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.
69 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
71 JH/12 Expand items in router/transport headers_add or headers_remove lists
72 individually rather than the list as a whole. Bug 1452.
74 Required for reasonable handling of multiple headers_ options when
75 they may be empty; requires that headers_remove items with embedded
76 colons must have them doubled (or the list-separator changed).
82 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
84 PP/02 Make -n do something, by making it not do something.
85 When combined with -bP, the name of an option is not output.
87 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
90 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
91 $sender_host_name and config options to manage this, and basic check
94 PP/05 DSCP support for outbound connections and control modifier for inbound.
96 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
97 (Only plugin which currently uses this is kerberos4, which nobody should
98 be using, but we should make it available and other future plugins might
99 conceivably use it, even though it would break NAT; stuff *should* be
100 using channel bindings instead).
102 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
103 name; added for Sendmail compatibility; requires admin caller.
104 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
105 just ignore it); requires trusted caller.
106 Also parse but ignore: -Ac -Am -X<logfile>
109 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
111 TL/02 Add +smtp_confirmation as a default logging option.
113 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
114 Patch by Magnus Holmgren from 2007-02-20.
116 TL/04 Bugzilla 1281 - Spec typo.
117 Bugzilla 1283 - Spec typo.
118 Bugzilla 1290 - Spec grammar fixes.
120 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
122 TL/06 Add Experimental DMARC support using libopendmarc libraries.
124 TL/07 Fix an out of order global option causing a segfault. Reported to dev
125 mailing list by by Dmitry Isaikin.
127 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
129 JH/02 Support "G" suffix to numbers in ${if comparisons.
131 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
133 NM/01 Bugzilla 1197 - Spec typo
134 Bugzilla 1196 - Spec examples corrections
136 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
138 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
139 gnutls_enable_pkcs11, but renamed to more accurately indicate its
142 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
143 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
145 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
146 "acl {{name}{arg}...}", and optional args on acl condition
149 JH/05 Permit multiple router/transport headers_add/remove lines.
151 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
153 JH/07 Avoid using a waiting database for a single-message-only transport.
154 Performance patch from Paul Fisher. Bugzilla 1262.
156 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
159 JH/09 Add $headers_added variable, with content from use of ACL modifier
160 add_header (but not yet added to the message). Bugzilla 199.
162 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
163 Pulled from Bugzilla 817 by Wolfgang Breyha.
165 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
167 (nb: this is the same fix as in Exim 4.80.1)
169 JH/11 Add A= logging on delivery lines, and a client_set_id option on
172 JH/12 Add optional authenticated_sender logging to A= and a log_selector
175 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
177 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
178 advertise SMTP AUTH mechanism to us, instead of a generic
179 protocol violation error. Also, make Exim more robust to bad
180 data from the Dovecot auth socket.
182 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
184 When a queue runner is handling a message, Exim first routes the
185 recipient addresses, during which it prunes them based on the retry
186 hints database. After that it attempts to deliver the message to
187 any remaining recipients. It then updates the hints database using
190 So if a recipient address works intermittently, it can get repeatedly
191 deferred at routing time. The retry hints record remains fresh so the
192 address never reaches the final cutoff time.
194 This is a fairly common occurrence when a user is bumping up against
195 their storage quota. Exim had some logic in its local delivery code
196 to deal with this. However it did not apply to per-recipient defers
197 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
199 This change adds a proper retry rule check during routing so that the
200 final cutoff time is checked against the message's age. We only do
201 this check if there is an address retry record and there is not a
202 domain retry record; this implies that previous attempts to handle
203 the address had the retry_use_local_parts option turned on. We use
204 this as an approximation for the destination being like a local
205 delivery, as in LMTP.
207 I suspect this new check makes the old local delivery cutoff check
208 redundant, but I have not verified this so I left the code in place.
210 TF/02 Correct gecos expansion when From: is a prefix of the username.
212 Test 0254 submits a message to Exim with the header
216 When I ran the test suite under the user fanf2, Exim expanded
217 the header to contain my full name, whereas it should have added
218 a Resent-Sender: header. It erroneously treats any prefix of the
219 username as equal to the username.
221 This change corrects that bug.
223 GF/01 DCC debug and logging tidyup
224 Error conditions log to paniclog rather than rejectlog.
225 Debug lines prefixed by "DCC: " to remove any ambiguity.
227 TF/03 Avoid unnecessary rebuilds of lookup-related code.
229 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
230 Bug spotted by Jeremy Harris; was flawed since initial commit.
231 Would have resulted in OCSP responses post-SNI triggering an Exim
232 NULL dereference and crash.
234 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
236 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
237 Bug detection, analysis and fix by Samuel Thibault.
238 Bugzilla 1331, Debian bug #698092.
240 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
242 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
243 Server implementation by Todd Lyons, client by JH.
244 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
245 config variable "prdr_enable" controls whether the server
246 advertises the facility. If the client requests PRDR a new
247 acl_data_smtp_prdr ACL is called once for each recipient, after
248 the body content is received and before the acl_smtp_data ACL.
249 The client is controlled by bolth of: a hosts_try_prdr option
250 on the smtp transport, and the server advertisement.
251 Default client logging of deliveries and rejections involving
252 PRDR are flagged with the string "PRDR".
254 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
255 fclose(). Diagnosis by Todd Lyons.
257 PP/17 Update configure.default to handle IPv6 localhost better.
258 Patch by Alain Williams (plus minor tweaks).
261 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
262 This is now consistent with GnuTLS, and is now documented: the
263 previous undocumented portable approach to treating the option as
264 unset was to force an expansion failure. That still works, and
265 an empty string is now equivalent.
267 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
268 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
269 not performing validation itself.
271 PP/20 Added force_command boolean option to pipe transport.
272 Patch from Nick Koston, of cPanel Inc.
274 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
277 TF/04 Added udpsend ACL modifer and hexquote expansion operator
279 PP/21 Fix eximon continuous updating with timestamped log-files.
280 Broken in a format-string cleanup in 4.80, missed when I repaired the
281 other false fix of the same issue.
282 Report and fix from Heiko Schlichting.
285 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
286 Report from Prashanth Katuri.
288 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
289 It's SecureTransport, so affects any MacOS clients which use the
290 system-integrated TLS libraries, including email clients.
292 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
293 using a MIME ACL for non-SMTP local injection.
294 Report and assistance in diagnosis by Warren Baker.
296 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
298 JH/16 Fix comparisons for 64b. Bugzilla 1385.
300 TL/09 Add expansion variable $authenticated_fail_id to keep track of
301 last id that failed so it may be referenced in subsequent ACL's.
303 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
306 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
307 ldap library initialization, allowing self-signed CA's to be
308 used. Also properly sets require_cert option later in code by
309 using NULL (global ldap config) instead of ldap handle (per
310 session). Bug diagnosis and testing by alxgomz.
312 TL/12 Enhanced documentation in the ratelimit.pl script provided in
313 the src/util/ subdirectory.
315 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
316 renamed to Transport Post Delivery Action by Jeremy Harris, as
319 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
320 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
321 redis_servers = needs to be configured which will be used by the redis
322 lookup. Patch from Warren Baker, of The Packet Hub.
324 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
326 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
327 hostname or reverse DNS when processing a host list. Used suggestions
328 from multiple comments on this bug.
330 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
332 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
333 Missed a few lines, added it to make the runtest require no keyboard
336 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
337 contains upper case chars. Make router use caseful_local_part.
339 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
340 support when GnuTLS has been built with p11-kit.
346 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
348 This, or similar/improved, will also be change PP/11 of 4.82.
354 PP/01 Handle short writes when writing local log-files.
355 In practice, only affects FreeBSD (8 onwards).
356 Bugzilla 1053, with thanks to Dmitry Isaikin.
358 NM/01 Bugzilla 949 - Documentation tweak
360 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
363 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
365 PP/02 Implemented gsasl authenticator.
367 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
369 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
370 `pkg-config foo` for cflags/libs.
372 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
373 with rest of GSASL and with heimdal_gssapi.
375 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
376 `pkg-config foo` for cflags/libs for the TLS implementation.
378 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
379 properties get this fed in as external SSF. A number of robustness
380 and debugging improvements to the cyrus_sasl authenticator.
382 PP/08 cyrus_sasl server now expands the server_realm option.
384 PP/09 Bugzilla 1214 - Log authentication information in reject log.
385 Patch by Jeremy Harris.
387 PP/10 Added dbmjz lookup type.
389 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
391 PP/12 MAIL args handles TAB as well as SP, for better interop with
392 non-compliant senders.
393 Analysis and variant patch by Todd Lyons.
395 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
396 Bug report from Lars Müller <lars@samba.org> (via SUSE),
397 Patch from Dirk Mueller <dmueller@suse.com>
399 PP/13 tls_peerdn now print-escaped for spool files.
400 Observed some $tls_peerdn in wild which contained \n, which resulted
401 in spool file corruption.
403 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
404 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
405 or write after TLS renegotiation, which otherwise led to messages
408 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
409 as a tracking header (ie: a signed header comes before the signature).
410 Patch from Wolfgang Breyha.
412 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
413 comma-sep list; embedded commas doubled.
415 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
417 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
419 Report and patch from Dmitry Banschikov.
421 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
422 Removed SSL_clear() after SSL_new() which led to protocol negotiation
423 failures. We appear to now support TLS1.1+ with Exim.
425 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
426 lets Exim select keys and certificates based upon TLS SNI from client.
427 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
428 before an outbound SMTP session. New log_selector, +tls_sni.
430 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
431 NULL dereference. Report and patch from Alun Jones.
433 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
434 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
435 Not seeing resolver debug output on NetBSD, but suspect this is a
436 resolver implementation change.
438 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
439 Left warnings. Added "eximon gdb" invocation mode.
441 PP/21 Defaulting "accept_8bitmime" to true, not false.
443 PP/22 Added -bw for inetd wait mode support.
445 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
446 locate the relevant includes and libraries. Made this the default.
448 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
449 Bugzilla 1246, report and most of solution from Tomasz Kusy.
451 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
452 This may cause build issues on older platforms.
454 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
455 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
456 gnutls_require_mac & gnutls_require_protocols (no longer supported).
457 Added SNI support via GnuTLS too.
458 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
460 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
462 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
463 Applied second patch from Janne, implementing suggestion to default
464 multiple-strings-in-record handling to match SPF spec.
466 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
468 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
469 read-only, out of scope).
470 Patch from Wolfgang Breyha, report from Stuart Northfield.
472 PP/29 Fix three issues highlighted by clang analyser static analysis.
473 Only crash-plausible issue would require the Cambridge-specific
474 iplookup router and a misconfiguration.
475 Report from Marcin Mirosław.
477 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
479 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
480 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
481 As part of this, removing so much warning spew let me fix some minor
482 real issues in debug logging.
484 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
485 assignment on my part. Fixed.
487 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
488 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
489 Janne Snabb (who went above and beyond: thank you).
491 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
492 string otherwise requires a connection and a bunch more work and it's
493 relatively easy to get wrong. Should also expose TLS library linkage
496 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
497 64-bit ${eval} (JH/03).
499 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
500 GNU libc to support some of the 64-bit stuff, should not lead to
501 conflicts. Defined before os.h is pulled in, so if a given platform
502 needs to override this, it can.
504 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
505 protection layer was required, which is not implemented.
506 Bugzilla 1254, patch from Wolfgang Breyha.
508 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
509 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
510 tls_dhparam take prime identifiers. Also unbreak combination of
511 OpenSSL+DH_params+TLSSNI.
513 PP/39 Disable SSLv2 by default in OpenSSL support.
519 PP/01 Solaris build fix for Oracle's LDAP libraries.
520 Bugzilla 1109, patch from Stephen Usher.
522 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
524 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
527 TF/02 Fix a couple more cases where we did not log the error message
528 when unlink() failed. See also change 4.74-TF/03.
530 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
531 lock up or crash if it happened to be inside a call to libc when it
532 got a SIGUSR1 from exiwhat.
534 The SIGUSR1 handler appends the current process status to the process
535 log which is later printed by exiwhat. It used to use the general
536 purpose logging code to do this, but several functions it calls are
537 not safe for signals.
539 The new output code in the SIGUSR1 handler is specific to the process
540 log, and simple enough that it's easy to inspect for signal safety.
541 Removing some special cases also simplifies the general logging code.
542 Removing the spurious timestamps from the process log simplifies
545 TF/04 Improved ratelimit ACL condition.
547 The /noupdate option has been deprecated in favour of /readonly which
548 has clearer semantics. The /leaky, /strict, and /readonly update modes
549 are mutually exclusive. The update mode is no longer included in the
550 database key; it just determines when the database is updated. (This
551 means that when you upgrde Exim will forget old rate measurements.)
553 Exim now checks that the per_* options are used with an update mode that
554 makes sense for the current ACL. For example, when Exim is processing a
555 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
556 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
557 must specify per_mail/readonly. If you omit the update mode it defaults to
558 /leaky where that makes sense (as before) or /readonly where required.
560 The /noupdate option is now undocumented but still supported for
561 backwards compatibility. It is equivalent to /readonly except that in
562 ACLs where /readonly is required you may specify /leaky/noupdate or
563 /strict/noupdate which are treated the same as /readonly.
565 A useful new feature is the /count= option. This is a generalization
566 of the per_byte option, so that you can measure the throughput of other
567 aggregate values. For example, the per_byte option is now equivalent
568 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
570 The per_rcpt option has been generalized using the /count= mechanism
571 (though it's more complicated than the per_byte equivalence). When it is
572 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
573 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
574 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
575 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
576 Note that using per_rcpt with a non-readonly update mode in more than
577 one ACL will cause the recipients to be double-counted. (The per_mail
578 and per_byte options don't have this problem.)
580 The handling of very low rates has changed slightly. If the computed rate
581 is less than the event's count (usually one) then this event is the first
582 after a long gap. In this case the rate is set to the same as this event's
583 count, so that the first message of a spam run is counted properly.
585 The major new feature is a mechanism for counting the rate of unique
586 events. The new per_addr option counts the number of different
587 recipients that someone has sent messages to in the last time period. It
588 behaves like per_rcpt if all the recipient addresses are different, but
589 duplicate recipient addresses do not increase the measured rate. Like
590 the /count= option this is a general mechanism, so the per_addr option
591 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
592 example, measure the rate that a client uses different sender addresses
593 with the options per_mail/unique=$sender_address. There are further
594 details in the main documentation.
596 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
598 TF/06 Removed a few PCRE remnants.
600 TF/07 Automatically extract Exim's version number from tags in the git
601 repository when doing development or release builds.
603 PP/02 Raise smtp_cmd_buffer_size to 16kB.
604 Bugzilla 879. Patch from Paul Fisher.
606 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
607 Heavily based on revision 40f9a89a from Simon Arlott's tree.
610 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
612 PP/05 Variable $av_failed, true if the AV scanner deferred.
613 Bugzilla 1078. Patch from John Horne.
615 PP/06 Stop make process more reliably on build failure.
616 Bugzilla 1087. Patch from Heiko Schlittermann.
618 PP/07 Make maildir_use_size_file an _expandable_ boolean.
619 Bugzilla 1089. Patch from Heiko Schlittermann.
621 PP/08 Handle ${run} returning more data than OS pipe buffer size.
622 Bugzilla 1131. Patch from Holger Weiß.
624 PP/09 Handle IPv6 addresses with SPF.
625 Bugzilla 860. Patch from Wolfgang Breyha.
627 PP/10 GnuTLS: support TLS 1.2 & 1.1.
629 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
632 PP/11 match_* no longer expand right-hand-side by default.
633 New compile-time build option, EXPAND_LISTMATCH_RHS.
634 New expansion conditions, "inlist", "inlisti".
636 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
638 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
640 PP/14 fix log_write() format string regression from TF/03.
641 Bugzilla 1152. Patch from Dmitry Isaikin.
647 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
649 PP/02 Harmonised TLS library version reporting; only show if debugging.
650 Layout now matches that introduced for other libraries in 4.74 PP/03.
652 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
654 PP/04 New "dns_use_edns0" global option.
656 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
659 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
660 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
662 TK/01 Updated PolarSSL code to 0.14.2.
663 Bugzilla 1097. Patch from Andreas Metzler.
665 PP/07 Catch divide-by-zero in ${eval:...}.
668 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
671 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
672 format-string attack -- SECURITY: remote arbitrary code execution.
674 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
675 time unintentionally subject to list matching rules, letting the header
676 cause arbitrary Exim lookups (of items which can occur in lists, *not*
677 arbitrary string expansion). This allowed for information disclosure.
679 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
680 INT_MIN/-1 -- value coerced to INT_MAX.
686 NM/01 Workround for PCRE version dependency in version reporting
689 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
690 This fixes portability to compilers other than gcc, notably
691 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
693 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
694 makefiles for portability to HP-UX and POSIX correctness.
696 PP/01 Permit LOOKUP_foo enabling on the make command-line.
697 Also via indented variable definition in the Makefile.
698 (Debugging by Oliver Heesakkers).
700 PP/02 Restore caching of spamd results with expanded spamd_address.
701 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
703 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
704 Improves build reliability. Fix from: Frank Elsner
706 NM/02 Fix wide character breakage in the rfc2047 coding
707 Fixes bug 1064. Patch from Andrey N. Oktyabrski
709 NM/03 Allow underscore in dnslist lookups
710 Fixes bug 1026. Patch from Graeme Fowler
712 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
713 Code patches from Adam Ciarcinski of NetBSD.
715 NM/04 Fixed exiqgrep to cope with mailq missing size issue
718 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
719 is logged, to avoid truncation. Patch from John Horne.
721 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
722 Patch from Jakob Hirsch.
724 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
725 SQL string expansion failure details.
726 Patch from Andrey Oktyabrski.
728 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
729 Patch from Simon Arlott.
731 PP/09 New lookups functionality failed to compile on old gcc which rejects
732 extern declarations in function scope.
733 Patch from Oliver Fleischmann
735 PP/10 Use sig_atomic_t for flags set from signal handlers.
736 Check getgroups() return and improve debugging.
737 Fixed developed for diagnosis in bug 927 (which turned out to be
740 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
741 Patch from Mark Zealey.
743 PP/12 Bugzilla 1056: Improved spamd server selection.
744 Patch from Mark Zealey.
746 PP/13 Bugzilla 1086: Deal with maildir quota file races.
747 Based on patch from Heiko Schlittermann.
749 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
750 Patch from Uwe Doering, sign-off by Michael Haardt.
752 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
753 variable declaration deep within a block. Bug and patch from
756 PP/15 lookups-Makefile IRIX compatibilty coercion.
758 PP/16 Make DISABLE_DKIM build knob functional.
760 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
761 Patch by Simon Arlott
763 TF/03 Fix valgrind.h portability to C89 compilers that do not support
764 variable argument macros. Our copy now differs from upstream.
770 TF/01 Failure to get a lock on a hints database can have serious
771 consequences so log it to the panic log.
773 TF/02 Log LMTP confirmation messages in the same way as SMTP,
774 controlled using the smtp_confirmation log selector.
776 TF/03 Include the error message when we fail to unlink a spool file.
778 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
779 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
780 for maintaining out-of-tree patches for some time.
782 PP/01 Bugzilla 139: Documentation and portability issues.
783 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
784 Handle per-OS dynamic-module compilation flags.
786 PP/02 Let /dev/null have normal permissions.
787 The 4.73 fixes were a little too stringent and complained about the
788 permissions on /dev/null. Exempt it from some checks.
789 Reported by Andreas M. Kirchwitz.
791 PP/03 Report version information for many libraries, including
792 Exim version information for dynamically loaded libraries. Created
793 version.h, now support a version extension string for distributors
794 who patch heavily. Dynamic module ABI change.
796 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
797 privilege escalation vulnerability whereby the Exim run-time user
798 can cause root to append content of the attacker's choosing to
801 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
804 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
805 If dropping privileges for untrusted macros, we disabled normal logging
806 on the basis that it would fail; for the Exim run-time user, this is not
807 the case, and it resulted in successful deliveries going unlogged.
808 Fixed. Reported by Andreas Metzler.
814 PP/01 Date: & Message-Id: revert to normally being appended to a message,
815 only prepend for the Resent-* case. Fixes regression introduced in
816 Exim 4.70 by NM/22 for Bugzilla 607.
818 PP/02 Include check_rfc2047_length in configure.default because we're seeing
819 increasing numbers of administrators be bitten by this.
821 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
823 PP/03 Bugzilla 994: added openssl_options main configuration option.
825 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
827 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
829 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
831 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
832 without a peer certificate, leading to a segfault because of an
833 assumption that peers always have certificates. Be a little more
834 paranoid. Problem reported by Martin Tscholak.
836 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
837 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
838 NB: ClamAV planning to remove STREAM in "middle of 2010".
839 CL also introduces -bmalware, various -d+acl logging additions and
840 more caution in buffer sizes.
842 PP/09 Implemented reverse_ip expansion operator.
844 PP/10 Bugzilla 937: provide a "debug" ACL control.
846 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
848 PP/12 Bugzilla 973: Implement --version.
850 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
852 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
854 PP/15 Bugzilla 816: support multiple condition rules on Routers.
856 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
857 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
858 ignore trailing whitespace.
860 JJ/02 prevent non-panic DKIM error from being sent to paniclog
862 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
865 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
866 Notification from Dr Andrew Aitchison.
868 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
869 ExtendedDetectionInfo response format.
870 Notification from John Horne.
872 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
875 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
876 XSL and documented dependency on system catalogs, with examples of how
879 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
882 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
883 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
884 configuration file which is writeable by the Exim user or group.
886 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
887 of configuration files to cover files specified with the -C option if
888 they are going to be used with root privileges, not just the default
891 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
892 option (effectively making it always true).
894 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
895 files to be used while preserving root privileges.
897 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
898 that rogue child processes cannot use them.
900 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
901 run-time user, instead of root.
903 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
904 Exim run-time user without dropping privileges.
906 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
907 result string, instead of calling string_vformat() twice with the same
910 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
911 for other users. Others should always drop root privileges if they use
912 -C on the command line, even for a whitelisted configure file.
914 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
916 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
922 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
923 $data_path, and $header_path variables; fixed documentation bugs and
926 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
927 exipick to access non-standard spools, including the "frozen" queue
930 NM/01 Bugzilla 965: Support mysql stored procedures.
931 Patch from Alain Williams
933 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
935 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
936 Patch from Andreas Metzler
938 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
939 Patch from Kirill Miazine
941 NM/05 Bugzilla 671: Added umask to procmail example.
943 JJ/03 installed exipick 20100323.0, fixing doc bug
945 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
946 directory. Notification and patch from Dan Rosenberg.
948 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
950 TK/02 Improve log output when DKIM signing operation fails.
952 MH/01 Treat the transport option dkim_domain as a colon separated
953 list, not as a single string, and sign the message with each element,
954 omitting multiple occurences of the same signer.
956 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
957 Bugzilla 985, 986. Patch by Simon Arlott
959 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
960 Patch by Simon Arlott
962 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
963 MBX locking. Notification from Dan Rosenberg.
969 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
971 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
973 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
975 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
977 NM/04 Bugzilla 915: Flag broken perl installation during build.
983 TK/01 Added patch by Johannes Berg that expands the main option
984 "spamd_address" if it starts with a dollar sign.
986 TK/02 Write list of recipients to X-Envelope-Sender header when building
987 the mbox-format spool file for content scanning (suggested by Jakob
990 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
991 (http://www.dcc-servers.net/) support via dccifd. Activated by
992 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
994 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
995 by Mark Daniel Reidel <mr@df.eu>.
997 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
998 When building exim an external PCRE library is now needed -
999 PCRE is a system library on the majority of modern systems.
1000 See entry on PCRE_LIBS in EDITME file.
1002 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1003 conversation. Added nologin parameter to request.
1004 Patch contributed by Kirill Miazine.
1006 TF/01 Do not log submission mode rewrites if they do not change the address.
1008 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1010 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1011 log files in place. Contributed by Roberto Lima.
1013 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1015 TF/03 Bugzilla 615: When checking the local_parts router precondition
1016 after a local_part_suffix or local_part_prefix option, Exim now
1017 does not use the address's named list lookup cache, since this
1018 contains cached lookups for the whole local part.
1020 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1021 Robert Millan. Documentation is in experimental-spec.txt.
1023 TF/04 Bugzilla 668: Fix parallel build (make -j).
1025 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1027 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1028 Patch provided by Jan Srzednicki.
1030 TF/05 Leading white space used to be stripped from $spam_report which
1031 wrecked the formatting. Now it is preserved.
1033 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1034 that they are available at delivery time.
1036 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1038 TF/08 TLS error reporting now respects the incoming_interface and
1039 incoming_port log selectors.
1041 TF/09 Produce a more useful error message if an SMTP transport's hosts
1042 setting expands to an empty string.
1044 NM/06 Bugzilla 744: EXPN did not work under TLS.
1045 Patch provided by Phil Pennock.
1047 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1048 Patch provided by Richard Godbee.
1050 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1051 acl_smtp_notquit, added index entry.
1053 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1054 Patch provided by Eugene Bujak.
1056 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1057 accept(). Patch provided by Maxim Dounin.
1059 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1060 Patch provided by Phil Pennock.
1062 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1064 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1065 Patch provided by Brad "anomie" Jorsch.
1067 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1068 Patch provided by Dean Brooks.
1070 TK/05 Add native DKIM support (does not depend on external libraries).
1072 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1073 Patch provided by Graeme Fowler.
1075 NM/16 Bugzilla 851: Documentation example syntax fix.
1077 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1079 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1082 NM/19 Bugzilla 745: TLS version reporting.
1083 Patch provided by Phil Pennock.
1085 NM/20 Bugzilla 167: bool: condition support.
1086 Patch provided by Phil Pennock.
1088 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1089 clients. Patch provided by Phil Pennock.
1091 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1092 Patch provided by Brad "anomie" Jorsch.
1094 NM/23 Bugzilla 687: Fix misparses in eximstats.
1095 Patch provided by Heiko Schlittermann.
1097 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1098 Patch provided by Heiko Schlittermann.
1100 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1101 plus update to original patch.
1103 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1105 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1106 Patch provided by David Brownlee.
1108 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1110 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1112 NM/30 Bugzilla 888: TLS documentation bugfixes.
1114 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1116 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1117 Unlike the original bugzilla I have changed all shell scripts in src tree.
1119 NM/33 Bugzilla 898: Transport filter timeout fix.
1120 Patch by Todd Rinaldo.
1122 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1123 Patch by Serge Demonchaux.
1125 NM/35 Bugzilla 39: Base64 decode bug fixes.
1126 Patch by Jakob Hirsch.
1128 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1130 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1132 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1134 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1140 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1141 ALT-N's libdkim that I have put here:
1142 http://duncanthrax.net/exim-experimental/
1144 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1145 were called 'true' and it seems that C99 defines that as a reserved
1146 keyword to be used with 'bool' variable types. That means you could
1147 not include C99-style headers which use bools without triggering
1148 build errors in sieve.c.
1150 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1151 as mailq or other aliases. Changed the --help handling significantly
1152 to do whats expected. exim_usage() emits usage/help information.
1154 SC/01 Added the -bylocaldomain option to eximstats.
1156 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1158 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1160 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1166 PH/01 Another patch from the Sieve maintainer.
1168 PH/02 When an IPv6 address is converted to a string for single-key lookup
1169 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1170 dots are used instead of colons so that keys in lsearch files need not
1171 contain colons. This was done some time before quoting was made available
1172 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1173 (notated using the quote facility) so as to distinguish them from IPv4
1174 keys. This meant that lookups for IP addresses in host lists did not work
1175 for iplsearch lookups.
1177 This has been fixed by arranging for IPv6 addresses to be expressed with
1178 colons if the lookup type is iplsearch. This is not incompatible, because
1179 previously such lookups could never work.
1181 The situation is now rather anomolous, since one *can* have colons in
1182 ordinary lsearch keys. However, making the change in all cases is
1183 incompatible and would probably break a number of configurations.
1185 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1188 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1189 conversion specification without a maximum field width, thereby enabling
1190 a rogue spamd server to cause a buffer overflow. While nobody in their
1191 right mind would setup Exim to query an untrusted spamd server, an
1192 attacker that gains access to a server running spamd could potentially
1193 exploit this vulnerability to run arbitrary code as the Exim user.
1195 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1196 $primary_hostname instead of what libspf2 thinks the hosts name is.
1198 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1199 a directory entry by the name of the lookup key. Previously, if a
1200 symlink pointed to a non-existing file or a file in a directory that
1201 Exim lacked permissions to read, a lookup for a key matching that
1202 symlink would fail. Now it is enough that a matching directory entry
1203 exists, symlink or not. (Bugzilla 503.)
1205 PH/03 The body_linecount and body_zerocount variables are now exported in the
1208 PH/04 Added the $dnslist_matched variable.
1210 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1211 This means they are set thereafter only if the connection becomes
1214 PH/06 Added the client_condition to authenticators so that some can be skipped
1215 by clients under certain conditions.
1217 PH/07 The error message for a badly-placed control=no_multiline_responses left
1218 "_responses" off the end of the name.
1220 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1222 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1223 (without spaces) instead of just copying the configuration text.
1225 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1227 PH/11 Added $max_received_linelength.
1229 PH/12 Added +ignore_defer and +include_defer to host lists.
1231 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1232 way in which PCRE > 7.0 is built.
1234 PH/14 Implemented queue_only_load_latch.
1236 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1237 MAIL command. The effect was to mangle the value on 64-bit systems.
1239 PH/16 Another patch from the Sieve maintainer.
1241 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1243 PH/18 If a system quota error occurred while trying to create the file for
1244 a maildir delivery, the message "Mailbox is full" was not appended to the
1245 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1246 only to a quota excession during the actual writing of the file.
1248 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1249 characters?) which causes problems in log lines. The DN values are now
1250 passed through string_printing() before being added to log lines.
1252 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1253 and InterBase are left for another time.)
1255 PH/21 Added message_body_newlines option.
1257 PH/22 Guard against possible overflow in moan_check_errorcopy().
1259 PH/23 POSIX allows open() to be a macro; guard against that.
1261 PH/24 If the recipient of an error message contained an @ in the local part
1262 (suitably quoted, of course), incorrect values were put in $domain and
1263 $local_part during the evaluation of errors_copy.
1269 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1270 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1273 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1274 issue a MAIL command.
1276 PH/02 In an ACL statement such as
1278 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1280 if a client was not listed at all, or was listed with a value other than
1281 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1282 the condition was not true (as it should be), so access was not denied.
1283 The bug was that the ! inversion was incorrectly passed on to the second
1284 item. This has been fixed.
1286 PH/03 Added additional dnslists conditions == and =& which are different from
1287 = and & when the dns lookup returns more than one IP address.
1289 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1290 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1292 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1293 FSYNC, which compiles an option called disable_fsync that allows for
1294 bypassing fsync(). The documentation is heavily laced with warnings.
1296 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1298 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1299 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1300 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1301 including adding "make clean"; (3) Added -fPIC when compiling the test
1302 dynamically loaded module, to get rid of a warning.
1304 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1305 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1306 = 0s. The bug is otherwise harmless.
1308 PH/07 There was a bug in the dovecot authenticator such that the value of
1309 $auth1 could be overwritten, and so not correctly preserved, after a
1310 successful authentication. This usually meant that the value preserved by
1311 the server_setid option was incorrect.
1313 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1315 PH/09 Installed PCRE release 7.0.
1317 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1318 run for batched SMTP input. It is now run at the start of every message
1319 in the batch. While fixing this I discovered that the process information
1320 (output by running exiwhat) was not always getting set for -bs and -bS
1321 input. This is fixed, and it now also says "batched" for BSMTP.
1323 PH/11 Added control=no_pipelining.
1325 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1326 patch, slightly modified), and move the expansion of helo_data till after
1327 the connection is made in the smtp transport (so it can use these
1330 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1332 PH/14 Added log_selector = +pid.
1334 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1336 PH/16 Add ${if forany and ${if forall.
1338 PH/17 Added dsn_from option to vary the From: line in DSNs.
1340 PH/18 Flush SMTP output before performing a callout, unless control =
1341 no_callout_flush is set.
1343 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1344 was true (the default) a successful delivery failed to delete the retry
1345 item, thus causing premature timeout of the address. The bug is now
1348 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1350 PH/21 Long custom messages for fakedefer and fakereject are now split up
1351 into multiline reponses in the same way that messages for "deny" and
1352 other ACL rejections are.
1354 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1355 with slight modification.
1357 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1358 draft, changing the syntax and factoring some duplicate code".
1360 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1361 for deliveries of the second and subsequent messages over the same SMTP
1364 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1365 ${reduce, with only minor "tidies".
1367 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1369 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1370 expansion side effects.
1372 PH/27 When a message times out after an over-quota error from an Exim-imposed
1373 quota, the bounce message says "mailbox is full". This message was not
1374 being given when it was a system quota that was exceeded. It now should
1377 MH/03 Made $recipients available in local_scan(). local_scan() already has
1378 better access to the recipient list through recipients_list[], but
1379 $recipients can be useful in postmaster-provided expansion strings.
1381 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1382 in the case of a MAIL command with additional options following the
1383 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1384 were accidentally chopped off.
1386 PH/29 SMTP synchronization checks are implemented when a command is read -
1387 there is a check that no more input is waiting when there shouldn't be
1388 any. However, for some commands, a delay in an ACL can mean that it is
1389 some time before the response is written. In this time, more input might
1390 arrive, invalidly. So now there are extra checks after an ACL has run for
1391 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1392 pipelining has not been advertised.
1394 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1396 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1397 returned in the SMTP response when smtp_return_error_details was false.
1398 This has been fixed.
1400 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1401 instead of the C I/O that was originally supplied, because problems were
1402 reported on Solaris.
1404 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1405 Exim which did not show up earlier: it was assuming that a call to
1406 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1407 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1408 was a macro that became an assignment, so it seemed to work. This has
1409 changed to a proper function call with a void return, hence the compile
1410 error. Exim's code has been fixed.
1412 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1415 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1417 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1419 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1420 because it turns out to be too big or there is a timeout) there is no
1421 "Completed" line in the log. When some messages of this type were
1422 selected by exigrep, they were listed as "not completed". Others were
1423 picked up by some special patterns. I have improved the selection
1424 criteria to be more general.
1426 PH/38 The host_find_failed option in the manualroute router can now be set
1427 to "ignore", to completely ignore a host whose IP address cannot be
1428 found. If all hosts are ignored, the behaviour is controlled by the new
1429 host_all_ignored option.
1431 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1432 homing or because of multiple MX records with /mx) generated more than
1433 one IP address, and the following item turned out to be the local host,
1434 all the secondary addresses of the first item were incorrectly removed
1435 from the list, along with the local host and any following hosts (which
1436 is what is supposed to happen).
1438 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1439 whoever called Exim into the -H file. In the case of the daemon it was
1440 behaving confusingly. When first started, it used values for whoever
1441 started the daemon, but after a SIGHUP it used the Exim user (because it
1442 calls itself on a restart). I have changed the code so that it now always
1445 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1446 message are rejected with the same error (e.g. no authentication or bad
1447 sender address), and a DATA command is nevertheless sent (as can happen
1448 with PIPELINING or a stupid MUA), the error message that was given to the
1449 RCPT commands is included in the rejection of the DATA command. This is
1450 intended to be helpful for MUAs that show only the final error to their
1453 PH/42 Another patch from the Sieve maintainer.
1455 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1456 Eximstats - Fixed some broken HTML links and added missing column headers
1458 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1459 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1461 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1463 PH/43 Yet another patch from the Sieve maintainer.
1465 PH/44 I found a way to check for a TCP/IP connection going away before sending
1466 the response to the final '.' that terminates a message, but only in the
1467 case where the client has not sent further data following the '.'
1468 (unfortunately, this is allowed). However, in many cases there won't be
1469 any further data because there won't be any more messages to send. A call
1470 to select() can be used: if it shows that the input is "ready", there is
1471 either input waiting, or the socket has been closed. An attempt to read
1472 the next input character can distinguish the two cases. Previously, Exim
1473 would have sent an OK response which the client would never have see.
1474 This could lead to message repetition. This fix should cure that, at
1475 least in a lot of common cases.
1477 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1478 advertised in response to EHLO.
1484 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1485 fixed by 4.65/MH/01 (is this a record?) are fixed:
1487 (i) An empty string was always treated as zero by the numeric comparison
1488 operators. This behaviour has been restored.
1490 (ii) It is documented that the numeric comparison operators always treat
1491 their arguments as decimal numbers. This was broken in that numbers
1492 starting with 0 were being interpreted as octal.
1494 While fixing these problems I realized that there was another issue that
1495 hadn't been noticed. Values of message_size_limit (both the global option
1496 and the transport option) were treated as octal if they started with 0.
1497 The documentation was vague. These values are now always treated as
1498 decimal, and I will make that clear in the documentation.
1504 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1505 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1508 MH/01 Don't check that the operands of numeric comparison operators are
1509 integers when their expansion is in "skipping" mode (fixes bug
1510 introduced by 4.64-PH/07).
1512 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1513 child addresses, Exim now panics and dies. Previously, because the count
1514 is held in a short int, deliveries were likely to be lost. As such a
1515 large number of recipients for a single message is ridiculous
1516 (performance will be very, very poor), I have chosen to impose a limit
1517 rather than extend the field.
1523 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1524 leftover -K file (the existence of which was triggered by #402).
1525 While we were at it, introduced process PID as part of the -K
1526 filename. This should rule out race conditions when creating
1529 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1530 processing considerably. Previous code took too long for large mails,
1531 triggering a timeout which in turn triggers #401.
1533 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1534 in the DK code in transports.c. sendfile() is not really portable,
1535 hence the _LINUX specificness.
1537 TF/01 In the add_headers option to the mail command in an Exim filter,
1538 there was a bug that Exim would claim a syntax error in any
1539 header after the first one which had an odd number of characters
1542 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1543 callout verification, Exim cached a "reject" for the entire domain. This
1544 is correct for most verifications, but it is not correct for a recipient
1545 verification with use_sender or use_postmaster set, because in that case
1546 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1547 case of MAIL FROM:<> rejection from other early rejections (e.g.
1548 rejection of HELO). When verifying a recipient using a non-null MAIL
1549 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1550 Whatever the result of the callout, the value of the domain cache is
1551 left unchanged (for any other kind of callout, getting as far as trying
1552 RCPT means that the domain itself is ok).
1554 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1557 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1558 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1559 session, the daemon ignores ECONNECT errors and logs others; it now
1560 ignores EPIPE as well.
1562 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1563 (quoted-printable decoding).
1565 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1566 later the small subsequent patch to fix an introduced bug.
1568 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1570 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1572 PH/08 An error is now given if message_size_limit is specified negative.
1574 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1575 to be given (somewhat) arbitrary names.
1577 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1580 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1581 miscellaneous code fixes
1583 PH/10 Added the log_reject_target ACL modifier to specify where to log
1586 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1587 hostname. This is wrong, because it relates to the incoming message (and
1588 probably the interface on which it is arriving) and not to the outgoing
1589 callout (which could be using a different interface). This has been
1590 changed to use the value of the helo_data option from the smtp transport
1591 instead - this is what is used when a message is actually being sent. If
1592 there is no remote transport (possible with a router that sets up host
1593 addresses), $smtp_active_hostname is used.
1595 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1596 tweaks were necessary in order to get it to work (see also 21 below):
1597 (a) The code assumed that strncpy() returns a negative number on buffer
1598 overflow, which isn't the case. Replaced with Exim's string_format()
1600 (b) There were several signed/unsigned issues. I just did the minimum
1601 hacking in of casts. There is scope for a larger refactoring.
1602 (c) The code used strcasecmp() which is not a standard C function.
1603 Replaced with Exim's strcmpic() function.
1604 (d) The code set only $1; it now sets $auth1 as well.
1605 (e) A simple test gave the error "authentication client didn't specify
1606 service in request". It would seem that Dovecot has changed its
1607 interface. Fortunately there's a specification; I followed it and
1608 changed what the client sends and it appears to be working now.
1610 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1613 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1614 address A is aliased to B and C, where B exists and C does not. Without
1615 -v the output is "A verified" because verification stops after a
1616 successful redirection if more than one address is generated. However,
1617 with -v the child addresses are also verified. Exim was outputting "A
1618 failed to verify" and then showing the successful verification for C,
1619 with its parentage. It now outputs "B failed to verify", showing B's
1620 parentage before showing the successful verification of C.
1622 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1623 look up a TXT record in a specific list after matching in a combined
1626 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1627 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1628 they consult the DNS. I had assumed they would set it the way they
1629 wanted; and indeed my experiments on Linux seem to show that in some
1630 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1631 To be on the safe side, however, I have now made the interface to
1632 host_find_byname() similar to host_find_bydns(), with an argument
1633 containing the DNS resolver options. The host_find_byname() function now
1634 sets these options at its start, just as host_find_bydns() does. The smtp
1635 transport options dns_qualify_single and dns_search_parents are passed to
1636 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1637 of host_find_byname() use the default settings of RES_DEFNAMES
1638 (qualify_single) but not RES_DNSRCH (search_parents).
1640 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1641 spool_read_header() do less string testing, by means of a preliminary
1642 switch on the second character of optional "-foo" lines. (This is
1643 overdue, caused by the large number of possibilities that now exist.
1644 Originally there were few.) While I was there, I also converted the
1645 str(n)cmp tests so they don't re-test the leading "-" and the first
1646 character, in the hope this might squeeze out yet more improvement.
1648 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1649 flag allowing group syntax was set by the header_syntax check but not
1650 turned off, possible causing trouble later; (2) The flag was not being
1651 set at all for the header_verify test, causing "group"-style headers to
1652 be rejected. I have now set it in this case, and also caused header_
1653 verify to ignore an empty address taken from a group. While doing this, I
1654 came across some other cases where the code for allowing group syntax
1655 while scanning a header line wasn't quite right (mostly, not resetting
1656 the flag correctly in the right place). These bugs could have caused
1657 trouble for malformed header lines. I hope it is now all correct.
1659 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1660 with the "reply" argument non-NULL. The code, however (which originally
1661 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1662 but it didn't always do it. This confused somebody who was copying the
1663 code for some other use. I have removed all the tests.
1665 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1666 feature that was used to support insecure browsers during the U.S. crypto
1667 embargo. It requires special client support, and Exim is probably the
1668 only MTA that supported it -- and would never use it because real RSA is
1669 always available. This code has been removed, because it had the bad
1670 effect of slowing Exim down by computing (never used) parameters for the
1671 RSA_EXPORT functionality.
1673 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1674 authenticator to fail if there's a tab character in the incoming data
1675 (there should never be unless someone is messing about, as it's supposed
1676 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1677 if the connection is using TLS or if the remote IP is the same as the
1678 local IP, and the "valid-client-cert option" if a client certificate has
1681 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1682 authenticators. This can be used for authorization after authentication
1683 succeeds. (In the case of plaintext, it servers for both authentication
1686 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1687 if any retry times were supplied.
1689 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1690 connection, where there is no client IP address to check. In this
1691 situation, the verify now always succeeds.
1693 PH/25 Applied John Jetmore's -Mset patch.
1695 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1697 PH/27 In a string expansion for a processed (not raw) header when multiple
1698 headers of the same name were present, leading whitespace was being
1699 removed from all of them, but trailing whitespace was being removed only
1700 from the last one. Now trailing whitespace is removed from each header
1701 before concatenation. Completely empty headers in a concatenation (as
1702 before) are ignored.
1704 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1705 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1707 PH/29 [Removed. This was a change that I later backed out, and forgot to
1708 correct the ChangeLog entry (that I had efficiently created) before
1709 committing the later change.]
1711 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1712 address errors (4xx response to RCPT) over the same connection as other
1713 messages routed to the same hosts. Such deliveries are always "forced",
1714 so retry times are not inspected. This resulted in far too many retries
1715 for the affected addresses. The effect occurred only when there were more
1716 hosts than the hosts_max_try setting in the smtp transport when it had
1717 the 4xx errors. Those hosts that it had tried were not added to the list
1718 of hosts for which the message was waiting, so if all were tried, there
1719 was no problem. Two fixes have been applied:
1721 (i) If there are any address or message errors in an SMTP delivery, none
1722 of the hosts (tried or untried) are now added to the list of hosts
1723 for which the message is waiting, so the message should not be a
1724 candidate for sending over the same connection that was used for a
1725 successful delivery of some other message. This seems entirely
1726 reasonable: after all the message is NOT "waiting for some host".
1727 This is so "obvious" that I'm not sure why it wasn't done
1728 previously. Hope I haven't missed anything, but it can't do any
1729 harm, as the worst effect is to miss an optimization.
1731 (ii) If, despite (i), such a delivery is accidentally attempted, the
1732 routing retry time is respected, so at least it doesn't keep
1733 hammering the server.
1735 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1736 in ${readsocket because some servers need this prod.
1738 PH/32 Added some extra debug output when updating a wait-xxx database.
1740 PH/33 The hint "could be header name not terminated by colon", which has been
1741 given for certain expansion errors for a long time, was not being given
1742 for the ${if def:h_colon_omitted{... case.
1744 PH/34 The spec says: "With one important exception, whenever a domain list is
1745 being scanned, $domain contains the subject domain." There was at least
1746 one case where this was not true.
1748 PH/35 The error "getsockname() failed: connection reset by peer" was being
1749 written to the panic log as well as the main log, but it isn't really
1750 panic-worthy as it just means the connection died rather early on. I have
1751 removed the panic log writing for the ECONNRESET error when getsockname()
1754 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1755 runs only) independently of the message's sender address. This meant
1756 that, if the 4xx error was in fact related to the sender, a different
1757 message to the same recipient with a different sender could confuse
1758 things. In particualar, this can happen when sending to a greylisting
1759 server, but other circumstances could also provoke similar problems.
1760 I have changed the default so that the retry time for these errors is now
1761 based a combination of the sender and recipient addresses. This change
1762 can be overridden by setting address_retry_include_sender=false in the
1765 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1766 remote server are returned as part of bounce messages. This was not
1767 happening for LMTP over a pipe (the lmtp transport), but now it is the
1768 same for both kinds of LMTP.
1770 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1771 in header lines that were in fact CNAMEs. This is no longer the case.
1773 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1774 and queue runs started by the daemon processed all messages. This has
1775 been fixed so that -R and -S can now usefully be given with -q<time>.
1777 PH/40 Import PCRE release 6.7 (fixes some bugs).
1779 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1781 PH/42 Give an error if -q is specified more than once.
1783 PH/43 Renamed the variables $interface_address and $interface_port as
1784 $received_ip_address and $received_port, to make it clear that these
1785 values apply to message reception, and not to the outgoing interface when
1786 a message is delivered. (The old names remain recognized, of course.)
1788 PH/44 There was no timeout on the connect() call when using a Unix domain
1789 socket in the ${readsocket expansion. There now is.
1791 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1792 be meaningful with "accept".
1794 SC/01 Eximstats V1.43
1795 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1797 SC/02 Eximstats V1.44
1798 Use a glob alias rather than an array ref in the generated
1799 parser. This improves both readability and performance.
1801 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1802 Collect SpamAssassin and rejection statistics.
1803 Don't display local sender or destination tables unless
1804 there is data to show.
1805 Added average volumes into the top table text output.
1807 SC/04 Eximstats V1.46
1808 Collect data on the number of addresses (recipients)
1809 as well as the number of messages.
1811 SC/05 Eximstats V1.47
1812 Added 'Message too big' to the list of mail rejection
1813 reasons (thanks to Marco Gaiarin).
1815 SC/06 Eximstats V1.48
1816 Mainlog lines which have GMT offsets and are too short to
1817 have a flag are now skipped.
1819 SC/07 Eximstats V1.49 (Alain Williams)
1820 Added the -emptyok flag.
1822 SC/08 Eximstats V1.50
1823 Fixes for obtaining the IP address from reject messages.
1825 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1826 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1827 whitesspace changes from 4.64-PH/27
1829 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1832 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1833 are found, allow negative numbers in numeric criteria)
1835 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1837 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1838 to match changes made in 4.64-PH/43
1840 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1842 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1843 "domain" is actually an IP address, and force a failure. This locks out
1844 those revolvers/nameservers that support "A-for-A" lookups, in
1845 contravention of the specifications.
1847 PH/48 When a host name was looked up from an IP address, and the subsequent
1848 forward lookup of the name timed out, the host name was left in
1849 $sender_host_name, contrary to the specification.
1851 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1852 restricted to single-key lookups, Exim was not diagnosing an error if
1853 * or *@ was used with a query-style lookup.
1855 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1857 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1858 long ago, but noone interested enough thought of it. Let's just say that
1859 the "1.1" means that there are some new functions that weren't there at
1860 some point in the past.
1862 PH/51 Error processing for expansion failure of helo_data from an smtp
1863 transport during callout processing was broken.
1865 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1866 tested/used via the -bh/-bhc/-bs options.
1868 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1869 bug, fixed in subsequent PCRE releases).
1871 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1872 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1874 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1880 SC/01 Use a glob alias rather than an array ref in eximstats generated
1881 parser. This improves both readability and performance.
1883 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1884 Don't display local sender or destination tables in eximstats unless
1885 there is data to show.
1886 Added average volumes into the eximstats top table text output.
1888 SC/03 Collect data on the number of addresses (recipients) as well
1889 as the number of messages in eximstats.
1891 TF/01 Correct an error in the documentation for the redirect router. Exim
1892 does (usually) call initgroups() when daemonizing.
1894 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1895 with consistent privilege compared to when running as a daemon.
1897 TF/03 Note in the spec that $authenticated_id is not set for local
1898 submissions from trusted users.
1900 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1901 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1903 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1904 by adding some example configuration directives to the default
1905 configuration file. A little bit of work is required to uncomment the
1906 directives and define how usernames and passwords are checked, but
1907 there is now a framework to start from.
1909 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1910 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1911 without this. I don't know how relevant this is to other LDAP libraries.
1913 PH/02 Add the verb name to the "unknown ACL verb" error.
1915 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1917 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1919 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1920 directory not expanded when it should be if an expanded home directory
1921 was set for the address (which is overridden by the transport).
1923 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1926 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1927 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1928 because it is too late at that time, and has no effect.
1930 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1931 security issue with \' (bugzilla #107). I could not use the
1932 PQescapeStringConn() function, because it needs a PGconn value as one of
1935 PH/08 When testing addresses using -bt, indicate those final addresses that
1936 are duplicates that would not cause an additional delivery. At least one
1937 person was confused, thinking that -bt output corresponded to deliveries.
1938 (Suppressing duplicates isn't a good idea as you lose the information
1939 about possibly different redirections that led to the duplicates.)
1941 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1942 systems where poll() doesn't work, in particular OS X.
1944 PH/10 Added more information to debugging output for retry time not reached.
1946 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1947 operations in malware.c.
1949 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1952 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1953 syslog_duplication set false, log lines that would normally be written
1954 both the the main log and to the reject log were not written to syslog at
1957 PH/14 In the default configuration, change the use of "message" in ACL warn
1958 statements to "add_header".
1960 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1961 not followed by a command (e.g. "seen endif").
1963 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1964 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1967 PH/17 Added extra conditions to the default value of delay_warning_condition
1971 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1972 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1973 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1976 The Auto-Submitted: and various List- headers are standardised, whereas I
1977 don't think Precedence: ever was.
1979 PH/18 Refactored debugging code in route_finduser() to show more information,
1980 in particular, the error code if getpwnam() issues one.
1982 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1983 This is apparently needed in addition to the PH/07 change above to avoid
1984 any possible encoding problems.
1986 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1987 but not after initializing Perl.
1989 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1990 output them only if debugging. By default they are written stderr,
1991 apparently, which is not desirable.
1993 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1996 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1999 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2001 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2002 authenticated or an ident call has been made. Suppress the default
2003 values for $authenticated_id and $authenticated_sender (but permit -oMai
2004 and -oMas) when testing with -bh.
2006 PH/24 Re-jigged the order of the tests in the default configuration so that the
2007 tests for valid domains and recipients precede the DNS black list and CSA
2008 tests, on the grounds that those ones are more expensive.
2010 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2011 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2012 This bug exists in every version of Exim that I still have, right back to
2015 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2016 However, an attempt to turn on case-sensitivity in a regex key by
2017 including (?-i) didn't work because the subject string was already
2018 lowercased, and the effects were non-intuitive. It turns out that a
2019 one-line patch can be used to allow (?-i) to work as expected.
2025 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2026 other effects) broke the use of negated acl sub-conditions.
2028 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2031 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2032 "Deny" causes Exim to reject the incoming connection with a 554 error.
2033 Unfortunately, if there is a major crisis, such as a disk failure,
2034 tcp-wrappers gives "deny", whereas what one would like would be some
2035 kind of temporary error. A kludge has been added to help with this.
2036 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2037 554 error is used if errno is still zero or contains ENOENT (which occurs
2038 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2041 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2043 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2044 errors. Otherwise a message that provokes a temporary error (when other
2045 messages do not) can cause a whole host to time out.
2047 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2048 addresses were routed directly to files or pipes from a redirect router.
2049 File deliveries just didn't batch; pipe deliveries might have suffered
2052 PH/06 A failure to get a lock for a hints database would erroneously always say
2053 "Failed to get write lock", even when it was really a read lock.
2055 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2056 of 0600. This has been changed to use the value of the lockfile_mode
2057 option (which defaults to 0600).
2059 PH/08 Applied small patch from the Sieve maintainer.
2061 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2062 folder from quota calculations, a direct delivery into this folder messed
2063 up the contents of the maildirsize file. This was because the regex was
2064 used only to exclude .Trash (or whatever) when the size of the mailbox
2065 was calculated. There was no check that a delivery was happening into an
2066 excluded directory. This bug has been fixed by ignoring all quota
2067 processing for deliveries into excluded directories.
2069 PH/10 Added the maildirfolder_create_regex option to appendfile.
2075 PH/01 The code for finding all the local interface addresses on a FreeBSD
2076 system running IPv6 was broken. This may well have applied to all BSD
2077 systems, as well as to others that have similar system calls. The broken
2078 code found IPv4 interfaces correctly, but gave incorrect values for the
2079 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2080 that it would not match correctly against @[] and not recognize the IPv6
2083 PH/02 The ipliteral router was not recognizing addresses of the form user@
2084 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2086 PH/03 Added disable_ipv6.
2088 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2089 decoded form, because it is most often used to construct To: headers
2090 lines in autoreplies, and the decoded form may well be syntactically
2091 invalid. However, $reply_address has leading white space removed, and all
2092 newlines turned into spaces so that the autoreply transport does not
2095 PH/05 If group was specified without a user on a router, and no group or user
2096 was specified on a transport, the group from the router was ignored.
2098 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2099 for visible compile-time settings that can be used to change these
2100 numbers, for those that want even more. Backwards compatibility with old
2101 spool files has been maintained. However, going back to a previous Exim
2102 release will lost any variables that are in spool files.
2104 PH/07 Two small changes when running in the test harness: increase delay when
2105 passing a TCP/IP connection to a new process, in case the original
2106 process has to generate a bounce, and remove special handling of
2107 127.0.0.2 (sic), which is no longer necessary.
2109 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2110 be the same on different OS.
2112 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2115 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2116 whether --show-vars was specified or not
2118 JJ/02 exipick: Added support for new ACL variable spool format introduced
2121 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2122 syntactically invalid From: or Reply-to: line, and a filter used this to
2123 generate an autoreply, and therefore failed to obtain an address for the
2124 autoreply, Exim could try to deliver to a non-existent relative file
2125 name, causing unrelated and misleading errors. What now happens is that
2126 it logs this as a hard delivery error, but does not attempt to create a
2129 PH/11 The exinext utility has a -C option for testing purposes, but although
2130 the given file was scanned by exinext itself; it wasn't being passed on
2131 when Exim was called.
2133 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2134 an end-of-file indication when reading a command response.
2136 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2137 compiled. In many other places in Exim, IPv6 addresses are always
2138 recognized, so I have changed this. It also means that IPv4 domain
2139 literals of the form [IPV4:n.n.n.n] are now always recognized.
2141 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2142 used if the router is not running as root, for example, when verifying at
2143 ACL time, or when using -bh. The debugging output from this situation was
2144 non-existent - all you got was a failure to exec. I have made two
2147 (a) Failures to set uid/gid, the current directory, or a process leader
2148 in a subprocess such as that created by queryprogram now generate
2149 suitable debugging ouput when -d is set.
2151 (b) The queryprogram router detects when it is not running as root,
2152 outputs suitable debugging information if -d is set, and then runs
2153 the subprocess without attempting to change uid/gid.
2155 PH/15 Minor change to Makefile for building test_host (undocumented testing
2158 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2159 additional section of a DNS packet that returns MX or SRV records.
2160 Instead, it always explicitly searches for A/AAAA records. This avoids
2161 major problems that occur when a DNS server includes only records of one
2162 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2163 fixed another bug: if SRV records were looked up and the corresponding
2164 address records were *not* found in the additional section, the port
2165 values from the SRV records were lost.
2167 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2168 using the correct key (the original address) when searching the retry
2169 rules in order to find which one to use for generating the retry hint.
2171 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2172 adding the default one. Similarly, if it contains a Reply-To: header, the
2173 errors_reply_to option, if set, is not used.
2175 PH/19 When calculating a retry time, Exim used to measure the "time since
2176 failure" by looking at the "first failed" field in the retry record. Now
2177 it does not use this if it is later than than the arrival time of the
2178 message. Instead it uses the arrival time. This makes for better
2179 behaviour in cases where some deliveries succeed, thus re-setting the
2180 "first failed" field. An example is a quota failure for a huge message
2181 when small messages continue to be delivered. Without this change, the
2182 "time since failure" will always be short, possible causing more frequent
2183 delivery attempts for the huge message than are intended.
2184 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2186 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2187 $1, $2, $3) because the numerical variables can be reset during some
2188 expansion items (e.g. "match"), thereby losing the authentication data.
2190 PH/21 Make -bV show the size of off_t variables so that the test suite can
2191 decide whether to run tests for quotas > 2G.
2193 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2194 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2195 filecount value is greater than 2G or if a quota value is greater than 2G
2196 on a system where the size of off_t is not greater than 4, a panic error
2199 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2200 never match. The debug and -bh output now contains an explicit error
2201 message indicating a malformed IPv4 address or mask.
2203 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2204 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2205 PH/23 above applies.
2207 PH/25 Do not write to syslog when running in the test harness. The only
2208 occasion when this arises is a failure to open the main or panic logs
2209 (for which there is an explicit test).
2211 PH/26 Added the /no_tell option to "control=freeze".
2213 PH/27 If a host name lookup failed very early in a connection, for example, if
2214 the IP address matched host_lookup and the reverse lookup yielded a name
2215 that did not have a forward lookup, an error message of the form "no IP
2216 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2217 could be logged. Now it outputs the IP address instead of "NULL".
2219 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2220 allows the sender and the authenticated sender to be set when
2221 submitting a message from within Exim. Since child_open_exim() is
2222 documented for local_scan(), the new function should be too.
2224 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2225 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2226 results in an empty string is now treated as unset.
2228 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2230 PH/31 Added sender_verify_fail logging option.
2232 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2233 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2234 tidied the source and removed it altogether.
2236 PH/33 When a queue run was abandoned because the load average was too high, a
2237 log line was always written; now it is written only if the queue_run log
2238 selector is set. In addition, the log line for abandonment now contains
2239 information about the queue run such as the pid. This is always present
2240 in "start" and "stop" lines but was omitted from the "abandon" line.
2242 PH/34 Omit spaces between a header name and the colon in the error message that
2243 is given when verify = headers_syntax fails (if there are lots of them,
2244 the message gets confusing).
2246 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2247 names, as there are now some PTR records that contain slashes. This check
2248 is only to protect against broken name servers that fall over on strange
2249 characters, so the fact that it applies to all lookups doesn't matter.
2251 PH/36 Now that the new test suite is complete, we can remove some of the
2252 special code in Exim that was needed for the old test suite. For example,
2253 sorting DNS records because real resolvers return them in an arbitrary
2254 order. The new test suite's fake resolver always returns records in the
2257 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2258 bounces) except when queue_only is set, to avoid logging races between
2259 the different processes.
2261 PH/38 Panic-die if .include specifies a non-absolute path.
2263 PH/39 A tweak to the "H" retry rule from its user.
2265 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2266 a label. They prevented compilation on older perls.
2268 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2269 a warning to be raised on newish perls.
2271 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2272 on queue. Changes to match documented behaviour of showing count of
2273 messages matching specified criteria.
2275 PH/40 Changed the default ident timeout from 30s to 5s.
2277 PH/41 Added support for the use of login_cap features, on those BSD systems
2278 that have them, for controlling the resources used by pipe deliveries.
2280 PH/42 The content-scanning code uses fopen() to create files in which to put
2281 message data. Previously it was not paying any attention to the mode of
2282 the files. Exim runs with umask(0) because the rest of the code creates
2283 files with open(), and sets the required mode explicitly. Thus, these
2284 files were ending up world-writeable. This was not a big issue, because,
2285 being within the spool directory, they were not world-accessible. I have
2286 created a function called modefopen, which takes an additional mode
2287 argument. It sets umask(777), creates the file, chmods it to the required
2288 mode, then resets the umask. All the relevant calls to fopen() in the
2289 content scanning code have been changed to use this function.
2291 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2292 to 24 hours. This avoids potential overflow problems when processing G
2293 and H retry rules. I suspect nobody ever tinkers with this value.
2295 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2297 PH/45 When the plaintext authenticator is running as a client, the server's
2298 challenges are checked to ensure they are valid base64 strings. By
2299 default, the authentication attempt is cancelled if an invalid string is
2300 received. Setting client_ignore_invalid_base64 true ignores these errors.
2301 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2302 they are received. Thus, the responses can be made to depend on the
2303 challenges. If an invalid string is ignored, an empty string is placed in
2306 PH/46 Messages that are created by the autoreply transport now contains a
2307 References: header, in accordance with RFCs 2822 and 3834.
2309 PH/47 Added authenticated_sender_force to the smtp transport.
2311 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2313 PH/49 Installed latest patch from the Sieve maintainer.
2315 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2316 was also set, the appendfile transport was unnecessarily scanning a
2317 directory of message files (e.g. for maildir delivery) to find the count
2318 of files (along with the size), even though it did not need this
2319 information. It now does the scan only if it needs to find either the
2320 size of the count of files.
2322 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2324 PH/52 Two bugs concerned with error handling when the smtp transport is
2327 (i) Exim was not creating retry information for temporary errors given
2328 for individual recipients after the DATA command when the smtp transport
2329 was used in LMTP mode. This meant that they could be retried too
2330 frequently, and not timed out correctly.
2332 (ii) Exim was setting the flag that allows error details to be returned
2333 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2334 recipients that were returned after the DATA command.
2336 PH/53 This is related to PH/52, but is more general: for any failing address,
2337 when detailed error information was permitted to be returned to the
2338 sender, but the error was temporary, then after the final timeout, only
2339 "retry timeout exceeded" was returned. Now it returns the full error as
2340 well as "retry timeout exceeded".
2342 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2343 do this, and (what is worse) MTAs that accept it.
2345 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2346 will now be deprecated.
2348 PH/56 New os.c-cygwin from the Cygwin maintainer.
2350 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2351 formats (previously only available in exim formats via -bpr, -bpru,
2352 and -bpra. Now also available in native and exiqgrep formats)
2354 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2355 with very large, slow to parse queues
2357 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2359 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2361 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2362 responses to authentication challenges, though it was showing the
2363 challenges; (ii) I've removed the CR characters from the debug output for
2366 PH/58 Allow for the insertion of a newline as well as a space when a string
2367 is turned into more than one encoded-word during RFC 2047 encoding. The
2368 Sieve code now uses this.
2370 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2371 data_4xx, lost_connection, tls_required.
2373 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2374 message was being sent as an SMTP response.
2376 PH/61 Add -l and -k options to exicyclog.
2378 PH/62 When verifying, if an address was redirected to one new address, so that
2379 verification continued, and the new address failed or deferred after
2380 having set something in $address_data, the value of $address_data was not
2381 passed back to the ACL. This was different to the case when no
2382 redirection occurred. The value is now passed back in both cases.
2384 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2385 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2386 use login_cap.h, so on its own it isn't the distinguishing feature. The
2387 new name refers directly to the setclassresources() function.
2389 PH/65 Added configuration files for NetBSD3.
2391 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2393 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2394 is preferred over IPv4.
2396 PH/68 The bounce_return_message and bounce_return_body options were not being
2397 honoured for bounces generated during the reception of non-SMTP messages.
2398 In particular, this applied to messages rejected by the ACL. This bug has
2399 been fixed. However, if bounce_return_message is true and bounce_return_
2400 body is false, the headers that are returned for a non-SMTP message
2401 include only those that have been read before the error was detected.
2402 (In the case of an ACL rejection, they have all been read.)
2404 PH/69 The HTML version of the specification is now built in a directory called
2405 spec_html instead of spec.html, because the latter looks like a path with
2406 a MIME-type, and this confuses some software.
2408 PH/70 Catch two compiler warnings in sieve.c.
2410 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2411 function verify_get_ident() calls ip_connect() to connect a socket, but
2412 if the "connect()" function timed out, ip_connect() used to close the
2413 socket. However, verify_get_ident() also closes the socket later, and in
2414 between Exim writes to the log, which may get opened at this point. When
2415 the socket was closed in ip_connect(), the log could get the same file
2416 descriptor number as the socket. This naturally causes chaos. The fix is
2417 not to close the socket in ip_connect(); the socket should be closed by
2418 the function that creates it. There was only one place in the code where
2419 this was missing, in the iplookup router, which I don't think anybody now
2420 uses, but I've fixed it anyway.
2422 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2423 well as to direct DNS lookups. Otherwise the handling of names in host
2424 lists is inconsistent and therefore confusing.
2430 PH/01 Two changes to the default runtime configuration:
2432 (1) Move the checks for relay_from_hosts and authenticated clients from
2433 after to before the (commented out) DNS black list checks.
2435 (2) Add control=submission to the relay_from_hosts and authenticated
2436 clients checks, on the grounds that messages accepted by these
2437 statements are most likely to be submissions.
2439 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2441 (1) Generate an error if the third argument for the ${prvs expansion is
2444 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2447 (3) Reset the variables that are obtained from the first argument of
2448 ${prvscheck and used in the second argument before leaving the code,
2449 because their memory is reclaimed, so using them afterwards may do
2452 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2453 one (it's much easier than Tom thought :-).
2455 (5) Because of (4), we can now allow for the use of $prvscheck_result
2456 inside the third argument.
2458 PH/03 For some reason, the default setting of PATH when running a command from
2459 a pipe transport was just "/usr/bin". I have changed it to
2462 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2463 anything to be listed in the output from -bV.
2465 PH/05 When a filter generated an autoreply, the entire To: header line was
2466 quoted in the delivery log line, like this:
2468 => >A.N.Other <ano@some.domain> <original@ddress> ...
2470 This has been changed so that it extracts the operative address. There
2471 may be more than one such address. If so, they are comma-separated, like
2474 => >ano@some.domain,ona@other.domain <original@ddress> ...
2476 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2477 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2478 not being looked up in the rDNS to get a host name, Exim was showing the
2479 IP address twice in Received: lines, even though the IP addresses were
2480 identical. For example:
2482 Received: from [1.2.3.4] (helo=[1.2.3.4])
2484 However, if the real host name was known, it was omitting the HELO data
2485 if it matched the actual IP address. This has been tidied up so that it
2486 doesn't show the same IP address twice.
2488 PH/07 When both +timestamp and +memory debugging was on, the value given by
2489 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2490 called by the debug printing, thereby overwriting the timestamp buffer.
2491 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2493 PH/08 When the original message was included in an autoreply transport, it
2494 always said "this is a copy of the message, including all the headers",
2495 even if body_only or headers_only was set. It now gives an appropriate
2498 PH/09 Applied a patch from the Sieve maintainer which:
2500 o fixes some comments
2501 o adds the (disabled) notify extension core
2502 o adds some debug output for the result of if/elsif tests
2503 o points to the current vacation draft in the documentation
2504 and documents the missing references header update
2508 o fixes a bug in processing the envelope test (when testing
2509 multiple envelope elements, the last element determinted the
2512 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2513 Electronic Mail") by including:
2515 Auto-submitted: auto-generated
2517 in the messages that it generates (bounce messages and others, such as
2518 warnings). In the case of bounce messages for non-SMTP mesages, there was
2519 also a typo: it was using "Auto_submitted" (underscore instead of
2520 hyphen). Since every message generated by Exim is necessarily in response
2521 to another message, thes have all been changed to:
2523 Auto-Submitted: auto-replied
2525 in accordance with these statements in the RFC:
2527 The auto-replied keyword:
2529 - SHOULD be used on messages sent in direct response to another
2530 message by an automatic process,
2532 - MUST NOT be used on manually-generated messages,
2534 - MAY be used on Delivery Status Notifications (DSNs) and Message
2535 Disposition Notifications (MDNs),
2537 - MUST NOT be used on messages generated by automatic or periodic
2538 processes, except for messages which are automatic responses to
2541 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2542 to the default Received: header definition.
2544 PH/12 Added log selector acl_warn_skipped (default on).
2546 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2547 variables because (a) they are in the wrong storage pool and (b) even if
2548 they were copied, it wouldn't work properly because of the caching.
2550 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2551 checking when decoding. Apparently there are clients that generate
2552 overlong encoded strings. Why am I not surprised?
2554 PH/15 If the first argument of "${if match_address" was not empty, but did not
2555 contain an "@" character, Exim crashed. Now it writes a panic log message
2556 and treats the condition as false.
2558 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2560 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2561 extension "envelope-auth". The code is finished and in agreement with
2562 other implementations, but there is no documentation so far and in fact,
2563 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2564 not changing the active code.
2566 Print executed "if" and "elsif" statements when debugging is used. This
2567 helps a great deal to understand what a filter does.
2569 Document more things not specified clearly in RFC3028. I had all this
2570 sorted out, when out of a sudden new issues came to my mind. Oops."
2572 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2575 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2576 canonical form (as documented). However, after a host name lookup from
2577 the IP address, check_host() was doing a simple string comparison with
2578 addresses acquired from the DNS when checking that the found name did
2579 have the original IP as one of its addresses. Since any found IPv6
2580 addresses are likely to be in abbreviated form, the comparison could
2581 fail. Luckily, there already exists a function for doing the comparison
2582 by converting both addresses to binary, so now that is used instead of
2583 the text comparison.
2585 PH/20 There was another similar case to PH/19, when a complete host name was
2586 given in a host list; looking up its IP address could give an abbreviated
2587 form, whereas the current host's name might or might not be abbreviated.
2588 The same fix has been applied.
2594 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2595 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2598 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2599 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2601 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2603 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2604 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2605 Also, for "Auto-Submitted", treat anything other than "no" as
2606 non-personal, in accordance with RFC 3834. (Previously it treated
2607 anything starting "auto-" as non-personal.)
2609 TF/01 The control=submission/name=... option had a problem with syntax
2610 errors if the name included a slash character. The /name= option
2611 now slurps the rest of the string, so it can include any characters
2612 but it must come last in the list of options (after /sender_retain
2615 PH/05 Some modifications to the interface to the fake nameserver for the new
2623 TK/01 Added the "success_on_redirect" address verification option. See
2624 NewStuff for rationale and an example.
2626 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2628 PH/02 Patch to exigrep to allow it to work on syslog lines.
2630 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2631 fread() to skip over the body file's header line, because in Cygwin the
2632 header line is locked and is inaccessible.
2634 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2635 co-exist for some time) to make it clear that it is the Exim ID that is
2636 referenced, not the Message-ID: header line.
2638 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2639 string_format() function, because snprintf() does not exist on all
2642 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2643 ${stat: expansion item.
2645 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2646 protocol synchronization error", to keep the pedants happy.
2648 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2649 well as for IRIX systems, when gcc is being used. See the host.c source
2652 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2654 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2657 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2658 transport and to the smtp transport in LMTP mode.
2660 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2662 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2663 run a filter in a subprocess. This could lead to confusion in subsequent
2664 lookups in the parent process. There should also be a search_tidyup() at
2665 the end of the subprocess.
2667 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2668 only if the host matched helo_try_verify_hosts, which caused the
2669 verification to occur when the EHLO/HELO command was issued. The ACL just
2670 tested the remembered result. Now, if a previous verification attempt has
2671 not happened, "verify = helo" does it there and then.
2673 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2675 TK/03 Fix log output including CR from clamd.
2677 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2678 exist provoked a memory error which could cause a segfault.
2680 PH/15 Installed PCRE 6.2
2682 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2684 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2685 of the problem. Specifically, suggested +O2 rather than +O1 for the
2688 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2690 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2691 record, the port was not correctly logged when the outgoing_port log
2692 selector was set (it logged the transort's default port).
2694 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2695 fallback_hosts, and "hosts" in the smtp transport.
2697 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2698 host errors such as "Connection refused".
2700 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2701 authentication with radiusclient 0.4.9:
2703 - Error returned from rc_read_config was caught wrongly
2704 - Username/password not passed on to radius server due to wrong length.
2706 The presumption is that some radiusclient API changes for 4.51/PH/17
2707 were not taken care of correctly. The code is still untested by me (my
2708 Linux distribution still has 0.3.2 of radiusclient), but it was
2709 contributed by a Radius user.
2711 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2712 expanding the "port" option of the smtp transport.
2714 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2715 while reading a MIME header. Thanks to Tom Hughes for a patch.
2717 PH/24 Include config.h inside local_scan.h so that configuration settings are
2720 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2721 that in an ACL for RCPT (for example), you can examine exactly what was
2724 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2725 commands, but it was not correctly comparing the address with the actual
2726 client host address. Thus, it would show the EHLO address in Received:
2727 header lines when this was not necessary.
2729 PH/27 Added the % operator to ${eval:}.
2731 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2732 it should be ignoring failures (because with -C, for example, it has lost
2733 privilege). It wasn't ignoring creation failures other than "already
2736 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2739 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2740 because an input line was too long, either on its own, or by virtue of
2741 too many continuations, the temporary file was not being removed, and the
2742 return code was incorrect.
2744 PH/31 Missing "BOOL" in function definition in filtertest.c.
2746 PH/32 Applied Sieve patches from the maintainer.
2748 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2750 PH/33 Added "verify = not_blind".
2752 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2753 Local/Makefile (with some defaults set). These are used in built scripts
2754 such as exicyclog, but they have never been used in the exim_install
2755 script (though there are many overriding facilities there). I have
2756 arranged that the exim_install script now takes note of these two
2759 PH/35 Installed configuration files for Dragonfly.
2761 PH/36 When a locally submitted message by a trusted user did not contain a
2762 From: header, and the sender address was obtained from -f or from an SMTP
2763 MAIL command, and the trusted user did not use -F to supply a sender
2764 name, $originator_name was incorrectly used when constructing a From:
2765 header. Furthermore, $originator_name was used for submission mode
2766 messages from external hosts without From: headers in a similar way,
2767 which is clearly wrong.
2769 PH/37 Added control=suppress_local_fixups.
2771 PH/38 When log_selector = +received_sender was set, and the addition of the
2772 sender made the log line's construction buffer exactly full, or one byte
2773 less than full, an overflow happened when the terminating "\n" was
2776 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2777 when the result of a list match is failure because a DNS lookup failed.
2779 PH/40 RM_COMMAND is now used in the building process.
2781 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2782 the "build-* directories that it finds.
2784 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2785 address in a domain literal was a prefix of an interface address.
2787 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2788 when verifying a sender address, unless rewrite_headers is false.
2790 PH/44 Wrote a long comment about why errors_to addresses are verified as
2791 recipients, not senders.
2793 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2794 the ratelimit ACL was added.
2796 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2798 PH/46 Added extra information about PostgreSQL errors to the error string.
2800 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2801 avoiding the need to install special zones in a real server. This is
2802 backwards compatible; if it can't find the fake resolver, it drops back.
2803 Thus, both old and new test suites can be run.
2805 TF/02 Added util/ratelimit.pl
2807 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2808 clock is set back in time.
2810 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2811 Candler <B.Candler@pobox.com>.
2813 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2814 OK for addresses that are the result of redirections.
2816 PH/48 A number of further additions for the benefit of the new test suite,
2817 including a fake gethostbyname() that interfaces to the fake DNS resolver
2820 TF/06 The fix for widen_domains has also been applied to qualify_single and
2821 search_parents which are the other dnslookup options that can cause
2824 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2827 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2829 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2830 service" if it encountered an MX record with an empty target hostname.
2831 The message is now "an MX or SRV record indicated no SMTP service".
2833 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2834 if the DNS of the sending site is misconfigured. This is quite a
2835 common situation. This change restores the behaviour of treating a
2836 helo verification defer as a failure.
2838 PH/51 If self=fail was set on a router, the bounce message did not include the
2839 actual error message.
2845 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2847 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2848 command itself ended in error, the underlying message about the transport
2849 filter timeout was being overwritten with the pipe command error. Now the
2850 underlying error message should be appended to the second error message.
2852 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2854 PH/02 Reduce the amount of output that "make" produces by default. Full output
2855 can still be requested.
2857 PH/03 The warning log line about a condition test deferring for a "warn" verb
2858 was being output only once per connection, rather than after each
2859 occurrence (because it was using the same function as for successful
2860 "warn" verbs). This seems wrong, so I have changed it.
2862 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2863 it should not have, which might have caused a crash in the right
2864 circumstances, but probably never did.
2866 PH/04 Installed a modified version of Tony Finch's patch to make submission
2867 mode fix the return path as well as the Sender: header line, and to
2868 add a /name= option so that you can make the user's friendly name appear
2871 TF/03 Added the control = fakedefer ACL modifier.
2873 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2874 Mark Lowes for thorough testing.
2876 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2878 TK/03 Merged latest SRS patch from Miles Wilton.
2880 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2881 with the definition in sysexits.h (which is #included earlier).
2882 Fortunately, Exim does not actually use EX_OK. The code used to try to
2883 preserve the sysexits.h value, by assumimg that macro definitions were
2884 scanned for macro replacements. I have been disabused of this notion,
2885 so now the code just undefines EX_OK before #including unistd.h.
2887 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2888 in the smtp transport. When a block could not be written in a single
2889 write() function, the timeout was being re-applied to each part-write.
2890 This seems wrong - if the receiver was accepting one byte at a time it
2891 would take for ever. The timeout is now adjusted when this happens. It
2892 doesn't have to be particularly precise.
2894 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2895 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2897 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2898 without a domain if the check to <postmaster@domain> fails.
2900 SC/01 Eximstats: added -xls and the ability to specify output files
2901 (patch written by Frank Heydlauf).
2903 SC/02 Eximstats: use FileHandles for outputing results.
2905 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2907 SC/04 Eximstats: fixed display of large numbers with -nvr option
2909 SC/05 Eximstats: fixed merging of reports with empty tables.
2911 SC/06 Eximstats: added the -include_original_destination flag
2913 SC/07 Eximstats: removed tabs and trailing whitespace.
2915 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2917 TK/06 MBOX spool code: Add real "From " MBOX separator line
2918 so the .eml file is really in mbox format (even though
2919 most programs do not really care). Patch from Alex Miller.
2921 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2922 The latter is generated from $received_to and is only set if the
2923 message has one envelope recipient. SA can use these headers,
2924 obviously out-of-the-box. Patch from Alex Miller.
2926 PH/08 The ${def test on a variable was returning false if the variable's
2927 value was "0", contrary to what the specification has always said!
2928 The result should be true unless the variable is empty.
2930 PH/09 The syntax error of a character other than { following "${if
2931 def:variable_name" (after optional whitespace) was not being diagnosed.
2932 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2933 accidental colon was present, for example, could give incorrect results.
2935 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2936 result is used (not including appendfile, where other changes are about
2939 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2940 This involved changing a lot of size variables from int to off_t. It
2941 should work with maildirs and everything.
2943 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2944 spamd dying while we are connected to it.
2946 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2949 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2950 function for BDB 4.3.
2952 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2954 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2955 the sources for PCRE have been re-arranged and more files are now
2958 PH/15 The code I had for printing potentially long long variables in PH/11
2959 above was not the best (it lost precision). The length of off_t variables
2960 is now inspected at build time, and an appropriate printing format (%ld
2961 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2962 to be "long long int" or "long int". This is needed for the internal
2963 formatting function string_vformat().
2965 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2966 the configuration file to be ":syslog", then the script "guesses" where
2967 the logs files are, rather than using the compiled in default. In our
2968 case the guess is not the same as the compiled default, so the script
2969 suddenly stopped working when I started to use syslog. The patch checks
2970 to see if log_file_path is "". If so, it attempts to read it from exim
2971 with no configuration file to get the compiled in version, before it
2972 falls back to the previous guessing code."
2974 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2975 implementing BATV in an Exim configuration. See NewStuff for the gory
2978 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2979 Makefile that are specific to HP-UX.
2981 PH/18 If the "use_postmaster" option was set for a recipient callout together
2982 with the "random" option, the postmaster address was used as the MAIL
2983 FROM address for the random test, but not for the subsequent recipient
2984 test. It is now used for both.
2986 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2987 patch removes a few documentation additions to RFC 3028, because the
2988 latest draft now contains them. It adds the new en;ascii-case comparator
2989 and a new error check for 8bit text in MIME parts. Comparator and
2990 require names are now matched exactly. I enabled the subaddress
2991 extension, but it is not well tested yet (read: it works for me)."
2993 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2994 rework some of the code of TK/09 above to avoid the hardwired use of
2995 "%lld" and "long long". Replaced the call to snprintf() with a call to
2998 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2999 records point to non-existent hosts", "retry timeout exceeded", and
3000 "retry time not reached for any host after a long failure period".
3002 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3003 experimental DomainKeys support:
3005 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3006 (2) On an error such as an illegally used "control", the wrong name for
3007 the control was given.
3009 These problems did NOT occur unless DomainKeys support was compiled.
3011 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3013 PH/24 Added ${if match_ip condition.
3015 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3016 databases so that it will be absolutely obvious if a crash occurs in the
3017 DB library. This is a regular occurrence (often caused by mis-matched
3020 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3021 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3022 picked up on a user's system that detects such things. There doesn't seem
3023 to be a gcc warning option for this - only an attribute that has to be
3024 put on the function's prototype. It seems that in Fedora Core 4 they have
3025 set this on a number of new functions. No doubt there will be more in due
3028 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3029 specify a transport. However, if an address that was verified by such a
3030 router was the subject of a callout, Exim crashed because it tried to
3031 read the rcpt_include_affixes from the non-existent transport. Now it
3032 just assumes that the setting of that option is false. This bug was
3033 introduced by 4.51/PH/31.
3035 PH/28 Changed -d+all to exclude +memory, because that information is very
3036 rarely of interest, but it makes the output a lot bigger. People tend to
3037 do -d+all out of habit.
3039 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3040 code in os-type was giving problems when libc.so lives in lib64, like on
3043 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3044 aren't the modern standard, and it seems that some systems' include files
3045 don't always have them. Exim was already checking for some of the newer
3046 ones like T_AAAA, and defining it itself. I've added checks for all the
3047 record types that Exim uses.
3049 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3050 not automatically generating a new one, as it is supposed to. This
3051 prevented TLS from working. If the file did exist, but contained invalid
3052 data, a new version was generated, as expected. It was only the case of a
3053 non-existent file that was broken.
3055 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3056 with a change in libdomainkeys > 0.64.
3058 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3059 from DNS. If the selector record carries the flag, it now has
3060 precedence over the domain-wide flag.
3062 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3064 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3065 the use of a port name that isn't defined in /etc/services) occurred, the
3066 message was deferred as in a normal delivery, and thus remained on the
3067 spool, instead of being failed because of the mua_wrapper setting. This
3068 is now fixed, and I tidied up some of the mua_wrapper messages at the
3071 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3072 the messages in a hash of arrays rather than using individual hashes.
3073 This is a bit cleaner and results in dramatic memory savings, albeit
3074 at a slight CPU cost.
3076 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3077 as requested by Marc Sherman.
3079 SC/10 Eximstats: added histograms for user specified patterns as requested
3082 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3084 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3085 fopen() in the content-scanning modules that did not already have it.
3091 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3092 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3094 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3096 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3098 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3099 to test Sieve filters that use "vacation".
3101 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3102 that changes the way the GnuTLS parameters are stored in the cache file.
3103 The new format can be generated externally. For backward compatibility,
3104 if the data in the cache doesn't make sense, Exim assumes it has read an
3105 old-format file, and it generates new data and writes a new file. This
3106 means that you can't go back to an older release without removing the
3109 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3110 work if there are any delivery delays because "one_time" forces the
3111 parent to be marked "delivered", so its unseen clone is never tried
3112 again. For this reason, Exim now forbids the simultaneous setting of
3115 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3116 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3117 entry if you want to know the details. The fix, however, neglected to
3118 consider the case when local delivery batching is involved. The test for
3119 "previously delivered" was not happening when checking to see if an
3120 address could be batched with a previous (undelivered) one; under
3121 certain circumstances this could lead to multiple deliveries to the same
3124 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3125 in its include files, and this causes problems building Exim.
3127 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3128 header_syntax) but Exim was just ignoring anything given after a slash.
3129 In particular, this caused confusion with an attempt to use "verify =
3130 reverse_host_lookup/defer_ok". An error is now given when options are
3131 supplied for verify items that do not have them. (Maybe reverse_host_
3132 lookup should have a defer_ok option, but that's a different point.)
3134 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3135 defined by RFC 821) to 2048, because there were problems with some AUTH
3136 commands, and RFC 1869 says the size should be increased for extended
3137 SMTP commands that take arguments.
3139 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3142 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3143 "unknown" error; now it says that the functionality isn't in the binary.
3145 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3146 an address' error message when a string expansion fails (syntax or
3147 whatever). Otherwise the password may appear in the log. Following change
3148 PH/42 below, there is no longer a chance of it appearing in a bounce
3151 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3153 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3154 of its addresses was ever tried. (Bugzilla bug #2.)
3156 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3157 the result incorrectly in the debug output. (It correctly added a newline
3158 to what was transported.)
3160 TF/01 Added $received_time.
3162 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3163 commented out examples of how to interface to a virus scanner and to
3164 SpamAssassin. Also added commented examples of av_scanner and
3165 spamd_address settings.
3167 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3168 and controls are allowed in which ACLs. There were a couple of minor
3169 errors. Some of the entries in the conditions table (which is a table of
3170 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3171 negation of where the condition IS allowed.
3173 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3175 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3176 header file does not have a version number, so I've had to invent a new
3177 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3178 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3179 radiusclient), but it was contributed by a Radius user.
3181 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3182 files or directories, for OpenSSL.
3184 PH/19 When an Exim process that is running as root has to create an Exim log
3185 file, it does so in a subprocess that runs as exim:exim so as to get the
3186 ownership right at creation (otherwise, other Exim processes might see
3187 the file with the wrong ownership). There was no test for failure of this
3188 fork() call, which would lead to the process getting stuck as it waited
3189 for a non-existent subprocess. Forks do occasionally fail when resources
3190 run out. I reviewed all the other calls to fork(); they all seem to check
3193 PH/20 When checking for unexpected SMTP input at connect time (before writing
3194 the banner), Exim was not dealing correctly with a non-positive return
3195 from the read() function. If the client had disconnected by this time,
3196 the result was a log entry for a synchronization error with an empty
3197 string after "input=" when read() returned zero. If read() returned -1
3198 (an event I could not check), uninitialized data bytes were printed.
3199 There were reports of junk text (parts of files, etc) appearing after
3202 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3204 PH/22 Added support for macro redefinition, and (re)definition in between
3205 driver and ACL definitions.
3207 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3208 forgetting to use the resulting value; it was using the unexpanded value.
3210 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3211 hadn't been configured. The fix is from Juergen Kreileder, who
3212 understands it better than I do:
3214 "Here's what I see happening with three configured cyrus_sasl
3215 authenticators configured (plain, login, cram-md5):
3217 On startup auth_cyrus_sasl_init() gets called for each of these.
3218 This means three calls to sasl_listmech() without a specified mech_list.
3219 => SASL tests which mechs of all available mechs actually work
3220 => three warnings about OTP not working
3221 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3223 With the patch, sasl_listmech() also gets called three times. But now
3224 SASL's mech_list option is set to the server_mech specified in the the
3225 authenticator. Or in other words, the answer from sasl_listmech()
3226 gets limited to just the mech you're testing for (which is different
3228 => the return list contains just 'plain' or 'login', 'cram-md5' or
3229 nothing depending on the value of ob->server_mech.
3231 I've just tested the patch: Authentication still works fine,
3232 unavailable mechs specified in the exim configuration are still
3233 caught, and the auth.log warnings about OTP are gone."
3235 PH/25 When debugging is enabled, the contents of the command line are added
3236 to the debugging output, even when log_selector=+arguments is not
3239 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3240 answer is "GNU", and only if the return is "GNU/something" is the answer
3243 PH/27 $acl_verify_message is now set immediately after the failure of a
3244 verification in an ACL, and so is available in subsequent modifiers. In
3245 particular, the message can be preserved by coding like this:
3247 warn !verify = sender
3248 set acl_m0 = $acl_verify_message
3250 Previously, $acl_verify_message was set only while expanding "message"
3251 and "log_message" when a very denied access.
3253 PH/28 Modified OS/os.c-Linux with
3255 -#ifndef OS_LOAD_AVERAGE
3256 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3258 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3259 nomenclature these days.)
3261 PH/29 Installed patch from the Sieve maintainer that adds the options
3262 sieve_useraddress and sieve_subaddress to the redirect router.
3264 PH/30 In these circumstances:
3265 . Two addresses routed to the same list of hosts;
3266 . First host does not offer TLS;
3267 . First host accepts first address;
3268 . First host gives temporary error to second address;
3269 . Second host offers TLS and a TLS session is established;
3270 . Second host accepts second address.
3271 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3272 and peerdn, if requested) that were in fact used only for the second
3275 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3276 attention to any local part prefix or suffix that was matched by the
3277 router that accepted the address. It now behaves in the same way as it
3278 does for delivery: the affixes are removed from the local part unless
3279 rcpt_include_affixes is set on the transport.
3281 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3282 timeout during the DATA phase of an incoming message.
3284 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3285 applied a patch sanctioned by the Sieve maintainer.
3287 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3288 the uid or gid is negative. A case of a negative gid caused this to be
3289 noticed. The fix allows for either to be negative.
3291 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3292 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3295 PH/36 The change PH/12 above was broken. Fixed it.
3297 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3298 the grounds that routing the same address twice would always produce the
3299 same answer. This might have been true once, but it is certainly no
3300 longer true now. Routing a child address may depend on the previous
3301 routing that produced that child. Some complicated redirection strategies
3302 went wrong when messages had multiple recipients, and made Exim's
3303 behaviour dependent on the order in which the addresses were given.
3305 I have moved the duplicate checking until after the routing is complete.
3306 Exim scans the addresses that are assigned to local and remote
3307 transports, and removes any duplicates. This means that more work will be
3308 done, as duplicates will always all be routed, but duplicates are
3309 presumably rare, so I don't expect this is of any significance.
3311 For deliveries to pipes, files, and autoreplies, the duplicate checking
3312 still happens during the routing process, since they are not going to be
3315 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3316 It corrects a timeout issue with spamd. This is Ian's comment: "The
3317 background is that sometimes spamd either never reads data from a
3318 connection it has accepted, or it never writes response data. The exiscan
3319 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3320 blindly assumes that writes won't block so it may never time out."
3322 PH/39 Allow G after quota size as well as K and M.
3324 PH/40 The value set for $authenticated_id in an authenticator may not contain
3325 binary zeroes or newlines because the value is written to log lines and
3326 to spool files. There was no check on this. Now the value is run through
3327 the string_printing() function so that such characters are converted to
3328 printable escape sequences.
3330 PH/41 $message_linecount is a new variable that contains the total number of
3331 lines in the message. Compare $body_linecount, which is the count for the
3334 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3335 bounce and delay warning messages, except in certain special cases, which
3338 (a) An SMTP error message from a remote host;
3339 (b) A message specified in a :fail: redirection;
3340 (c) A message specified in a "fail" command in a system filter;
3341 (d) A message specified in a FAIL return from the queryprogram router;
3342 (e) A message specified by the cannot_route_message router option.
3344 In these cases only, Exim does include the error details in bounce and
3345 warning messages. There are also a few cases where bland messages such
3346 as "unrouteable address" or "local delivery error" are given.
3348 PH/43 $value is now also set for the "else" part of a ${run expansion.
3350 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3351 being worked on, but at least Exim now implements the latest version to
3354 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3355 process to complete was treated as a delivery failure, a timeout while
3356 writing the message to the pipe was logged, but erroneously treated as a
3357 successful delivery. Such timeouts include transport filter timeouts. For
3358 consistency with the overall process timeout, these timeouts are now
3359 treated as errors, giving rise to delivery failures by default. However,
3360 there is now a new Boolean option for the pipe transport called
3361 timeout_defer, which, if set TRUE, converts the failures into defers for
3362 both kinds of timeout. A transport filter timeout is now identified in
3365 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3366 systems where "make" and "gmake" are different, calling "gmake" at top
3367 level broke things. I've arranged for the value of $(MAKE) to be passed
3368 from the Makefile to this script so that it can call the same version of
3372 A note about Exim versions 4.44 and 4.50
3373 ----------------------------------------
3375 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3376 changes of various kinds. As a consequence, a big documentation update was
3377 needed. This delayed the release for rather longer than seemed good, especially
3378 in the light of a couple of (minor) security issues. Therefore, the changes
3379 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3380 release. So 4.44 and 4.50 are in effect two different branches that both start
3383 I have left the 4.50 change log unchanged; it contains all the changes since
3384 4.43. The change log for 4.44 is below; many of its items are identical to
3385 those for 4.50. This seems to be the most sensible way to preserve the
3386 historical information.
3392 1. Minor wording change to the doc/README.SIEVE file.
3394 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3395 computation of the current number of files was incorrect.
3397 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3398 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3401 4. Give more explanation in the error message when the command for a transport
3402 filter fails to execute.
3404 5. There are several places where Exim runs a non-Exim command in a
3405 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3406 was being done only for the command run by the queryprogram router. It is
3407 now done for all such subprocesses. The other cases are: ${run, transport
3408 filters, and the commands run by the lmtp and pipe transports.
3410 6. Added CONFIGURE_GROUP build-time option.
3412 7. Some older OS have a limit of 256 on the maximum number of file
3413 descriptors. Exim was using setrlimit() to set 1000 as a large value
3414 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3415 systems. I've change it so that if it can't get 1000, it tries for 256.
3417 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3418 was an oversight, and furthermore, ever since the addition of extra
3419 controls (e.g. 4.43/32), the checks on when to allow different forms of
3420 "control" were broken. There should now be diagnostics for all cases when a
3421 control that does not make sense is encountered.
3423 9. Added the /retain_sender option to "control=submission".
3425 10. $recipients is now available in the predata ACL (oversight).
3427 11. Tidy the search cache before the fork to do a delivery from a message
3428 received from the command line. Otherwise the child will trigger a lookup
3429 failure and thereby defer the delivery if it tries to use (for example) a
3430 cached ldap connection that the parent has called unbind on.
3432 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3433 of $address_data from the recipient verification was clobbered by the
3434 sender verification.
3436 13. The value of address_data from a sender verification is now available in
3437 $sender_address_data in subsequent conditions in the ACL statement.
3439 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3441 15. Added a new option "connect=<time>" to callout options, to set a different
3444 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3445 was its contents. (It was OK if the option was not defined at all.)
3447 17. A "Completed" log line is now written for messages that are removed from
3448 the spool by the -Mrm option.
3450 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3451 information about exactly what failed.
3453 19. Added -dd to debug only the daemon process.
3455 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3456 handles timeouts, both on the server side and network timeouts. Renamed the
3457 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3459 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3460 It is now set to "smtps".
3462 22. $host_address is now set to the target address during the checking of
3463 ignore_target_hosts.
3465 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3466 being passed; this would have caused $sender_host_name to have been used if
3467 matching the list had actually called for a host name (not very likely,
3468 since this list is usually IP addresses). A host name is now passed as
3471 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3472 code that specifies a non-restarting handler (typically sigaction() in
3473 modern systems) in an attempt to fix a rare and obscure crash bug.
3475 25. Narrowed the window for a race in the daemon that could cause it to ignore
3476 SIGCHLD signals. This is not a major problem, because they are used only to
3477 wake it up if nothing else does.
3479 26. A malformed maildirsize file could cause Exim to calculate negative values
3480 for the mailbox size or file count. Odd effects could occur as a result.
3481 The maildirsize information is now recalculated if the size or filecount
3484 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3485 support for a long time. Removed HAVE_SYS_VFS_H.
3487 28. Installed the latest version of exipick from John Jetmore.
3489 29. In an address list, if the pattern was not a regular expression, an empty
3490 subject address (from a bounce message) matched only if the pattern was an
3491 empty string. Non-empty patterns were not even tested. This was the wrong
3492 because it is perfectly reasonable to use an empty address as part of a
3493 database query. An empty address is now tested by patterns that are
3494 lookups. However, all the other forms of pattern expect the subject to
3495 contain a local part and a domain, and therefore, for them, an empty
3496 address still always fails if the pattern is not itself empty.
3498 30. Exim went into a mad DNS loop when attempting to do a callout where the
3499 host was specified on an smtp transport, and looking it up yielded more
3500 than one IP address.
3502 31. Re-factored the code for checking spool and log partition space into a
3503 function that finds that data and another that does the check. The former
3504 is then used to implement four new variables: $spool_space, $log_space,
3505 $spool_inodes, and $log_inodes.
3507 32. The RFC2047 encoding function was originally intended for short strings
3508 such as real names; it was not keeping to the 75-character limit for
3509 encoded words that the RFC imposes. It now respects the limit, and
3510 generates multiple encoded words if necessary. To be on the safe side, I
3511 have increased the buffer size for the ${rfc2047: expansion operator from
3514 33. It is now permitted to omit both strings after an "if" condition; if the
3515 condition is true, the result is "true". As before, when the second string
3516 is omitted, a false condition yields an empty string. This makes it less
3517 cumbersome to write custom ACL and router conditions.
3519 34. Failure to deliver a bounce message always caused it to be frozen, even if
3520 there was an errors_to setting on the router. The errors_to setting is now
3523 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3524 canonical form (fully expanded) before being placed in
3525 $sender_host_address.
3527 36. The table in the code that translates DNS record types into text (T_A to
3528 "A" for instance) was missing entries for NS and CNAME. It is just possible
3529 that this could have caused confusion if both these types were looked up
3530 for the same domain, because the text type is used as part of Exim's
3531 per-process caching. But the chance of anyone hitting this buglet seems
3534 37. The dnsdb lookup has been extended in a number of ways.
3536 (1) There is a new type, "zns", which walks up the domain tree until it
3537 finds some nameserver records. It should be used with care.
3539 (2) There is a new type, "mxh", which is like "mx" except that it returns
3540 just the host names, not the priorities.
3542 (3) It is now possible to give a list of domains (or IP addresses) to be
3543 looked up. The behaviour when one of the lookups defers can be
3544 controlled by a keyword.
3546 (4) It is now possible to specify the separator character for use when
3547 multiple records are returned.
3549 38. The dnslists ACL condition has been extended: it is now possible to supply
3550 a list of IP addresses and/or domains to be looked up in a particular DNS
3553 39. Added log_selector=+queue_time_overall.
3555 40. When running the queue in the test harness, wait just a tad after forking a
3556 delivery process, to get repeatability of debugging output.
3558 41. Include certificate and key file names in error message when GnuTLS fails
3559 to set them up, because the GnuTLS error message doesn't include the name
3560 of the failing file when there is a problem reading it.
3562 42. Allow both -bf and -bF in the same test run.
3564 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3566 44. The "Exiscan patch" is now merged into the mainline Exim source.
3568 45. Sometimes the final signoff response after QUIT could fail to get
3569 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3570 < 0 before doing a fflush(). This bug looks as though it goes back to the
3571 introduction of TLS in release 3.20, but "sometimes" must have been rare
3572 because the tests only now provoked it.
3574 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3575 (this can affect the format of dates).
3577 47. exim_tidydb, when checking for the continued existence of a message for
3578 which it has found a message-specific retry record, was not finding
3579 messages that were in split spool directories. Consequently, it was
3580 deleting retry records that should have stayed in existence.
3582 48. Steve fixed some bugs in eximstats.
3584 49. The SPA authentication driver was not abandoning authentication and moving
3585 on to the next authenticator when an expansion was forced to fail,
3586 contradicting the general specification for all authenticators. Instead it
3587 was generating a temporary error. It now behaves as specified.
3589 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3590 (the order specifies the preference for clients). The order is now AES256,
3591 AES128, 3DES, ARCFOUR128.
3593 51. Small patch to Sieve code - explicitly set From: when generating an
3596 52. Exim crashed if a remote delivery caused a very long error message to be
3597 recorded - for instance if somebody sent an entire SpamAssassin report back
3598 as a large number of 550 error lines. This bug was coincidentally fixed by
3599 increasing the size of one of Exim's internal buffers (big_buffer) that
3600 happened as part of the Exiscan merge. However, to be on the safe side, I
3601 have made the code more robust (and fixed the comments that describe what
3604 53. Now that there can be additional text after "Completed" in log lines (if
3605 the queue_time_overall log selector is set), a one-byte patch to exigrep
3606 was needed to allow it to recognize "Completed" as not the last thing in
3609 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3610 patch that reportedly fixes this has been added. I am not expert enough to
3611 create a test for it. This is what the patch creator wrote:
3613 "I found a little strange behaviour of ldap code when working with
3614 Windows 2003 AD Domain, where users was placed in more than one
3615 Organization Units. When I tried to give exim partial DN, the exit code
3616 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3617 But simultaneously result of request was absolutely normal ldap result,
3618 so I produce this patch..."
3620 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3621 REFERENCE, so I have modified the code to exclude the patch when that macro
3624 55. Some experimental protocols are using DNS PTR records for new purposes. The
3625 keys for these records are domain names, not reversed IP addresses. The
3626 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3627 leaves it alone. Component reversal etc. now happens only for IP addresses.
3630 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3632 57. Double the size of the debug message buffer (to 2048) so that more of very
3633 long debug lines gets shown.
3635 58. The exicyclog utility now does better if the number of log files to keep
3636 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3638 59. Two changes related to the smtp_active_hostname option:
3640 (1) $smtp_active_hostname is now available as a variable.
3641 (2) The default for smtp_banner uses $smtp_active_hostname instead
3642 of $primary_hostname.
3644 60. The host_aton() function is supposed to be passed a string that is known
3645 to be a valid IP address. However, in the case of IPv6 addresses, it was
3646 not checking this. This is a hostage to fortune. Exim now panics and dies
3647 if the condition is not met. A case was found where this could be provoked
3648 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3649 components; fortuitously, this particular loophole had already been fixed
3650 by change 4.50/55 above.
3652 If there are any other similar loopholes, the new check in host_aton()
3653 itself should stop them being exploited. The report I received stated that
3654 data on the command line could provoke the exploit when Exim was running as
3655 exim, but did not say which command line option was involved. All I could
3656 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3657 running as the user.
3660 61. There was a buffer overflow vulnerability in the SPA authentication code
3661 (which came originally from the Samba project). I have added a test to the
3662 spa_base64_to_bits() function which I hope fixes it.
3665 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3666 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3668 63. The daemon start-up calls getloadavg() while still root for those OS that
3669 need the first call to be done as root, but it missed one case: when
3670 deliver_queue_load_max is set with deliver_drop_privilege. This is
3671 necessary for the benefit of the queue runner, because there is no re-exec
3672 when deliver_drop_privilege is set.
3674 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3675 This has been fixed.
3677 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3678 was in use, was not putting the data itself into the right store pool;
3679 consequently, it could be overwritten for a subsequent message in the same
3680 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3683 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3685 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3686 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3687 the function were treating the return as a boolean value, which happened to
3688 work because 0=false and not-0=true, but is not correct code.
3690 68. The host_aton() function was not handling scoped IPv6 addresses (those
3691 with, for example, "%eth0" on the end) correctly.
3693 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3694 negated items (that is, ~something) in unsigned ints. Some compilers
3695 apparently mutter when there is no cast.
3697 70. If an address verification called from an ACL failed, and did not produce a
3698 user-specific message (i.e. there was only a "system" message), nothing was
3699 put in $acl_verify_message. In this situation, it now puts the system
3702 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3703 session; change 4.31/43 added the unwanted input to the log line - except
3704 that it did not do this in the start of session case. It now does.
3706 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3707 This is wrong and can cause the other end to generate a synchronization
3708 error if it is another Exim or anything else that does the synchronization
3709 check. A QUIT command is no longer sent after a timeout.
3711 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3712 during host lookups.
3714 74. The defer_ok option of callout verification was not working if it was used
3715 when verifying addresses in header lines, that is, for this case:
3717 verify = header_sender/callout=defer_ok
3719 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3720 those file descriptors could be used for SMTP connections. If anything
3721 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3722 could be sent to the SMTP client, causing chaos. The daemon now opens
3723 stdin, stdout, and stderr to /dev/null when it puts itself into the
3726 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3727 log by default. The user can override this with suitable Perl magic.
3729 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3730 the log message when discard triggers, was not working for the DATA ACL or
3731 for the non-SMTP ACL.
3733 78. Error message wording change in sieve.c.
3735 79. If smtp_accept_max_per_host was set, the number of connections could be
3736 restricted to fewer than expected, because the daemon was trying to set up
3737 a new connection before checking whether the processes handling previous
3738 connections had finished. The check for completed processes is now done
3739 earlier. On busy systems, this bug wouldn't be noticed because something
3740 else would have woken the daemon, and it would have reaped the completed
3743 80. If a message was submitted locally by a user whose login name contained one
3744 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3745 It caused a spool format error. I have fixed the spool reading code. A
3746 related problem was that the "from" clause in the Received: line became
3747 illegal because of the space(s). It is now covered by ${quote_local_part.
3749 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3752 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3754 83. Patch from Sieve maintainer to fix unterminated string problem in
3755 "vacation" handling.
3757 84. Some minor changes to the Linux configuration files to help with other
3758 OS variants using glibc.
3760 85. One more patch for Sieve to update vacation handling to latest spec.
3763 ----------------------------------------------------
3764 See the note above about the 4.44 and 4.50 releases.
3765 ----------------------------------------------------
3771 1. Change 4.43/35 introduced a bug that caused file counts to be
3772 incorrectly computed when quota_filecount was set in an appendfile
3775 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3776 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3779 3. Give more explanation in the error message when the command for a transport
3780 filter fails to execute.
3782 4. There are several places where Exim runs a non-Exim command in a
3783 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3784 was being done only for the command run by the queryprogram router. It is
3785 now done for all such subprocesses. The other cases are: ${run, transport
3786 filters, and the commands run by the lmtp and pipe transports.
3788 5. Some older OS have a limit of 256 on the maximum number of file
3789 descriptors. Exim was using setrlimit() to set 1000 as a large value
3790 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3791 systems. I've change it so that if it can't get 1000, it tries for 256.
3793 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3794 was an oversight, and furthermore, ever since the addition of extra
3795 controls (e.g. 4.43/32), the checks on when to allow different forms of
3796 "control" were broken. There should now be diagnostics for all cases when a
3797 control that does not make sense is encountered.
3799 7. $recipients is now available in the predata ACL (oversight).
3801 8. Tidy the search cache before the fork to do a delivery from a message
3802 received from the command line. Otherwise the child will trigger a lookup
3803 failure and thereby defer the delivery if it tries to use (for example) a
3804 cached ldap connection that the parent has called unbind on.
3806 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3807 of $address_data from the recipient verification was clobbered by the
3808 sender verification.
3810 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3811 was its contents. (It was OK if the option was not defined at all.)
3813 11. A "Completed" log line is now written for messages that are removed from
3814 the spool by the -Mrm option.
3816 12. $host_address is now set to the target address during the checking of
3817 ignore_target_hosts.
3819 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3820 being passed; this would have caused $sender_host_name to have been used if
3821 matching the list had actually called for a host name (not very likely,
3822 since this list is usually IP addresses). A host name is now passed as
3825 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3826 code that specifies a non-restarting handler (typically sigaction() in
3827 modern systems) in an attempt to fix a rare and obscure crash bug.
3829 15. Narrowed the window for a race in the daemon that could cause it to ignore
3830 SIGCHLD signals. This is not a major problem, because they are used only to
3831 wake it up if nothing else does.
3833 16. A malformed maildirsize file could cause Exim to calculate negative values
3834 for the mailbox size or file count. Odd effects could occur as a result.
3835 The maildirsize information is now recalculated if the size or filecount
3838 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3839 support for a long time. Removed HAVE_SYS_VFS_H.
3841 18. Updated exipick to current release from John Jetmore.
3843 19. Allow an empty sender to be matched against a lookup in an address list.
3844 Previously the only cases considered were a regular expression, or an
3847 20. Exim went into a mad DNS lookup loop when doing a callout where the
3848 host was specified on the transport, if the DNS lookup yielded more than
3851 21. The RFC2047 encoding function was originally intended for short strings
3852 such as real names; it was not keeping to the 75-character limit for
3853 encoded words that the RFC imposes. It now respects the limit, and
3854 generates multiple encoded words if necessary. To be on the safe side, I
3855 have increased the buffer size for the ${rfc2047: expansion operator from
3858 22. Failure to deliver a bounce message always caused it to be frozen, even if
3859 there was an errors_to setting on the router. The errors_to setting is now
3862 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3863 canonical form (fully expanded) before being placed in
3864 $sender_host_address.
3866 24. Updated eximstats to version 1.33
3868 25. Include certificate and key file names in error message when GnuTLS fails
3869 to set them up, because the GnuTLS error message doesn't include the name
3870 of the failing file when there is a problem reading it.
3872 26. Expand error message when OpenSSL has problems setting up cert/key files.
3875 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3876 (this can affect the format of dates).
3878 28. exim_tidydb, when checking for the continued existence of a message for
3879 which it has found a message-specific retry record, was not finding
3880 messages that were in split spool directories. Consequently, it was
3881 deleting retry records that should have stayed in existence.
3883 29. eximstats updated to version 1.35
3884 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3885 1.35 - bugfix such that pie charts by volume are generated correctly
3887 30. The SPA authentication driver was not abandoning authentication and moving
3888 on to the next authenticator when an expansion was forced to fail,
3889 contradicting the general specification for all authenticators. Instead it
3890 was generating a temporary error. It now behaves as specified.
3892 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3893 (the order specifies the preference for clients). The order is now AES256,
3894 AES128, 3DES, ARCFOUR128.
3896 31. Small patch to Sieve code - explicitly set From: when generating an
3899 32. Exim crashed if a remote delivery caused a very long error message to be
3900 recorded - for instance if somebody sent an entire SpamAssassin report back
3901 as a large number of 550 error lines. This bug was coincidentally fixed by
3902 increasing the size of one of Exim's internal buffers (big_buffer) that
3903 happened as part of the Exiscan merge. However, to be on the safe side, I
3904 have made the code more robust (and fixed the comments that describe what
3907 33. Some experimental protocols are using DNS PTR records for new purposes. The
3908 keys for these records are domain names, not reversed IP addresses. The
3909 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3910 leaves it alone. Component reversal etc. now happens only for IP addresses.
3913 34. The host_aton() function is supposed to be passed a string that is known
3914 to be a valid IP address. However, in the case of IPv6 addresses, it was
3915 not checking this. This is a hostage to fortune. Exim now panics and dies
3916 if the condition is not met. A case was found where this could be provoked
3917 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3918 components; fortuitously, this particular loophole had already been fixed
3919 by change 4.50/55 or 4.44/33 above.
3921 If there are any other similar loopholes, the new check in host_aton()
3922 itself should stop them being exploited. The report I received stated that
3923 data on the command line could provoke the exploit when Exim was running as
3924 exim, but did not say which command line option was involved. All I could
3925 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3926 running as the user.
3929 35. There was a buffer overflow vulnerability in the SPA authentication code
3930 (which came originally from the Samba project). I have added a test to the
3931 spa_base64_to_bits() function which I hope fixes it.
3934 36. The daemon start-up calls getloadavg() while still root for those OS that
3935 need the first call to be done as root, but it missed one case: when
3936 deliver_queue_load_max is set with deliver_drop_privilege. This is
3937 necessary for the benefit of the queue runner, because there is no re-exec
3938 when deliver_drop_privilege is set.
3940 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3941 was in use, was not putting the data itself into the right store pool;
3942 consequently, it could be overwritten for a subsequent message in the same
3943 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3946 38. Sometimes the final signoff response after QUIT could fail to get
3947 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3948 < 0 before doing a fflush(). This bug looks as though it goes back to the
3949 introduction of TLS in release 3.20, but "sometimes" must have been rare
3950 because the tests only now provoked it.
3956 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3957 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3958 Now it returns an integer. A number of calls were still expecting a T/F
3959 return. Fortuitously, in all cases, the tests worked in OK situations,
3960 which is the norm. However, things would have gone wrong on any write
3961 failures on the smtp file descriptor. This function is used when sending
3962 messages over SMTP and also when doing verify callouts.
3964 2. When Exim is called to do synchronous delivery of a locally submitted
3965 message (the -odf or -odi options), it no longer closes stderr before doing
3968 3. Implemented the mua_wrapper option.
3970 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3972 5. Implemented the functions header_remove(), header_testname(),
3973 header_add_at_position(), and receive_remove_recipient(), and exported them
3976 6. If an ACL "warn" statement specified the addition of headers, Exim already
3977 inserted X-ACL-Warn: at the start if there was no header name. However, it
3978 was not making this test for the second and subsequent header lines if
3979 there were newlines in the string. This meant that an invalid header could
3980 be inserted if Exim was badly configured.
3982 7. Allow an ACL "warn" statement to add header lines at the start or after all
3983 the Received: headers, as well as at the end.
3985 8. Added the rcpt_4xx retry error code.
3987 9. Added postmaster_mailfrom=xxx to callout verification option.
3989 10. Added mailfrom=xxxx to the callout verification option, for verify=
3992 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3993 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3995 12. Inserted some casts to stop certain compilers warning when using pointer
3996 differences as field lengths or precisions in printf-type calls (mostly
3997 affecting debugging statements).
3999 13. Added optional readline() support for -be (dynamically loaded).
4001 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4002 same clock tick as a message's arrival, so that its received time was the
4003 same as the "first fail" time on the retry record, and that message
4004 remained on the queue past the ultimate address timeout, every queue runner
4005 would try a delivery (because it was past the ultimate address timeout) but
4006 after another failure, the ultimate address timeout, which should have then
4007 bounced the address, did not kick in. This was a "< instead of <=" error;
4008 in most cases the first failure would have been in the next clock tick
4009 after the received time, and all would be well.
4011 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4012 being recognized when the domain list was tested by the match_domain
4013 condition in an expansion string.
4015 16. Added the ${str2b64: operator.
4017 17. Exim was always calling setrlimit() to set a large limit for the number of
4018 processes, without checking whether the existing limit was already
4019 adequate. (It did check for the limit on file descriptors.) Furthermore,
4020 errors from getrlimit() and setrlimit() were being ignored. Now they are
4021 logged to the main and panic logs, but Exim does carry on, to try to do its
4022 job under whatever limits there are.
4024 18. Imported PCRE 5.0.
4026 19. Trivial typo in log message " temporarily refused connection" (the leading
4029 20. If the log selector return_path_on_delivery was set and an address was
4030 redirected to /dev/null, the delivery process crashed because it assumed
4031 that a return path would always be set for a "successful" delivery. In this
4032 case, the whole delivery is bypassed as an optimization, and therefore no
4035 21. Internal re-arrangement: the function for sending a challenge and reading
4036 a response while authentication was assuming a zero-terminated challenge
4037 string. It's now changed to take a pointer and a length, to allow for
4038 binary data in such strings.
4040 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4042 23. Exim was not respecting finduser_retries when seeking the login of the
4043 uid under which it was called; it was always trying 10 times. (The default
4044 setting of finduser_retries is zero.) Also, it was sleeping after the final
4045 failure, which is pointless.
4047 24. Implemented tls_on_connect_ports.
4049 25. Implemented acl_smtp_predata.
4051 26. If the domain in control=submission is set empty, Exim assumes that the
4052 authenticated id is a complete email address when it generates From: or
4053 Sender: header lines.
4055 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4056 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4057 chown and chgrp in /bin and hostname in /usr/bin.
4059 28. Exim was keeping the "process log" file open after each use, just as it
4060 does for the main log. This opens the possibility of it remaining open for
4061 long periods when the USR1 signal hits a daemon. Occasional processlog
4062 errors were reported, that could have been caused by this. Anyway, it seems
4063 much more sensible not to leave this file open at all, so that is what now
4066 29. The long-running daemon process does not normally write to the log once it
4067 has entered its main loop, and it closes the log before doing so. This is
4068 so that log files can straightforwardly be renamed and moved. However,
4069 there are a couple of unusual error situations where the daemon does write
4070 log entries, and I had neglected to close the log afterwards.
4072 30. The text of an SMTP error response that was received during a remote
4073 delivery was being truncated at 512 bytes. This is too short for some of
4074 the long messages that one sometimes sees. I've increased the limit to
4077 31. It is now possible to make retry rules that apply only when a message has a
4078 specific sender, in particular, an empty sender.
4080 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4081 it possible to be selective about when SMTP synchronization is enforced.
4083 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4085 32. Implemented hosts_connection_nolog.
4087 33. Added an ACL for QUIT.
4089 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4092 35. Added mailbox_size and mailbox_filecount to appendfile.
4094 36. Added control = no_multiline_responses to ACLs.
4096 37. There was a bug in the logic of the code that waits for the clock to tick
4097 in the case where the clock went backwards by a substantial amount such
4098 that the microsecond fraction of "now" was more than the microsecond
4099 fraction of "then" (but the whole seconds number was less).
4101 38. Added support for the libradius Radius client library this is found on
4102 FreeBSD (previously only the radiusclient library was supported).
4108 1. When certain lookups returned multiple values in the form name=value, the
4109 quoting of the values was not always being done properly. Specifically:
4110 (a) If the value started with a double quote, but contained no whitespace,
4112 (b) If the value contained whitespace other than a space character (i.e.
4113 tabs or newlines or carriage returns) it was not quoted.
4114 This fix has been applied to the mysql and pgsql lookups by writing a
4115 separate quoting function and calling it from the lookup code. The fix
4116 should probably also be applied to nisplus, ibase and oracle lookups, but
4117 since I cannot test any of those, I have not disturbed their existing code.
4119 2. A hit in the callout cache for a specific address caused a log line with no
4120 reason for rejecting RCPT. Now it says "Previous (cached) callout
4121 verification failure".
4123 3. There was an off-by-one bug in the queryprogram router. An over-long
4124 return line was truncated at 256 instead of 255 characters, thereby
4125 overflowing its buffer with the terminating zero. As well as fixing this, I
4126 have increased the buffer size to 1024 (and made a note to document this).
4128 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4129 when Exim is waiting for an SMTP response from a remote server, Exim
4130 restarts its select() call on the socket, thereby resetting its timeout.
4131 This is not a problem when such interrupts are rare. Somebody set up a cron
4132 job to run exiwhat every 2 minutes, which is less than the normal select()
4133 timeout (5 or 10 minutes). This meant that the select() timeout never
4134 kicked in because it was always reset. I have fixed this by comparing the
4135 time when an interrupt arrives with the time at the start of the first call
4136 to select(). If more time than the timeout has elapsed, the interrupt is
4137 treated as a timeout.
4139 5. Some internal re-factoring in preparation for the addition of Sieve
4140 extensions (by MH). In particular, the "personal" test is moved to a
4141 separate function, and given an option for scanning Cc: and Bcc: (which is
4142 not set for Exim filters).
4144 6. When Exim created an email address using the login of the caller as the
4145 local part (e.g. when creating a From: or Sender: header line), it was not
4146 quoting the local part when it contained special characters such as @.
4148 7. Installed new OpenBSD configuration files.
4150 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4151 try to make them clearer.
4153 9. Callout options, other than the timeout value, were being ignored when
4154 verifying sender addresses in header lines. For example, when using
4156 verify = header_sender/callout=no_cache
4158 the cache was (incorrectly) being used.
4160 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4161 only the Cygwin environment.
4163 11. When return_path_on_delivery was set as a log selector, if different remote
4164 addresses in the same message used different return paths and parallel
4165 remote delivery occurred, the wrong values would sometimes be logged.
4166 (Whenever a remote delivery process finished, the return path value from
4167 the most recently started remote delivery process was logged.)
4169 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4170 lines when AUTH and/or TLS are in use. This is the "received protocol"
4171 field. Exim used to use "asmtp" for authenticated SMTP, without any
4172 indication (in the protocol name) for TLS use. Now it follows the RFC and
4173 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4174 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4175 names appear in log lines as well as in Received: header lines.
4177 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4178 extensions, and comparison tests, and to fix some bugs.
4180 14. Changes to the "personal" filter test:
4182 (1) The test was buggy in that it was just doing the equivalent of
4183 "contains" tests on header lines. For example, if a user's address was
4184 anne@some.where, the "personal" test would incorrectly be true for
4186 To: susanne@some.where
4188 This test is now done by extracting each address from the header in turn,
4189 and checking the entire address. Other tests that are part of "personal"
4190 are now done using regular expressions (for example, to check local parts
4191 of addresses in From: header lines).
4193 (2) The list of non-personal local parts in From: addresses has been
4194 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4195 taken from the Sieve specification recommendations.
4197 (3) If the message contains any header line starting with "List-" it is
4198 treated as non-personal.
4200 (4) The test for "circular" in the Subject: header line has been removed
4201 because it now seems ill-conceived.
4203 15. Minor typos in src/EDITME comments corrected.
4205 16. Installed latest exipick from John Jetmore.
4207 17. If headers_add on a router specified a text string that was too long for
4208 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4209 of string_sprintf() is now avoided.
4211 18. $message_body_size was not set (it was always zero) when running the DATA
4212 ACL and the local_scan() function.
4214 19. For the "mail" command in an Exim filter, no default was being set for
4215 the once_repeat time, causing a random time value to be used if "once" was
4216 specified. (If the value happened to be <= 0, no repeat happened.) The
4217 default is now 0s, meaning "never repeat". The "vacation" command was OK
4218 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4219 (I found it when inspecting the code).
4221 20. There is now an overall timeout for performing a callout verification. It
4222 defaults to 4 times the callout timeout, which applies to individual SMTP
4223 commands during the callout. The overall timeout applies when there is more
4224 than one host that can be tried. The timeout is checked before trying the
4225 next host. This prevents very long delays if there are a large number of
4226 hosts and all are timing out (e.g. when the network connections are timing
4227 out). The value of the overall timeout can be changed by specifying an
4228 additional sub-option for "callout", called "maxwait". For example:
4230 verify = sender/callout=5s,maxwait=20s
4232 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4233 to the end before writing, but this should make it even safer).
4235 22. Exim was forgetting that it had advertised PIPELINING for the second and
4236 subsequent messages on an SMTP connection. It was also not resetting its
4237 memory on STARTTLS and an internal HELO.
4239 23. When Exim logs an SMTP synchronization error within a session, it now
4240 records whether PIPELINING has been advertised or not.
4242 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4243 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4244 rather than long int.
4246 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4248 26. Added the never_mail option to autoreply.
4254 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4255 crash if the getsockname() call failed; this can happen if a connection is
4256 closed very soon after it is established. The problem was simply in the
4257 order in which certain operations were done, causing Exim to try to write
4258 to the SMTP stream before it had set up the file descriptor. The bug has
4259 been fixed by making things happen in the correct order.
4265 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4266 before the connection was closed, thus losing the rejection response.
4268 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4269 some early Solaris releases, but causes trouble in current releases where
4270 socklen_t is defined.
4272 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4275 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4278 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4279 that releasing the top of it at the end releases what was used for sub-
4280 expansions (unless the block got too big). However, discard this block if
4281 the first thing is a variable or header, so that we can use its block when
4282 it is dynamic (useful for very large $message_headers, for example).
4284 6. Lookups now cache *every* query, not just the most recent. A new, separate
4285 store pool is used for this. It can be recovered when all lookup caches are
4286 flushed. Lookups now release memory at the end of their result strings.
4287 This has involved some general refactoring of the lookup sources.
4289 7. Some code has been added to the store_xxx() functions to reduce the amount
4290 of flapping under certain conditions.
4292 8. log_incoming_interface used to affect only the <= reception log lines. Now
4293 it causes the local interface and port to be added to several more SMTP log
4294 lines, for example "SMTP connection from", and rejection lines.
4296 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4298 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4300 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4301 could be overwritten at the end of the current message (or the start of a
4302 new message if it was set in a HELO ACL). The value is now preserved for
4303 the duration of the SMTP connection.
4305 12. If a transport had a headers_rewrite setting, and a matching header line
4306 contained an unqualified address, that address was qualified, even if it
4307 did not match any rewriting rules. The underlying bug was that the values
4308 of the flags that permit the existence of unqualified sender and recipient
4309 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4310 non-local messages, and by -bnq for local messages) were not being
4311 preserved with the message after it was received.
4313 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4314 "next input=" as part of the text comprising the host identity instead of
4315 the correct text. The code was using the same buffer for two different
4316 strings. However, depending on which order the printing function evaluated
4317 its arguments, the bug did not always show up. Under Linux, for example, my
4318 test suite worked just fine.
4320 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4321 doesn't work with some older versions of Perl. It has been changed to "my",
4322 which in any case is probably the better facility to use.
4324 15. A really picky compiler found some instances of statements for creating
4325 error messages that either had too many or two few arguments for the format
4328 16. The size of the buffer for calls to the DNS resolver has been increased
4329 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4330 for addresses that have a lot of PTR records. This alleviates a problem; it
4331 does not fully solve it.
4333 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4334 buffer now truncates the list and logs the incident, which is the same
4335 action as happens when Exim is looking up a host name and its aliases.
4336 Previously in this situation something unpredictable would happen;
4337 sometimes it was "internal error: store_reset failed".
4339 18. If a server dropped the connection unexpectedly when an Exim client was
4340 using GnuTLS and trying to read a response, the client delivery process
4341 crashed while trying to generate an error log message.
4343 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4344 string, for example:
4346 warn message = H1: something\nH2: something
4348 the text was added as a single header line from Exim's point of view
4349 though it ended up OK in the delivered message. However, searching for the
4350 second and subsequent header lines using $h_h2: did not work. This has been
4351 fixed. Similarly, if a system filter added multiple headers in this way,
4352 the routers could not see them.
4354 20. Expanded the error message when iplsearch is called with an invalid key to
4355 suggest using net-iplsearch in a host list.
4357 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4358 ACLs are no longer actually imposed (and a message to that effect is
4361 22. If a "gecos" field in a passwd entry contained escaped characters, in
4362 particular, if it contained a \" sequence, Exim got it wrong when building
4363 a From: or a Sender: header from that name. A second bug also caused
4364 incorrect handling when an unquoted " was present following a character
4365 that needed quoting.
4367 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4368 was not being matched caselessly.
4370 24. Arranged for all hyphens in the exim.8 source to be escaped with
4373 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4374 properly. Recipient callout cache records were still being keyed to include
4375 the sender, even when use_sender was set false. This led to far more
4376 callouts that were necessary. The sender is no longer included in the key
4377 when use_sender is false.
4379 26. Added "control = submission" modifier to ACLs.
4381 27. Added the ${base62d: operator to decode base 62 numbers.
4383 28. dnsdb lookups can now access SRV records.
4385 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4386 the configuration file.
4388 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4389 (-v) mode. This makes the output for a verbose queue run more intelligible.
4391 31. Added a use_postmaster feature to recipient callouts.
4393 32. Added the $body_zerocount variable, containing the number of binary zero
4394 bytes in the message body.
4396 33. The time of last modification of the "new" subdirectory is now used as the
4397 "mailbox time last read" when there is a quota error for a maildir
4400 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4402 35. Added +ignore_unknown as a special item in host lists.
4404 36. Code for decoding IPv6 addresses in host lists is now included, even if
4405 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4406 address was recognized as an IP address, but was then not correctly decoded
4407 into binary, causing unexpected and incorrect effects when compared with
4414 1. Very minor rewording of debugging text in manualroute to say "list of
4415 hosts" instead of "hostlist".
4417 2. If verify=header_syntax was set, and a header line with an unqualified
4418 address (no domain) and a large number of spaces between the end of the
4419 name and the colon was received, the reception process suffered a buffer
4420 overflow, and (when I tested it) crashed. This was caused by some obsolete
4421 code that should have been removed. The fix is to remove it!
4423 3. When running in the test harness, delay a bit after writing a bounce
4424 message to get a bit more predictability in the log output.
4426 4. Added a call to search_tidyup() just before forking a reception process. In
4427 theory, someone could use a lookup in the expansion of smtp_accept_max_
4428 per_host which, without the tidyup, could leave open a database connection.
4430 5. Added the variables $recipient_data and $sender_data which get set from a
4431 lookup success in an ACL "recipients" or "senders" condition, or a router
4432 "senders" option, similar to $domain_data and $local_part_data.
4434 6. Moved the writing of debug_print from before to after the "senders" test
4437 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4438 problems for message scanning, either using a data ACL, or using
4439 local_scan() because the Received: header was not generated till after they
4440 were called (in order to set the time as the time of reception completion).
4441 I have revised the way this works. The header is now generated after the
4442 body is received, but before the ACL or local_scan() are called. After they
4443 are run, the timestamp in the header is updated.
4449 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4450 before starting a queue runner without re-exec. This happened only when
4451 deliver_drop_privilege was set or when the Exim user was set to root. The
4452 effect of the bug was that timeouts during subsequent deliveries caused
4453 crashes instead of being properly handled. The handler is now left at its
4454 default (and expected) setting.
4456 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4457 message, again when deliver_drop_privilege is set or Exim is run as root.
4458 The bug described in (1) was not present in this case, but the tidying up
4459 of the other signals was missing. I have made the two cases consistent.
4461 3. The ignore_target_hosts setting on a manualroute router was being ignored
4462 for hosts that were looked up using the /MX notation.
4464 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4467 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4468 operated on the sender address. After changing the $sender_address to <>
4469 for the sender address verify, Exim was re-instated it as the original
4470 (before rewriting) address, but remembering that it had rewritten it, so it
4471 wasn't rewriting it again. This bug also had the effect of breaking the
4472 sender address verification caching when the sender address was rewritten.
4474 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4475 This has been changed so that if the ip literal address matches
4476 ignore_target_hosts, the router declines.
4478 7. Added expansion conditions match_domain, match_address, and match_local_
4479 part (NOT match_host).
4481 8. The placeholder for the Received: header didn't have a length field set.
4483 9. Added code to Exim itself and to exim_lock to test for a specific race
4484 condition that could lead to file corruption when using MBX delivery. The
4485 issue is with the lockfile that is created in /tmp. If this file is removed
4486 after a process has opened it but before that process has acquired a lock,
4487 there is the potential for a second process to recreate the file and also
4488 acquire a lock. This could lead to two Exim processes writing to the file
4489 at the same time. The added code performs the same test as UW imapd; it
4490 checks after acquiring the lock that its file descriptor still refers to
4491 the same named file.
4493 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4494 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4497 11. Added the smtp_active_hostname option. If used, this will typically be made
4498 to depend on the incoming interface address. Because $interface_address is
4499 not set up until the daemon has forked a reception process, error responses
4500 that can happen earlier (such as "too many connections") no longer contain
4503 12. If an expansion in a condition on a "warn" statement fails because a lookup
4504 defers, the "warn" statement is abandoned, and the next ACL statement is
4505 processed. Previously this caused the whole ACL to be aborted.
4507 13. Added the iplsearch lookup type.
4509 14. Added ident_timeout as a log selector.
4511 15. Added tls_certificate_verified as a log selector.
4513 16. Added a global option tls_require_ciphers (compare the smtp transport
4514 option of the same name). This controls incoming TLS connections.
4516 17. I finally figured out how to make tls_require_ciphers do a similar thing
4517 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4518 before starting the TLS session.
4520 18. Tabs are now shown as \t in -bP output.
4522 19. If the log selector return_path_on_delivery was set, Exim crashed when
4523 bouncing a message because it had too many Received: header lines.
4525 20. If two routers both had headers_remove settings, and the first one included
4526 a superfluous trailing colon, the final name in the first list and the
4527 first name in the second list were incorrectly joined into one item (with a
4528 colon in the middle).
4534 1. Added -C and -D options to the exinext utility, mainly to make it easier
4535 to include in the automated testing, but these could be helpful when
4536 multiple configurations are in use.
4538 2. The exinext utility was not formatting the output nicely when there was
4539 an alternate port involved in the retry record key, nor when there was a
4540 message id as well (for retries that were specific to a specific message
4541 and a specific host). It was also confused by IPv6 addresses, because of
4542 the additional colons they contain. I have fixed the IPv4 problem, and
4543 patched it up to do a reasonable job for IPv6.
4545 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4546 delivery, the log line now contains "pipelined" if PIPELINING was used.
4548 4. An SMTP transport process used to panic and die if the bind() call to set
4549 an explicit outgoing interface failed. This has been changed; it is now
4550 treated in the same way as a connect() failure.
4552 5. A reference to $sender_host_name in the part of a conditional expansion
4553 that was being skipped was still causing a DNS lookup. This no longer
4556 6. The def: expansion condition was not recognizing references to header lines
4557 that used bh_ and bheader_.
4559 7. Added the _cache feature to named lists.
4561 8. The code for checking quota_filecount in the appendfile transport was
4562 allowing one more file than it should have been.
4564 9. For compatibility with Sendmail, the command line option
4572 and sets the incoming protocol and host name (for trusted callers). The
4573 host name and its colon can be omitted when only the protocol is to be set.
4574 Note the Exim already has two private options, -pd and -ps, that refer to
4575 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4576 "s", but I don't think that's a major issue.
4578 10. A number of refactoring changes to the code, none of which should affect
4581 (a) The number of logging options was getting close to filling up the
4582 32-bit word that was used as a bit map. I have split them into two classes:
4583 those that are passed in the argument to log_write(), and those that are
4584 only ever tested independently outside of that function. These are now in
4585 separate 32-bit words, so there is plenty of room for expansion again.
4586 There is no change in the user interface or the logging behaviour.
4588 (b) When building, for example, log lines, the code previously used a
4589 macro that called string_cat() twice, in order to add two strings. This is
4590 not really sufficiently general. Furthermore, there was one instance where
4591 it was actually wrong because one of the argument was used twice, and in
4592 one call a function was used. (As it happened, calling the function twice
4593 did not affect the overall behaviour.) The macro has been replaced by a
4594 function that can join an arbitrary number of extra strings onto a growing
4597 (c) The code for expansion conditions now uses a table and a binary chop
4598 instead of a serial search (which was left over from when there were very
4599 few conditions). Also, it now recognizes conditions like "pam" even when
4600 the relevant support is not compiled in: a suitably worded error message is
4601 given if an attempt is made to use such a condition.
4603 11. Added ${time_interval:xxxxx}.
4605 12. A bug was causing one of the ddress fields not to be passed back correctly
4606 from remote delivery subprocesses. The field in question was not being
4607 subsequently used, so this caused to problems in practice.
4609 13. Added new log selectors queue_time and deliver_time.
4611 14. Might have fixed a bug in maildirsizefile handling that threw up
4612 "unexpected character" debug warnings, and recalculated the data
4613 unnecessarily. In any case, I expanded the warning message to give more
4616 15. Added the message "Restricted characters in address" to the statements in
4617 the default ACL that block characters like @ and % in local parts.
4619 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4620 Three changes have been made:
4622 (a) There was a serious bug; a negative response to MAIL caused the whole
4623 recipient domain to be cached as invalid, thereby blocking all messages
4624 to all local parts at the same domain, from all senders. This bug has
4625 been fixed. The domain is no longer cached after a negative response to
4626 MAIL if the sender used is not empty.
4628 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4631 (c) A new callout option, "use_sender" has been added for people who want
4632 the modified behaviour.
4638 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4641 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4642 indeed breaks things for older releases.
4644 3. Added additional logging to the case where there is a problem reading data
4645 from a filter that is running in a subprocess using a pipe, in order to
4646 try to track down a specific problem.
4648 4. Testing facility fudge: when running in the test harness and attempting
4649 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4650 getting "No route to host". Convert this to a timeout.
4652 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4655 6. Some OS don't have socklen_t but use size_t instead. This affects the
4656 fifth argument of getsockopt() amongst other things. This is now
4657 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4658 can be set for individual OS. I have set it for SunOS5, OSF1, and
4659 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4660 some earlier ones do not.
4662 7. Change 4.30/15 was not doing the test caselessly.
4664 8. The standard form for an IPv6 address literal was being rejected by address
4665 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4666 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4667 this, as well as the form without the "IPv6" on the front (but only when
4668 address literals are enabled, of course).
4670 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4672 10. Exim crashed if a message with an empty sender address specified by -f
4673 encountered a router with an errors_to setting. This could be provoked only
4674 by a command such as
4678 where an empty string was supplied; "<>" did not hit this bug.
4680 11. Installed PCRE release 4.5.
4682 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4683 remained set. It is now erased.
4685 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4686 times from message ids (which are base 36 rather than the normal 62).
4688 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4689 were being counted as actual protocol errors, and logged if the log
4690 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4691 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4692 and DATA following a set of rejected RCPTs do not count as protocol errors.
4693 In other words, Exim assumes they were pipelined, though this may not
4694 actually be the case. Of course, in all cases the client gets an
4695 appropriate error code.
4697 15. If a lookup fails in an ACL condition, a message about the failure may
4698 be available; it is used if testing the ACL cannot continue, because most
4699 such messages specify what the cause of the deferral is. However, some
4700 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4701 that caused an old message to be retained and used if a later statement
4702 caused a defer, replacing the real cause of the deferral.
4704 16. If an IP address had so many PTR records that the DNS lookup buffer
4705 was not large enough to hold them, Exim could crash while trying to process
4706 the truncated data. It now detects and logs this case.
4708 17. Further to 4.21/58, another change has been made: if (and only if) the
4709 first line of a message (the first header line) ends with CRLF, a bare LF
4710 in a subsequent header line has a space inserted after it, so as not to
4711 terminate the header.
4713 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4714 unnecessarily, used atoi() instead of strtol(), and didn't check the
4715 termination when getting file sizes from file names by regex.
4717 19. Completely re-implemented the support for maildirsize files, in the light
4718 of a number of problems with the previous contributed implementation
4719 (4.30/29). In particular:
4721 . If the quota is zero, the maildirsize file is maintained, but no quota is
4724 . If the maildir directory does not exist, it is created before any attempt
4725 to write a maildirsize file.
4727 . The quota value in the file is just a cache; if the quota is changed in
4728 the transport, the new value overrides.
4730 . A regular expression is available for excluding directories from the
4733 20. The autoreply transport checks the characters in options that define the
4734 message's headers; it allows continued headers, but it was checking with
4735 isspace() after an embedded newline instead of explicitly looking for a
4738 21. If all the "regular" hosts to which an address was routed had passed their
4739 expiry times, and had not reached their retry times, the address was
4740 bounced, even if fallback hosts were defined. Now Exim should go on to try
4743 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4744 equivalent code in the SMTP transport. Some hosts send humungous responses
4745 to HELO/EHLO, more than 1024 it seems.
4747 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4748 gives compiler warnings in some environments. I've now done it "properly",
4751 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4752 (because of problems with the built-in one) was declared to return uschar *
4753 instead of char *, causing compiler failure.
4755 25. Fixed a file descriptor leak when processing alias/forward files.
4757 26. Fixed a minor format string issue in dbfn.c.
4759 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4761 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4762 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4765 29. When a sender address is verified, it is cached, to save repeating the test
4766 when there is more than one recipient in a message. However, when the
4767 verification involves a callout, it is possible for different callout
4768 options to be set for different recipients. It is too complicated to keep
4769 track of this in the cache, so now Exim always runs a verification when a
4770 callout is required, relying on the callout cache for the optimization.
4771 The overhead is duplication of the address routing, but this should not be
4774 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4775 to be verified with callout=postmaster, and the main callout worked but the
4776 postmaster check failed, the verification correctly failed. However, if a
4777 subsequent RCPT command asked for sender verification *without* the
4778 postmaster check, incorrect caching caused this verification also to fail,
4781 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4782 it was not caching the DNS options (qualify_single, search_parents) that
4783 were used when the lookup failed. A subsequent lookup with different
4784 options therefore always gave the same answer, though there were cases
4785 where it should not have. (Example: a "domains = !$mx_any" option on a
4786 dnslookup router: the "domains" option is always processed without any
4787 widening, but the router might have qualify_single set.) Now Exim uses the
4788 cached value only when the same options are set.
4790 32. Added John Jetmore's "exipick" utility to the distribution.
4792 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4793 than a timeout (e.g. a certificate is required, and is not provided), an
4794 Exim server now closes the connection immediately. Previously it waited for
4795 the client to close - but if the client is SSL, it seems that they each
4796 wait for each other, leading to a delay before one of them times out.
4798 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4799 maintained 0.8.x compatibility because I don't think many are using it, and
4800 it is clearly obsolete.
4802 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4805 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4806 client certificate was expired. A simple patch fixes this, though I don't
4807 understand the full logic of why the verify callback is called multiple
4810 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4811 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4812 which causes problems with some clients (such as the Certicom SSL Plus
4813 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4814 disables the coutermeasure allowing Eudora to connect."
4816 38. Exim was not checking that a write() to a log file succeeded. This could
4817 lead to Bad Things if a log got too big, in particular if it hit a file
4818 size limit. Exim now panics and dies if it cannot write to a log file, just
4819 as it does if it cannot open a log file.
4821 39. Modified OS/Makefile-Linux so that it now contains
4823 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4825 The two -D definitions ensure that Exim is compiled with large file
4826 support, which makes it possible to handle log files that are bigger than
4829 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4830 instance) a domain was checked against a named list that involved a lookup,
4831 causing $domain_data to be set, then another domain was checked against the
4832 same list, then the first domain was re-checked, the value of $domain_data
4833 after the final check could be wrong. In particular, if the second check
4834 failed, it could be set empty. This bug probably also applied to
4837 41. The strip_trailing_dot option was not being applied to the address given
4838 with the -f command-line option.
4840 42. The code for reading a message's header from the spool was incrementing
4841 $received_count, but never initializing it. This meant that the value was
4842 incorrect (doubled) while delivering a message in the same process in which
4843 it was received. In the most common configuration of Exim, this never
4844 happens - a fresh exec is done - but it can happen when
4845 deliver_drop_privilege is set.
4847 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4848 it now includes up to 150 characters of the unexpected data in the log
4851 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4852 and building data strings. The size of both of these buffers was 10 000
4853 bytes - far larger than anybody would *ever* want, thought I. Needless to
4854 say, somebody hit the limit. I have increased the maximum line length to
4855 20 000 and the maximum data length of concatenated lines to 100 000. I have
4856 also fixed two bugs, because there was no checking on these buffers. Tsk,
4857 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4858 buffer is too small.
4860 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4861 lsearch lookups. Now it does.
4863 46. When parsing a route_list item in a manualroute router, a fixed-length
4864 buffer was used for the list of hosts. I made this 1024 bytes long,
4865 thinking that nobody would ever have a list of hosts that long. Wrong.
4866 Somebody had a whole pile of complicated expansion conditions, and the
4867 string was silently truncated, leading to an expansion error. It turns out
4868 that it is easier to change to an unlimited length (owing to other changes
4869 that have happened since this code was originally written) than to build
4870 structure for giving a limitation error. The length of the item that
4871 expands into the list of hosts is now unlimited.
4873 47. The lsearch lookup could not handle data where the length of text line was
4874 more than 4095 characters. Such lines were truncated, leading to shortened
4875 data being returned. It should now handle lines of any length.
4877 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4878 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4881 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4882 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4883 no longer happens in comment lines. A list of replacements is now placed
4884 at the head of all of the source files, except those whose only change is
4885 to replace PERL_COMMAND in the very first #! line.
4887 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4888 messages on the queue, with a bottom-up merge sort, using code contributed
4889 by Michael Haardt. This should make operations like -bp somewhat faster on
4890 large queues. It won't affect queue runners, except when queue_run_in_order
4893 51. Installed eximstats 1.31 in the distribution.
4895 52. Added support for SRV lookups to the dnslookup router.
4897 53. If an ACL referred to $message_body or $message_body_end, the value was not
4898 reset for any messages that followed in the same SMTP session.
4900 54. The store-handling optimization for building very long strings was not
4901 differentiating between the different store pools. I don't think this
4902 actually made any difference in practice, but I've tidied it.
4904 55. While running the routers to verify a sender address, $sender_address
4905 was still set to the sender address. This is wrong, because when routing to
4906 send a bounce to the sender, it would be empty. Therefore, I have changed
4907 it so that, while verifying a sender address, $sender_address is set to <>.
4908 (There is no change to what happens when verifying a recipient address.)
4910 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4911 target A or AAAA records (if not already returned) without resetting the
4912 qualify_single or search_parents options of the DNS resolver. These are
4913 inappropriate in this case because the targets of MX and SRV records must
4914 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4915 target that, when qualified, matched something in the local domain. These
4916 two options are now turned off when doing these lookups.
4918 57. It seems that at least some releases of Reiserfs (which does not have the
4919 concept of a fixed number of inodes) returns zero and not -1 for the
4920 number of available inodes. This interacted badly with check_spool_inodes,
4921 which assumed that -1 was the "no such thing" setting. What I have done is
4922 to check that the total number of inodes is greater than zero before doing
4923 the test of how many are available.
4925 58. When a "warn" ACL statement has a log_message modifier, the message is
4926 remembered, and not repeated. This is to avoid a lot of repetition when a
4927 message has many recipients that cause the same warning to be written.
4928 Howewer, Exim was preserving the list of already written lines for an
4929 entire SMTP session, which doesn't seem right. The memory is now reset if a
4930 new message is started.
4932 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4933 debugging output unless log_rewrite was also set.
4935 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4936 of (int)(handle) when we know that handle contains (void *)(-1).
4938 61. The Exim daemon panic-logs an error return when it closes the incoming
4939 connection. However "connection reset by peer" seems to be common, and
4940 isn't really an error worthy of noting specially, so that particular error
4943 62. When Exim is trying to find all the local interfaces, it used to panic and
4944 die if the ioctl to get the interface flags failed. However, it seems that
4945 on at least one OS (Solaris 9) it is possible to have an interface that is
4946 included in the list of interfaces, but for which you get a failure error
4947 for this call. This happens when the interface is not "plumbed" into a
4948 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4949 failure of the "get flags" call assumes that the interface is down.
4951 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4952 makes life easier for people who are doing arithmetic on fields extracted
4953 from dates, where you often get leading zeros that should not be
4954 interpreted as octal.
4956 64. Added qualify_domain to the redirect router, to override the global
4959 65. If a pathologically long header line contained very many addresses (the
4960 report of this problem mentioned 10 000) and each of them was rewritten,
4961 Exim could use up a very large amount of memory. (It kept on making new
4962 copies of the header line as it rewrote, and never released the old ones.)
4963 At the expense of a bit more processing, the header rewriting function has
4964 been changed so that it no longer eats memory in this way.
4966 66. The generation of the Received: header has been moved from the time that a
4967 message starts to be received, to the time that it finishes. The timestamp
4968 in the Received: header should now be very close to that of the <= log
4969 line. There are two side-effects of this change:
4971 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4972 logged header lines no longer include the local Received: line, because
4973 it has not yet been created. The same applies to a copy of the message
4974 that is returned to a non-SMTP sender when a message is rejected.
4976 (b) When a filter file is tested using -bf, no additional Received: header
4977 is added to the test message. After some thought, I decided that this
4980 This change does not affect the value of $received_for. It is still set
4981 after address rewriting, but before local_scan() is called.
4983 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4985 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4986 gave an unhelpful panic error message, and a defer error. I have managed to
4987 change this behaviour so that it now rejects any supplied certificate,
4988 which seems right, as the list of acceptable certificates is empty.
4990 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4991 gave an unhelpful defer error. I have not managed to make this reject any
4992 supplied certificates, but the error message it gives is "no certificate
4993 supplied", which is not helpful.
4995 70. exigrep's output now also includes lines that are not associated with any
4996 message, but which match the given pattern. Implemented by a patch from
4997 Martin Sluka, which also tidied up the Perl a bit.
4999 71. Recipient callout verification, like sender verification, was using <> in
5000 the MAIL FROM command. This isn't really the right thing, since the actual
5001 sender may affect whether the remote host accepts the recipient or not. I
5002 have changed it to use the actual sender in the callout; this means that
5003 the cache record is now keyed on a recipient/sender pair, not just the
5004 recipient address. There doesn't seem to be a real danger of callout loops,
5005 since a callout by the remote host to check the sender would use <>.
5006 [SEE ABOVE: changed after hitting problems.]
5008 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5009 temporary errors. However, in the case of such a code being given after
5010 the end of a data transmission (i.e. after ".") Exim was failing to write
5011 a retry record for the message. (Yes, there was some broken host that was
5012 actually sending 8xx at this point.)
5014 73. An unknown lookup type in a host list could cause Exim to panic-die when
5015 the list was checked. (An example that provoked this was putting <; in the
5016 middle of a list instead of at the start.) If this happened during a DATA
5017 ACL check, a -D file could be left lying around. This kind of configuration
5018 error no longer causes Exim to die; instead it causes a defer errror. The
5019 incident is still logged to the main and panic logs.
5021 74. Buglet left over from Exim 3 conversion. The message "too many messages
5022 in one connection" was written to the rejectlog but not the mainlog, except
5023 when address rewriting (yes!) was being logged.
5025 75. Added write_rejectlog option.
5027 76. When a system filter was run not as root (that is, when system_filter_user
5028 was set), the values of the $n variables were not being returned to the
5029 main process; thus, they were not subsequently available in the $sn
5032 77. Added +return_path_on_delivery log selector.
5034 78. A connection timeout was being treated differently from recipients deferred
5035 when testing hosts_max_try with a message that was older than the host's
5036 retry timeout. (The host should not be counted, thus allowing all hosts to
5037 be tried at least once before bouncing.) This may have been the cause of an
5038 occasionally reported bug whereby a message would remain on the queue
5039 longer than the retry timeout, but would be bounced if a delivery was
5040 forced. I say "may" because I never totally pinned down the problem;
5041 setting up timeout/retry tests is difficult. See also the next item.
5043 79. The ultimate address timeout was not being applied to errors that involved
5044 a combination of host plus message (for example, a timeout on a MAIL
5045 command). When an address resolved to a number of possible hosts, and they
5046 were not all tried for each delivery (e.g. because of hosts_max_try), a
5047 message could remain on the queue longer than the retry timeout.
5049 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5052 81. Fixed an obscure SMTP outgoing bug which required at least the following
5053 conditions: (a) there was another message waiting for the same server;
5054 (b) the server returned 5xx to all RCPT commands in the first message so
5055 that the message was not completed; (c) the server dropped the connection
5056 or gave a negative response to the RSET that Exim sends to abort the
5057 transaction. The observed case was a dropped connection after DATA that had
5058 been sent in pipelining mode. That is, the server had advertised PIPELINING
5059 but was not implementing it correctly. The effect of the bug was incorrect
5060 behaviour, such as trying another host, and this could lead to a crash.
5066 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5067 and daemon.c were passed as pointers to ints; they should have been
5068 pointers to socklen_t variables (which are typically unsigned ints).
5070 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5073 3. Fixed a really odd bug that affected only the testing scheme; patching a
5074 certain fixed string in the binary changed the value of another string that
5075 happened to be identical to the end of the original first string.
5077 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5078 name", it returns that address as the IP address. On some operating
5079 systems (e.g. Solaris), it also passes back the IP address string as the
5080 "host name". However, on others (e.g. Linux), it passes back an empty
5081 string. Exim wasn't checking for this, and was changing the host name to an
5082 empty string, assuming it had been canonicized.
5084 5. Although rare, it is permitted to have more than one PTR record for a given
5085 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5086 all the names associated with an address, because they do in Solaris.
5087 However, it seems that they do not in Linux for data that comes from the
5088 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5089 I found this out when I moved to a new Linux workstation and tried to run
5090 the Exim test suite.
5092 To get round this problem I have changed the code so that it now does its
5093 own call to the DNS to look up PTR records when searching for a host name.
5094 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5095 addresses that are only in /etc/hosts are still found.
5097 This behaviour is, however, controlled by an option called host_lookup_
5098 order, which defaults to "bydns:byaddr". If people want to use the other
5099 order, or indeed, just use one or the other means of lookup, they can
5100 specify it in this variable.
5102 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5103 some operating systems, this comes back from gethostbyaddr() as an empty
5104 string, and this is what Exim used to test for. However, it seems that in
5105 other systems, "." is yielded. Exim now tests for this case too.
5107 7. The values of check_spool_space and check_log_space are now held internally
5108 as a number of kilobytes instead of an absolute number of bytes. If a
5109 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5110 kilobyte. This means that much larger values can be stored.
5112 8. Exim monitor: an attempt to get the action menu when not actually pointing
5113 at a message produces an empty menu entitled "No message selected". This
5114 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5115 no entries in it ("Shell widget menu has zero width and/or height"). So I
5116 have added a single, blank menu entry in this case.
5118 9. Added ${quote_local_part.
5120 10. MIME decoding is now applied to the contents of Subject: header lines when
5123 11. Now that a reference to $sender_host_address automatically causes a reverse
5124 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5125 host lookup before query-style lookups in lists that might use this
5126 variable. This has therefore been abolished, and the "net-" prefix is no
5127 longer necessary for query-style lookups.
5129 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5130 have been a typo for LFLAGS, so it has been changed.
5132 13. The install script calls Exim with "-C /dev/null" in order to find the
5133 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5134 to be output. Howeve, since Exim outputs its version number before the
5135 error, it didn't break the script. It just looked ugly. I fixed this by
5136 always allowing "-C /dev/null" if the caller is root.
5138 14. Ignore overlarge ACL variable number when reading spool file - insurance
5139 against a later release with more variables having written the file.
5141 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5142 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5143 this, as well as the form without the "IPv6" on the front.
5145 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5146 OS/Makefile-Darwin file.
5148 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5149 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5151 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5152 variable. However, some OS use ints for inodes. I've added cast to long int
5153 to get rid of the compiler warning.
5155 19. I had forgotten to lock out "/../" in configuration file names when
5156 ALT_CONFIG_PREFIX was set.
5158 20. Routers used for verification do not need to specify transports. However,
5159 if such a router generated a host list, and callout was configured, Exim
5160 crashed, because it could not find a port number from the (non-existent)
5161 transport. It now assumes port 25 in this circumstance.
5163 21. Added the -t option to exigrep.
5165 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5166 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5167 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5168 them is set, LOOKUP_LSEARCH is forced.
5170 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5172 24. Added sender and host information to the "rejected by local_scan()" log
5173 line; previously there was no indication of these.
5175 25. Added .include_if_exists.
5177 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5178 Linux. It turns out that not all file systems support this. Apparently some
5179 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5180 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5181 not supported on the file descriptor, is now ignored when Exim is trying to
5182 sync a directory. This applies only to Linux.
5184 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5186 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5187 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5190 29. Added support for maildirsize files from supplied patch (modified a bit).
5192 30. The use of :fail: followed by an empty string could lead Exim to respond to
5193 sender verification failures with (e.g.):
5195 550 Verification failed for <xxx>
5196 550 Sender verify failed
5198 where the first response line was missing the '-' that indicates it is not
5199 the final line of the response.
5201 31. The loop for finding the name of the user that called Exim had a hardwired
5202 limit of 10; it now uses the value of finduser_retries, which is used for
5203 all other user lookups.
5205 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5208 33. Exim was neglecting to zero errno before one call of strtol() when
5209 expanding a string and expecting an integer value. On some systems this
5210 resulted in spurious "integer overflow" errors. Also, it was casting the
5211 result into an int without checking.
5213 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5214 did not work. The code looks as if it has *never* worked, though it appears
5215 to have been documented since at least releast 1.62. I have made it work.
5217 35. The "timeout_DNS" error in retry rules, also documented since at least
5218 1.62, also never worked. As it isn't clear exactly what this means, and
5219 clearly it isn't a major issue, I have abolished the feature by treating it
5220 as "timeout", and writing a warning to the main and panic logs.
5222 36. The display of retry rules for -brt wasn't always showing the error code
5225 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5226 timeout_connect_A, timeout_connect_MX.
5228 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5229 to the empty sender.
5231 39. The daemon was not analysing the content of -oX till after it had closed
5232 stderr and disconnected from the controlling terminal. This meant that any
5233 syntax errors were only noted on the panic log, and the return code from
5234 the command was 0. By re-arranging the code a little, I've made the
5235 decoding happen first, so such errors now appear on stderr, and the return
5236 code is 1. However, the actual setting up of the sockets still happens in
5237 the disconnected process, so errors there are still only recorded on the
5240 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5241 connections (as happens on some IP stacks) was logged at start up time as
5242 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5243 it from "IPv6 and IPv4", which means that two separate sockets are being
5246 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5247 says whether AF_INET or AF_INET6 was passed as an argument.
5249 42. Exiwhat output was messed up when time zones were included in log
5252 43. Exiwhat now gives more information about the daemon's listening ports,
5253 and whether -tls-on-connect was used.
5255 44. The "port" option of the smtp transport is now expanded.
5257 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5258 silently ignored. Now an error message is written to the main and panic
5261 46. There's a new ACL modifier called "logwrite" which writes to a log file
5262 as soon as it is encountered.
5264 47. Added $local_user_uid and $local_user_gid at routing time.
5266 48. Exim crashed when trying to verify a sender address that was being
5269 49. Exim was recognizing only a space character after ".include". It now also
5270 recognizes a tab character.
5272 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5273 extracting the relevant information from the specification. The man page no
5274 longer contains scrambled data for the -d option, and I've added a section
5275 at the front about calling Exim under different names.
5277 51. Added "extra_headers" argument to the "mail" command in filter files.
5279 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5282 53. Installed eximstats 1.29.
5284 54. Added transport_filter_timeout as a generic transport option.
5286 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5287 Cc: header lines. This was required by RFC 822, but it not required by RFC
5290 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5291 incoming messages that did not have them. Now it does so only if the
5292 message originates locally, that is, if there is no associated remote host
5293 address. When Resent- header lines are present, this applies to the Resent-
5294 lines rather than the non-Resent- lines.
5296 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5297 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5299 58. Messages for configuration errors now include the name of the main
5300 configuration file - useful now that there may be more than one file in a
5301 list (.included file names were always shown).
5303 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5304 for those rare installations that do not start the daemon as root or run it
5305 setuid root. I've cut out the call to initgroups() if the daemon is not
5308 60. The Exim user and group can now be bound into the binary as text strings
5309 that are looked up at the start of Exim's processing.
5311 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5313 62. Added $mailstore_basename variable.
5315 63. Installed patch to sieve.c from Michael Haardt.
5317 64. When Exim failed to open the panic log after failing to open the main log,
5318 the original message it was trying to log was written to stderr and debug
5319 output, but if they were not available (the usual case in production), it
5320 was lost. Now it is written to syslog before the two lines that record the
5321 failures to open the logs.
5323 65. Users' Exim filters run in subprocesses under the user's uid. It is
5324 possible for a "deliver" command or an alias in a "personal" command to
5325 provoke an address rewrite. If logging of address rewriting is configured,
5326 this fails because the process is not running as root or exim. There may be
5327 a better way of dealing with this, but for the moment (because 4.30 needs
5328 to be released), I have disabled address rewrite logging when running a
5329 filter in a non-root, non-exim process.
5335 1. The buildconfig auxiliary program wasn't quoting the value set for
5336 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5337 not defined. This bug was masked in 4.22 by the effect that was fixed in
5340 2. Some messages that were rejected after a message id was allocated were
5341 shown as "incomplete" by exigrep. It no longer does this for messages that
5342 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5344 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5345 have allow_domain_literals set, the ID did not get logged in the <= line.
5346 Domain literals are now always recognized in Message-ID: header lines.
5348 4. The first argument for a ${extract expansion item is the key name or field
5349 number. Leading and trailing spaces in this item were not being ignored,
5350 causing some misleading effects.
5352 5. When deliver_drop_privilege was set, single queue runner processes started
5353 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5354 same command in the process it spins off) were not dropping privilege.
5356 6. When the daemon running as "exim" started a queue runner, it always
5357 re-executed Exim in the spun-off process. This is a waste of effort when
5358 deliver_drop_privilege is set. The new process now just calls the
5359 queue-runner function directly.
5365 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5368 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5369 ignored. Though the use of -oP was forcing the writing of a pid file, it
5370 was always written to the default place.
5372 3. If the message "no IP address found for host xxxx" is generated during
5373 incoming verification, it is now followed by identification of the incoming
5374 connection (so you can more easily find what provoked it).
5376 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5378 5. Added some features to "harden" Exim a bit more against certain attacks:
5380 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5381 be put in Local/Makefile. This is like the never_users runtime option,
5382 but it cannot be overridden. The default setting is "root".
5384 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5385 prefix string with which any file named in a -C command line option
5388 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5389 is retained for -C and -D only if the caller of Exim is root. Without
5390 it, the exim user may also use -C and -D and retain privilege.
5392 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5393 command line option is disabled.
5395 6. Macro names set by the -D option must start with an upper case letter, just
5396 like macro names defined in the configuration file.
5398 7. Added "dereference=" facility to LDAP.
5400 8. Two instances of the typo "uknown" in the source files are fixed.
5402 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5403 the Configure-Makefile script screwed up while processing it.
5405 10. Incorporated PCRE 4.4.
5407 11. The SMTP synchronization check was not operating right at the start of an
5408 SMTP session. For example, it could not catch a HELO sent before the client
5409 waited for the greeting. There is now a check for outstanding input at the
5410 point when the greeting is written. Because of the duplex, asynchronous
5411 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5412 way, but not yet received, when the check is performed.
5414 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5415 on TCP/IP sockets, because this apparently causes some broken clients to
5418 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5419 unchanged) from the Cygwin maintainer.
5421 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5422 was supported instead of testing for mbx. Effectively a typo.
5424 15. The spa authenticator server code was not checking that the input it
5425 received was valid base64.
5427 16. The debug output line for the "set" modifier in ACLs was not showing the
5428 name of the variable that was being set.
5430 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5432 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5433 lookup on its own. Something else was needed to trigger the lookup. For
5434 example, a match in host_lookup or the need for a host name in a host list.
5435 Now, if $sender_host_name is referenced and the host name has not yet been
5436 looked up, a lookup is performed. If the lookup fails, the variable remains
5437 empty, and $host_lookup_failed is set to "1".
5439 19. Added "eqi" as a case-independent comparison operator.
5441 20. The saslauthd authentication condition could segfault if neither service
5442 nor realm was specified.
5444 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5445 error message that was logged was misleading, and incoming SMTP
5446 connections were dropped. The message is now more accurate, and temporary
5447 errors are given to SMTP connections.
5449 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5450 (except RSET) with a 421 error, until QUIT is received. However, it was
5451 failing to send a response to QUIT.
5453 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5454 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5455 ACL is now run after the helo_try_verify_hosts code.
5457 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5458 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5459 case-independent, so other case variants are also recognized). Apparently
5460 some systems use these upper case variants.
5462 25. If more than two messages were waiting for the same host, and a transport
5463 filter was specified for the transport, Exim sent two messages over the
5464 same TCP/IP connection, and then failed with "socket operation on non-
5465 socket" when it tried to send the third.
5467 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5469 27. The extern definition of crypt16() in expand.c was not being excluded when
5470 the OS had its own crypt16() function.
5472 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5473 as a preferred synonym for return_size_limit, both as an option and as an
5476 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5478 30. Changed the default configuration ACL to relax the local part checking rule
5479 for addresses that are not in any local domains. For these addresses,
5480 slashes and pipe symbols are allowed within local parts, but the sequence
5481 /../ is explicitly forbidden.
5483 31. SPA server authentication was not clearing the challenge buffer before
5486 32. log_message in a "warn" ACL statement was writing to the reject log as
5487 well as to the main log, which contradicts the documentation and doesn't
5488 seem right (because no rejection is happening). So I have stopped it.
5490 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5491 However, I am unable to do any testing of this.
5493 34. Fixed an infelicity in the appendfile transport. When checking directories
5494 for a mailbox, to see if any needed to be created, it was accidentally
5495 using path names with one or more superfluous leading slashes; tracing
5496 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5498 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5499 contents are added to the log line that is written for every discarded
5500 recipient. (Previously a log_message setting was ignored.)
5502 36. The ${quote: operator now quotes the string if it is empty.
5504 37. The install script runs exim in order to find its version number. If for
5505 some reason other than non-existence or emptiness, which it checks, it
5506 could not run './exim', it was installing it with an empty version number,
5507 i.e. as "exim-". This error state is now caught, and the installation is
5510 38. An argument was missing from the function that creates an error message
5511 when Exim fails to connect to the socket for saslauthd authentication.
5512 This could cause Exim to crash, or give a corrupted message.
5514 39. Added isip, isip4, and isip6 to ${if conditions.
5516 40. The ACL variables $acl_xx are now saved with the message, and can be
5517 accessed later in routers, transports, and filters.
5519 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5520 strings in the file are not string-expanded.
5522 42. If a MAIL command specified a SIZE value that was too large to fit into an
5523 int variable, the check against message_size_limit failed. Such values are
5524 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5525 day this will have to be increased, but I don't think I want to be around
5526 when emails are that large.
5533 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5534 iconv() is not standard in FreeBSD.
5536 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5537 IPv6 enabled. The observed symptom was a segmentation fault on return from
5538 the function os_common_find_running_interfaces() in src/os.c.
5540 3. In the check_special_case() function in daemon.c I had used "errno" as an
5541 argument name, which causes warnings on some systems. This was basically a
5542 typo, since it was named "eno" in the comments!
5544 4. The code that waits for the clock to tick (at a resolution of some fraction
5545 of a second) so as to ensure message-id uniqueness was always waiting for
5546 at least one whole tick, when it could have waited for less. [This is
5547 almost certainly not relevant at current processor speeds, where it is
5548 unlikely to ever wait at all. But we try to future-proof.]
5550 5. The function that sleeps for a time interval that includes fractions of a
5551 second contained a race. It did not block SIGALRM between setting the
5552 timer, and suspending (a couple of lines later). If the interval was short
5553 and the sigsuspend() was delayed until after it had expired, the suspension
5554 never ended. On busy systems this could lead to processes getting stuck for
5557 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5558 process, before it forks any delivery processes. The open lookup caching
5559 mechanism meant that the open file or database connection was passed into
5560 the delivery process. The problem was that delivery processes always tidy
5561 up cached lookup data. This could cause a problem for the next delivery
5562 process started by the queue runner, because the external queue runner
5563 process does not know about the closure. So the next delivery process
5564 still has data in the lookup cache. In the case of a file lookup, there was
5565 no problem because closing a file descriptor in a subprocess doesn't affect
5566 the parent. However, if the lookup was caching a connection to a database,
5567 the connection was closed, and the second delivery process was likely to
5568 see errors such as "PGSQL: query failed: server closed the connection
5569 unexpectedly". The problem has been fixed by closing all cached lookups
5570 in a queue runner before running a delivery process.
5572 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5573 seem to have the "const" qualifier which it has on other OS. I've
5576 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5577 *of the same type* (client or server) with the same public name that an
5578 error should be diagnosed.
5580 9. When Exim looked up a host name for an IP address, but failed to find the
5581 original IP address when looking up the host name (a safety check), it
5582 output the message "<ip address> does not match any IP for NULL", which was
5583 confusing, to say the least. The bug was that the host name should have
5584 appeared instead of "NULL".
5586 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5587 user that is built into the binary, and the -C or -D options is used, root
5588 privilege is dropped before the configuration file is read. In addition,
5589 logging is switched to stderr instead of the normal log files. If the
5590 configuration then re-defines the Exim user, the unprivileged environment
5591 is probably not what is expected, so Exim logs a panic warning message (but
5594 However, if deliver_drop_privilege is set, the unprivileged state may well
5595 be exactly what is intended, so the warning has been cut out in that case,
5596 and Exim is allowed to try to write to its normal log files.
5602 1. smtp_return_error_details was not giving details for temporary sender
5603 or receiver verification errors.
5605 2. Diagnose a configuration error if two authenticators have the same public
5608 3. Exim used not to create the message log file for a message until the first
5609 delivery attempt. This could be confusing when incoming messages were held
5610 for policy or load reasons. The message log file is now created at the time
5611 the message is received, and an initial "Received" line is written to it.
5613 4. The automatically generated man page for command line options had a minor
5614 bug that caused no ill effects; however, a more serious problem was that
5615 the procedure for building the man page automatically didn't always
5616 operate. Consequently, release 4.20 contains an out-of-date version. This
5617 shouldn't happen again.
5619 5. When building Exim with embedded Perl support, the script that builds the
5620 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5621 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5623 6. The freeze_tell option was not being used for messages that were frozen on
5624 arrival, either by an ACL or by local_scan().
5626 7. Added the smtp_incomplete_transaction log selector.
5628 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5629 was accepting AUTH without a new EHLO.
5631 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5632 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5635 10. Logging of TCP/IP connections (when configured) now happens in the main
5636 daemon process instead of the child process, so that the TCP/IP connection
5637 count is more accurate (but it can never be perfect).
5639 11. The use of "drop" in a nested ACL was not being handled correctly in the
5640 outer ACL. Now, if condition failure induced by the nested "drop" causes
5641 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5642 or "require"), the connection is dropped.
5644 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5645 that yield "discard" can now be used with an "accept" or a "discard" verb,
5646 but an error is generated for any others (because I can't see a useful way
5647 to define what should happen).
5649 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5650 are now processed in the same way as lines in the Exim configuration file.
5651 In particular, continuation lines are supported.
5653 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5655 15. Added -ti meaning -t -i.
5657 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5658 domains, and warn by logging if others are found.
5660 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5661 returned from GIFCONF when Exim is trying to find the list of interfaces on
5662 a host. The code in os.c has been modified to copy each ifreq to an aligned
5663 structure in all cases.
5665 Also, in some cases, the returned ifreq's were being copied to a 'struct
5666 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5667 means the last couple of bytes of an IPv6 address could be chopped if the
5668 ifreq contained only a normal sockaddr (14 bytes storage).
5670 18. Named domain lists were not supported in the hosts_treat_as_local option.
5671 An entry such as +xxxx was not recognized, and was treated as a literal
5674 19. Ensure that header lines added by a DATA ACL are included in the reject log
5675 if the ACL subsequently rejects the message.
5677 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5678 MD5 (which is deprecated).
5680 21. When testing a filter file using -bf, Exim was writing a message when it
5681 took the sender from a "From " line in the message, but it was not doing so
5682 when it took $return_path from a Return-Path: header line. It now does.
5684 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5685 with a valid header line field name (a series of printing characters
5686 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5688 23. Changed "disc" in the source to "disk" to conform to the documentation and
5689 the book and for uniformity.
5691 24. Ignore Sendmail's -Ooption=value command line item.
5693 25. When execve() failed while trying to run a command in a pipe transport,
5694 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5695 could be confused with a return value of 69 from the command itself. This
5696 has been changed to 127, the value the shell returns if it is asked to run
5697 a non-existent command. The wording for the related log line suggests a
5698 non-existent command as the problem.
5700 26. If received_header_text expands to an empty string, do not add a Received:
5701 header line to the message. (Well, it adds a token one on the spool, but
5702 marks it "old" so that it doesn't get used or transmitted.)
5704 27. Installed eximstats 1.28 (addition of -nt option).
5706 28. There was no check for failure on the call to getsockname() in the daemon
5707 code. This can fail if there is a shortage of resources on the system, with
5708 ENOMEM, for example. A temporary error is now given on failure.
5710 29. Contrary to the C standard, it seems that in some environments, the
5711 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5712 program. Exim now does this explicitly; it affects the formatting of
5713 timestamps using strftime().
5715 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5716 complaints. I've now initialized all the variables, to avoid this.
5718 32. Header lines added by a system filter were not being "seen" during
5719 transport-time rewrites.
5721 33. The info_callback() function passed to OpenSSL is set up with type void
5722 (*)(SSL *, int, int), as described somewhere. However, when calling the
5723 function (actually a macro) that sets it up, the type void(*)() is
5724 expected. I've put in a cast to prevent warnings from picky compilers.
5726 34. If a DNS black list lookup found a CNAME record, but there were no A
5727 records associated with the domain it pointed at, Exim crashed.
5729 35. If a DNS black list lookup returned more than one A record, Exim ignored
5730 all but the first. It now scans all returned addresses if a particular IP
5731 value is being sought. In this situation, the contents of the
5732 $dnslist_value variable are a list of all the addresses, separated by a
5735 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5736 to accept a host name and all its aliases if the forward lookup for any of
5737 them yielded the IP address of the incoming connection. Now it accepts only
5738 those names whose forward lookup yields the correct IP address. Any other
5739 names are discarded. This closes a loophole whereby a rogue DNS
5740 administrator could create reverse DNS records to break through a
5741 wildcarded host restriction in an ACL.
5743 37. If a user filter or a system filter that ran in a subprocess used any of
5744 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5745 the wrong values were passed to the pipe command ($thisaddress had the
5746 value of $0, $0 had the value of $1, etc). This bug was introduced by
5747 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5749 38. Improved the line breaking for long SMTP error messages from ACLs.
5750 Previously, if there was no break point between 40 and 75 characters, Exim
5751 left the rest of the message alone. Two changes have been made: (a) I've
5752 reduced the minimum length to 35 characters; (b) if it can't find a break
5753 point between 35 and 75 characters, it looks ahead and uses the first one
5754 that it finds. This may give the occasional overlong line, but at least the
5755 remaining text gets split now.
5757 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5758 file descriptors might be low, and that setting 1000 would always raise it.
5759 It turns out that in some environments, the limit is already over 1000 and
5760 that lowering it causes trouble. So now Exim takes care not to decrease it.
5762 40. When delivering a message, the value of $return_path is set to $sender_
5763 address at the start of routing (routers may change the value). By an
5764 oversight, this default was not being set up when an address was tested by
5765 -bt or -bv, which affected the outcome if any router or filter referred to
5768 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5769 log only when "message" or "log_message" are set. However, if one of the
5770 conditions was an address verification, or a call to a nested ACL, the
5771 messages generated by the underlying test were being passed through. This
5772 no longer happens. The underlying message is available in $acl_verify_
5773 message for both "message" and "log_message" expansions, so it can be
5774 passed through if needed.
5776 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5777 new expansion $bh_ to give the encoded byte string without charset
5778 translation. Translation happens only if iconv() is available; HAVE_ICONV
5779 indicates this at build time. HEADERS_CHARSET gives the charset to
5780 translate to; headers_charset can change it in the configuration, and
5781 "headers charset" can change it in an individual filter file.
5783 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5784 that creates RFC 2047 encoded "words" labels them as that charset instead
5785 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5786 expansion operator; (ii) when Exim creates a From: line for a local
5787 message; (iii) when a header line is rewritten to include a "phrase" part.
5789 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5790 buggy, causing it to skip the first lines of messages whose message ID
5791 ended in 'D'. This would not have bitten before Exim release 4.14, because
5792 message IDs were unlikely to end in 'D' before then. The effect was to have
5793 incorrect size information for certain domains.
5795 45. #include "config.h" was missing at the start of the crypt16.c module. This
5796 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5799 46. If there was a timeout during a "random" callout check, Exim treated it as
5800 a failure of the random address, and carried on sending RSET and the real
5801 address. If the delay was just some slowness somewhere, the response to the
5802 original RCPT would be taken as a response to RSET and so on, causing
5803 mayhem of various kinds.
5805 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5806 when I implemented it. It didn't allow for the fact that some option values
5807 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5808 the right test for positive values.
5810 48. Domain names in DNS records are case-independent. Exim always looks them up
5811 in lower case. Some resolvers return domain names in exactly the case they
5812 appear in the zone file, that is, they may contain uppercase letters. Not
5813 all resolvers do this - some return always lower case. Exim was treating a
5814 change of case by a resolver as a change of domain, similar to a widening
5815 of a domain abbreviation. This triggered its re-routing code and so it was
5816 trying to route what was effectively the same domain again. This normally
5817 caused routing to fail (because the router wouldn't handle the domain
5818 twice). Now Exim checks for this case specially, and just changes the
5819 casing of the domain that it ultimately uses when it transmits the message
5822 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5825 50. If a filter generated a file delivery with a non-absolute name (possible if
5826 no home directory exists for the router), the forbid_file option was not
5829 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5830 the existing equality matching.
5832 52. Exim was using ints instead of ino_t variables in some places where it was
5833 dealing with inode numbers.
5835 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5836 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5837 called TMPDIR, and if it finds it is different, it changes its value.
5839 54. The smtp_printf() function is now made available to local_scan() so
5840 additional output lines can be written before returning. There is also an
5841 smtp_fflush() function to enable the detection of a dropped connection.
5842 The variables smtp_input and smtp_batched_input are exported to
5845 55. Changed the default runtime configuration: the message "Unknown user"
5846 has been removed from the ACL, and instead placed on the localuser router,
5847 using the cannot_route_message feature. This means that any verification
5848 failures that generate their own messages won't get overridden. Similarly,
5849 the "Unrouteable address" message that was in the ACL for unverifiable
5850 relay addresses has also been removed.
5852 56. Added hosts_avoid_esmtp to the smtp transport.
5854 57. The exicyclog script was not checking for the esoteric option
5855 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5856 will work only if exicyclog is run under the appropriate euid.
5858 58. Following a discussion on the list, the rules by which Exim recognises line
5859 endings on incoming messages have been changed. The -dropcr and drop_cr
5860 options are now no-ops, retained only for backwards compatibility. The
5861 following line terminators are recognized: LF CRLF CR. However, special
5862 processing applies to CR:
5864 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5865 nor a local message in the state where . is a terminator.
5867 (ii) If a bare CR is encountered in a header line, an extra space is added
5868 after the line terminator so as not to end the header. The reasoning
5869 behind this is that bare CRs in header lines are most likely either
5870 to be mistakes, or people trying to play silly games.
5872 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5873 was being incorrectly given as 18 bytes larger than it should have been.
5874 This is a VOB (very old bug).
5876 60. This may never have affected anything current, but just in case it has:
5877 When the local host is found other than at the start of a list of hosts,
5878 the local host, those with the same MX, and any that follow, are discarded.
5879 When the list in question was part of a longer list of hosts, the following
5880 hosts (not currently being processed) were also being discarded. This no
5881 longer happens. I'm not sure if this situation could ever has previously
5884 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5887 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5888 header. This is something that is recommended in a new Internet Draft, and
5889 is something that is documented as being done by Sendmail. There are two
5890 possible values. For messages generated by the autoreply transport, Exim
5893 Auto-Submitted: auto-replied
5895 whereas for all other generated messages (e.g. bounces) it adds
5897 Auto-Submitted: auto-generated
5899 63. The "personal" condition in filters now includes a test for the
5900 Auto-Submitted: header. If it contains the string "auto-" the message it
5901 not considered personal.
5903 64. Added rcpt_include_affixes as a generic transport option.
5905 65. Added queue_only_override (default true).
5907 66. Added the syslog_duplication option.
5909 67. If what should have been the first header line of a message consisted of
5910 a space followed by a colon, Exim was mis-interpreting it as a header line.
5911 It isn't of course - it is syntactically invalid and should therefore be
5912 treated as the start of the message body. The misbehaviour could have
5913 caused a number of strange effects, including loss of data in subsequent
5914 header lines, and spool format errors.
5916 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5917 client host had authenticated. This control can now be exercised by an ACL
5918 for more flexibility.
5920 69. By default, callouts do not happen when testing with -bh. There is now a
5921 variant, -bhc, which does actually run the callout code, including
5922 consulting and updating the callout cache.
5924 70. Added support for saslauthd authentication, courtesy of Alexander
5927 71. If statvfs() failed on the spool or log directories while checking their
5928 size for availability, Exim confusingly gave the error "space shortage".
5929 Furthermore, in debugging mode it crashed with a floating point exception.
5930 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5931 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5932 serious problem, Exim now writes to the main and panic logs when this
5933 happens, with details of the failure. It then refuses to accept the
5934 incoming message, giving the message "spool directory problem" or "log
5935 directory problem" with a 421 code for SMTP messages.
5937 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5938 0, 1, and 2 exist, because some OS complain for execs without them (see
5939 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5940 descriptors. However, the code omitted to check that the open succeeded,
5941 causing mysterious errors if for some reason the permissions on /dev/null
5942 got screwed. Now Exim writes a message to the main and panic logs, and
5943 bombs out if it can't open /dev/null.
5945 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5946 interact so that it is all more flexible. It is supposed to remain
5947 backwards compatible. Also added extra_local_interfaces.
5949 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5950 to bomb out with an assertion failure - to the client this appears as a
5951 connection drop. This problem occurs in the part of the code that was taken
5952 from the Samba project. Fortunately, the assertion is in a very simple
5953 function, so I have fixed this by reproducing the function inline in the
5954 one place where it is called, and arranging for authentication to fail
5955 instead of killing the process with assert().
5957 75. The SPA client code was not working when the server requested OEM rather
5958 than Unicode encoding.
5960 76. Added code to make require_files with a specific uid setting more usable in
5961 the case where statting the file as root fails - usually a non-root-mounted
5962 NFS file system. When this happens and the failure is EACCES, Exim now
5963 forks a subprocess and does the per-uid checking as the relevant uid.
5965 77. Added process_log_path.
5967 78. If log_file_path was not explicitly set, a setting of check_log_space or
5968 check_log_inodes was ignored.
5970 79. If a space check for the spool or log partitions fails, the incident is now
5971 logged. Of course, in the latter case the data may get lost...
5973 80. Added the %p formatting code to string_format() so that it can be used to
5974 print addresses in debug_print(). Adjusted all the address printing in the
5975 debugging in store.c to use %p rather than %d.
5977 81. There was a concern that a line of code in smtp_in.c could overflow a
5978 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5979 initially expressed, the concern was not well-founded, because trailing
5980 spaces are removed early. However, if the trailing spaces were followed by
5981 a NULL, they did not get removed, so the overflow was possible. Two fixes
5984 (a) I re-wrote the offending code in a cleaner fashion.
5985 (b) If an incoming SMTP command contains a NULL character, it is rejected
5988 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5989 runs initgroups(), so that if the Exim user is in any additional groups,
5990 they will be used during message reception.
5996 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob