, incorrect caching caused this verification also to fail,
4772 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4773 it was not caching the DNS options (qualify_single, search_parents) that
4774 were used when the lookup failed. A subsequent lookup with different
4775 options therefore always gave the same answer, though there were cases
4776 where it should not have. (Example: a "domains = !$mx_any" option on a
4777 dnslookup router: the "domains" option is always processed without any
4778 widening, but the router might have qualify_single set.) Now Exim uses the
4779 cached value only when the same options are set.
4781 32. Added John Jetmore's "exipick" utility to the distribution.
4783 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4784 than a timeout (e.g. a certificate is required, and is not provided), an
4785 Exim server now closes the connection immediately. Previously it waited for
4786 the client to close - but if the client is SSL, it seems that they each
4787 wait for each other, leading to a delay before one of them times out.
4789 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4790 maintained 0.8.x compatibility because I don't think many are using it, and
4791 it is clearly obsolete.
4793 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4796 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4797 client certificate was expired. A simple patch fixes this, though I don't
4798 understand the full logic of why the verify callback is called multiple
4801 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4802 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4803 which causes problems with some clients (such as the Certicom SSL Plus
4804 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4805 disables the coutermeasure allowing Eudora to connect."
4807 38. Exim was not checking that a write() to a log file succeeded. This could
4808 lead to Bad Things if a log got too big, in particular if it hit a file
4809 size limit. Exim now panics and dies if it cannot write to a log file, just
4810 as it does if it cannot open a log file.
4812 39. Modified OS/Makefile-Linux so that it now contains
4814 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4816 The two -D definitions ensure that Exim is compiled with large file
4817 support, which makes it possible to handle log files that are bigger than
4820 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4821 instance) a domain was checked against a named list that involved a lookup,
4822 causing $domain_data to be set, then another domain was checked against the
4823 same list, then the first domain was re-checked, the value of $domain_data
4824 after the final check could be wrong. In particular, if the second check
4825 failed, it could be set empty. This bug probably also applied to
4828 41. The strip_trailing_dot option was not being applied to the address given
4829 with the -f command-line option.
4831 42. The code for reading a message's header from the spool was incrementing
4832 $received_count, but never initializing it. This meant that the value was
4833 incorrect (doubled) while delivering a message in the same process in which
4834 it was received. In the most common configuration of Exim, this never
4835 happens - a fresh exec is done - but it can happen when
4836 deliver_drop_privilege is set.
4838 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4839 it now includes up to 150 characters of the unexpected data in the log
4842 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4843 and building data strings. The size of both of these buffers was 10 000
4844 bytes - far larger than anybody would *ever* want, thought I. Needless to
4845 say, somebody hit the limit. I have increased the maximum line length to
4846 20 000 and the maximum data length of concatenated lines to 100 000. I have
4847 also fixed two bugs, because there was no checking on these buffers. Tsk,
4848 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4849 buffer is too small.
4851 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4852 lsearch lookups. Now it does.
4854 46. When parsing a route_list item in a manualroute router, a fixed-length
4855 buffer was used for the list of hosts. I made this 1024 bytes long,
4856 thinking that nobody would ever have a list of hosts that long. Wrong.
4857 Somebody had a whole pile of complicated expansion conditions, and the
4858 string was silently truncated, leading to an expansion error. It turns out
4859 that it is easier to change to an unlimited length (owing to other changes
4860 that have happened since this code was originally written) than to build
4861 structure for giving a limitation error. The length of the item that
4862 expands into the list of hosts is now unlimited.
4864 47. The lsearch lookup could not handle data where the length of text line was
4865 more than 4095 characters. Such lines were truncated, leading to shortened
4866 data being returned. It should now handle lines of any length.
4868 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4869 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4872 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4873 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4874 no longer happens in comment lines. A list of replacements is now placed
4875 at the head of all of the source files, except those whose only change is
4876 to replace PERL_COMMAND in the very first #! line.
4878 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4879 messages on the queue, with a bottom-up merge sort, using code contributed
4880 by Michael Haardt. This should make operations like -bp somewhat faster on
4881 large queues. It won't affect queue runners, except when queue_run_in_order
4884 51. Installed eximstats 1.31 in the distribution.
4886 52. Added support for SRV lookups to the dnslookup router.
4888 53. If an ACL referred to $message_body or $message_body_end, the value was not
4889 reset for any messages that followed in the same SMTP session.
4891 54. The store-handling optimization for building very long strings was not
4892 differentiating between the different store pools. I don't think this
4893 actually made any difference in practice, but I've tidied it.
4895 55. While running the routers to verify a sender address, $sender_address
4896 was still set to the sender address. This is wrong, because when routing to
4897 send a bounce to the sender, it would be empty. Therefore, I have changed
4898 it so that, while verifying a sender address, $sender_address is set to <>.
4899 (There is no change to what happens when verifying a recipient address.)
4901 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4902 target A or AAAA records (if not already returned) without resetting the
4903 qualify_single or search_parents options of the DNS resolver. These are
4904 inappropriate in this case because the targets of MX and SRV records must
4905 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4906 target that, when qualified, matched something in the local domain. These
4907 two options are now turned off when doing these lookups.
4909 57. It seems that at least some releases of Reiserfs (which does not have the
4910 concept of a fixed number of inodes) returns zero and not -1 for the
4911 number of available inodes. This interacted badly with check_spool_inodes,
4912 which assumed that -1 was the "no such thing" setting. What I have done is
4913 to check that the total number of inodes is greater than zero before doing
4914 the test of how many are available.
4916 58. When a "warn" ACL statement has a log_message modifier, the message is
4917 remembered, and not repeated. This is to avoid a lot of repetition when a
4918 message has many recipients that cause the same warning to be written.
4919 Howewer, Exim was preserving the list of already written lines for an
4920 entire SMTP session, which doesn't seem right. The memory is now reset if a
4921 new message is started.
4923 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4924 debugging output unless log_rewrite was also set.
4926 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4927 of (int)(handle) when we know that handle contains (void *)(-1).
4929 61. The Exim daemon panic-logs an error return when it closes the incoming
4930 connection. However "connection reset by peer" seems to be common, and
4931 isn't really an error worthy of noting specially, so that particular error
4934 62. When Exim is trying to find all the local interfaces, it used to panic and
4935 die if the ioctl to get the interface flags failed. However, it seems that
4936 on at least one OS (Solaris 9) it is possible to have an interface that is
4937 included in the list of interfaces, but for which you get a failure error
4938 for this call. This happens when the interface is not "plumbed" into a
4939 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4940 failure of the "get flags" call assumes that the interface is down.
4942 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4943 makes life easier for people who are doing arithmetic on fields extracted
4944 from dates, where you often get leading zeros that should not be
4945 interpreted as octal.
4947 64. Added qualify_domain to the redirect router, to override the global
4950 65. If a pathologically long header line contained very many addresses (the
4951 report of this problem mentioned 10 000) and each of them was rewritten,
4952 Exim could use up a very large amount of memory. (It kept on making new
4953 copies of the header line as it rewrote, and never released the old ones.)
4954 At the expense of a bit more processing, the header rewriting function has
4955 been changed so that it no longer eats memory in this way.
4957 66. The generation of the Received: header has been moved from the time that a
4958 message starts to be received, to the time that it finishes. The timestamp
4959 in the Received: header should now be very close to that of the <= log
4960 line. There are two side-effects of this change:
4962 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4963 logged header lines no longer include the local Received: line, because
4964 it has not yet been created. The same applies to a copy of the message
4965 that is returned to a non-SMTP sender when a message is rejected.
4967 (b) When a filter file is tested using -bf, no additional Received: header
4968 is added to the test message. After some thought, I decided that this
4971 This change does not affect the value of $received_for. It is still set
4972 after address rewriting, but before local_scan() is called.
4974 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4976 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4977 gave an unhelpful panic error message, and a defer error. I have managed to
4978 change this behaviour so that it now rejects any supplied certificate,
4979 which seems right, as the list of acceptable certificates is empty.
4981 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4982 gave an unhelpful defer error. I have not managed to make this reject any
4983 supplied certificates, but the error message it gives is "no certificate
4984 supplied", which is not helpful.
4986 70. exigrep's output now also includes lines that are not associated with any
4987 message, but which match the given pattern. Implemented by a patch from
4988 Martin Sluka, which also tidied up the Perl a bit.
4990 71. Recipient callout verification, like sender verification, was using <> in
4991 the MAIL FROM command. This isn't really the right thing, since the actual
4992 sender may affect whether the remote host accepts the recipient or not. I
4993 have changed it to use the actual sender in the callout; this means that
4994 the cache record is now keyed on a recipient/sender pair, not just the
4995 recipient address. There doesn't seem to be a real danger of callout loops,
4996 since a callout by the remote host to check the sender would use <>.
4997 [SEE ABOVE: changed after hitting problems.]
4999 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5000 temporary errors. However, in the case of such a code being given after
5001 the end of a data transmission (i.e. after ".") Exim was failing to write
5002 a retry record for the message. (Yes, there was some broken host that was
5003 actually sending 8xx at this point.)
5005 73. An unknown lookup type in a host list could cause Exim to panic-die when
5006 the list was checked. (An example that provoked this was putting <; in the
5007 middle of a list instead of at the start.) If this happened during a DATA
5008 ACL check, a -D file could be left lying around. This kind of configuration
5009 error no longer causes Exim to die; instead it causes a defer errror. The
5010 incident is still logged to the main and panic logs.
5012 74. Buglet left over from Exim 3 conversion. The message "too many messages
5013 in one connection" was written to the rejectlog but not the mainlog, except
5014 when address rewriting (yes!) was being logged.
5016 75. Added write_rejectlog option.
5018 76. When a system filter was run not as root (that is, when system_filter_user
5019 was set), the values of the $n variables were not being returned to the
5020 main process; thus, they were not subsequently available in the $sn
5023 77. Added +return_path_on_delivery log selector.
5025 78. A connection timeout was being treated differently from recipients deferred
5026 when testing hosts_max_try with a message that was older than the host's
5027 retry timeout. (The host should not be counted, thus allowing all hosts to
5028 be tried at least once before bouncing.) This may have been the cause of an
5029 occasionally reported bug whereby a message would remain on the queue
5030 longer than the retry timeout, but would be bounced if a delivery was
5031 forced. I say "may" because I never totally pinned down the problem;
5032 setting up timeout/retry tests is difficult. See also the next item.
5034 79. The ultimate address timeout was not being applied to errors that involved
5035 a combination of host plus message (for example, a timeout on a MAIL
5036 command). When an address resolved to a number of possible hosts, and they
5037 were not all tried for each delivery (e.g. because of hosts_max_try), a
5038 message could remain on the queue longer than the retry timeout.
5040 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5043 81. Fixed an obscure SMTP outgoing bug which required at least the following
5044 conditions: (a) there was another message waiting for the same server;
5045 (b) the server returned 5xx to all RCPT commands in the first message so
5046 that the message was not completed; (c) the server dropped the connection
5047 or gave a negative response to the RSET that Exim sends to abort the
5048 transaction. The observed case was a dropped connection after DATA that had
5049 been sent in pipelining mode. That is, the server had advertised PIPELINING
5050 but was not implementing it correctly. The effect of the bug was incorrect
5051 behaviour, such as trying another host, and this could lead to a crash.
5057 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5058 and daemon.c were passed as pointers to ints; they should have been
5059 pointers to socklen_t variables (which are typically unsigned ints).
5061 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5064 3. Fixed a really odd bug that affected only the testing scheme; patching a
5065 certain fixed string in the binary changed the value of another string that
5066 happened to be identical to the end of the original first string.
5068 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5069 name", it returns that address as the IP address. On some operating
5070 systems (e.g. Solaris), it also passes back the IP address string as the
5071 "host name". However, on others (e.g. Linux), it passes back an empty
5072 string. Exim wasn't checking for this, and was changing the host name to an
5073 empty string, assuming it had been canonicized.
5075 5. Although rare, it is permitted to have more than one PTR record for a given
5076 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5077 all the names associated with an address, because they do in Solaris.
5078 However, it seems that they do not in Linux for data that comes from the
5079 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5080 I found this out when I moved to a new Linux workstation and tried to run
5081 the Exim test suite.
5083 To get round this problem I have changed the code so that it now does its
5084 own call to the DNS to look up PTR records when searching for a host name.
5085 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5086 addresses that are only in /etc/hosts are still found.
5088 This behaviour is, however, controlled by an option called host_lookup_
5089 order, which defaults to "bydns:byaddr". If people want to use the other
5090 order, or indeed, just use one or the other means of lookup, they can
5091 specify it in this variable.
5093 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5094 some operating systems, this comes back from gethostbyaddr() as an empty
5095 string, and this is what Exim used to test for. However, it seems that in
5096 other systems, "." is yielded. Exim now tests for this case too.
5098 7. The values of check_spool_space and check_log_space are now held internally
5099 as a number of kilobytes instead of an absolute number of bytes. If a
5100 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5101 kilobyte. This means that much larger values can be stored.
5103 8. Exim monitor: an attempt to get the action menu when not actually pointing
5104 at a message produces an empty menu entitled "No message selected". This
5105 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5106 no entries in it ("Shell widget menu has zero width and/or height"). So I
5107 have added a single, blank menu entry in this case.
5109 9. Added ${quote_local_part.
5111 10. MIME decoding is now applied to the contents of Subject: header lines when
5114 11. Now that a reference to $sender_host_address automatically causes a reverse
5115 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5116 host lookup before query-style lookups in lists that might use this
5117 variable. This has therefore been abolished, and the "net-" prefix is no
5118 longer necessary for query-style lookups.
5120 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5121 have been a typo for LFLAGS, so it has been changed.
5123 13. The install script calls Exim with "-C /dev/null" in order to find the
5124 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5125 to be output. Howeve, since Exim outputs its version number before the
5126 error, it didn't break the script. It just looked ugly. I fixed this by
5127 always allowing "-C /dev/null" if the caller is root.
5129 14. Ignore overlarge ACL variable number when reading spool file - insurance
5130 against a later release with more variables having written the file.
5132 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5133 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5134 this, as well as the form without the "IPv6" on the front.
5136 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5137 OS/Makefile-Darwin file.
5139 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5140 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5142 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5143 variable. However, some OS use ints for inodes. I've added cast to long int
5144 to get rid of the compiler warning.
5146 19. I had forgotten to lock out "/../" in configuration file names when
5147 ALT_CONFIG_PREFIX was set.
5149 20. Routers used for verification do not need to specify transports. However,
5150 if such a router generated a host list, and callout was configured, Exim
5151 crashed, because it could not find a port number from the (non-existent)
5152 transport. It now assumes port 25 in this circumstance.
5154 21. Added the -t option to exigrep.
5156 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5157 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5158 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5159 them is set, LOOKUP_LSEARCH is forced.
5161 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5163 24. Added sender and host information to the "rejected by local_scan()" log
5164 line; previously there was no indication of these.
5166 25. Added .include_if_exists.
5168 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5169 Linux. It turns out that not all file systems support this. Apparently some
5170 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5171 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5172 not supported on the file descriptor, is now ignored when Exim is trying to
5173 sync a directory. This applies only to Linux.
5175 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5177 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5178 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5181 29. Added support for maildirsize files from supplied patch (modified a bit).
5183 30. The use of :fail: followed by an empty string could lead Exim to respond to
5184 sender verification failures with (e.g.):
5186 550 Verification failed for <xxx>
5187 550 Sender verify failed
5189 where the first response line was missing the '-' that indicates it is not
5190 the final line of the response.
5192 31. The loop for finding the name of the user that called Exim had a hardwired
5193 limit of 10; it now uses the value of finduser_retries, which is used for
5194 all other user lookups.
5196 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5199 33. Exim was neglecting to zero errno before one call of strtol() when
5200 expanding a string and expecting an integer value. On some systems this
5201 resulted in spurious "integer overflow" errors. Also, it was casting the
5202 result into an int without checking.
5204 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5205 did not work. The code looks as if it has *never* worked, though it appears
5206 to have been documented since at least releast 1.62. I have made it work.
5208 35. The "timeout_DNS" error in retry rules, also documented since at least
5209 1.62, also never worked. As it isn't clear exactly what this means, and
5210 clearly it isn't a major issue, I have abolished the feature by treating it
5211 as "timeout", and writing a warning to the main and panic logs.
5213 36. The display of retry rules for -brt wasn't always showing the error code
5216 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5217 timeout_connect_A, timeout_connect_MX.
5219 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5220 to the empty sender.
5222 39. The daemon was not analysing the content of -oX till after it had closed
5223 stderr and disconnected from the controlling terminal. This meant that any
5224 syntax errors were only noted on the panic log, and the return code from
5225 the command was 0. By re-arranging the code a little, I've made the
5226 decoding happen first, so such errors now appear on stderr, and the return
5227 code is 1. However, the actual setting up of the sockets still happens in
5228 the disconnected process, so errors there are still only recorded on the
5231 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5232 connections (as happens on some IP stacks) was logged at start up time as
5233 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5234 it from "IPv6 and IPv4", which means that two separate sockets are being
5237 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5238 says whether AF_INET or AF_INET6 was passed as an argument.
5240 42. Exiwhat output was messed up when time zones were included in log
5243 43. Exiwhat now gives more information about the daemon's listening ports,
5244 and whether -tls-on-connect was used.
5246 44. The "port" option of the smtp transport is now expanded.
5248 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5249 silently ignored. Now an error message is written to the main and panic
5252 46. There's a new ACL modifier called "logwrite" which writes to a log file
5253 as soon as it is encountered.
5255 47. Added $local_user_uid and $local_user_gid at routing time.
5257 48. Exim crashed when trying to verify a sender address that was being
5260 49. Exim was recognizing only a space character after ".include". It now also
5261 recognizes a tab character.
5263 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5264 extracting the relevant information from the specification. The man page no
5265 longer contains scrambled data for the -d option, and I've added a section
5266 at the front about calling Exim under different names.
5268 51. Added "extra_headers" argument to the "mail" command in filter files.
5270 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5273 53. Installed eximstats 1.29.
5275 54. Added transport_filter_timeout as a generic transport option.
5277 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5278 Cc: header lines. This was required by RFC 822, but it not required by RFC
5281 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5282 incoming messages that did not have them. Now it does so only if the
5283 message originates locally, that is, if there is no associated remote host
5284 address. When Resent- header lines are present, this applies to the Resent-
5285 lines rather than the non-Resent- lines.
5287 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5288 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5290 58. Messages for configuration errors now include the name of the main
5291 configuration file - useful now that there may be more than one file in a
5292 list (.included file names were always shown).
5294 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5295 for those rare installations that do not start the daemon as root or run it
5296 setuid root. I've cut out the call to initgroups() if the daemon is not
5299 60. The Exim user and group can now be bound into the binary as text strings
5300 that are looked up at the start of Exim's processing.
5302 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5304 62. Added $mailstore_basename variable.
5306 63. Installed patch to sieve.c from Michael Haardt.
5308 64. When Exim failed to open the panic log after failing to open the main log,
5309 the original message it was trying to log was written to stderr and debug
5310 output, but if they were not available (the usual case in production), it
5311 was lost. Now it is written to syslog before the two lines that record the
5312 failures to open the logs.
5314 65. Users' Exim filters run in subprocesses under the user's uid. It is
5315 possible for a "deliver" command or an alias in a "personal" command to
5316 provoke an address rewrite. If logging of address rewriting is configured,
5317 this fails because the process is not running as root or exim. There may be
5318 a better way of dealing with this, but for the moment (because 4.30 needs
5319 to be released), I have disabled address rewrite logging when running a
5320 filter in a non-root, non-exim process.
5326 1. The buildconfig auxiliary program wasn't quoting the value set for
5327 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5328 not defined. This bug was masked in 4.22 by the effect that was fixed in
5331 2. Some messages that were rejected after a message id was allocated were
5332 shown as "incomplete" by exigrep. It no longer does this for messages that
5333 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5335 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5336 have allow_domain_literals set, the ID did not get logged in the <= line.
5337 Domain literals are now always recognized in Message-ID: header lines.
5339 4. The first argument for a ${extract expansion item is the key name or field
5340 number. Leading and trailing spaces in this item were not being ignored,
5341 causing some misleading effects.
5343 5. When deliver_drop_privilege was set, single queue runner processes started
5344 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5345 same command in the process it spins off) were not dropping privilege.
5347 6. When the daemon running as "exim" started a queue runner, it always
5348 re-executed Exim in the spun-off process. This is a waste of effort when
5349 deliver_drop_privilege is set. The new process now just calls the
5350 queue-runner function directly.
5356 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5359 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5360 ignored. Though the use of -oP was forcing the writing of a pid file, it
5361 was always written to the default place.
5363 3. If the message "no IP address found for host xxxx" is generated during
5364 incoming verification, it is now followed by identification of the incoming
5365 connection (so you can more easily find what provoked it).
5367 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5369 5. Added some features to "harden" Exim a bit more against certain attacks:
5371 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5372 be put in Local/Makefile. This is like the never_users runtime option,
5373 but it cannot be overridden. The default setting is "root".
5375 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5376 prefix string with which any file named in a -C command line option
5379 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5380 is retained for -C and -D only if the caller of Exim is root. Without
5381 it, the exim user may also use -C and -D and retain privilege.
5383 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5384 command line option is disabled.
5386 6. Macro names set by the -D option must start with an upper case letter, just
5387 like macro names defined in the configuration file.
5389 7. Added "dereference=" facility to LDAP.
5391 8. Two instances of the typo "uknown" in the source files are fixed.
5393 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5394 the Configure-Makefile script screwed up while processing it.
5396 10. Incorporated PCRE 4.4.
5398 11. The SMTP synchronization check was not operating right at the start of an
5399 SMTP session. For example, it could not catch a HELO sent before the client
5400 waited for the greeting. There is now a check for outstanding input at the
5401 point when the greeting is written. Because of the duplex, asynchronous
5402 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5403 way, but not yet received, when the check is performed.
5405 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5406 on TCP/IP sockets, because this apparently causes some broken clients to
5409 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5410 unchanged) from the Cygwin maintainer.
5412 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5413 was supported instead of testing for mbx. Effectively a typo.
5415 15. The spa authenticator server code was not checking that the input it
5416 received was valid base64.
5418 16. The debug output line for the "set" modifier in ACLs was not showing the
5419 name of the variable that was being set.
5421 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5423 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5424 lookup on its own. Something else was needed to trigger the lookup. For
5425 example, a match in host_lookup or the need for a host name in a host list.
5426 Now, if $sender_host_name is referenced and the host name has not yet been
5427 looked up, a lookup is performed. If the lookup fails, the variable remains
5428 empty, and $host_lookup_failed is set to "1".
5430 19. Added "eqi" as a case-independent comparison operator.
5432 20. The saslauthd authentication condition could segfault if neither service
5433 nor realm was specified.
5435 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5436 error message that was logged was misleading, and incoming SMTP
5437 connections were dropped. The message is now more accurate, and temporary
5438 errors are given to SMTP connections.
5440 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5441 (except RSET) with a 421 error, until QUIT is received. However, it was
5442 failing to send a response to QUIT.
5444 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5445 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5446 ACL is now run after the helo_try_verify_hosts code.
5448 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5449 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5450 case-independent, so other case variants are also recognized). Apparently
5451 some systems use these upper case variants.
5453 25. If more than two messages were waiting for the same host, and a transport
5454 filter was specified for the transport, Exim sent two messages over the
5455 same TCP/IP connection, and then failed with "socket operation on non-
5456 socket" when it tried to send the third.
5458 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5460 27. The extern definition of crypt16() in expand.c was not being excluded when
5461 the OS had its own crypt16() function.
5463 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5464 as a preferred synonym for return_size_limit, both as an option and as an
5467 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5469 30. Changed the default configuration ACL to relax the local part checking rule
5470 for addresses that are not in any local domains. For these addresses,
5471 slashes and pipe symbols are allowed within local parts, but the sequence
5472 /../ is explicitly forbidden.
5474 31. SPA server authentication was not clearing the challenge buffer before
5477 32. log_message in a "warn" ACL statement was writing to the reject log as
5478 well as to the main log, which contradicts the documentation and doesn't
5479 seem right (because no rejection is happening). So I have stopped it.
5481 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5482 However, I am unable to do any testing of this.
5484 34. Fixed an infelicity in the appendfile transport. When checking directories
5485 for a mailbox, to see if any needed to be created, it was accidentally
5486 using path names with one or more superfluous leading slashes; tracing
5487 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5489 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5490 contents are added to the log line that is written for every discarded
5491 recipient. (Previously a log_message setting was ignored.)
5493 36. The ${quote: operator now quotes the string if it is empty.
5495 37. The install script runs exim in order to find its version number. If for
5496 some reason other than non-existence or emptiness, which it checks, it
5497 could not run './exim', it was installing it with an empty version number,
5498 i.e. as "exim-". This error state is now caught, and the installation is
5501 38. An argument was missing from the function that creates an error message
5502 when Exim fails to connect to the socket for saslauthd authentication.
5503 This could cause Exim to crash, or give a corrupted message.
5505 39. Added isip, isip4, and isip6 to ${if conditions.
5507 40. The ACL variables $acl_xx are now saved with the message, and can be
5508 accessed later in routers, transports, and filters.
5510 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5511 strings in the file are not string-expanded.
5513 42. If a MAIL command specified a SIZE value that was too large to fit into an
5514 int variable, the check against message_size_limit failed. Such values are
5515 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5516 day this will have to be increased, but I don't think I want to be around
5517 when emails are that large.
5524 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5525 iconv() is not standard in FreeBSD.
5527 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5528 IPv6 enabled. The observed symptom was a segmentation fault on return from
5529 the function os_common_find_running_interfaces() in src/os.c.
5531 3. In the check_special_case() function in daemon.c I had used "errno" as an
5532 argument name, which causes warnings on some systems. This was basically a
5533 typo, since it was named "eno" in the comments!
5535 4. The code that waits for the clock to tick (at a resolution of some fraction
5536 of a second) so as to ensure message-id uniqueness was always waiting for
5537 at least one whole tick, when it could have waited for less. [This is
5538 almost certainly not relevant at current processor speeds, where it is
5539 unlikely to ever wait at all. But we try to future-proof.]
5541 5. The function that sleeps for a time interval that includes fractions of a
5542 second contained a race. It did not block SIGALRM between setting the
5543 timer, and suspending (a couple of lines later). If the interval was short
5544 and the sigsuspend() was delayed until after it had expired, the suspension
5545 never ended. On busy systems this could lead to processes getting stuck for
5548 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5549 process, before it forks any delivery processes. The open lookup caching
5550 mechanism meant that the open file or database connection was passed into
5551 the delivery process. The problem was that delivery processes always tidy
5552 up cached lookup data. This could cause a problem for the next delivery
5553 process started by the queue runner, because the external queue runner
5554 process does not know about the closure. So the next delivery process
5555 still has data in the lookup cache. In the case of a file lookup, there was
5556 no problem because closing a file descriptor in a subprocess doesn't affect
5557 the parent. However, if the lookup was caching a connection to a database,
5558 the connection was closed, and the second delivery process was likely to
5559 see errors such as "PGSQL: query failed: server closed the connection
5560 unexpectedly". The problem has been fixed by closing all cached lookups
5561 in a queue runner before running a delivery process.
5563 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5564 seem to have the "const" qualifier which it has on other OS. I've
5567 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5568 *of the same type* (client or server) with the same public name that an
5569 error should be diagnosed.
5571 9. When Exim looked up a host name for an IP address, but failed to find the
5572 original IP address when looking up the host name (a safety check), it
5573 output the message "<ip address> does not match any IP for NULL", which was
5574 confusing, to say the least. The bug was that the host name should have
5575 appeared instead of "NULL".
5577 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5578 user that is built into the binary, and the -C or -D options is used, root
5579 privilege is dropped before the configuration file is read. In addition,
5580 logging is switched to stderr instead of the normal log files. If the
5581 configuration then re-defines the Exim user, the unprivileged environment
5582 is probably not what is expected, so Exim logs a panic warning message (but
5585 However, if deliver_drop_privilege is set, the unprivileged state may well
5586 be exactly what is intended, so the warning has been cut out in that case,
5587 and Exim is allowed to try to write to its normal log files.
5593 1. smtp_return_error_details was not giving details for temporary sender
5594 or receiver verification errors.
5596 2. Diagnose a configuration error if two authenticators have the same public
5599 3. Exim used not to create the message log file for a message until the first
5600 delivery attempt. This could be confusing when incoming messages were held
5601 for policy or load reasons. The message log file is now created at the time
5602 the message is received, and an initial "Received" line is written to it.
5604 4. The automatically generated man page for command line options had a minor
5605 bug that caused no ill effects; however, a more serious problem was that
5606 the procedure for building the man page automatically didn't always
5607 operate. Consequently, release 4.20 contains an out-of-date version. This
5608 shouldn't happen again.
5610 5. When building Exim with embedded Perl support, the script that builds the
5611 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5612 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5614 6. The freeze_tell option was not being used for messages that were frozen on
5615 arrival, either by an ACL or by local_scan().
5617 7. Added the smtp_incomplete_transaction log selector.
5619 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5620 was accepting AUTH without a new EHLO.
5622 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5623 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5626 10. Logging of TCP/IP connections (when configured) now happens in the main
5627 daemon process instead of the child process, so that the TCP/IP connection
5628 count is more accurate (but it can never be perfect).
5630 11. The use of "drop" in a nested ACL was not being handled correctly in the
5631 outer ACL. Now, if condition failure induced by the nested "drop" causes
5632 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5633 or "require"), the connection is dropped.
5635 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5636 that yield "discard" can now be used with an "accept" or a "discard" verb,
5637 but an error is generated for any others (because I can't see a useful way
5638 to define what should happen).
5640 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5641 are now processed in the same way as lines in the Exim configuration file.
5642 In particular, continuation lines are supported.
5644 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5646 15. Added -ti meaning -t -i.
5648 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5649 domains, and warn by logging if others are found.
5651 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5652 returned from GIFCONF when Exim is trying to find the list of interfaces on
5653 a host. The code in os.c has been modified to copy each ifreq to an aligned
5654 structure in all cases.
5656 Also, in some cases, the returned ifreq's were being copied to a 'struct
5657 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5658 means the last couple of bytes of an IPv6 address could be chopped if the
5659 ifreq contained only a normal sockaddr (14 bytes storage).
5661 18. Named domain lists were not supported in the hosts_treat_as_local option.
5662 An entry such as +xxxx was not recognized, and was treated as a literal
5665 19. Ensure that header lines added by a DATA ACL are included in the reject log
5666 if the ACL subsequently rejects the message.
5668 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5669 MD5 (which is deprecated).
5671 21. When testing a filter file using -bf, Exim was writing a message when it
5672 took the sender from a "From " line in the message, but it was not doing so
5673 when it took $return_path from a Return-Path: header line. It now does.
5675 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5676 with a valid header line field name (a series of printing characters
5677 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5679 23. Changed "disc" in the source to "disk" to conform to the documentation and
5680 the book and for uniformity.
5682 24. Ignore Sendmail's -Ooption=value command line item.
5684 25. When execve() failed while trying to run a command in a pipe transport,
5685 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5686 could be confused with a return value of 69 from the command itself. This
5687 has been changed to 127, the value the shell returns if it is asked to run
5688 a non-existent command. The wording for the related log line suggests a
5689 non-existent command as the problem.
5691 26. If received_header_text expands to an empty string, do not add a Received:
5692 header line to the message. (Well, it adds a token one on the spool, but
5693 marks it "old" so that it doesn't get used or transmitted.)
5695 27. Installed eximstats 1.28 (addition of -nt option).
5697 28. There was no check for failure on the call to getsockname() in the daemon
5698 code. This can fail if there is a shortage of resources on the system, with
5699 ENOMEM, for example. A temporary error is now given on failure.
5701 29. Contrary to the C standard, it seems that in some environments, the
5702 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5703 program. Exim now does this explicitly; it affects the formatting of
5704 timestamps using strftime().
5706 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5707 complaints. I've now initialized all the variables, to avoid this.
5709 32. Header lines added by a system filter were not being "seen" during
5710 transport-time rewrites.
5712 33. The info_callback() function passed to OpenSSL is set up with type void
5713 (*)(SSL *, int, int), as described somewhere. However, when calling the
5714 function (actually a macro) that sets it up, the type void(*)() is
5715 expected. I've put in a cast to prevent warnings from picky compilers.
5717 34. If a DNS black list lookup found a CNAME record, but there were no A
5718 records associated with the domain it pointed at, Exim crashed.
5720 35. If a DNS black list lookup returned more than one A record, Exim ignored
5721 all but the first. It now scans all returned addresses if a particular IP
5722 value is being sought. In this situation, the contents of the
5723 $dnslist_value variable are a list of all the addresses, separated by a
5726 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5727 to accept a host name and all its aliases if the forward lookup for any of
5728 them yielded the IP address of the incoming connection. Now it accepts only
5729 those names whose forward lookup yields the correct IP address. Any other
5730 names are discarded. This closes a loophole whereby a rogue DNS
5731 administrator could create reverse DNS records to break through a
5732 wildcarded host restriction in an ACL.
5734 37. If a user filter or a system filter that ran in a subprocess used any of
5735 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5736 the wrong values were passed to the pipe command ($thisaddress had the
5737 value of $0, $0 had the value of $1, etc). This bug was introduced by
5738 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5740 38. Improved the line breaking for long SMTP error messages from ACLs.
5741 Previously, if there was no break point between 40 and 75 characters, Exim
5742 left the rest of the message alone. Two changes have been made: (a) I've
5743 reduced the minimum length to 35 characters; (b) if it can't find a break
5744 point between 35 and 75 characters, it looks ahead and uses the first one
5745 that it finds. This may give the occasional overlong line, but at least the
5746 remaining text gets split now.
5748 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5749 file descriptors might be low, and that setting 1000 would always raise it.
5750 It turns out that in some environments, the limit is already over 1000 and
5751 that lowering it causes trouble. So now Exim takes care not to decrease it.
5753 40. When delivering a message, the value of $return_path is set to $sender_
5754 address at the start of routing (routers may change the value). By an
5755 oversight, this default was not being set up when an address was tested by
5756 -bt or -bv, which affected the outcome if any router or filter referred to
5759 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5760 log only when "message" or "log_message" are set. However, if one of the
5761 conditions was an address verification, or a call to a nested ACL, the
5762 messages generated by the underlying test were being passed through. This
5763 no longer happens. The underlying message is available in $acl_verify_
5764 message for both "message" and "log_message" expansions, so it can be
5765 passed through if needed.
5767 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5768 new expansion $bh_ to give the encoded byte string without charset
5769 translation. Translation happens only if iconv() is available; HAVE_ICONV
5770 indicates this at build time. HEADERS_CHARSET gives the charset to
5771 translate to; headers_charset can change it in the configuration, and
5772 "headers charset" can change it in an individual filter file.
5774 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5775 that creates RFC 2047 encoded "words" labels them as that charset instead
5776 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5777 expansion operator; (ii) when Exim creates a From: line for a local
5778 message; (iii) when a header line is rewritten to include a "phrase" part.
5780 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5781 buggy, causing it to skip the first lines of messages whose message ID
5782 ended in 'D'. This would not have bitten before Exim release 4.14, because
5783 message IDs were unlikely to end in 'D' before then. The effect was to have
5784 incorrect size information for certain domains.
5786 45. #include "config.h" was missing at the start of the crypt16.c module. This
5787 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5790 46. If there was a timeout during a "random" callout check, Exim treated it as
5791 a failure of the random address, and carried on sending RSET and the real
5792 address. If the delay was just some slowness somewhere, the response to the
5793 original RCPT would be taken as a response to RSET and so on, causing
5794 mayhem of various kinds.
5796 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5797 when I implemented it. It didn't allow for the fact that some option values
5798 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5799 the right test for positive values.
5801 48. Domain names in DNS records are case-independent. Exim always looks them up
5802 in lower case. Some resolvers return domain names in exactly the case they
5803 appear in the zone file, that is, they may contain uppercase letters. Not
5804 all resolvers do this - some return always lower case. Exim was treating a
5805 change of case by a resolver as a change of domain, similar to a widening
5806 of a domain abbreviation. This triggered its re-routing code and so it was
5807 trying to route what was effectively the same domain again. This normally
5808 caused routing to fail (because the router wouldn't handle the domain
5809 twice). Now Exim checks for this case specially, and just changes the
5810 casing of the domain that it ultimately uses when it transmits the message
5813 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5816 50. If a filter generated a file delivery with a non-absolute name (possible if
5817 no home directory exists for the router), the forbid_file option was not
5820 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5821 the existing equality matching.
5823 52. Exim was using ints instead of ino_t variables in some places where it was
5824 dealing with inode numbers.
5826 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5827 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5828 called TMPDIR, and if it finds it is different, it changes its value.
5830 54. The smtp_printf() function is now made available to local_scan() so
5831 additional output lines can be written before returning. There is also an
5832 smtp_fflush() function to enable the detection of a dropped connection.
5833 The variables smtp_input and smtp_batched_input are exported to
5836 55. Changed the default runtime configuration: the message "Unknown user"
5837 has been removed from the ACL, and instead placed on the localuser router,
5838 using the cannot_route_message feature. This means that any verification
5839 failures that generate their own messages won't get overridden. Similarly,
5840 the "Unrouteable address" message that was in the ACL for unverifiable
5841 relay addresses has also been removed.
5843 56. Added hosts_avoid_esmtp to the smtp transport.
5845 57. The exicyclog script was not checking for the esoteric option
5846 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5847 will work only if exicyclog is run under the appropriate euid.
5849 58. Following a discussion on the list, the rules by which Exim recognises line
5850 endings on incoming messages have been changed. The -dropcr and drop_cr
5851 options are now no-ops, retained only for backwards compatibility. The
5852 following line terminators are recognized: LF CRLF CR. However, special
5853 processing applies to CR:
5855 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5856 nor a local message in the state where . is a terminator.
5858 (ii) If a bare CR is encountered in a header line, an extra space is added
5859 after the line terminator so as not to end the header. The reasoning
5860 behind this is that bare CRs in header lines are most likely either
5861 to be mistakes, or people trying to play silly games.
5863 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5864 was being incorrectly given as 18 bytes larger than it should have been.
5865 This is a VOB (very old bug).
5867 60. This may never have affected anything current, but just in case it has:
5868 When the local host is found other than at the start of a list of hosts,
5869 the local host, those with the same MX, and any that follow, are discarded.
5870 When the list in question was part of a longer list of hosts, the following
5871 hosts (not currently being processed) were also being discarded. This no
5872 longer happens. I'm not sure if this situation could ever has previously
5875 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5878 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5879 header. This is something that is recommended in a new Internet Draft, and
5880 is something that is documented as being done by Sendmail. There are two
5881 possible values. For messages generated by the autoreply transport, Exim
5884 Auto-Submitted: auto-replied
5886 whereas for all other generated messages (e.g. bounces) it adds
5888 Auto-Submitted: auto-generated
5890 63. The "personal" condition in filters now includes a test for the
5891 Auto-Submitted: header. If it contains the string "auto-" the message it
5892 not considered personal.
5894 64. Added rcpt_include_affixes as a generic transport option.
5896 65. Added queue_only_override (default true).
5898 66. Added the syslog_duplication option.
5900 67. If what should have been the first header line of a message consisted of
5901 a space followed by a colon, Exim was mis-interpreting it as a header line.
5902 It isn't of course - it is syntactically invalid and should therefore be
5903 treated as the start of the message body. The misbehaviour could have
5904 caused a number of strange effects, including loss of data in subsequent
5905 header lines, and spool format errors.
5907 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5908 client host had authenticated. This control can now be exercised by an ACL
5909 for more flexibility.
5911 69. By default, callouts do not happen when testing with -bh. There is now a
5912 variant, -bhc, which does actually run the callout code, including
5913 consulting and updating the callout cache.
5915 70. Added support for saslauthd authentication, courtesy of Alexander
5918 71. If statvfs() failed on the spool or log directories while checking their
5919 size for availability, Exim confusingly gave the error "space shortage".
5920 Furthermore, in debugging mode it crashed with a floating point exception.
5921 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5922 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5923 serious problem, Exim now writes to the main and panic logs when this
5924 happens, with details of the failure. It then refuses to accept the
5925 incoming message, giving the message "spool directory problem" or "log
5926 directory problem" with a 421 code for SMTP messages.
5928 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5929 0, 1, and 2 exist, because some OS complain for execs without them (see
5930 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5931 descriptors. However, the code omitted to check that the open succeeded,
5932 causing mysterious errors if for some reason the permissions on /dev/null
5933 got screwed. Now Exim writes a message to the main and panic logs, and
5934 bombs out if it can't open /dev/null.
5936 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5937 interact so that it is all more flexible. It is supposed to remain
5938 backwards compatible. Also added extra_local_interfaces.
5940 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5941 to bomb out with an assertion failure - to the client this appears as a
5942 connection drop. This problem occurs in the part of the code that was taken
5943 from the Samba project. Fortunately, the assertion is in a very simple
5944 function, so I have fixed this by reproducing the function inline in the
5945 one place where it is called, and arranging for authentication to fail
5946 instead of killing the process with assert().
5948 75. The SPA client code was not working when the server requested OEM rather
5949 than Unicode encoding.
5951 76. Added code to make require_files with a specific uid setting more usable in
5952 the case where statting the file as root fails - usually a non-root-mounted
5953 NFS file system. When this happens and the failure is EACCES, Exim now
5954 forks a subprocess and does the per-uid checking as the relevant uid.
5956 77. Added process_log_path.
5958 78. If log_file_path was not explicitly set, a setting of check_log_space or
5959 check_log_inodes was ignored.
5961 79. If a space check for the spool or log partitions fails, the incident is now
5962 logged. Of course, in the latter case the data may get lost...
5964 80. Added the %p formatting code to string_format() so that it can be used to
5965 print addresses in debug_print(). Adjusted all the address printing in the
5966 debugging in store.c to use %p rather than %d.
5968 81. There was a concern that a line of code in smtp_in.c could overflow a
5969 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5970 initially expressed, the concern was not well-founded, because trailing
5971 spaces are removed early. However, if the trailing spaces were followed by
5972 a NULL, they did not get removed, so the overflow was possible. Two fixes
5975 (a) I re-wrote the offending code in a cleaner fashion.
5976 (b) If an incoming SMTP command contains a NULL character, it is rejected
5979 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5980 runs initgroups(), so that if the Exim user is in any additional groups,
5981 they will be used during message reception.
5987 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob