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. OpenSSL only.
65 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
67 PP/02 Make -n do something, by making it not do something.
68 When combined with -bP, the name of an option is not output.
70 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
73 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
74 $sender_host_name and config options to manage this, and basic check
77 PP/05 DSCP support for outbound connections and control modifier for inbound.
79 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
80 (Only plugin which currently uses this is kerberos4, which nobody should
81 be using, but we should make it available and other future plugins might
82 conceivably use it, even though it would break NAT; stuff *should* be
83 using channel bindings instead).
85 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
86 name; added for Sendmail compatibility; requires admin caller.
87 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
88 just ignore it); requires trusted caller.
89 Also parse but ignore: -Ac -Am -X<logfile>
92 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
94 TL/02 Add +smtp_confirmation as a default logging option.
96 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
97 Patch by Magnus Holmgren from 2007-02-20.
99 TL/04 Bugzilla 1281 - Spec typo.
100 Bugzilla 1283 - Spec typo.
101 Bugzilla 1290 - Spec grammar fixes.
103 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
105 TL/06 Add Experimental DMARC support using libopendmarc libraries.
107 TL/07 Fix an out of order global option causing a segfault. Reported to dev
108 mailing list by by Dmitry Isaikin.
110 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
112 JH/02 Support "G" suffix to numbers in ${if comparisons.
114 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
116 NM/01 Bugzilla 1197 - Spec typo
117 Bugzilla 1196 - Spec examples corrections
119 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
121 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
122 gnutls_enable_pkcs11, but renamed to more accurately indicate its
125 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
126 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
128 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
129 "acl {{name}{arg}...}", and optional args on acl condition
132 JH/05 Permit multiple router/transport headers_add/remove lines.
134 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
136 JH/07 Avoid using a waiting database for a single-message-only transport.
137 Performance patch from Paul Fisher. Bugzilla 1262.
139 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
142 JH/09 Add $headers_added variable, with content from use of ACL modifier
143 add_header (but not yet added to the message). Bugzilla 199.
145 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
146 Pulled from Bugzilla 817 by Wolfgang Breyha.
148 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
150 (nb: this is the same fix as in Exim 4.80.1)
152 JH/11 Add A= logging on delivery lines, and a client_set_id option on
155 JH/12 Add optional authenticated_sender logging to A= and a log_selector
158 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
160 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
161 advertise SMTP AUTH mechanism to us, instead of a generic
162 protocol violation error. Also, make Exim more robust to bad
163 data from the Dovecot auth socket.
165 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
167 When a queue runner is handling a message, Exim first routes the
168 recipient addresses, during which it prunes them based on the retry
169 hints database. After that it attempts to deliver the message to
170 any remaining recipients. It then updates the hints database using
173 So if a recipient address works intermittently, it can get repeatedly
174 deferred at routing time. The retry hints record remains fresh so the
175 address never reaches the final cutoff time.
177 This is a fairly common occurrence when a user is bumping up against
178 their storage quota. Exim had some logic in its local delivery code
179 to deal with this. However it did not apply to per-recipient defers
180 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
182 This change adds a proper retry rule check during routing so that the
183 final cutoff time is checked against the message's age. We only do
184 this check if there is an address retry record and there is not a
185 domain retry record; this implies that previous attempts to handle
186 the address had the retry_use_local_parts option turned on. We use
187 this as an approximation for the destination being like a local
188 delivery, as in LMTP.
190 I suspect this new check makes the old local delivery cutoff check
191 redundant, but I have not verified this so I left the code in place.
193 TF/02 Correct gecos expansion when From: is a prefix of the username.
195 Test 0254 submits a message to Exim with the header
199 When I ran the test suite under the user fanf2, Exim expanded
200 the header to contain my full name, whereas it should have added
201 a Resent-Sender: header. It erroneously treats any prefix of the
202 username as equal to the username.
204 This change corrects that bug.
206 GF/01 DCC debug and logging tidyup
207 Error conditions log to paniclog rather than rejectlog.
208 Debug lines prefixed by "DCC: " to remove any ambiguity.
210 TF/03 Avoid unnecessary rebuilds of lookup-related code.
212 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
213 Bug spotted by Jeremy Harris; was flawed since initial commit.
214 Would have resulted in OCSP responses post-SNI triggering an Exim
215 NULL dereference and crash.
217 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
219 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
220 Bug detection, analysis and fix by Samuel Thibault.
221 Bugzilla 1331, Debian bug #698092.
223 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
225 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
226 Server implementation by Todd Lyons, client by JH.
227 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
228 config variable "prdr_enable" controls whether the server
229 advertises the facility. If the client requests PRDR a new
230 acl_data_smtp_prdr ACL is called once for each recipient, after
231 the body content is received and before the acl_smtp_data ACL.
232 The client is controlled by bolth of: a hosts_try_prdr option
233 on the smtp transport, and the server advertisement.
234 Default client logging of deliveries and rejections involving
235 PRDR are flagged with the string "PRDR".
237 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
238 fclose(). Diagnosis by Todd Lyons.
240 PP/17 Update configure.default to handle IPv6 localhost better.
241 Patch by Alain Williams (plus minor tweaks).
244 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
245 This is now consistent with GnuTLS, and is now documented: the
246 previous undocumented portable approach to treating the option as
247 unset was to force an expansion failure. That still works, and
248 an empty string is now equivalent.
250 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
251 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
252 not performing validation itself.
254 PP/20 Added force_command boolean option to pipe transport.
255 Patch from Nick Koston, of cPanel Inc.
257 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
260 TF/04 Added udpsend ACL modifer and hexquote expansion operator
262 PP/21 Fix eximon continuous updating with timestamped log-files.
263 Broken in a format-string cleanup in 4.80, missed when I repaired the
264 other false fix of the same issue.
265 Report and fix from Heiko Schlichting.
268 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
269 Report from Prashanth Katuri.
271 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
272 It's SecureTransport, so affects any MacOS clients which use the
273 system-integrated TLS libraries, including email clients.
275 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
276 using a MIME ACL for non-SMTP local injection.
277 Report and assistance in diagnosis by Warren Baker.
279 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
281 JH/16 Fix comparisons for 64b. Bugzilla 1385.
283 TL/09 Add expansion variable $authenticated_fail_id to keep track of
284 last id that failed so it may be referenced in subsequent ACL's.
286 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
289 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
290 ldap library initialization, allowing self-signed CA's to be
291 used. Also properly sets require_cert option later in code by
292 using NULL (global ldap config) instead of ldap handle (per
293 session). Bug diagnosis and testing by alxgomz.
295 TL/12 Enhanced documentation in the ratelimit.pl script provided in
296 the src/util/ subdirectory.
298 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
299 renamed to Transport Post Delivery Action by Jeremy Harris, as
302 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
303 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
304 redis_servers = needs to be configured which will be used by the redis
305 lookup. Patch from Warren Baker, of The Packet Hub.
307 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
309 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
310 hostname or reverse DNS when processing a host list. Used suggestions
311 from multiple comments on this bug.
313 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
315 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
316 Missed a few lines, added it to make the runtest require no keyboard
319 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
320 contains upper case chars. Make router use caseful_local_part.
322 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
323 support when GnuTLS has been built with p11-kit.
329 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
331 This, or similar/improved, will also be change PP/11 of 4.82.
337 PP/01 Handle short writes when writing local log-files.
338 In practice, only affects FreeBSD (8 onwards).
339 Bugzilla 1053, with thanks to Dmitry Isaikin.
341 NM/01 Bugzilla 949 - Documentation tweak
343 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
346 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
348 PP/02 Implemented gsasl authenticator.
350 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
352 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
353 `pkg-config foo` for cflags/libs.
355 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
356 with rest of GSASL and with heimdal_gssapi.
358 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
359 `pkg-config foo` for cflags/libs for the TLS implementation.
361 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
362 properties get this fed in as external SSF. A number of robustness
363 and debugging improvements to the cyrus_sasl authenticator.
365 PP/08 cyrus_sasl server now expands the server_realm option.
367 PP/09 Bugzilla 1214 - Log authentication information in reject log.
368 Patch by Jeremy Harris.
370 PP/10 Added dbmjz lookup type.
372 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
374 PP/12 MAIL args handles TAB as well as SP, for better interop with
375 non-compliant senders.
376 Analysis and variant patch by Todd Lyons.
378 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
379 Bug report from Lars Müller <lars@samba.org> (via SUSE),
380 Patch from Dirk Mueller <dmueller@suse.com>
382 PP/13 tls_peerdn now print-escaped for spool files.
383 Observed some $tls_peerdn in wild which contained \n, which resulted
384 in spool file corruption.
386 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
387 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
388 or write after TLS renegotiation, which otherwise led to messages
391 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
392 as a tracking header (ie: a signed header comes before the signature).
393 Patch from Wolfgang Breyha.
395 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
396 comma-sep list; embedded commas doubled.
398 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
400 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
402 Report and patch from Dmitry Banschikov.
404 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
405 Removed SSL_clear() after SSL_new() which led to protocol negotiation
406 failures. We appear to now support TLS1.1+ with Exim.
408 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
409 lets Exim select keys and certificates based upon TLS SNI from client.
410 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
411 before an outbound SMTP session. New log_selector, +tls_sni.
413 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
414 NULL dereference. Report and patch from Alun Jones.
416 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
417 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
418 Not seeing resolver debug output on NetBSD, but suspect this is a
419 resolver implementation change.
421 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
422 Left warnings. Added "eximon gdb" invocation mode.
424 PP/21 Defaulting "accept_8bitmime" to true, not false.
426 PP/22 Added -bw for inetd wait mode support.
428 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
429 locate the relevant includes and libraries. Made this the default.
431 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
432 Bugzilla 1246, report and most of solution from Tomasz Kusy.
434 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
435 This may cause build issues on older platforms.
437 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
438 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
439 gnutls_require_mac & gnutls_require_protocols (no longer supported).
440 Added SNI support via GnuTLS too.
441 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
443 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
445 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
446 Applied second patch from Janne, implementing suggestion to default
447 multiple-strings-in-record handling to match SPF spec.
449 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
451 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
452 read-only, out of scope).
453 Patch from Wolfgang Breyha, report from Stuart Northfield.
455 PP/29 Fix three issues highlighted by clang analyser static analysis.
456 Only crash-plausible issue would require the Cambridge-specific
457 iplookup router and a misconfiguration.
458 Report from Marcin Mirosław.
460 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
462 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
463 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
464 As part of this, removing so much warning spew let me fix some minor
465 real issues in debug logging.
467 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
468 assignment on my part. Fixed.
470 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
471 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
472 Janne Snabb (who went above and beyond: thank you).
474 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
475 string otherwise requires a connection and a bunch more work and it's
476 relatively easy to get wrong. Should also expose TLS library linkage
479 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
480 64-bit ${eval} (JH/03).
482 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
483 GNU libc to support some of the 64-bit stuff, should not lead to
484 conflicts. Defined before os.h is pulled in, so if a given platform
485 needs to override this, it can.
487 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
488 protection layer was required, which is not implemented.
489 Bugzilla 1254, patch from Wolfgang Breyha.
491 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
492 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
493 tls_dhparam take prime identifiers. Also unbreak combination of
494 OpenSSL+DH_params+TLSSNI.
496 PP/39 Disable SSLv2 by default in OpenSSL support.
502 PP/01 Solaris build fix for Oracle's LDAP libraries.
503 Bugzilla 1109, patch from Stephen Usher.
505 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
507 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
510 TF/02 Fix a couple more cases where we did not log the error message
511 when unlink() failed. See also change 4.74-TF/03.
513 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
514 lock up or crash if it happened to be inside a call to libc when it
515 got a SIGUSR1 from exiwhat.
517 The SIGUSR1 handler appends the current process status to the process
518 log which is later printed by exiwhat. It used to use the general
519 purpose logging code to do this, but several functions it calls are
520 not safe for signals.
522 The new output code in the SIGUSR1 handler is specific to the process
523 log, and simple enough that it's easy to inspect for signal safety.
524 Removing some special cases also simplifies the general logging code.
525 Removing the spurious timestamps from the process log simplifies
528 TF/04 Improved ratelimit ACL condition.
530 The /noupdate option has been deprecated in favour of /readonly which
531 has clearer semantics. The /leaky, /strict, and /readonly update modes
532 are mutually exclusive. The update mode is no longer included in the
533 database key; it just determines when the database is updated. (This
534 means that when you upgrde Exim will forget old rate measurements.)
536 Exim now checks that the per_* options are used with an update mode that
537 makes sense for the current ACL. For example, when Exim is processing a
538 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
539 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
540 must specify per_mail/readonly. If you omit the update mode it defaults to
541 /leaky where that makes sense (as before) or /readonly where required.
543 The /noupdate option is now undocumented but still supported for
544 backwards compatibility. It is equivalent to /readonly except that in
545 ACLs where /readonly is required you may specify /leaky/noupdate or
546 /strict/noupdate which are treated the same as /readonly.
548 A useful new feature is the /count= option. This is a generalization
549 of the per_byte option, so that you can measure the throughput of other
550 aggregate values. For example, the per_byte option is now equivalent
551 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
553 The per_rcpt option has been generalized using the /count= mechanism
554 (though it's more complicated than the per_byte equivalence). When it is
555 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
556 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
557 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
558 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
559 Note that using per_rcpt with a non-readonly update mode in more than
560 one ACL will cause the recipients to be double-counted. (The per_mail
561 and per_byte options don't have this problem.)
563 The handling of very low rates has changed slightly. If the computed rate
564 is less than the event's count (usually one) then this event is the first
565 after a long gap. In this case the rate is set to the same as this event's
566 count, so that the first message of a spam run is counted properly.
568 The major new feature is a mechanism for counting the rate of unique
569 events. The new per_addr option counts the number of different
570 recipients that someone has sent messages to in the last time period. It
571 behaves like per_rcpt if all the recipient addresses are different, but
572 duplicate recipient addresses do not increase the measured rate. Like
573 the /count= option this is a general mechanism, so the per_addr option
574 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
575 example, measure the rate that a client uses different sender addresses
576 with the options per_mail/unique=$sender_address. There are further
577 details in the main documentation.
579 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
581 TF/06 Removed a few PCRE remnants.
583 TF/07 Automatically extract Exim's version number from tags in the git
584 repository when doing development or release builds.
586 PP/02 Raise smtp_cmd_buffer_size to 16kB.
587 Bugzilla 879. Patch from Paul Fisher.
589 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
590 Heavily based on revision 40f9a89a from Simon Arlott's tree.
593 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
595 PP/05 Variable $av_failed, true if the AV scanner deferred.
596 Bugzilla 1078. Patch from John Horne.
598 PP/06 Stop make process more reliably on build failure.
599 Bugzilla 1087. Patch from Heiko Schlittermann.
601 PP/07 Make maildir_use_size_file an _expandable_ boolean.
602 Bugzilla 1089. Patch from Heiko Schlittermann.
604 PP/08 Handle ${run} returning more data than OS pipe buffer size.
605 Bugzilla 1131. Patch from Holger Weiß.
607 PP/09 Handle IPv6 addresses with SPF.
608 Bugzilla 860. Patch from Wolfgang Breyha.
610 PP/10 GnuTLS: support TLS 1.2 & 1.1.
612 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
615 PP/11 match_* no longer expand right-hand-side by default.
616 New compile-time build option, EXPAND_LISTMATCH_RHS.
617 New expansion conditions, "inlist", "inlisti".
619 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
621 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
623 PP/14 fix log_write() format string regression from TF/03.
624 Bugzilla 1152. Patch from Dmitry Isaikin.
630 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
632 PP/02 Harmonised TLS library version reporting; only show if debugging.
633 Layout now matches that introduced for other libraries in 4.74 PP/03.
635 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
637 PP/04 New "dns_use_edns0" global option.
639 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
642 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
643 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
645 TK/01 Updated PolarSSL code to 0.14.2.
646 Bugzilla 1097. Patch from Andreas Metzler.
648 PP/07 Catch divide-by-zero in ${eval:...}.
651 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
654 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
655 format-string attack -- SECURITY: remote arbitrary code execution.
657 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
658 time unintentionally subject to list matching rules, letting the header
659 cause arbitrary Exim lookups (of items which can occur in lists, *not*
660 arbitrary string expansion). This allowed for information disclosure.
662 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
663 INT_MIN/-1 -- value coerced to INT_MAX.
669 NM/01 Workround for PCRE version dependency in version reporting
672 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
673 This fixes portability to compilers other than gcc, notably
674 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
676 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
677 makefiles for portability to HP-UX and POSIX correctness.
679 PP/01 Permit LOOKUP_foo enabling on the make command-line.
680 Also via indented variable definition in the Makefile.
681 (Debugging by Oliver Heesakkers).
683 PP/02 Restore caching of spamd results with expanded spamd_address.
684 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
686 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
687 Improves build reliability. Fix from: Frank Elsner
689 NM/02 Fix wide character breakage in the rfc2047 coding
690 Fixes bug 1064. Patch from Andrey N. Oktyabrski
692 NM/03 Allow underscore in dnslist lookups
693 Fixes bug 1026. Patch from Graeme Fowler
695 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
696 Code patches from Adam Ciarcinski of NetBSD.
698 NM/04 Fixed exiqgrep to cope with mailq missing size issue
701 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
702 is logged, to avoid truncation. Patch from John Horne.
704 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
705 Patch from Jakob Hirsch.
707 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
708 SQL string expansion failure details.
709 Patch from Andrey Oktyabrski.
711 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
712 Patch from Simon Arlott.
714 PP/09 New lookups functionality failed to compile on old gcc which rejects
715 extern declarations in function scope.
716 Patch from Oliver Fleischmann
718 PP/10 Use sig_atomic_t for flags set from signal handlers.
719 Check getgroups() return and improve debugging.
720 Fixed developed for diagnosis in bug 927 (which turned out to be
723 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
724 Patch from Mark Zealey.
726 PP/12 Bugzilla 1056: Improved spamd server selection.
727 Patch from Mark Zealey.
729 PP/13 Bugzilla 1086: Deal with maildir quota file races.
730 Based on patch from Heiko Schlittermann.
732 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
733 Patch from Uwe Doering, sign-off by Michael Haardt.
735 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
736 variable declaration deep within a block. Bug and patch from
739 PP/15 lookups-Makefile IRIX compatibilty coercion.
741 PP/16 Make DISABLE_DKIM build knob functional.
743 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
744 Patch by Simon Arlott
746 TF/03 Fix valgrind.h portability to C89 compilers that do not support
747 variable argument macros. Our copy now differs from upstream.
753 TF/01 Failure to get a lock on a hints database can have serious
754 consequences so log it to the panic log.
756 TF/02 Log LMTP confirmation messages in the same way as SMTP,
757 controlled using the smtp_confirmation log selector.
759 TF/03 Include the error message when we fail to unlink a spool file.
761 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
762 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
763 for maintaining out-of-tree patches for some time.
765 PP/01 Bugzilla 139: Documentation and portability issues.
766 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
767 Handle per-OS dynamic-module compilation flags.
769 PP/02 Let /dev/null have normal permissions.
770 The 4.73 fixes were a little too stringent and complained about the
771 permissions on /dev/null. Exempt it from some checks.
772 Reported by Andreas M. Kirchwitz.
774 PP/03 Report version information for many libraries, including
775 Exim version information for dynamically loaded libraries. Created
776 version.h, now support a version extension string for distributors
777 who patch heavily. Dynamic module ABI change.
779 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
780 privilege escalation vulnerability whereby the Exim run-time user
781 can cause root to append content of the attacker's choosing to
784 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
787 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
788 If dropping privileges for untrusted macros, we disabled normal logging
789 on the basis that it would fail; for the Exim run-time user, this is not
790 the case, and it resulted in successful deliveries going unlogged.
791 Fixed. Reported by Andreas Metzler.
797 PP/01 Date: & Message-Id: revert to normally being appended to a message,
798 only prepend for the Resent-* case. Fixes regression introduced in
799 Exim 4.70 by NM/22 for Bugzilla 607.
801 PP/02 Include check_rfc2047_length in configure.default because we're seeing
802 increasing numbers of administrators be bitten by this.
804 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
806 PP/03 Bugzilla 994: added openssl_options main configuration option.
808 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
810 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
812 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
814 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
815 without a peer certificate, leading to a segfault because of an
816 assumption that peers always have certificates. Be a little more
817 paranoid. Problem reported by Martin Tscholak.
819 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
820 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
821 NB: ClamAV planning to remove STREAM in "middle of 2010".
822 CL also introduces -bmalware, various -d+acl logging additions and
823 more caution in buffer sizes.
825 PP/09 Implemented reverse_ip expansion operator.
827 PP/10 Bugzilla 937: provide a "debug" ACL control.
829 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
831 PP/12 Bugzilla 973: Implement --version.
833 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
835 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
837 PP/15 Bugzilla 816: support multiple condition rules on Routers.
839 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
840 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
841 ignore trailing whitespace.
843 JJ/02 prevent non-panic DKIM error from being sent to paniclog
845 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
848 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
849 Notification from Dr Andrew Aitchison.
851 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
852 ExtendedDetectionInfo response format.
853 Notification from John Horne.
855 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
858 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
859 XSL and documented dependency on system catalogs, with examples of how
862 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
865 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
866 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
867 configuration file which is writeable by the Exim user or group.
869 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
870 of configuration files to cover files specified with the -C option if
871 they are going to be used with root privileges, not just the default
874 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
875 option (effectively making it always true).
877 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
878 files to be used while preserving root privileges.
880 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
881 that rogue child processes cannot use them.
883 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
884 run-time user, instead of root.
886 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
887 Exim run-time user without dropping privileges.
889 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
890 result string, instead of calling string_vformat() twice with the same
893 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
894 for other users. Others should always drop root privileges if they use
895 -C on the command line, even for a whitelisted configure file.
897 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
899 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
905 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
906 $data_path, and $header_path variables; fixed documentation bugs and
909 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
910 exipick to access non-standard spools, including the "frozen" queue
913 NM/01 Bugzilla 965: Support mysql stored procedures.
914 Patch from Alain Williams
916 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
918 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
919 Patch from Andreas Metzler
921 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
922 Patch from Kirill Miazine
924 NM/05 Bugzilla 671: Added umask to procmail example.
926 JJ/03 installed exipick 20100323.0, fixing doc bug
928 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
929 directory. Notification and patch from Dan Rosenberg.
931 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
933 TK/02 Improve log output when DKIM signing operation fails.
935 MH/01 Treat the transport option dkim_domain as a colon separated
936 list, not as a single string, and sign the message with each element,
937 omitting multiple occurences of the same signer.
939 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
940 Bugzilla 985, 986. Patch by Simon Arlott
942 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
943 Patch by Simon Arlott
945 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
946 MBX locking. Notification from Dan Rosenberg.
952 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
954 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
956 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
958 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
960 NM/04 Bugzilla 915: Flag broken perl installation during build.
966 TK/01 Added patch by Johannes Berg that expands the main option
967 "spamd_address" if it starts with a dollar sign.
969 TK/02 Write list of recipients to X-Envelope-Sender header when building
970 the mbox-format spool file for content scanning (suggested by Jakob
973 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
974 (http://www.dcc-servers.net/) support via dccifd. Activated by
975 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
977 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
978 by Mark Daniel Reidel <mr@df.eu>.
980 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
981 When building exim an external PCRE library is now needed -
982 PCRE is a system library on the majority of modern systems.
983 See entry on PCRE_LIBS in EDITME file.
985 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
986 conversation. Added nologin parameter to request.
987 Patch contributed by Kirill Miazine.
989 TF/01 Do not log submission mode rewrites if they do not change the address.
991 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
993 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
994 log files in place. Contributed by Roberto Lima.
996 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
998 TF/03 Bugzilla 615: When checking the local_parts router precondition
999 after a local_part_suffix or local_part_prefix option, Exim now
1000 does not use the address's named list lookup cache, since this
1001 contains cached lookups for the whole local part.
1003 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1004 Robert Millan. Documentation is in experimental-spec.txt.
1006 TF/04 Bugzilla 668: Fix parallel build (make -j).
1008 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1010 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1011 Patch provided by Jan Srzednicki.
1013 TF/05 Leading white space used to be stripped from $spam_report which
1014 wrecked the formatting. Now it is preserved.
1016 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1017 that they are available at delivery time.
1019 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1021 TF/08 TLS error reporting now respects the incoming_interface and
1022 incoming_port log selectors.
1024 TF/09 Produce a more useful error message if an SMTP transport's hosts
1025 setting expands to an empty string.
1027 NM/06 Bugzilla 744: EXPN did not work under TLS.
1028 Patch provided by Phil Pennock.
1030 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1031 Patch provided by Richard Godbee.
1033 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1034 acl_smtp_notquit, added index entry.
1036 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1037 Patch provided by Eugene Bujak.
1039 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1040 accept(). Patch provided by Maxim Dounin.
1042 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1043 Patch provided by Phil Pennock.
1045 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1047 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1048 Patch provided by Brad "anomie" Jorsch.
1050 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1051 Patch provided by Dean Brooks.
1053 TK/05 Add native DKIM support (does not depend on external libraries).
1055 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1056 Patch provided by Graeme Fowler.
1058 NM/16 Bugzilla 851: Documentation example syntax fix.
1060 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1062 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1065 NM/19 Bugzilla 745: TLS version reporting.
1066 Patch provided by Phil Pennock.
1068 NM/20 Bugzilla 167: bool: condition support.
1069 Patch provided by Phil Pennock.
1071 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1072 clients. Patch provided by Phil Pennock.
1074 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1075 Patch provided by Brad "anomie" Jorsch.
1077 NM/23 Bugzilla 687: Fix misparses in eximstats.
1078 Patch provided by Heiko Schlittermann.
1080 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1081 Patch provided by Heiko Schlittermann.
1083 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1084 plus update to original patch.
1086 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1088 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1089 Patch provided by David Brownlee.
1091 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1093 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1095 NM/30 Bugzilla 888: TLS documentation bugfixes.
1097 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1099 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1100 Unlike the original bugzilla I have changed all shell scripts in src tree.
1102 NM/33 Bugzilla 898: Transport filter timeout fix.
1103 Patch by Todd Rinaldo.
1105 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1106 Patch by Serge Demonchaux.
1108 NM/35 Bugzilla 39: Base64 decode bug fixes.
1109 Patch by Jakob Hirsch.
1111 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1113 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1115 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1117 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1123 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1124 ALT-N's libdkim that I have put here:
1125 http://duncanthrax.net/exim-experimental/
1127 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1128 were called 'true' and it seems that C99 defines that as a reserved
1129 keyword to be used with 'bool' variable types. That means you could
1130 not include C99-style headers which use bools without triggering
1131 build errors in sieve.c.
1133 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1134 as mailq or other aliases. Changed the --help handling significantly
1135 to do whats expected. exim_usage() emits usage/help information.
1137 SC/01 Added the -bylocaldomain option to eximstats.
1139 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1141 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1143 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1149 PH/01 Another patch from the Sieve maintainer.
1151 PH/02 When an IPv6 address is converted to a string for single-key lookup
1152 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1153 dots are used instead of colons so that keys in lsearch files need not
1154 contain colons. This was done some time before quoting was made available
1155 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1156 (notated using the quote facility) so as to distinguish them from IPv4
1157 keys. This meant that lookups for IP addresses in host lists did not work
1158 for iplsearch lookups.
1160 This has been fixed by arranging for IPv6 addresses to be expressed with
1161 colons if the lookup type is iplsearch. This is not incompatible, because
1162 previously such lookups could never work.
1164 The situation is now rather anomolous, since one *can* have colons in
1165 ordinary lsearch keys. However, making the change in all cases is
1166 incompatible and would probably break a number of configurations.
1168 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1171 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1172 conversion specification without a maximum field width, thereby enabling
1173 a rogue spamd server to cause a buffer overflow. While nobody in their
1174 right mind would setup Exim to query an untrusted spamd server, an
1175 attacker that gains access to a server running spamd could potentially
1176 exploit this vulnerability to run arbitrary code as the Exim user.
1178 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1179 $primary_hostname instead of what libspf2 thinks the hosts name is.
1181 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1182 a directory entry by the name of the lookup key. Previously, if a
1183 symlink pointed to a non-existing file or a file in a directory that
1184 Exim lacked permissions to read, a lookup for a key matching that
1185 symlink would fail. Now it is enough that a matching directory entry
1186 exists, symlink or not. (Bugzilla 503.)
1188 PH/03 The body_linecount and body_zerocount variables are now exported in the
1191 PH/04 Added the $dnslist_matched variable.
1193 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1194 This means they are set thereafter only if the connection becomes
1197 PH/06 Added the client_condition to authenticators so that some can be skipped
1198 by clients under certain conditions.
1200 PH/07 The error message for a badly-placed control=no_multiline_responses left
1201 "_responses" off the end of the name.
1203 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1205 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1206 (without spaces) instead of just copying the configuration text.
1208 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1210 PH/11 Added $max_received_linelength.
1212 PH/12 Added +ignore_defer and +include_defer to host lists.
1214 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1215 way in which PCRE > 7.0 is built.
1217 PH/14 Implemented queue_only_load_latch.
1219 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1220 MAIL command. The effect was to mangle the value on 64-bit systems.
1222 PH/16 Another patch from the Sieve maintainer.
1224 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1226 PH/18 If a system quota error occurred while trying to create the file for
1227 a maildir delivery, the message "Mailbox is full" was not appended to the
1228 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1229 only to a quota excession during the actual writing of the file.
1231 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1232 characters?) which causes problems in log lines. The DN values are now
1233 passed through string_printing() before being added to log lines.
1235 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1236 and InterBase are left for another time.)
1238 PH/21 Added message_body_newlines option.
1240 PH/22 Guard against possible overflow in moan_check_errorcopy().
1242 PH/23 POSIX allows open() to be a macro; guard against that.
1244 PH/24 If the recipient of an error message contained an @ in the local part
1245 (suitably quoted, of course), incorrect values were put in $domain and
1246 $local_part during the evaluation of errors_copy.
1252 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1253 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1256 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1257 issue a MAIL command.
1259 PH/02 In an ACL statement such as
1261 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1263 if a client was not listed at all, or was listed with a value other than
1264 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1265 the condition was not true (as it should be), so access was not denied.
1266 The bug was that the ! inversion was incorrectly passed on to the second
1267 item. This has been fixed.
1269 PH/03 Added additional dnslists conditions == and =& which are different from
1270 = and & when the dns lookup returns more than one IP address.
1272 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1273 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1275 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1276 FSYNC, which compiles an option called disable_fsync that allows for
1277 bypassing fsync(). The documentation is heavily laced with warnings.
1279 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1281 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1282 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1283 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1284 including adding "make clean"; (3) Added -fPIC when compiling the test
1285 dynamically loaded module, to get rid of a warning.
1287 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1288 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1289 = 0s. The bug is otherwise harmless.
1291 PH/07 There was a bug in the dovecot authenticator such that the value of
1292 $auth1 could be overwritten, and so not correctly preserved, after a
1293 successful authentication. This usually meant that the value preserved by
1294 the server_setid option was incorrect.
1296 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1298 PH/09 Installed PCRE release 7.0.
1300 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1301 run for batched SMTP input. It is now run at the start of every message
1302 in the batch. While fixing this I discovered that the process information
1303 (output by running exiwhat) was not always getting set for -bs and -bS
1304 input. This is fixed, and it now also says "batched" for BSMTP.
1306 PH/11 Added control=no_pipelining.
1308 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1309 patch, slightly modified), and move the expansion of helo_data till after
1310 the connection is made in the smtp transport (so it can use these
1313 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1315 PH/14 Added log_selector = +pid.
1317 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1319 PH/16 Add ${if forany and ${if forall.
1321 PH/17 Added dsn_from option to vary the From: line in DSNs.
1323 PH/18 Flush SMTP output before performing a callout, unless control =
1324 no_callout_flush is set.
1326 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1327 was true (the default) a successful delivery failed to delete the retry
1328 item, thus causing premature timeout of the address. The bug is now
1331 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1333 PH/21 Long custom messages for fakedefer and fakereject are now split up
1334 into multiline reponses in the same way that messages for "deny" and
1335 other ACL rejections are.
1337 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1338 with slight modification.
1340 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1341 draft, changing the syntax and factoring some duplicate code".
1343 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1344 for deliveries of the second and subsequent messages over the same SMTP
1347 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1348 ${reduce, with only minor "tidies".
1350 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1352 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1353 expansion side effects.
1355 PH/27 When a message times out after an over-quota error from an Exim-imposed
1356 quota, the bounce message says "mailbox is full". This message was not
1357 being given when it was a system quota that was exceeded. It now should
1360 MH/03 Made $recipients available in local_scan(). local_scan() already has
1361 better access to the recipient list through recipients_list[], but
1362 $recipients can be useful in postmaster-provided expansion strings.
1364 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1365 in the case of a MAIL command with additional options following the
1366 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1367 were accidentally chopped off.
1369 PH/29 SMTP synchronization checks are implemented when a command is read -
1370 there is a check that no more input is waiting when there shouldn't be
1371 any. However, for some commands, a delay in an ACL can mean that it is
1372 some time before the response is written. In this time, more input might
1373 arrive, invalidly. So now there are extra checks after an ACL has run for
1374 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1375 pipelining has not been advertised.
1377 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1379 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1380 returned in the SMTP response when smtp_return_error_details was false.
1381 This has been fixed.
1383 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1384 instead of the C I/O that was originally supplied, because problems were
1385 reported on Solaris.
1387 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1388 Exim which did not show up earlier: it was assuming that a call to
1389 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1390 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1391 was a macro that became an assignment, so it seemed to work. This has
1392 changed to a proper function call with a void return, hence the compile
1393 error. Exim's code has been fixed.
1395 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1398 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1400 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1402 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1403 because it turns out to be too big or there is a timeout) there is no
1404 "Completed" line in the log. When some messages of this type were
1405 selected by exigrep, they were listed as "not completed". Others were
1406 picked up by some special patterns. I have improved the selection
1407 criteria to be more general.
1409 PH/38 The host_find_failed option in the manualroute router can now be set
1410 to "ignore", to completely ignore a host whose IP address cannot be
1411 found. If all hosts are ignored, the behaviour is controlled by the new
1412 host_all_ignored option.
1414 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1415 homing or because of multiple MX records with /mx) generated more than
1416 one IP address, and the following item turned out to be the local host,
1417 all the secondary addresses of the first item were incorrectly removed
1418 from the list, along with the local host and any following hosts (which
1419 is what is supposed to happen).
1421 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1422 whoever called Exim into the -H file. In the case of the daemon it was
1423 behaving confusingly. When first started, it used values for whoever
1424 started the daemon, but after a SIGHUP it used the Exim user (because it
1425 calls itself on a restart). I have changed the code so that it now always
1428 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1429 message are rejected with the same error (e.g. no authentication or bad
1430 sender address), and a DATA command is nevertheless sent (as can happen
1431 with PIPELINING or a stupid MUA), the error message that was given to the
1432 RCPT commands is included in the rejection of the DATA command. This is
1433 intended to be helpful for MUAs that show only the final error to their
1436 PH/42 Another patch from the Sieve maintainer.
1438 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1439 Eximstats - Fixed some broken HTML links and added missing column headers
1441 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1442 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1444 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1446 PH/43 Yet another patch from the Sieve maintainer.
1448 PH/44 I found a way to check for a TCP/IP connection going away before sending
1449 the response to the final '.' that terminates a message, but only in the
1450 case where the client has not sent further data following the '.'
1451 (unfortunately, this is allowed). However, in many cases there won't be
1452 any further data because there won't be any more messages to send. A call
1453 to select() can be used: if it shows that the input is "ready", there is
1454 either input waiting, or the socket has been closed. An attempt to read
1455 the next input character can distinguish the two cases. Previously, Exim
1456 would have sent an OK response which the client would never have see.
1457 This could lead to message repetition. This fix should cure that, at
1458 least in a lot of common cases.
1460 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1461 advertised in response to EHLO.
1467 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1468 fixed by 4.65/MH/01 (is this a record?) are fixed:
1470 (i) An empty string was always treated as zero by the numeric comparison
1471 operators. This behaviour has been restored.
1473 (ii) It is documented that the numeric comparison operators always treat
1474 their arguments as decimal numbers. This was broken in that numbers
1475 starting with 0 were being interpreted as octal.
1477 While fixing these problems I realized that there was another issue that
1478 hadn't been noticed. Values of message_size_limit (both the global option
1479 and the transport option) were treated as octal if they started with 0.
1480 The documentation was vague. These values are now always treated as
1481 decimal, and I will make that clear in the documentation.
1487 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1488 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1491 MH/01 Don't check that the operands of numeric comparison operators are
1492 integers when their expansion is in "skipping" mode (fixes bug
1493 introduced by 4.64-PH/07).
1495 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1496 child addresses, Exim now panics and dies. Previously, because the count
1497 is held in a short int, deliveries were likely to be lost. As such a
1498 large number of recipients for a single message is ridiculous
1499 (performance will be very, very poor), I have chosen to impose a limit
1500 rather than extend the field.
1506 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1507 leftover -K file (the existence of which was triggered by #402).
1508 While we were at it, introduced process PID as part of the -K
1509 filename. This should rule out race conditions when creating
1512 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1513 processing considerably. Previous code took too long for large mails,
1514 triggering a timeout which in turn triggers #401.
1516 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1517 in the DK code in transports.c. sendfile() is not really portable,
1518 hence the _LINUX specificness.
1520 TF/01 In the add_headers option to the mail command in an Exim filter,
1521 there was a bug that Exim would claim a syntax error in any
1522 header after the first one which had an odd number of characters
1525 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1526 callout verification, Exim cached a "reject" for the entire domain. This
1527 is correct for most verifications, but it is not correct for a recipient
1528 verification with use_sender or use_postmaster set, because in that case
1529 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1530 case of MAIL FROM:<> rejection from other early rejections (e.g.
1531 rejection of HELO). When verifying a recipient using a non-null MAIL
1532 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1533 Whatever the result of the callout, the value of the domain cache is
1534 left unchanged (for any other kind of callout, getting as far as trying
1535 RCPT means that the domain itself is ok).
1537 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1540 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1541 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1542 session, the daemon ignores ECONNECT errors and logs others; it now
1543 ignores EPIPE as well.
1545 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1546 (quoted-printable decoding).
1548 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1549 later the small subsequent patch to fix an introduced bug.
1551 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1553 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1555 PH/08 An error is now given if message_size_limit is specified negative.
1557 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1558 to be given (somewhat) arbitrary names.
1560 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1563 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1564 miscellaneous code fixes
1566 PH/10 Added the log_reject_target ACL modifier to specify where to log
1569 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1570 hostname. This is wrong, because it relates to the incoming message (and
1571 probably the interface on which it is arriving) and not to the outgoing
1572 callout (which could be using a different interface). This has been
1573 changed to use the value of the helo_data option from the smtp transport
1574 instead - this is what is used when a message is actually being sent. If
1575 there is no remote transport (possible with a router that sets up host
1576 addresses), $smtp_active_hostname is used.
1578 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1579 tweaks were necessary in order to get it to work (see also 21 below):
1580 (a) The code assumed that strncpy() returns a negative number on buffer
1581 overflow, which isn't the case. Replaced with Exim's string_format()
1583 (b) There were several signed/unsigned issues. I just did the minimum
1584 hacking in of casts. There is scope for a larger refactoring.
1585 (c) The code used strcasecmp() which is not a standard C function.
1586 Replaced with Exim's strcmpic() function.
1587 (d) The code set only $1; it now sets $auth1 as well.
1588 (e) A simple test gave the error "authentication client didn't specify
1589 service in request". It would seem that Dovecot has changed its
1590 interface. Fortunately there's a specification; I followed it and
1591 changed what the client sends and it appears to be working now.
1593 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1596 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1597 address A is aliased to B and C, where B exists and C does not. Without
1598 -v the output is "A verified" because verification stops after a
1599 successful redirection if more than one address is generated. However,
1600 with -v the child addresses are also verified. Exim was outputting "A
1601 failed to verify" and then showing the successful verification for C,
1602 with its parentage. It now outputs "B failed to verify", showing B's
1603 parentage before showing the successful verification of C.
1605 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1606 look up a TXT record in a specific list after matching in a combined
1609 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1610 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1611 they consult the DNS. I had assumed they would set it the way they
1612 wanted; and indeed my experiments on Linux seem to show that in some
1613 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1614 To be on the safe side, however, I have now made the interface to
1615 host_find_byname() similar to host_find_bydns(), with an argument
1616 containing the DNS resolver options. The host_find_byname() function now
1617 sets these options at its start, just as host_find_bydns() does. The smtp
1618 transport options dns_qualify_single and dns_search_parents are passed to
1619 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1620 of host_find_byname() use the default settings of RES_DEFNAMES
1621 (qualify_single) but not RES_DNSRCH (search_parents).
1623 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1624 spool_read_header() do less string testing, by means of a preliminary
1625 switch on the second character of optional "-foo" lines. (This is
1626 overdue, caused by the large number of possibilities that now exist.
1627 Originally there were few.) While I was there, I also converted the
1628 str(n)cmp tests so they don't re-test the leading "-" and the first
1629 character, in the hope this might squeeze out yet more improvement.
1631 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1632 flag allowing group syntax was set by the header_syntax check but not
1633 turned off, possible causing trouble later; (2) The flag was not being
1634 set at all for the header_verify test, causing "group"-style headers to
1635 be rejected. I have now set it in this case, and also caused header_
1636 verify to ignore an empty address taken from a group. While doing this, I
1637 came across some other cases where the code for allowing group syntax
1638 while scanning a header line wasn't quite right (mostly, not resetting
1639 the flag correctly in the right place). These bugs could have caused
1640 trouble for malformed header lines. I hope it is now all correct.
1642 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1643 with the "reply" argument non-NULL. The code, however (which originally
1644 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1645 but it didn't always do it. This confused somebody who was copying the
1646 code for some other use. I have removed all the tests.
1648 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1649 feature that was used to support insecure browsers during the U.S. crypto
1650 embargo. It requires special client support, and Exim is probably the
1651 only MTA that supported it -- and would never use it because real RSA is
1652 always available. This code has been removed, because it had the bad
1653 effect of slowing Exim down by computing (never used) parameters for the
1654 RSA_EXPORT functionality.
1656 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1657 authenticator to fail if there's a tab character in the incoming data
1658 (there should never be unless someone is messing about, as it's supposed
1659 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1660 if the connection is using TLS or if the remote IP is the same as the
1661 local IP, and the "valid-client-cert option" if a client certificate has
1664 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1665 authenticators. This can be used for authorization after authentication
1666 succeeds. (In the case of plaintext, it servers for both authentication
1669 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1670 if any retry times were supplied.
1672 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1673 connection, where there is no client IP address to check. In this
1674 situation, the verify now always succeeds.
1676 PH/25 Applied John Jetmore's -Mset patch.
1678 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1680 PH/27 In a string expansion for a processed (not raw) header when multiple
1681 headers of the same name were present, leading whitespace was being
1682 removed from all of them, but trailing whitespace was being removed only
1683 from the last one. Now trailing whitespace is removed from each header
1684 before concatenation. Completely empty headers in a concatenation (as
1685 before) are ignored.
1687 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1688 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1690 PH/29 [Removed. This was a change that I later backed out, and forgot to
1691 correct the ChangeLog entry (that I had efficiently created) before
1692 committing the later change.]
1694 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1695 address errors (4xx response to RCPT) over the same connection as other
1696 messages routed to the same hosts. Such deliveries are always "forced",
1697 so retry times are not inspected. This resulted in far too many retries
1698 for the affected addresses. The effect occurred only when there were more
1699 hosts than the hosts_max_try setting in the smtp transport when it had
1700 the 4xx errors. Those hosts that it had tried were not added to the list
1701 of hosts for which the message was waiting, so if all were tried, there
1702 was no problem. Two fixes have been applied:
1704 (i) If there are any address or message errors in an SMTP delivery, none
1705 of the hosts (tried or untried) are now added to the list of hosts
1706 for which the message is waiting, so the message should not be a
1707 candidate for sending over the same connection that was used for a
1708 successful delivery of some other message. This seems entirely
1709 reasonable: after all the message is NOT "waiting for some host".
1710 This is so "obvious" that I'm not sure why it wasn't done
1711 previously. Hope I haven't missed anything, but it can't do any
1712 harm, as the worst effect is to miss an optimization.
1714 (ii) If, despite (i), such a delivery is accidentally attempted, the
1715 routing retry time is respected, so at least it doesn't keep
1716 hammering the server.
1718 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1719 in ${readsocket because some servers need this prod.
1721 PH/32 Added some extra debug output when updating a wait-xxx database.
1723 PH/33 The hint "could be header name not terminated by colon", which has been
1724 given for certain expansion errors for a long time, was not being given
1725 for the ${if def:h_colon_omitted{... case.
1727 PH/34 The spec says: "With one important exception, whenever a domain list is
1728 being scanned, $domain contains the subject domain." There was at least
1729 one case where this was not true.
1731 PH/35 The error "getsockname() failed: connection reset by peer" was being
1732 written to the panic log as well as the main log, but it isn't really
1733 panic-worthy as it just means the connection died rather early on. I have
1734 removed the panic log writing for the ECONNRESET error when getsockname()
1737 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1738 runs only) independently of the message's sender address. This meant
1739 that, if the 4xx error was in fact related to the sender, a different
1740 message to the same recipient with a different sender could confuse
1741 things. In particualar, this can happen when sending to a greylisting
1742 server, but other circumstances could also provoke similar problems.
1743 I have changed the default so that the retry time for these errors is now
1744 based a combination of the sender and recipient addresses. This change
1745 can be overridden by setting address_retry_include_sender=false in the
1748 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1749 remote server are returned as part of bounce messages. This was not
1750 happening for LMTP over a pipe (the lmtp transport), but now it is the
1751 same for both kinds of LMTP.
1753 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1754 in header lines that were in fact CNAMEs. This is no longer the case.
1756 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1757 and queue runs started by the daemon processed all messages. This has
1758 been fixed so that -R and -S can now usefully be given with -q<time>.
1760 PH/40 Import PCRE release 6.7 (fixes some bugs).
1762 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1764 PH/42 Give an error if -q is specified more than once.
1766 PH/43 Renamed the variables $interface_address and $interface_port as
1767 $received_ip_address and $received_port, to make it clear that these
1768 values apply to message reception, and not to the outgoing interface when
1769 a message is delivered. (The old names remain recognized, of course.)
1771 PH/44 There was no timeout on the connect() call when using a Unix domain
1772 socket in the ${readsocket expansion. There now is.
1774 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1775 be meaningful with "accept".
1777 SC/01 Eximstats V1.43
1778 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1780 SC/02 Eximstats V1.44
1781 Use a glob alias rather than an array ref in the generated
1782 parser. This improves both readability and performance.
1784 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1785 Collect SpamAssassin and rejection statistics.
1786 Don't display local sender or destination tables unless
1787 there is data to show.
1788 Added average volumes into the top table text output.
1790 SC/04 Eximstats V1.46
1791 Collect data on the number of addresses (recipients)
1792 as well as the number of messages.
1794 SC/05 Eximstats V1.47
1795 Added 'Message too big' to the list of mail rejection
1796 reasons (thanks to Marco Gaiarin).
1798 SC/06 Eximstats V1.48
1799 Mainlog lines which have GMT offsets and are too short to
1800 have a flag are now skipped.
1802 SC/07 Eximstats V1.49 (Alain Williams)
1803 Added the -emptyok flag.
1805 SC/08 Eximstats V1.50
1806 Fixes for obtaining the IP address from reject messages.
1808 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1809 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1810 whitesspace changes from 4.64-PH/27
1812 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1815 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1816 are found, allow negative numbers in numeric criteria)
1818 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1820 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1821 to match changes made in 4.64-PH/43
1823 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1825 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1826 "domain" is actually an IP address, and force a failure. This locks out
1827 those revolvers/nameservers that support "A-for-A" lookups, in
1828 contravention of the specifications.
1830 PH/48 When a host name was looked up from an IP address, and the subsequent
1831 forward lookup of the name timed out, the host name was left in
1832 $sender_host_name, contrary to the specification.
1834 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1835 restricted to single-key lookups, Exim was not diagnosing an error if
1836 * or *@ was used with a query-style lookup.
1838 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1840 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1841 long ago, but noone interested enough thought of it. Let's just say that
1842 the "1.1" means that there are some new functions that weren't there at
1843 some point in the past.
1845 PH/51 Error processing for expansion failure of helo_data from an smtp
1846 transport during callout processing was broken.
1848 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1849 tested/used via the -bh/-bhc/-bs options.
1851 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1852 bug, fixed in subsequent PCRE releases).
1854 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1855 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1857 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1863 SC/01 Use a glob alias rather than an array ref in eximstats generated
1864 parser. This improves both readability and performance.
1866 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1867 Don't display local sender or destination tables in eximstats unless
1868 there is data to show.
1869 Added average volumes into the eximstats top table text output.
1871 SC/03 Collect data on the number of addresses (recipients) as well
1872 as the number of messages in eximstats.
1874 TF/01 Correct an error in the documentation for the redirect router. Exim
1875 does (usually) call initgroups() when daemonizing.
1877 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1878 with consistent privilege compared to when running as a daemon.
1880 TF/03 Note in the spec that $authenticated_id is not set for local
1881 submissions from trusted users.
1883 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1884 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1886 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1887 by adding some example configuration directives to the default
1888 configuration file. A little bit of work is required to uncomment the
1889 directives and define how usernames and passwords are checked, but
1890 there is now a framework to start from.
1892 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1893 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1894 without this. I don't know how relevant this is to other LDAP libraries.
1896 PH/02 Add the verb name to the "unknown ACL verb" error.
1898 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1900 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1902 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1903 directory not expanded when it should be if an expanded home directory
1904 was set for the address (which is overridden by the transport).
1906 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1909 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1910 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1911 because it is too late at that time, and has no effect.
1913 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1914 security issue with \' (bugzilla #107). I could not use the
1915 PQescapeStringConn() function, because it needs a PGconn value as one of
1918 PH/08 When testing addresses using -bt, indicate those final addresses that
1919 are duplicates that would not cause an additional delivery. At least one
1920 person was confused, thinking that -bt output corresponded to deliveries.
1921 (Suppressing duplicates isn't a good idea as you lose the information
1922 about possibly different redirections that led to the duplicates.)
1924 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1925 systems where poll() doesn't work, in particular OS X.
1927 PH/10 Added more information to debugging output for retry time not reached.
1929 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1930 operations in malware.c.
1932 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1935 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1936 syslog_duplication set false, log lines that would normally be written
1937 both the the main log and to the reject log were not written to syslog at
1940 PH/14 In the default configuration, change the use of "message" in ACL warn
1941 statements to "add_header".
1943 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1944 not followed by a command (e.g. "seen endif").
1946 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1947 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1950 PH/17 Added extra conditions to the default value of delay_warning_condition
1954 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1955 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1956 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1959 The Auto-Submitted: and various List- headers are standardised, whereas I
1960 don't think Precedence: ever was.
1962 PH/18 Refactored debugging code in route_finduser() to show more information,
1963 in particular, the error code if getpwnam() issues one.
1965 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1966 This is apparently needed in addition to the PH/07 change above to avoid
1967 any possible encoding problems.
1969 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1970 but not after initializing Perl.
1972 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1973 output them only if debugging. By default they are written stderr,
1974 apparently, which is not desirable.
1976 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1979 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1982 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1984 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1985 authenticated or an ident call has been made. Suppress the default
1986 values for $authenticated_id and $authenticated_sender (but permit -oMai
1987 and -oMas) when testing with -bh.
1989 PH/24 Re-jigged the order of the tests in the default configuration so that the
1990 tests for valid domains and recipients precede the DNS black list and CSA
1991 tests, on the grounds that those ones are more expensive.
1993 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1994 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1995 This bug exists in every version of Exim that I still have, right back to
1998 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1999 However, an attempt to turn on case-sensitivity in a regex key by
2000 including (?-i) didn't work because the subject string was already
2001 lowercased, and the effects were non-intuitive. It turns out that a
2002 one-line patch can be used to allow (?-i) to work as expected.
2008 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2009 other effects) broke the use of negated acl sub-conditions.
2011 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2014 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2015 "Deny" causes Exim to reject the incoming connection with a 554 error.
2016 Unfortunately, if there is a major crisis, such as a disk failure,
2017 tcp-wrappers gives "deny", whereas what one would like would be some
2018 kind of temporary error. A kludge has been added to help with this.
2019 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2020 554 error is used if errno is still zero or contains ENOENT (which occurs
2021 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2024 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2026 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2027 errors. Otherwise a message that provokes a temporary error (when other
2028 messages do not) can cause a whole host to time out.
2030 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2031 addresses were routed directly to files or pipes from a redirect router.
2032 File deliveries just didn't batch; pipe deliveries might have suffered
2035 PH/06 A failure to get a lock for a hints database would erroneously always say
2036 "Failed to get write lock", even when it was really a read lock.
2038 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2039 of 0600. This has been changed to use the value of the lockfile_mode
2040 option (which defaults to 0600).
2042 PH/08 Applied small patch from the Sieve maintainer.
2044 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2045 folder from quota calculations, a direct delivery into this folder messed
2046 up the contents of the maildirsize file. This was because the regex was
2047 used only to exclude .Trash (or whatever) when the size of the mailbox
2048 was calculated. There was no check that a delivery was happening into an
2049 excluded directory. This bug has been fixed by ignoring all quota
2050 processing for deliveries into excluded directories.
2052 PH/10 Added the maildirfolder_create_regex option to appendfile.
2058 PH/01 The code for finding all the local interface addresses on a FreeBSD
2059 system running IPv6 was broken. This may well have applied to all BSD
2060 systems, as well as to others that have similar system calls. The broken
2061 code found IPv4 interfaces correctly, but gave incorrect values for the
2062 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2063 that it would not match correctly against @[] and not recognize the IPv6
2066 PH/02 The ipliteral router was not recognizing addresses of the form user@
2067 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2069 PH/03 Added disable_ipv6.
2071 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2072 decoded form, because it is most often used to construct To: headers
2073 lines in autoreplies, and the decoded form may well be syntactically
2074 invalid. However, $reply_address has leading white space removed, and all
2075 newlines turned into spaces so that the autoreply transport does not
2078 PH/05 If group was specified without a user on a router, and no group or user
2079 was specified on a transport, the group from the router was ignored.
2081 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2082 for visible compile-time settings that can be used to change these
2083 numbers, for those that want even more. Backwards compatibility with old
2084 spool files has been maintained. However, going back to a previous Exim
2085 release will lost any variables that are in spool files.
2087 PH/07 Two small changes when running in the test harness: increase delay when
2088 passing a TCP/IP connection to a new process, in case the original
2089 process has to generate a bounce, and remove special handling of
2090 127.0.0.2 (sic), which is no longer necessary.
2092 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2093 be the same on different OS.
2095 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2098 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2099 whether --show-vars was specified or not
2101 JJ/02 exipick: Added support for new ACL variable spool format introduced
2104 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2105 syntactically invalid From: or Reply-to: line, and a filter used this to
2106 generate an autoreply, and therefore failed to obtain an address for the
2107 autoreply, Exim could try to deliver to a non-existent relative file
2108 name, causing unrelated and misleading errors. What now happens is that
2109 it logs this as a hard delivery error, but does not attempt to create a
2112 PH/11 The exinext utility has a -C option for testing purposes, but although
2113 the given file was scanned by exinext itself; it wasn't being passed on
2114 when Exim was called.
2116 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2117 an end-of-file indication when reading a command response.
2119 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2120 compiled. In many other places in Exim, IPv6 addresses are always
2121 recognized, so I have changed this. It also means that IPv4 domain
2122 literals of the form [IPV4:n.n.n.n] are now always recognized.
2124 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2125 used if the router is not running as root, for example, when verifying at
2126 ACL time, or when using -bh. The debugging output from this situation was
2127 non-existent - all you got was a failure to exec. I have made two
2130 (a) Failures to set uid/gid, the current directory, or a process leader
2131 in a subprocess such as that created by queryprogram now generate
2132 suitable debugging ouput when -d is set.
2134 (b) The queryprogram router detects when it is not running as root,
2135 outputs suitable debugging information if -d is set, and then runs
2136 the subprocess without attempting to change uid/gid.
2138 PH/15 Minor change to Makefile for building test_host (undocumented testing
2141 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2142 additional section of a DNS packet that returns MX or SRV records.
2143 Instead, it always explicitly searches for A/AAAA records. This avoids
2144 major problems that occur when a DNS server includes only records of one
2145 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2146 fixed another bug: if SRV records were looked up and the corresponding
2147 address records were *not* found in the additional section, the port
2148 values from the SRV records were lost.
2150 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2151 using the correct key (the original address) when searching the retry
2152 rules in order to find which one to use for generating the retry hint.
2154 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2155 adding the default one. Similarly, if it contains a Reply-To: header, the
2156 errors_reply_to option, if set, is not used.
2158 PH/19 When calculating a retry time, Exim used to measure the "time since
2159 failure" by looking at the "first failed" field in the retry record. Now
2160 it does not use this if it is later than than the arrival time of the
2161 message. Instead it uses the arrival time. This makes for better
2162 behaviour in cases where some deliveries succeed, thus re-setting the
2163 "first failed" field. An example is a quota failure for a huge message
2164 when small messages continue to be delivered. Without this change, the
2165 "time since failure" will always be short, possible causing more frequent
2166 delivery attempts for the huge message than are intended.
2167 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2169 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2170 $1, $2, $3) because the numerical variables can be reset during some
2171 expansion items (e.g. "match"), thereby losing the authentication data.
2173 PH/21 Make -bV show the size of off_t variables so that the test suite can
2174 decide whether to run tests for quotas > 2G.
2176 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2177 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2178 filecount value is greater than 2G or if a quota value is greater than 2G
2179 on a system where the size of off_t is not greater than 4, a panic error
2182 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2183 never match. The debug and -bh output now contains an explicit error
2184 message indicating a malformed IPv4 address or mask.
2186 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2187 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2188 PH/23 above applies.
2190 PH/25 Do not write to syslog when running in the test harness. The only
2191 occasion when this arises is a failure to open the main or panic logs
2192 (for which there is an explicit test).
2194 PH/26 Added the /no_tell option to "control=freeze".
2196 PH/27 If a host name lookup failed very early in a connection, for example, if
2197 the IP address matched host_lookup and the reverse lookup yielded a name
2198 that did not have a forward lookup, an error message of the form "no IP
2199 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2200 could be logged. Now it outputs the IP address instead of "NULL".
2202 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2203 allows the sender and the authenticated sender to be set when
2204 submitting a message from within Exim. Since child_open_exim() is
2205 documented for local_scan(), the new function should be too.
2207 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2208 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2209 results in an empty string is now treated as unset.
2211 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2213 PH/31 Added sender_verify_fail logging option.
2215 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2216 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2217 tidied the source and removed it altogether.
2219 PH/33 When a queue run was abandoned because the load average was too high, a
2220 log line was always written; now it is written only if the queue_run log
2221 selector is set. In addition, the log line for abandonment now contains
2222 information about the queue run such as the pid. This is always present
2223 in "start" and "stop" lines but was omitted from the "abandon" line.
2225 PH/34 Omit spaces between a header name and the colon in the error message that
2226 is given when verify = headers_syntax fails (if there are lots of them,
2227 the message gets confusing).
2229 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2230 names, as there are now some PTR records that contain slashes. This check
2231 is only to protect against broken name servers that fall over on strange
2232 characters, so the fact that it applies to all lookups doesn't matter.
2234 PH/36 Now that the new test suite is complete, we can remove some of the
2235 special code in Exim that was needed for the old test suite. For example,
2236 sorting DNS records because real resolvers return them in an arbitrary
2237 order. The new test suite's fake resolver always returns records in the
2240 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2241 bounces) except when queue_only is set, to avoid logging races between
2242 the different processes.
2244 PH/38 Panic-die if .include specifies a non-absolute path.
2246 PH/39 A tweak to the "H" retry rule from its user.
2248 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2249 a label. They prevented compilation on older perls.
2251 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2252 a warning to be raised on newish perls.
2254 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2255 on queue. Changes to match documented behaviour of showing count of
2256 messages matching specified criteria.
2258 PH/40 Changed the default ident timeout from 30s to 5s.
2260 PH/41 Added support for the use of login_cap features, on those BSD systems
2261 that have them, for controlling the resources used by pipe deliveries.
2263 PH/42 The content-scanning code uses fopen() to create files in which to put
2264 message data. Previously it was not paying any attention to the mode of
2265 the files. Exim runs with umask(0) because the rest of the code creates
2266 files with open(), and sets the required mode explicitly. Thus, these
2267 files were ending up world-writeable. This was not a big issue, because,
2268 being within the spool directory, they were not world-accessible. I have
2269 created a function called modefopen, which takes an additional mode
2270 argument. It sets umask(777), creates the file, chmods it to the required
2271 mode, then resets the umask. All the relevant calls to fopen() in the
2272 content scanning code have been changed to use this function.
2274 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2275 to 24 hours. This avoids potential overflow problems when processing G
2276 and H retry rules. I suspect nobody ever tinkers with this value.
2278 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2280 PH/45 When the plaintext authenticator is running as a client, the server's
2281 challenges are checked to ensure they are valid base64 strings. By
2282 default, the authentication attempt is cancelled if an invalid string is
2283 received. Setting client_ignore_invalid_base64 true ignores these errors.
2284 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2285 they are received. Thus, the responses can be made to depend on the
2286 challenges. If an invalid string is ignored, an empty string is placed in
2289 PH/46 Messages that are created by the autoreply transport now contains a
2290 References: header, in accordance with RFCs 2822 and 3834.
2292 PH/47 Added authenticated_sender_force to the smtp transport.
2294 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2296 PH/49 Installed latest patch from the Sieve maintainer.
2298 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2299 was also set, the appendfile transport was unnecessarily scanning a
2300 directory of message files (e.g. for maildir delivery) to find the count
2301 of files (along with the size), even though it did not need this
2302 information. It now does the scan only if it needs to find either the
2303 size of the count of files.
2305 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2307 PH/52 Two bugs concerned with error handling when the smtp transport is
2310 (i) Exim was not creating retry information for temporary errors given
2311 for individual recipients after the DATA command when the smtp transport
2312 was used in LMTP mode. This meant that they could be retried too
2313 frequently, and not timed out correctly.
2315 (ii) Exim was setting the flag that allows error details to be returned
2316 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2317 recipients that were returned after the DATA command.
2319 PH/53 This is related to PH/52, but is more general: for any failing address,
2320 when detailed error information was permitted to be returned to the
2321 sender, but the error was temporary, then after the final timeout, only
2322 "retry timeout exceeded" was returned. Now it returns the full error as
2323 well as "retry timeout exceeded".
2325 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2326 do this, and (what is worse) MTAs that accept it.
2328 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2329 will now be deprecated.
2331 PH/56 New os.c-cygwin from the Cygwin maintainer.
2333 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2334 formats (previously only available in exim formats via -bpr, -bpru,
2335 and -bpra. Now also available in native and exiqgrep formats)
2337 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2338 with very large, slow to parse queues
2340 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2342 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2344 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2345 responses to authentication challenges, though it was showing the
2346 challenges; (ii) I've removed the CR characters from the debug output for
2349 PH/58 Allow for the insertion of a newline as well as a space when a string
2350 is turned into more than one encoded-word during RFC 2047 encoding. The
2351 Sieve code now uses this.
2353 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2354 data_4xx, lost_connection, tls_required.
2356 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2357 message was being sent as an SMTP response.
2359 PH/61 Add -l and -k options to exicyclog.
2361 PH/62 When verifying, if an address was redirected to one new address, so that
2362 verification continued, and the new address failed or deferred after
2363 having set something in $address_data, the value of $address_data was not
2364 passed back to the ACL. This was different to the case when no
2365 redirection occurred. The value is now passed back in both cases.
2367 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2368 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2369 use login_cap.h, so on its own it isn't the distinguishing feature. The
2370 new name refers directly to the setclassresources() function.
2372 PH/65 Added configuration files for NetBSD3.
2374 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2376 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2377 is preferred over IPv4.
2379 PH/68 The bounce_return_message and bounce_return_body options were not being
2380 honoured for bounces generated during the reception of non-SMTP messages.
2381 In particular, this applied to messages rejected by the ACL. This bug has
2382 been fixed. However, if bounce_return_message is true and bounce_return_
2383 body is false, the headers that are returned for a non-SMTP message
2384 include only those that have been read before the error was detected.
2385 (In the case of an ACL rejection, they have all been read.)
2387 PH/69 The HTML version of the specification is now built in a directory called
2388 spec_html instead of spec.html, because the latter looks like a path with
2389 a MIME-type, and this confuses some software.
2391 PH/70 Catch two compiler warnings in sieve.c.
2393 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2394 function verify_get_ident() calls ip_connect() to connect a socket, but
2395 if the "connect()" function timed out, ip_connect() used to close the
2396 socket. However, verify_get_ident() also closes the socket later, and in
2397 between Exim writes to the log, which may get opened at this point. When
2398 the socket was closed in ip_connect(), the log could get the same file
2399 descriptor number as the socket. This naturally causes chaos. The fix is
2400 not to close the socket in ip_connect(); the socket should be closed by
2401 the function that creates it. There was only one place in the code where
2402 this was missing, in the iplookup router, which I don't think anybody now
2403 uses, but I've fixed it anyway.
2405 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2406 well as to direct DNS lookups. Otherwise the handling of names in host
2407 lists is inconsistent and therefore confusing.
2413 PH/01 Two changes to the default runtime configuration:
2415 (1) Move the checks for relay_from_hosts and authenticated clients from
2416 after to before the (commented out) DNS black list checks.
2418 (2) Add control=submission to the relay_from_hosts and authenticated
2419 clients checks, on the grounds that messages accepted by these
2420 statements are most likely to be submissions.
2422 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2424 (1) Generate an error if the third argument for the ${prvs expansion is
2427 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2430 (3) Reset the variables that are obtained from the first argument of
2431 ${prvscheck and used in the second argument before leaving the code,
2432 because their memory is reclaimed, so using them afterwards may do
2435 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2436 one (it's much easier than Tom thought :-).
2438 (5) Because of (4), we can now allow for the use of $prvscheck_result
2439 inside the third argument.
2441 PH/03 For some reason, the default setting of PATH when running a command from
2442 a pipe transport was just "/usr/bin". I have changed it to
2445 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2446 anything to be listed in the output from -bV.
2448 PH/05 When a filter generated an autoreply, the entire To: header line was
2449 quoted in the delivery log line, like this:
2451 => >A.N.Other <ano@some.domain> <original@ddress> ...
2453 This has been changed so that it extracts the operative address. There
2454 may be more than one such address. If so, they are comma-separated, like
2457 => >ano@some.domain,ona@other.domain <original@ddress> ...
2459 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2460 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2461 not being looked up in the rDNS to get a host name, Exim was showing the
2462 IP address twice in Received: lines, even though the IP addresses were
2463 identical. For example:
2465 Received: from [1.2.3.4] (helo=[1.2.3.4])
2467 However, if the real host name was known, it was omitting the HELO data
2468 if it matched the actual IP address. This has been tidied up so that it
2469 doesn't show the same IP address twice.
2471 PH/07 When both +timestamp and +memory debugging was on, the value given by
2472 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2473 called by the debug printing, thereby overwriting the timestamp buffer.
2474 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2476 PH/08 When the original message was included in an autoreply transport, it
2477 always said "this is a copy of the message, including all the headers",
2478 even if body_only or headers_only was set. It now gives an appropriate
2481 PH/09 Applied a patch from the Sieve maintainer which:
2483 o fixes some comments
2484 o adds the (disabled) notify extension core
2485 o adds some debug output for the result of if/elsif tests
2486 o points to the current vacation draft in the documentation
2487 and documents the missing references header update
2491 o fixes a bug in processing the envelope test (when testing
2492 multiple envelope elements, the last element determinted the
2495 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2496 Electronic Mail") by including:
2498 Auto-submitted: auto-generated
2500 in the messages that it generates (bounce messages and others, such as
2501 warnings). In the case of bounce messages for non-SMTP mesages, there was
2502 also a typo: it was using "Auto_submitted" (underscore instead of
2503 hyphen). Since every message generated by Exim is necessarily in response
2504 to another message, thes have all been changed to:
2506 Auto-Submitted: auto-replied
2508 in accordance with these statements in the RFC:
2510 The auto-replied keyword:
2512 - SHOULD be used on messages sent in direct response to another
2513 message by an automatic process,
2515 - MUST NOT be used on manually-generated messages,
2517 - MAY be used on Delivery Status Notifications (DSNs) and Message
2518 Disposition Notifications (MDNs),
2520 - MUST NOT be used on messages generated by automatic or periodic
2521 processes, except for messages which are automatic responses to
2524 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2525 to the default Received: header definition.
2527 PH/12 Added log selector acl_warn_skipped (default on).
2529 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2530 variables because (a) they are in the wrong storage pool and (b) even if
2531 they were copied, it wouldn't work properly because of the caching.
2533 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2534 checking when decoding. Apparently there are clients that generate
2535 overlong encoded strings. Why am I not surprised?
2537 PH/15 If the first argument of "${if match_address" was not empty, but did not
2538 contain an "@" character, Exim crashed. Now it writes a panic log message
2539 and treats the condition as false.
2541 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2543 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2544 extension "envelope-auth". The code is finished and in agreement with
2545 other implementations, but there is no documentation so far and in fact,
2546 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2547 not changing the active code.
2549 Print executed "if" and "elsif" statements when debugging is used. This
2550 helps a great deal to understand what a filter does.
2552 Document more things not specified clearly in RFC3028. I had all this
2553 sorted out, when out of a sudden new issues came to my mind. Oops."
2555 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2558 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2559 canonical form (as documented). However, after a host name lookup from
2560 the IP address, check_host() was doing a simple string comparison with
2561 addresses acquired from the DNS when checking that the found name did
2562 have the original IP as one of its addresses. Since any found IPv6
2563 addresses are likely to be in abbreviated form, the comparison could
2564 fail. Luckily, there already exists a function for doing the comparison
2565 by converting both addresses to binary, so now that is used instead of
2566 the text comparison.
2568 PH/20 There was another similar case to PH/19, when a complete host name was
2569 given in a host list; looking up its IP address could give an abbreviated
2570 form, whereas the current host's name might or might not be abbreviated.
2571 The same fix has been applied.
2577 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2578 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2581 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2582 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2584 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2586 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2587 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2588 Also, for "Auto-Submitted", treat anything other than "no" as
2589 non-personal, in accordance with RFC 3834. (Previously it treated
2590 anything starting "auto-" as non-personal.)
2592 TF/01 The control=submission/name=... option had a problem with syntax
2593 errors if the name included a slash character. The /name= option
2594 now slurps the rest of the string, so it can include any characters
2595 but it must come last in the list of options (after /sender_retain
2598 PH/05 Some modifications to the interface to the fake nameserver for the new
2606 TK/01 Added the "success_on_redirect" address verification option. See
2607 NewStuff for rationale and an example.
2609 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2611 PH/02 Patch to exigrep to allow it to work on syslog lines.
2613 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2614 fread() to skip over the body file's header line, because in Cygwin the
2615 header line is locked and is inaccessible.
2617 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2618 co-exist for some time) to make it clear that it is the Exim ID that is
2619 referenced, not the Message-ID: header line.
2621 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2622 string_format() function, because snprintf() does not exist on all
2625 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2626 ${stat: expansion item.
2628 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2629 protocol synchronization error", to keep the pedants happy.
2631 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2632 well as for IRIX systems, when gcc is being used. See the host.c source
2635 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2637 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2640 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2641 transport and to the smtp transport in LMTP mode.
2643 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2645 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2646 run a filter in a subprocess. This could lead to confusion in subsequent
2647 lookups in the parent process. There should also be a search_tidyup() at
2648 the end of the subprocess.
2650 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2651 only if the host matched helo_try_verify_hosts, which caused the
2652 verification to occur when the EHLO/HELO command was issued. The ACL just
2653 tested the remembered result. Now, if a previous verification attempt has
2654 not happened, "verify = helo" does it there and then.
2656 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2658 TK/03 Fix log output including CR from clamd.
2660 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2661 exist provoked a memory error which could cause a segfault.
2663 PH/15 Installed PCRE 6.2
2665 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2667 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2668 of the problem. Specifically, suggested +O2 rather than +O1 for the
2671 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2673 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2674 record, the port was not correctly logged when the outgoing_port log
2675 selector was set (it logged the transort's default port).
2677 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2678 fallback_hosts, and "hosts" in the smtp transport.
2680 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2681 host errors such as "Connection refused".
2683 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2684 authentication with radiusclient 0.4.9:
2686 - Error returned from rc_read_config was caught wrongly
2687 - Username/password not passed on to radius server due to wrong length.
2689 The presumption is that some radiusclient API changes for 4.51/PH/17
2690 were not taken care of correctly. The code is still untested by me (my
2691 Linux distribution still has 0.3.2 of radiusclient), but it was
2692 contributed by a Radius user.
2694 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2695 expanding the "port" option of the smtp transport.
2697 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2698 while reading a MIME header. Thanks to Tom Hughes for a patch.
2700 PH/24 Include config.h inside local_scan.h so that configuration settings are
2703 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2704 that in an ACL for RCPT (for example), you can examine exactly what was
2707 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2708 commands, but it was not correctly comparing the address with the actual
2709 client host address. Thus, it would show the EHLO address in Received:
2710 header lines when this was not necessary.
2712 PH/27 Added the % operator to ${eval:}.
2714 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2715 it should be ignoring failures (because with -C, for example, it has lost
2716 privilege). It wasn't ignoring creation failures other than "already
2719 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2722 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2723 because an input line was too long, either on its own, or by virtue of
2724 too many continuations, the temporary file was not being removed, and the
2725 return code was incorrect.
2727 PH/31 Missing "BOOL" in function definition in filtertest.c.
2729 PH/32 Applied Sieve patches from the maintainer.
2731 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2733 PH/33 Added "verify = not_blind".
2735 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2736 Local/Makefile (with some defaults set). These are used in built scripts
2737 such as exicyclog, but they have never been used in the exim_install
2738 script (though there are many overriding facilities there). I have
2739 arranged that the exim_install script now takes note of these two
2742 PH/35 Installed configuration files for Dragonfly.
2744 PH/36 When a locally submitted message by a trusted user did not contain a
2745 From: header, and the sender address was obtained from -f or from an SMTP
2746 MAIL command, and the trusted user did not use -F to supply a sender
2747 name, $originator_name was incorrectly used when constructing a From:
2748 header. Furthermore, $originator_name was used for submission mode
2749 messages from external hosts without From: headers in a similar way,
2750 which is clearly wrong.
2752 PH/37 Added control=suppress_local_fixups.
2754 PH/38 When log_selector = +received_sender was set, and the addition of the
2755 sender made the log line's construction buffer exactly full, or one byte
2756 less than full, an overflow happened when the terminating "\n" was
2759 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2760 when the result of a list match is failure because a DNS lookup failed.
2762 PH/40 RM_COMMAND is now used in the building process.
2764 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2765 the "build-* directories that it finds.
2767 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2768 address in a domain literal was a prefix of an interface address.
2770 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2771 when verifying a sender address, unless rewrite_headers is false.
2773 PH/44 Wrote a long comment about why errors_to addresses are verified as
2774 recipients, not senders.
2776 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2777 the ratelimit ACL was added.
2779 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2781 PH/46 Added extra information about PostgreSQL errors to the error string.
2783 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2784 avoiding the need to install special zones in a real server. This is
2785 backwards compatible; if it can't find the fake resolver, it drops back.
2786 Thus, both old and new test suites can be run.
2788 TF/02 Added util/ratelimit.pl
2790 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2791 clock is set back in time.
2793 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2794 Candler <B.Candler@pobox.com>.
2796 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2797 OK for addresses that are the result of redirections.
2799 PH/48 A number of further additions for the benefit of the new test suite,
2800 including a fake gethostbyname() that interfaces to the fake DNS resolver
2803 TF/06 The fix for widen_domains has also been applied to qualify_single and
2804 search_parents which are the other dnslookup options that can cause
2807 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2810 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2812 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2813 service" if it encountered an MX record with an empty target hostname.
2814 The message is now "an MX or SRV record indicated no SMTP service".
2816 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2817 if the DNS of the sending site is misconfigured. This is quite a
2818 common situation. This change restores the behaviour of treating a
2819 helo verification defer as a failure.
2821 PH/51 If self=fail was set on a router, the bounce message did not include the
2822 actual error message.
2828 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2830 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2831 command itself ended in error, the underlying message about the transport
2832 filter timeout was being overwritten with the pipe command error. Now the
2833 underlying error message should be appended to the second error message.
2835 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2837 PH/02 Reduce the amount of output that "make" produces by default. Full output
2838 can still be requested.
2840 PH/03 The warning log line about a condition test deferring for a "warn" verb
2841 was being output only once per connection, rather than after each
2842 occurrence (because it was using the same function as for successful
2843 "warn" verbs). This seems wrong, so I have changed it.
2845 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2846 it should not have, which might have caused a crash in the right
2847 circumstances, but probably never did.
2849 PH/04 Installed a modified version of Tony Finch's patch to make submission
2850 mode fix the return path as well as the Sender: header line, and to
2851 add a /name= option so that you can make the user's friendly name appear
2854 TF/03 Added the control = fakedefer ACL modifier.
2856 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2857 Mark Lowes for thorough testing.
2859 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2861 TK/03 Merged latest SRS patch from Miles Wilton.
2863 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2864 with the definition in sysexits.h (which is #included earlier).
2865 Fortunately, Exim does not actually use EX_OK. The code used to try to
2866 preserve the sysexits.h value, by assumimg that macro definitions were
2867 scanned for macro replacements. I have been disabused of this notion,
2868 so now the code just undefines EX_OK before #including unistd.h.
2870 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2871 in the smtp transport. When a block could not be written in a single
2872 write() function, the timeout was being re-applied to each part-write.
2873 This seems wrong - if the receiver was accepting one byte at a time it
2874 would take for ever. The timeout is now adjusted when this happens. It
2875 doesn't have to be particularly precise.
2877 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2878 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2880 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2881 without a domain if the check to <postmaster@domain> fails.
2883 SC/01 Eximstats: added -xls and the ability to specify output files
2884 (patch written by Frank Heydlauf).
2886 SC/02 Eximstats: use FileHandles for outputing results.
2888 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2890 SC/04 Eximstats: fixed display of large numbers with -nvr option
2892 SC/05 Eximstats: fixed merging of reports with empty tables.
2894 SC/06 Eximstats: added the -include_original_destination flag
2896 SC/07 Eximstats: removed tabs and trailing whitespace.
2898 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2900 TK/06 MBOX spool code: Add real "From " MBOX separator line
2901 so the .eml file is really in mbox format (even though
2902 most programs do not really care). Patch from Alex Miller.
2904 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2905 The latter is generated from $received_to and is only set if the
2906 message has one envelope recipient. SA can use these headers,
2907 obviously out-of-the-box. Patch from Alex Miller.
2909 PH/08 The ${def test on a variable was returning false if the variable's
2910 value was "0", contrary to what the specification has always said!
2911 The result should be true unless the variable is empty.
2913 PH/09 The syntax error of a character other than { following "${if
2914 def:variable_name" (after optional whitespace) was not being diagnosed.
2915 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2916 accidental colon was present, for example, could give incorrect results.
2918 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2919 result is used (not including appendfile, where other changes are about
2922 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2923 This involved changing a lot of size variables from int to off_t. It
2924 should work with maildirs and everything.
2926 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2927 spamd dying while we are connected to it.
2929 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2932 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2933 function for BDB 4.3.
2935 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2937 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2938 the sources for PCRE have been re-arranged and more files are now
2941 PH/15 The code I had for printing potentially long long variables in PH/11
2942 above was not the best (it lost precision). The length of off_t variables
2943 is now inspected at build time, and an appropriate printing format (%ld
2944 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2945 to be "long long int" or "long int". This is needed for the internal
2946 formatting function string_vformat().
2948 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2949 the configuration file to be ":syslog", then the script "guesses" where
2950 the logs files are, rather than using the compiled in default. In our
2951 case the guess is not the same as the compiled default, so the script
2952 suddenly stopped working when I started to use syslog. The patch checks
2953 to see if log_file_path is "". If so, it attempts to read it from exim
2954 with no configuration file to get the compiled in version, before it
2955 falls back to the previous guessing code."
2957 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2958 implementing BATV in an Exim configuration. See NewStuff for the gory
2961 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2962 Makefile that are specific to HP-UX.
2964 PH/18 If the "use_postmaster" option was set for a recipient callout together
2965 with the "random" option, the postmaster address was used as the MAIL
2966 FROM address for the random test, but not for the subsequent recipient
2967 test. It is now used for both.
2969 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2970 patch removes a few documentation additions to RFC 3028, because the
2971 latest draft now contains them. It adds the new en;ascii-case comparator
2972 and a new error check for 8bit text in MIME parts. Comparator and
2973 require names are now matched exactly. I enabled the subaddress
2974 extension, but it is not well tested yet (read: it works for me)."
2976 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2977 rework some of the code of TK/09 above to avoid the hardwired use of
2978 "%lld" and "long long". Replaced the call to snprintf() with a call to
2981 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2982 records point to non-existent hosts", "retry timeout exceeded", and
2983 "retry time not reached for any host after a long failure period".
2985 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2986 experimental DomainKeys support:
2988 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2989 (2) On an error such as an illegally used "control", the wrong name for
2990 the control was given.
2992 These problems did NOT occur unless DomainKeys support was compiled.
2994 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2996 PH/24 Added ${if match_ip condition.
2998 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2999 databases so that it will be absolutely obvious if a crash occurs in the
3000 DB library. This is a regular occurrence (often caused by mis-matched
3003 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3004 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3005 picked up on a user's system that detects such things. There doesn't seem
3006 to be a gcc warning option for this - only an attribute that has to be
3007 put on the function's prototype. It seems that in Fedora Core 4 they have
3008 set this on a number of new functions. No doubt there will be more in due
3011 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3012 specify a transport. However, if an address that was verified by such a
3013 router was the subject of a callout, Exim crashed because it tried to
3014 read the rcpt_include_affixes from the non-existent transport. Now it
3015 just assumes that the setting of that option is false. This bug was
3016 introduced by 4.51/PH/31.
3018 PH/28 Changed -d+all to exclude +memory, because that information is very
3019 rarely of interest, but it makes the output a lot bigger. People tend to
3020 do -d+all out of habit.
3022 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3023 code in os-type was giving problems when libc.so lives in lib64, like on
3026 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3027 aren't the modern standard, and it seems that some systems' include files
3028 don't always have them. Exim was already checking for some of the newer
3029 ones like T_AAAA, and defining it itself. I've added checks for all the
3030 record types that Exim uses.
3032 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3033 not automatically generating a new one, as it is supposed to. This
3034 prevented TLS from working. If the file did exist, but contained invalid
3035 data, a new version was generated, as expected. It was only the case of a
3036 non-existent file that was broken.
3038 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3039 with a change in libdomainkeys > 0.64.
3041 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3042 from DNS. If the selector record carries the flag, it now has
3043 precedence over the domain-wide flag.
3045 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3047 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3048 the use of a port name that isn't defined in /etc/services) occurred, the
3049 message was deferred as in a normal delivery, and thus remained on the
3050 spool, instead of being failed because of the mua_wrapper setting. This
3051 is now fixed, and I tidied up some of the mua_wrapper messages at the
3054 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3055 the messages in a hash of arrays rather than using individual hashes.
3056 This is a bit cleaner and results in dramatic memory savings, albeit
3057 at a slight CPU cost.
3059 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3060 as requested by Marc Sherman.
3062 SC/10 Eximstats: added histograms for user specified patterns as requested
3065 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3067 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3068 fopen() in the content-scanning modules that did not already have it.
3074 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3075 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3077 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3079 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3081 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3082 to test Sieve filters that use "vacation".
3084 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3085 that changes the way the GnuTLS parameters are stored in the cache file.
3086 The new format can be generated externally. For backward compatibility,
3087 if the data in the cache doesn't make sense, Exim assumes it has read an
3088 old-format file, and it generates new data and writes a new file. This
3089 means that you can't go back to an older release without removing the
3092 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3093 work if there are any delivery delays because "one_time" forces the
3094 parent to be marked "delivered", so its unseen clone is never tried
3095 again. For this reason, Exim now forbids the simultaneous setting of
3098 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3099 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3100 entry if you want to know the details. The fix, however, neglected to
3101 consider the case when local delivery batching is involved. The test for
3102 "previously delivered" was not happening when checking to see if an
3103 address could be batched with a previous (undelivered) one; under
3104 certain circumstances this could lead to multiple deliveries to the same
3107 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3108 in its include files, and this causes problems building Exim.
3110 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3111 header_syntax) but Exim was just ignoring anything given after a slash.
3112 In particular, this caused confusion with an attempt to use "verify =
3113 reverse_host_lookup/defer_ok". An error is now given when options are
3114 supplied for verify items that do not have them. (Maybe reverse_host_
3115 lookup should have a defer_ok option, but that's a different point.)
3117 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3118 defined by RFC 821) to 2048, because there were problems with some AUTH
3119 commands, and RFC 1869 says the size should be increased for extended
3120 SMTP commands that take arguments.
3122 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3125 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3126 "unknown" error; now it says that the functionality isn't in the binary.
3128 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3129 an address' error message when a string expansion fails (syntax or
3130 whatever). Otherwise the password may appear in the log. Following change
3131 PH/42 below, there is no longer a chance of it appearing in a bounce
3134 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3136 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3137 of its addresses was ever tried. (Bugzilla bug #2.)
3139 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3140 the result incorrectly in the debug output. (It correctly added a newline
3141 to what was transported.)
3143 TF/01 Added $received_time.
3145 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3146 commented out examples of how to interface to a virus scanner and to
3147 SpamAssassin. Also added commented examples of av_scanner and
3148 spamd_address settings.
3150 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3151 and controls are allowed in which ACLs. There were a couple of minor
3152 errors. Some of the entries in the conditions table (which is a table of
3153 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3154 negation of where the condition IS allowed.
3156 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3158 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3159 header file does not have a version number, so I've had to invent a new
3160 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3161 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3162 radiusclient), but it was contributed by a Radius user.
3164 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3165 files or directories, for OpenSSL.
3167 PH/19 When an Exim process that is running as root has to create an Exim log
3168 file, it does so in a subprocess that runs as exim:exim so as to get the
3169 ownership right at creation (otherwise, other Exim processes might see
3170 the file with the wrong ownership). There was no test for failure of this
3171 fork() call, which would lead to the process getting stuck as it waited
3172 for a non-existent subprocess. Forks do occasionally fail when resources
3173 run out. I reviewed all the other calls to fork(); they all seem to check
3176 PH/20 When checking for unexpected SMTP input at connect time (before writing
3177 the banner), Exim was not dealing correctly with a non-positive return
3178 from the read() function. If the client had disconnected by this time,
3179 the result was a log entry for a synchronization error with an empty
3180 string after "input=" when read() returned zero. If read() returned -1
3181 (an event I could not check), uninitialized data bytes were printed.
3182 There were reports of junk text (parts of files, etc) appearing after
3185 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3187 PH/22 Added support for macro redefinition, and (re)definition in between
3188 driver and ACL definitions.
3190 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3191 forgetting to use the resulting value; it was using the unexpanded value.
3193 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3194 hadn't been configured. The fix is from Juergen Kreileder, who
3195 understands it better than I do:
3197 "Here's what I see happening with three configured cyrus_sasl
3198 authenticators configured (plain, login, cram-md5):
3200 On startup auth_cyrus_sasl_init() gets called for each of these.
3201 This means three calls to sasl_listmech() without a specified mech_list.
3202 => SASL tests which mechs of all available mechs actually work
3203 => three warnings about OTP not working
3204 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3206 With the patch, sasl_listmech() also gets called three times. But now
3207 SASL's mech_list option is set to the server_mech specified in the the
3208 authenticator. Or in other words, the answer from sasl_listmech()
3209 gets limited to just the mech you're testing for (which is different
3211 => the return list contains just 'plain' or 'login', 'cram-md5' or
3212 nothing depending on the value of ob->server_mech.
3214 I've just tested the patch: Authentication still works fine,
3215 unavailable mechs specified in the exim configuration are still
3216 caught, and the auth.log warnings about OTP are gone."
3218 PH/25 When debugging is enabled, the contents of the command line are added
3219 to the debugging output, even when log_selector=+arguments is not
3222 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3223 answer is "GNU", and only if the return is "GNU/something" is the answer
3226 PH/27 $acl_verify_message is now set immediately after the failure of a
3227 verification in an ACL, and so is available in subsequent modifiers. In
3228 particular, the message can be preserved by coding like this:
3230 warn !verify = sender
3231 set acl_m0 = $acl_verify_message
3233 Previously, $acl_verify_message was set only while expanding "message"
3234 and "log_message" when a very denied access.
3236 PH/28 Modified OS/os.c-Linux with
3238 -#ifndef OS_LOAD_AVERAGE
3239 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3241 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3242 nomenclature these days.)
3244 PH/29 Installed patch from the Sieve maintainer that adds the options
3245 sieve_useraddress and sieve_subaddress to the redirect router.
3247 PH/30 In these circumstances:
3248 . Two addresses routed to the same list of hosts;
3249 . First host does not offer TLS;
3250 . First host accepts first address;
3251 . First host gives temporary error to second address;
3252 . Second host offers TLS and a TLS session is established;
3253 . Second host accepts second address.
3254 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3255 and peerdn, if requested) that were in fact used only for the second
3258 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3259 attention to any local part prefix or suffix that was matched by the
3260 router that accepted the address. It now behaves in the same way as it
3261 does for delivery: the affixes are removed from the local part unless
3262 rcpt_include_affixes is set on the transport.
3264 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3265 timeout during the DATA phase of an incoming message.
3267 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3268 applied a patch sanctioned by the Sieve maintainer.
3270 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3271 the uid or gid is negative. A case of a negative gid caused this to be
3272 noticed. The fix allows for either to be negative.
3274 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3275 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3278 PH/36 The change PH/12 above was broken. Fixed it.
3280 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3281 the grounds that routing the same address twice would always produce the
3282 same answer. This might have been true once, but it is certainly no
3283 longer true now. Routing a child address may depend on the previous
3284 routing that produced that child. Some complicated redirection strategies
3285 went wrong when messages had multiple recipients, and made Exim's
3286 behaviour dependent on the order in which the addresses were given.
3288 I have moved the duplicate checking until after the routing is complete.
3289 Exim scans the addresses that are assigned to local and remote
3290 transports, and removes any duplicates. This means that more work will be
3291 done, as duplicates will always all be routed, but duplicates are
3292 presumably rare, so I don't expect this is of any significance.
3294 For deliveries to pipes, files, and autoreplies, the duplicate checking
3295 still happens during the routing process, since they are not going to be
3298 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3299 It corrects a timeout issue with spamd. This is Ian's comment: "The
3300 background is that sometimes spamd either never reads data from a
3301 connection it has accepted, or it never writes response data. The exiscan
3302 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3303 blindly assumes that writes won't block so it may never time out."
3305 PH/39 Allow G after quota size as well as K and M.
3307 PH/40 The value set for $authenticated_id in an authenticator may not contain
3308 binary zeroes or newlines because the value is written to log lines and
3309 to spool files. There was no check on this. Now the value is run through
3310 the string_printing() function so that such characters are converted to
3311 printable escape sequences.
3313 PH/41 $message_linecount is a new variable that contains the total number of
3314 lines in the message. Compare $body_linecount, which is the count for the
3317 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3318 bounce and delay warning messages, except in certain special cases, which
3321 (a) An SMTP error message from a remote host;
3322 (b) A message specified in a :fail: redirection;
3323 (c) A message specified in a "fail" command in a system filter;
3324 (d) A message specified in a FAIL return from the queryprogram router;
3325 (e) A message specified by the cannot_route_message router option.
3327 In these cases only, Exim does include the error details in bounce and
3328 warning messages. There are also a few cases where bland messages such
3329 as "unrouteable address" or "local delivery error" are given.
3331 PH/43 $value is now also set for the "else" part of a ${run expansion.
3333 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3334 being worked on, but at least Exim now implements the latest version to
3337 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3338 process to complete was treated as a delivery failure, a timeout while
3339 writing the message to the pipe was logged, but erroneously treated as a
3340 successful delivery. Such timeouts include transport filter timeouts. For
3341 consistency with the overall process timeout, these timeouts are now
3342 treated as errors, giving rise to delivery failures by default. However,
3343 there is now a new Boolean option for the pipe transport called
3344 timeout_defer, which, if set TRUE, converts the failures into defers for
3345 both kinds of timeout. A transport filter timeout is now identified in
3348 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3349 systems where "make" and "gmake" are different, calling "gmake" at top
3350 level broke things. I've arranged for the value of $(MAKE) to be passed
3351 from the Makefile to this script so that it can call the same version of
3355 A note about Exim versions 4.44 and 4.50
3356 ----------------------------------------
3358 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3359 changes of various kinds. As a consequence, a big documentation update was
3360 needed. This delayed the release for rather longer than seemed good, especially
3361 in the light of a couple of (minor) security issues. Therefore, the changes
3362 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3363 release. So 4.44 and 4.50 are in effect two different branches that both start
3366 I have left the 4.50 change log unchanged; it contains all the changes since
3367 4.43. The change log for 4.44 is below; many of its items are identical to
3368 those for 4.50. This seems to be the most sensible way to preserve the
3369 historical information.
3375 1. Minor wording change to the doc/README.SIEVE file.
3377 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3378 computation of the current number of files was incorrect.
3380 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3381 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3384 4. Give more explanation in the error message when the command for a transport
3385 filter fails to execute.
3387 5. There are several places where Exim runs a non-Exim command in a
3388 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3389 was being done only for the command run by the queryprogram router. It is
3390 now done for all such subprocesses. The other cases are: ${run, transport
3391 filters, and the commands run by the lmtp and pipe transports.
3393 6. Added CONFIGURE_GROUP build-time option.
3395 7. Some older OS have a limit of 256 on the maximum number of file
3396 descriptors. Exim was using setrlimit() to set 1000 as a large value
3397 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3398 systems. I've change it so that if it can't get 1000, it tries for 256.
3400 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3401 was an oversight, and furthermore, ever since the addition of extra
3402 controls (e.g. 4.43/32), the checks on when to allow different forms of
3403 "control" were broken. There should now be diagnostics for all cases when a
3404 control that does not make sense is encountered.
3406 9. Added the /retain_sender option to "control=submission".
3408 10. $recipients is now available in the predata ACL (oversight).
3410 11. Tidy the search cache before the fork to do a delivery from a message
3411 received from the command line. Otherwise the child will trigger a lookup
3412 failure and thereby defer the delivery if it tries to use (for example) a
3413 cached ldap connection that the parent has called unbind on.
3415 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3416 of $address_data from the recipient verification was clobbered by the
3417 sender verification.
3419 13. The value of address_data from a sender verification is now available in
3420 $sender_address_data in subsequent conditions in the ACL statement.
3422 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3424 15. Added a new option "connect=<time>" to callout options, to set a different
3427 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3428 was its contents. (It was OK if the option was not defined at all.)
3430 17. A "Completed" log line is now written for messages that are removed from
3431 the spool by the -Mrm option.
3433 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3434 information about exactly what failed.
3436 19. Added -dd to debug only the daemon process.
3438 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3439 handles timeouts, both on the server side and network timeouts. Renamed the
3440 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3442 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3443 It is now set to "smtps".
3445 22. $host_address is now set to the target address during the checking of
3446 ignore_target_hosts.
3448 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3449 being passed; this would have caused $sender_host_name to have been used if
3450 matching the list had actually called for a host name (not very likely,
3451 since this list is usually IP addresses). A host name is now passed as
3454 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3455 code that specifies a non-restarting handler (typically sigaction() in
3456 modern systems) in an attempt to fix a rare and obscure crash bug.
3458 25. Narrowed the window for a race in the daemon that could cause it to ignore
3459 SIGCHLD signals. This is not a major problem, because they are used only to
3460 wake it up if nothing else does.
3462 26. A malformed maildirsize file could cause Exim to calculate negative values
3463 for the mailbox size or file count. Odd effects could occur as a result.
3464 The maildirsize information is now recalculated if the size or filecount
3467 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3468 support for a long time. Removed HAVE_SYS_VFS_H.
3470 28. Installed the latest version of exipick from John Jetmore.
3472 29. In an address list, if the pattern was not a regular expression, an empty
3473 subject address (from a bounce message) matched only if the pattern was an
3474 empty string. Non-empty patterns were not even tested. This was the wrong
3475 because it is perfectly reasonable to use an empty address as part of a
3476 database query. An empty address is now tested by patterns that are
3477 lookups. However, all the other forms of pattern expect the subject to
3478 contain a local part and a domain, and therefore, for them, an empty
3479 address still always fails if the pattern is not itself empty.
3481 30. Exim went into a mad DNS loop when attempting to do a callout where the
3482 host was specified on an smtp transport, and looking it up yielded more
3483 than one IP address.
3485 31. Re-factored the code for checking spool and log partition space into a
3486 function that finds that data and another that does the check. The former
3487 is then used to implement four new variables: $spool_space, $log_space,
3488 $spool_inodes, and $log_inodes.
3490 32. The RFC2047 encoding function was originally intended for short strings
3491 such as real names; it was not keeping to the 75-character limit for
3492 encoded words that the RFC imposes. It now respects the limit, and
3493 generates multiple encoded words if necessary. To be on the safe side, I
3494 have increased the buffer size for the ${rfc2047: expansion operator from
3497 33. It is now permitted to omit both strings after an "if" condition; if the
3498 condition is true, the result is "true". As before, when the second string
3499 is omitted, a false condition yields an empty string. This makes it less
3500 cumbersome to write custom ACL and router conditions.
3502 34. Failure to deliver a bounce message always caused it to be frozen, even if
3503 there was an errors_to setting on the router. The errors_to setting is now
3506 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3507 canonical form (fully expanded) before being placed in
3508 $sender_host_address.
3510 36. The table in the code that translates DNS record types into text (T_A to
3511 "A" for instance) was missing entries for NS and CNAME. It is just possible
3512 that this could have caused confusion if both these types were looked up
3513 for the same domain, because the text type is used as part of Exim's
3514 per-process caching. But the chance of anyone hitting this buglet seems
3517 37. The dnsdb lookup has been extended in a number of ways.
3519 (1) There is a new type, "zns", which walks up the domain tree until it
3520 finds some nameserver records. It should be used with care.
3522 (2) There is a new type, "mxh", which is like "mx" except that it returns
3523 just the host names, not the priorities.
3525 (3) It is now possible to give a list of domains (or IP addresses) to be
3526 looked up. The behaviour when one of the lookups defers can be
3527 controlled by a keyword.
3529 (4) It is now possible to specify the separator character for use when
3530 multiple records are returned.
3532 38. The dnslists ACL condition has been extended: it is now possible to supply
3533 a list of IP addresses and/or domains to be looked up in a particular DNS
3536 39. Added log_selector=+queue_time_overall.
3538 40. When running the queue in the test harness, wait just a tad after forking a
3539 delivery process, to get repeatability of debugging output.
3541 41. Include certificate and key file names in error message when GnuTLS fails
3542 to set them up, because the GnuTLS error message doesn't include the name
3543 of the failing file when there is a problem reading it.
3545 42. Allow both -bf and -bF in the same test run.
3547 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3549 44. The "Exiscan patch" is now merged into the mainline Exim source.
3551 45. Sometimes the final signoff response after QUIT could fail to get
3552 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3553 < 0 before doing a fflush(). This bug looks as though it goes back to the
3554 introduction of TLS in release 3.20, but "sometimes" must have been rare
3555 because the tests only now provoked it.
3557 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3558 (this can affect the format of dates).
3560 47. exim_tidydb, when checking for the continued existence of a message for
3561 which it has found a message-specific retry record, was not finding
3562 messages that were in split spool directories. Consequently, it was
3563 deleting retry records that should have stayed in existence.
3565 48. Steve fixed some bugs in eximstats.
3567 49. The SPA authentication driver was not abandoning authentication and moving
3568 on to the next authenticator when an expansion was forced to fail,
3569 contradicting the general specification for all authenticators. Instead it
3570 was generating a temporary error. It now behaves as specified.
3572 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3573 (the order specifies the preference for clients). The order is now AES256,
3574 AES128, 3DES, ARCFOUR128.
3576 51. Small patch to Sieve code - explicitly set From: when generating an
3579 52. Exim crashed if a remote delivery caused a very long error message to be
3580 recorded - for instance if somebody sent an entire SpamAssassin report back
3581 as a large number of 550 error lines. This bug was coincidentally fixed by
3582 increasing the size of one of Exim's internal buffers (big_buffer) that
3583 happened as part of the Exiscan merge. However, to be on the safe side, I
3584 have made the code more robust (and fixed the comments that describe what
3587 53. Now that there can be additional text after "Completed" in log lines (if
3588 the queue_time_overall log selector is set), a one-byte patch to exigrep
3589 was needed to allow it to recognize "Completed" as not the last thing in
3592 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3593 patch that reportedly fixes this has been added. I am not expert enough to
3594 create a test for it. This is what the patch creator wrote:
3596 "I found a little strange behaviour of ldap code when working with
3597 Windows 2003 AD Domain, where users was placed in more than one
3598 Organization Units. When I tried to give exim partial DN, the exit code
3599 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3600 But simultaneously result of request was absolutely normal ldap result,
3601 so I produce this patch..."
3603 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3604 REFERENCE, so I have modified the code to exclude the patch when that macro
3607 55. Some experimental protocols are using DNS PTR records for new purposes. The
3608 keys for these records are domain names, not reversed IP addresses. The
3609 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3610 leaves it alone. Component reversal etc. now happens only for IP addresses.
3613 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3615 57. Double the size of the debug message buffer (to 2048) so that more of very
3616 long debug lines gets shown.
3618 58. The exicyclog utility now does better if the number of log files to keep
3619 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3621 59. Two changes related to the smtp_active_hostname option:
3623 (1) $smtp_active_hostname is now available as a variable.
3624 (2) The default for smtp_banner uses $smtp_active_hostname instead
3625 of $primary_hostname.
3627 60. The host_aton() function is supposed to be passed a string that is known
3628 to be a valid IP address. However, in the case of IPv6 addresses, it was
3629 not checking this. This is a hostage to fortune. Exim now panics and dies
3630 if the condition is not met. A case was found where this could be provoked
3631 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3632 components; fortuitously, this particular loophole had already been fixed
3633 by change 4.50/55 above.
3635 If there are any other similar loopholes, the new check in host_aton()
3636 itself should stop them being exploited. The report I received stated that
3637 data on the command line could provoke the exploit when Exim was running as
3638 exim, but did not say which command line option was involved. All I could
3639 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3640 running as the user.
3643 61. There was a buffer overflow vulnerability in the SPA authentication code
3644 (which came originally from the Samba project). I have added a test to the
3645 spa_base64_to_bits() function which I hope fixes it.
3648 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3649 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3651 63. The daemon start-up calls getloadavg() while still root for those OS that
3652 need the first call to be done as root, but it missed one case: when
3653 deliver_queue_load_max is set with deliver_drop_privilege. This is
3654 necessary for the benefit of the queue runner, because there is no re-exec
3655 when deliver_drop_privilege is set.
3657 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3658 This has been fixed.
3660 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3661 was in use, was not putting the data itself into the right store pool;
3662 consequently, it could be overwritten for a subsequent message in the same
3663 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3666 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3668 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3669 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3670 the function were treating the return as a boolean value, which happened to
3671 work because 0=false and not-0=true, but is not correct code.
3673 68. The host_aton() function was not handling scoped IPv6 addresses (those
3674 with, for example, "%eth0" on the end) correctly.
3676 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3677 negated items (that is, ~something) in unsigned ints. Some compilers
3678 apparently mutter when there is no cast.
3680 70. If an address verification called from an ACL failed, and did not produce a
3681 user-specific message (i.e. there was only a "system" message), nothing was
3682 put in $acl_verify_message. In this situation, it now puts the system
3685 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3686 session; change 4.31/43 added the unwanted input to the log line - except
3687 that it did not do this in the start of session case. It now does.
3689 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3690 This is wrong and can cause the other end to generate a synchronization
3691 error if it is another Exim or anything else that does the synchronization
3692 check. A QUIT command is no longer sent after a timeout.
3694 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3695 during host lookups.
3697 74. The defer_ok option of callout verification was not working if it was used
3698 when verifying addresses in header lines, that is, for this case:
3700 verify = header_sender/callout=defer_ok
3702 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3703 those file descriptors could be used for SMTP connections. If anything
3704 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3705 could be sent to the SMTP client, causing chaos. The daemon now opens
3706 stdin, stdout, and stderr to /dev/null when it puts itself into the
3709 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3710 log by default. The user can override this with suitable Perl magic.
3712 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3713 the log message when discard triggers, was not working for the DATA ACL or
3714 for the non-SMTP ACL.
3716 78. Error message wording change in sieve.c.
3718 79. If smtp_accept_max_per_host was set, the number of connections could be
3719 restricted to fewer than expected, because the daemon was trying to set up
3720 a new connection before checking whether the processes handling previous
3721 connections had finished. The check for completed processes is now done
3722 earlier. On busy systems, this bug wouldn't be noticed because something
3723 else would have woken the daemon, and it would have reaped the completed
3726 80. If a message was submitted locally by a user whose login name contained one
3727 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3728 It caused a spool format error. I have fixed the spool reading code. A
3729 related problem was that the "from" clause in the Received: line became
3730 illegal because of the space(s). It is now covered by ${quote_local_part.
3732 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3735 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3737 83. Patch from Sieve maintainer to fix unterminated string problem in
3738 "vacation" handling.
3740 84. Some minor changes to the Linux configuration files to help with other
3741 OS variants using glibc.
3743 85. One more patch for Sieve to update vacation handling to latest spec.
3746 ----------------------------------------------------
3747 See the note above about the 4.44 and 4.50 releases.
3748 ----------------------------------------------------
3754 1. Change 4.43/35 introduced a bug that caused file counts to be
3755 incorrectly computed when quota_filecount was set in an appendfile
3758 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3759 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3762 3. Give more explanation in the error message when the command for a transport
3763 filter fails to execute.
3765 4. There are several places where Exim runs a non-Exim command in a
3766 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3767 was being done only for the command run by the queryprogram router. It is
3768 now done for all such subprocesses. The other cases are: ${run, transport
3769 filters, and the commands run by the lmtp and pipe transports.
3771 5. Some older OS have a limit of 256 on the maximum number of file
3772 descriptors. Exim was using setrlimit() to set 1000 as a large value
3773 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3774 systems. I've change it so that if it can't get 1000, it tries for 256.
3776 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3777 was an oversight, and furthermore, ever since the addition of extra
3778 controls (e.g. 4.43/32), the checks on when to allow different forms of
3779 "control" were broken. There should now be diagnostics for all cases when a
3780 control that does not make sense is encountered.
3782 7. $recipients is now available in the predata ACL (oversight).
3784 8. Tidy the search cache before the fork to do a delivery from a message
3785 received from the command line. Otherwise the child will trigger a lookup
3786 failure and thereby defer the delivery if it tries to use (for example) a
3787 cached ldap connection that the parent has called unbind on.
3789 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3790 of $address_data from the recipient verification was clobbered by the
3791 sender verification.
3793 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3794 was its contents. (It was OK if the option was not defined at all.)
3796 11. A "Completed" log line is now written for messages that are removed from
3797 the spool by the -Mrm option.
3799 12. $host_address is now set to the target address during the checking of
3800 ignore_target_hosts.
3802 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3803 being passed; this would have caused $sender_host_name to have been used if
3804 matching the list had actually called for a host name (not very likely,
3805 since this list is usually IP addresses). A host name is now passed as
3808 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3809 code that specifies a non-restarting handler (typically sigaction() in
3810 modern systems) in an attempt to fix a rare and obscure crash bug.
3812 15. Narrowed the window for a race in the daemon that could cause it to ignore
3813 SIGCHLD signals. This is not a major problem, because they are used only to
3814 wake it up if nothing else does.
3816 16. A malformed maildirsize file could cause Exim to calculate negative values
3817 for the mailbox size or file count. Odd effects could occur as a result.
3818 The maildirsize information is now recalculated if the size or filecount
3821 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3822 support for a long time. Removed HAVE_SYS_VFS_H.
3824 18. Updated exipick to current release from John Jetmore.
3826 19. Allow an empty sender to be matched against a lookup in an address list.
3827 Previously the only cases considered were a regular expression, or an
3830 20. Exim went into a mad DNS lookup loop when doing a callout where the
3831 host was specified on the transport, if the DNS lookup yielded more than
3834 21. The RFC2047 encoding function was originally intended for short strings
3835 such as real names; it was not keeping to the 75-character limit for
3836 encoded words that the RFC imposes. It now respects the limit, and
3837 generates multiple encoded words if necessary. To be on the safe side, I
3838 have increased the buffer size for the ${rfc2047: expansion operator from
3841 22. Failure to deliver a bounce message always caused it to be frozen, even if
3842 there was an errors_to setting on the router. The errors_to setting is now
3845 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3846 canonical form (fully expanded) before being placed in
3847 $sender_host_address.
3849 24. Updated eximstats to version 1.33
3851 25. Include certificate and key file names in error message when GnuTLS fails
3852 to set them up, because the GnuTLS error message doesn't include the name
3853 of the failing file when there is a problem reading it.
3855 26. Expand error message when OpenSSL has problems setting up cert/key files.
3858 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3859 (this can affect the format of dates).
3861 28. exim_tidydb, when checking for the continued existence of a message for
3862 which it has found a message-specific retry record, was not finding
3863 messages that were in split spool directories. Consequently, it was
3864 deleting retry records that should have stayed in existence.
3866 29. eximstats updated to version 1.35
3867 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3868 1.35 - bugfix such that pie charts by volume are generated correctly
3870 30. The SPA authentication driver was not abandoning authentication and moving
3871 on to the next authenticator when an expansion was forced to fail,
3872 contradicting the general specification for all authenticators. Instead it
3873 was generating a temporary error. It now behaves as specified.
3875 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3876 (the order specifies the preference for clients). The order is now AES256,
3877 AES128, 3DES, ARCFOUR128.
3879 31. Small patch to Sieve code - explicitly set From: when generating an
3882 32. Exim crashed if a remote delivery caused a very long error message to be
3883 recorded - for instance if somebody sent an entire SpamAssassin report back
3884 as a large number of 550 error lines. This bug was coincidentally fixed by
3885 increasing the size of one of Exim's internal buffers (big_buffer) that
3886 happened as part of the Exiscan merge. However, to be on the safe side, I
3887 have made the code more robust (and fixed the comments that describe what
3890 33. Some experimental protocols are using DNS PTR records for new purposes. The
3891 keys for these records are domain names, not reversed IP addresses. The
3892 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3893 leaves it alone. Component reversal etc. now happens only for IP addresses.
3896 34. The host_aton() function is supposed to be passed a string that is known
3897 to be a valid IP address. However, in the case of IPv6 addresses, it was
3898 not checking this. This is a hostage to fortune. Exim now panics and dies
3899 if the condition is not met. A case was found where this could be provoked
3900 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3901 components; fortuitously, this particular loophole had already been fixed
3902 by change 4.50/55 or 4.44/33 above.
3904 If there are any other similar loopholes, the new check in host_aton()
3905 itself should stop them being exploited. The report I received stated that
3906 data on the command line could provoke the exploit when Exim was running as
3907 exim, but did not say which command line option was involved. All I could
3908 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3909 running as the user.
3912 35. There was a buffer overflow vulnerability in the SPA authentication code
3913 (which came originally from the Samba project). I have added a test to the
3914 spa_base64_to_bits() function which I hope fixes it.
3917 36. The daemon start-up calls getloadavg() while still root for those OS that
3918 need the first call to be done as root, but it missed one case: when
3919 deliver_queue_load_max is set with deliver_drop_privilege. This is
3920 necessary for the benefit of the queue runner, because there is no re-exec
3921 when deliver_drop_privilege is set.
3923 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3924 was in use, was not putting the data itself into the right store pool;
3925 consequently, it could be overwritten for a subsequent message in the same
3926 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3929 38. Sometimes the final signoff response after QUIT could fail to get
3930 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3931 < 0 before doing a fflush(). This bug looks as though it goes back to the
3932 introduction of TLS in release 3.20, but "sometimes" must have been rare
3933 because the tests only now provoked it.
3939 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3940 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3941 Now it returns an integer. A number of calls were still expecting a T/F
3942 return. Fortuitously, in all cases, the tests worked in OK situations,
3943 which is the norm. However, things would have gone wrong on any write
3944 failures on the smtp file descriptor. This function is used when sending
3945 messages over SMTP and also when doing verify callouts.
3947 2. When Exim is called to do synchronous delivery of a locally submitted
3948 message (the -odf or -odi options), it no longer closes stderr before doing
3951 3. Implemented the mua_wrapper option.
3953 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3955 5. Implemented the functions header_remove(), header_testname(),
3956 header_add_at_position(), and receive_remove_recipient(), and exported them
3959 6. If an ACL "warn" statement specified the addition of headers, Exim already
3960 inserted X-ACL-Warn: at the start if there was no header name. However, it
3961 was not making this test for the second and subsequent header lines if
3962 there were newlines in the string. This meant that an invalid header could
3963 be inserted if Exim was badly configured.
3965 7. Allow an ACL "warn" statement to add header lines at the start or after all
3966 the Received: headers, as well as at the end.
3968 8. Added the rcpt_4xx retry error code.
3970 9. Added postmaster_mailfrom=xxx to callout verification option.
3972 10. Added mailfrom=xxxx to the callout verification option, for verify=
3975 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3976 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3978 12. Inserted some casts to stop certain compilers warning when using pointer
3979 differences as field lengths or precisions in printf-type calls (mostly
3980 affecting debugging statements).
3982 13. Added optional readline() support for -be (dynamically loaded).
3984 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3985 same clock tick as a message's arrival, so that its received time was the
3986 same as the "first fail" time on the retry record, and that message
3987 remained on the queue past the ultimate address timeout, every queue runner
3988 would try a delivery (because it was past the ultimate address timeout) but
3989 after another failure, the ultimate address timeout, which should have then
3990 bounced the address, did not kick in. This was a "< instead of <=" error;
3991 in most cases the first failure would have been in the next clock tick
3992 after the received time, and all would be well.
3994 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3995 being recognized when the domain list was tested by the match_domain
3996 condition in an expansion string.
3998 16. Added the ${str2b64: operator.
4000 17. Exim was always calling setrlimit() to set a large limit for the number of
4001 processes, without checking whether the existing limit was already
4002 adequate. (It did check for the limit on file descriptors.) Furthermore,
4003 errors from getrlimit() and setrlimit() were being ignored. Now they are
4004 logged to the main and panic logs, but Exim does carry on, to try to do its
4005 job under whatever limits there are.
4007 18. Imported PCRE 5.0.
4009 19. Trivial typo in log message " temporarily refused connection" (the leading
4012 20. If the log selector return_path_on_delivery was set and an address was
4013 redirected to /dev/null, the delivery process crashed because it assumed
4014 that a return path would always be set for a "successful" delivery. In this
4015 case, the whole delivery is bypassed as an optimization, and therefore no
4018 21. Internal re-arrangement: the function for sending a challenge and reading
4019 a response while authentication was assuming a zero-terminated challenge
4020 string. It's now changed to take a pointer and a length, to allow for
4021 binary data in such strings.
4023 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4025 23. Exim was not respecting finduser_retries when seeking the login of the
4026 uid under which it was called; it was always trying 10 times. (The default
4027 setting of finduser_retries is zero.) Also, it was sleeping after the final
4028 failure, which is pointless.
4030 24. Implemented tls_on_connect_ports.
4032 25. Implemented acl_smtp_predata.
4034 26. If the domain in control=submission is set empty, Exim assumes that the
4035 authenticated id is a complete email address when it generates From: or
4036 Sender: header lines.
4038 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4039 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4040 chown and chgrp in /bin and hostname in /usr/bin.
4042 28. Exim was keeping the "process log" file open after each use, just as it
4043 does for the main log. This opens the possibility of it remaining open for
4044 long periods when the USR1 signal hits a daemon. Occasional processlog
4045 errors were reported, that could have been caused by this. Anyway, it seems
4046 much more sensible not to leave this file open at all, so that is what now
4049 29. The long-running daemon process does not normally write to the log once it
4050 has entered its main loop, and it closes the log before doing so. This is
4051 so that log files can straightforwardly be renamed and moved. However,
4052 there are a couple of unusual error situations where the daemon does write
4053 log entries, and I had neglected to close the log afterwards.
4055 30. The text of an SMTP error response that was received during a remote
4056 delivery was being truncated at 512 bytes. This is too short for some of
4057 the long messages that one sometimes sees. I've increased the limit to
4060 31. It is now possible to make retry rules that apply only when a message has a
4061 specific sender, in particular, an empty sender.
4063 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4064 it possible to be selective about when SMTP synchronization is enforced.
4066 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4068 32. Implemented hosts_connection_nolog.
4070 33. Added an ACL for QUIT.
4072 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4075 35. Added mailbox_size and mailbox_filecount to appendfile.
4077 36. Added control = no_multiline_responses to ACLs.
4079 37. There was a bug in the logic of the code that waits for the clock to tick
4080 in the case where the clock went backwards by a substantial amount such
4081 that the microsecond fraction of "now" was more than the microsecond
4082 fraction of "then" (but the whole seconds number was less).
4084 38. Added support for the libradius Radius client library this is found on
4085 FreeBSD (previously only the radiusclient library was supported).
4091 1. When certain lookups returned multiple values in the form name=value, the
4092 quoting of the values was not always being done properly. Specifically:
4093 (a) If the value started with a double quote, but contained no whitespace,
4095 (b) If the value contained whitespace other than a space character (i.e.
4096 tabs or newlines or carriage returns) it was not quoted.
4097 This fix has been applied to the mysql and pgsql lookups by writing a
4098 separate quoting function and calling it from the lookup code. The fix
4099 should probably also be applied to nisplus, ibase and oracle lookups, but
4100 since I cannot test any of those, I have not disturbed their existing code.
4102 2. A hit in the callout cache for a specific address caused a log line with no
4103 reason for rejecting RCPT. Now it says "Previous (cached) callout
4104 verification failure".
4106 3. There was an off-by-one bug in the queryprogram router. An over-long
4107 return line was truncated at 256 instead of 255 characters, thereby
4108 overflowing its buffer with the terminating zero. As well as fixing this, I
4109 have increased the buffer size to 1024 (and made a note to document this).
4111 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4112 when Exim is waiting for an SMTP response from a remote server, Exim
4113 restarts its select() call on the socket, thereby resetting its timeout.
4114 This is not a problem when such interrupts are rare. Somebody set up a cron
4115 job to run exiwhat every 2 minutes, which is less than the normal select()
4116 timeout (5 or 10 minutes). This meant that the select() timeout never
4117 kicked in because it was always reset. I have fixed this by comparing the
4118 time when an interrupt arrives with the time at the start of the first call
4119 to select(). If more time than the timeout has elapsed, the interrupt is
4120 treated as a timeout.
4122 5. Some internal re-factoring in preparation for the addition of Sieve
4123 extensions (by MH). In particular, the "personal" test is moved to a
4124 separate function, and given an option for scanning Cc: and Bcc: (which is
4125 not set for Exim filters).
4127 6. When Exim created an email address using the login of the caller as the
4128 local part (e.g. when creating a From: or Sender: header line), it was not
4129 quoting the local part when it contained special characters such as @.
4131 7. Installed new OpenBSD configuration files.
4133 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4134 try to make them clearer.
4136 9. Callout options, other than the timeout value, were being ignored when
4137 verifying sender addresses in header lines. For example, when using
4139 verify = header_sender/callout=no_cache
4141 the cache was (incorrectly) being used.
4143 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4144 only the Cygwin environment.
4146 11. When return_path_on_delivery was set as a log selector, if different remote
4147 addresses in the same message used different return paths and parallel
4148 remote delivery occurred, the wrong values would sometimes be logged.
4149 (Whenever a remote delivery process finished, the return path value from
4150 the most recently started remote delivery process was logged.)
4152 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4153 lines when AUTH and/or TLS are in use. This is the "received protocol"
4154 field. Exim used to use "asmtp" for authenticated SMTP, without any
4155 indication (in the protocol name) for TLS use. Now it follows the RFC and
4156 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4157 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4158 names appear in log lines as well as in Received: header lines.
4160 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4161 extensions, and comparison tests, and to fix some bugs.
4163 14. Changes to the "personal" filter test:
4165 (1) The test was buggy in that it was just doing the equivalent of
4166 "contains" tests on header lines. For example, if a user's address was
4167 anne@some.where, the "personal" test would incorrectly be true for
4169 To: susanne@some.where
4171 This test is now done by extracting each address from the header in turn,
4172 and checking the entire address. Other tests that are part of "personal"
4173 are now done using regular expressions (for example, to check local parts
4174 of addresses in From: header lines).
4176 (2) The list of non-personal local parts in From: addresses has been
4177 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4178 taken from the Sieve specification recommendations.
4180 (3) If the message contains any header line starting with "List-" it is
4181 treated as non-personal.
4183 (4) The test for "circular" in the Subject: header line has been removed
4184 because it now seems ill-conceived.
4186 15. Minor typos in src/EDITME comments corrected.
4188 16. Installed latest exipick from John Jetmore.
4190 17. If headers_add on a router specified a text string that was too long for
4191 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4192 of string_sprintf() is now avoided.
4194 18. $message_body_size was not set (it was always zero) when running the DATA
4195 ACL and the local_scan() function.
4197 19. For the "mail" command in an Exim filter, no default was being set for
4198 the once_repeat time, causing a random time value to be used if "once" was
4199 specified. (If the value happened to be <= 0, no repeat happened.) The
4200 default is now 0s, meaning "never repeat". The "vacation" command was OK
4201 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4202 (I found it when inspecting the code).
4204 20. There is now an overall timeout for performing a callout verification. It
4205 defaults to 4 times the callout timeout, which applies to individual SMTP
4206 commands during the callout. The overall timeout applies when there is more
4207 than one host that can be tried. The timeout is checked before trying the
4208 next host. This prevents very long delays if there are a large number of
4209 hosts and all are timing out (e.g. when the network connections are timing
4210 out). The value of the overall timeout can be changed by specifying an
4211 additional sub-option for "callout", called "maxwait". For example:
4213 verify = sender/callout=5s,maxwait=20s
4215 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4216 to the end before writing, but this should make it even safer).
4218 22. Exim was forgetting that it had advertised PIPELINING for the second and
4219 subsequent messages on an SMTP connection. It was also not resetting its
4220 memory on STARTTLS and an internal HELO.
4222 23. When Exim logs an SMTP synchronization error within a session, it now
4223 records whether PIPELINING has been advertised or not.
4225 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4226 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4227 rather than long int.
4229 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4231 26. Added the never_mail option to autoreply.
4237 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4238 crash if the getsockname() call failed; this can happen if a connection is
4239 closed very soon after it is established. The problem was simply in the
4240 order in which certain operations were done, causing Exim to try to write
4241 to the SMTP stream before it had set up the file descriptor. The bug has
4242 been fixed by making things happen in the correct order.
4248 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4249 before the connection was closed, thus losing the rejection response.
4251 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4252 some early Solaris releases, but causes trouble in current releases where
4253 socklen_t is defined.
4255 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4258 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4261 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4262 that releasing the top of it at the end releases what was used for sub-
4263 expansions (unless the block got too big). However, discard this block if
4264 the first thing is a variable or header, so that we can use its block when
4265 it is dynamic (useful for very large $message_headers, for example).
4267 6. Lookups now cache *every* query, not just the most recent. A new, separate
4268 store pool is used for this. It can be recovered when all lookup caches are
4269 flushed. Lookups now release memory at the end of their result strings.
4270 This has involved some general refactoring of the lookup sources.
4272 7. Some code has been added to the store_xxx() functions to reduce the amount
4273 of flapping under certain conditions.
4275 8. log_incoming_interface used to affect only the <= reception log lines. Now
4276 it causes the local interface and port to be added to several more SMTP log
4277 lines, for example "SMTP connection from", and rejection lines.
4279 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4281 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4283 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4284 could be overwritten at the end of the current message (or the start of a
4285 new message if it was set in a HELO ACL). The value is now preserved for
4286 the duration of the SMTP connection.
4288 12. If a transport had a headers_rewrite setting, and a matching header line
4289 contained an unqualified address, that address was qualified, even if it
4290 did not match any rewriting rules. The underlying bug was that the values
4291 of the flags that permit the existence of unqualified sender and recipient
4292 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4293 non-local messages, and by -bnq for local messages) were not being
4294 preserved with the message after it was received.
4296 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4297 "next input=" as part of the text comprising the host identity instead of
4298 the correct text. The code was using the same buffer for two different
4299 strings. However, depending on which order the printing function evaluated
4300 its arguments, the bug did not always show up. Under Linux, for example, my
4301 test suite worked just fine.
4303 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4304 doesn't work with some older versions of Perl. It has been changed to "my",
4305 which in any case is probably the better facility to use.
4307 15. A really picky compiler found some instances of statements for creating
4308 error messages that either had too many or two few arguments for the format
4311 16. The size of the buffer for calls to the DNS resolver has been increased
4312 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4313 for addresses that have a lot of PTR records. This alleviates a problem; it
4314 does not fully solve it.
4316 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4317 buffer now truncates the list and logs the incident, which is the same
4318 action as happens when Exim is looking up a host name and its aliases.
4319 Previously in this situation something unpredictable would happen;
4320 sometimes it was "internal error: store_reset failed".
4322 18. If a server dropped the connection unexpectedly when an Exim client was
4323 using GnuTLS and trying to read a response, the client delivery process
4324 crashed while trying to generate an error log message.
4326 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4327 string, for example:
4329 warn message = H1: something\nH2: something
4331 the text was added as a single header line from Exim's point of view
4332 though it ended up OK in the delivered message. However, searching for the
4333 second and subsequent header lines using $h_h2: did not work. This has been
4334 fixed. Similarly, if a system filter added multiple headers in this way,
4335 the routers could not see them.
4337 20. Expanded the error message when iplsearch is called with an invalid key to
4338 suggest using net-iplsearch in a host list.
4340 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4341 ACLs are no longer actually imposed (and a message to that effect is
4344 22. If a "gecos" field in a passwd entry contained escaped characters, in
4345 particular, if it contained a \" sequence, Exim got it wrong when building
4346 a From: or a Sender: header from that name. A second bug also caused
4347 incorrect handling when an unquoted " was present following a character
4348 that needed quoting.
4350 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4351 was not being matched caselessly.
4353 24. Arranged for all hyphens in the exim.8 source to be escaped with
4356 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4357 properly. Recipient callout cache records were still being keyed to include
4358 the sender, even when use_sender was set false. This led to far more
4359 callouts that were necessary. The sender is no longer included in the key
4360 when use_sender is false.
4362 26. Added "control = submission" modifier to ACLs.
4364 27. Added the ${base62d: operator to decode base 62 numbers.
4366 28. dnsdb lookups can now access SRV records.
4368 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4369 the configuration file.
4371 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4372 (-v) mode. This makes the output for a verbose queue run more intelligible.
4374 31. Added a use_postmaster feature to recipient callouts.
4376 32. Added the $body_zerocount variable, containing the number of binary zero
4377 bytes in the message body.
4379 33. The time of last modification of the "new" subdirectory is now used as the
4380 "mailbox time last read" when there is a quota error for a maildir
4383 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4385 35. Added +ignore_unknown as a special item in host lists.
4387 36. Code for decoding IPv6 addresses in host lists is now included, even if
4388 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4389 address was recognized as an IP address, but was then not correctly decoded
4390 into binary, causing unexpected and incorrect effects when compared with
4397 1. Very minor rewording of debugging text in manualroute to say "list of
4398 hosts" instead of "hostlist".
4400 2. If verify=header_syntax was set, and a header line with an unqualified
4401 address (no domain) and a large number of spaces between the end of the
4402 name and the colon was received, the reception process suffered a buffer
4403 overflow, and (when I tested it) crashed. This was caused by some obsolete
4404 code that should have been removed. The fix is to remove it!
4406 3. When running in the test harness, delay a bit after writing a bounce
4407 message to get a bit more predictability in the log output.
4409 4. Added a call to search_tidyup() just before forking a reception process. In
4410 theory, someone could use a lookup in the expansion of smtp_accept_max_
4411 per_host which, without the tidyup, could leave open a database connection.
4413 5. Added the variables $recipient_data and $sender_data which get set from a
4414 lookup success in an ACL "recipients" or "senders" condition, or a router
4415 "senders" option, similar to $domain_data and $local_part_data.
4417 6. Moved the writing of debug_print from before to after the "senders" test
4420 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4421 problems for message scanning, either using a data ACL, or using
4422 local_scan() because the Received: header was not generated till after they
4423 were called (in order to set the time as the time of reception completion).
4424 I have revised the way this works. The header is now generated after the
4425 body is received, but before the ACL or local_scan() are called. After they
4426 are run, the timestamp in the header is updated.
4432 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4433 before starting a queue runner without re-exec. This happened only when
4434 deliver_drop_privilege was set or when the Exim user was set to root. The
4435 effect of the bug was that timeouts during subsequent deliveries caused
4436 crashes instead of being properly handled. The handler is now left at its
4437 default (and expected) setting.
4439 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4440 message, again when deliver_drop_privilege is set or Exim is run as root.
4441 The bug described in (1) was not present in this case, but the tidying up
4442 of the other signals was missing. I have made the two cases consistent.
4444 3. The ignore_target_hosts setting on a manualroute router was being ignored
4445 for hosts that were looked up using the /MX notation.
4447 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4450 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4451 operated on the sender address. After changing the $sender_address to <>
4452 for the sender address verify, Exim was re-instated it as the original
4453 (before rewriting) address, but remembering that it had rewritten it, so it
4454 wasn't rewriting it again. This bug also had the effect of breaking the
4455 sender address verification caching when the sender address was rewritten.
4457 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4458 This has been changed so that if the ip literal address matches
4459 ignore_target_hosts, the router declines.
4461 7. Added expansion conditions match_domain, match_address, and match_local_
4462 part (NOT match_host).
4464 8. The placeholder for the Received: header didn't have a length field set.
4466 9. Added code to Exim itself and to exim_lock to test for a specific race
4467 condition that could lead to file corruption when using MBX delivery. The
4468 issue is with the lockfile that is created in /tmp. If this file is removed
4469 after a process has opened it but before that process has acquired a lock,
4470 there is the potential for a second process to recreate the file and also
4471 acquire a lock. This could lead to two Exim processes writing to the file
4472 at the same time. The added code performs the same test as UW imapd; it
4473 checks after acquiring the lock that its file descriptor still refers to
4474 the same named file.
4476 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4477 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4480 11. Added the smtp_active_hostname option. If used, this will typically be made
4481 to depend on the incoming interface address. Because $interface_address is
4482 not set up until the daemon has forked a reception process, error responses
4483 that can happen earlier (such as "too many connections") no longer contain
4486 12. If an expansion in a condition on a "warn" statement fails because a lookup
4487 defers, the "warn" statement is abandoned, and the next ACL statement is
4488 processed. Previously this caused the whole ACL to be aborted.
4490 13. Added the iplsearch lookup type.
4492 14. Added ident_timeout as a log selector.
4494 15. Added tls_certificate_verified as a log selector.
4496 16. Added a global option tls_require_ciphers (compare the smtp transport
4497 option of the same name). This controls incoming TLS connections.
4499 17. I finally figured out how to make tls_require_ciphers do a similar thing
4500 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4501 before starting the TLS session.
4503 18. Tabs are now shown as \t in -bP output.
4505 19. If the log selector return_path_on_delivery was set, Exim crashed when
4506 bouncing a message because it had too many Received: header lines.
4508 20. If two routers both had headers_remove settings, and the first one included
4509 a superfluous trailing colon, the final name in the first list and the
4510 first name in the second list were incorrectly joined into one item (with a
4511 colon in the middle).
4517 1. Added -C and -D options to the exinext utility, mainly to make it easier
4518 to include in the automated testing, but these could be helpful when
4519 multiple configurations are in use.
4521 2. The exinext utility was not formatting the output nicely when there was
4522 an alternate port involved in the retry record key, nor when there was a
4523 message id as well (for retries that were specific to a specific message
4524 and a specific host). It was also confused by IPv6 addresses, because of
4525 the additional colons they contain. I have fixed the IPv4 problem, and
4526 patched it up to do a reasonable job for IPv6.
4528 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4529 delivery, the log line now contains "pipelined" if PIPELINING was used.
4531 4. An SMTP transport process used to panic and die if the bind() call to set
4532 an explicit outgoing interface failed. This has been changed; it is now
4533 treated in the same way as a connect() failure.
4535 5. A reference to $sender_host_name in the part of a conditional expansion
4536 that was being skipped was still causing a DNS lookup. This no longer
4539 6. The def: expansion condition was not recognizing references to header lines
4540 that used bh_ and bheader_.
4542 7. Added the _cache feature to named lists.
4544 8. The code for checking quota_filecount in the appendfile transport was
4545 allowing one more file than it should have been.
4547 9. For compatibility with Sendmail, the command line option
4555 and sets the incoming protocol and host name (for trusted callers). The
4556 host name and its colon can be omitted when only the protocol is to be set.
4557 Note the Exim already has two private options, -pd and -ps, that refer to
4558 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4559 "s", but I don't think that's a major issue.
4561 10. A number of refactoring changes to the code, none of which should affect
4564 (a) The number of logging options was getting close to filling up the
4565 32-bit word that was used as a bit map. I have split them into two classes:
4566 those that are passed in the argument to log_write(), and those that are
4567 only ever tested independently outside of that function. These are now in
4568 separate 32-bit words, so there is plenty of room for expansion again.
4569 There is no change in the user interface or the logging behaviour.
4571 (b) When building, for example, log lines, the code previously used a
4572 macro that called string_cat() twice, in order to add two strings. This is
4573 not really sufficiently general. Furthermore, there was one instance where
4574 it was actually wrong because one of the argument was used twice, and in
4575 one call a function was used. (As it happened, calling the function twice
4576 did not affect the overall behaviour.) The macro has been replaced by a
4577 function that can join an arbitrary number of extra strings onto a growing
4580 (c) The code for expansion conditions now uses a table and a binary chop
4581 instead of a serial search (which was left over from when there were very
4582 few conditions). Also, it now recognizes conditions like "pam" even when
4583 the relevant support is not compiled in: a suitably worded error message is
4584 given if an attempt is made to use such a condition.
4586 11. Added ${time_interval:xxxxx}.
4588 12. A bug was causing one of the ddress fields not to be passed back correctly
4589 from remote delivery subprocesses. The field in question was not being
4590 subsequently used, so this caused to problems in practice.
4592 13. Added new log selectors queue_time and deliver_time.
4594 14. Might have fixed a bug in maildirsizefile handling that threw up
4595 "unexpected character" debug warnings, and recalculated the data
4596 unnecessarily. In any case, I expanded the warning message to give more
4599 15. Added the message "Restricted characters in address" to the statements in
4600 the default ACL that block characters like @ and % in local parts.
4602 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4603 Three changes have been made:
4605 (a) There was a serious bug; a negative response to MAIL caused the whole
4606 recipient domain to be cached as invalid, thereby blocking all messages
4607 to all local parts at the same domain, from all senders. This bug has
4608 been fixed. The domain is no longer cached after a negative response to
4609 MAIL if the sender used is not empty.
4611 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4614 (c) A new callout option, "use_sender" has been added for people who want
4615 the modified behaviour.
4621 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4624 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4625 indeed breaks things for older releases.
4627 3. Added additional logging to the case where there is a problem reading data
4628 from a filter that is running in a subprocess using a pipe, in order to
4629 try to track down a specific problem.
4631 4. Testing facility fudge: when running in the test harness and attempting
4632 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4633 getting "No route to host". Convert this to a timeout.
4635 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4638 6. Some OS don't have socklen_t but use size_t instead. This affects the
4639 fifth argument of getsockopt() amongst other things. This is now
4640 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4641 can be set for individual OS. I have set it for SunOS5, OSF1, and
4642 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4643 some earlier ones do not.
4645 7. Change 4.30/15 was not doing the test caselessly.
4647 8. The standard form for an IPv6 address literal was being rejected by address
4648 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4649 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4650 this, as well as the form without the "IPv6" on the front (but only when
4651 address literals are enabled, of course).
4653 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4655 10. Exim crashed if a message with an empty sender address specified by -f
4656 encountered a router with an errors_to setting. This could be provoked only
4657 by a command such as
4661 where an empty string was supplied; "<>" did not hit this bug.
4663 11. Installed PCRE release 4.5.
4665 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4666 remained set. It is now erased.
4668 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4669 times from message ids (which are base 36 rather than the normal 62).
4671 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4672 were being counted as actual protocol errors, and logged if the log
4673 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4674 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4675 and DATA following a set of rejected RCPTs do not count as protocol errors.
4676 In other words, Exim assumes they were pipelined, though this may not
4677 actually be the case. Of course, in all cases the client gets an
4678 appropriate error code.
4680 15. If a lookup fails in an ACL condition, a message about the failure may
4681 be available; it is used if testing the ACL cannot continue, because most
4682 such messages specify what the cause of the deferral is. However, some
4683 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4684 that caused an old message to be retained and used if a later statement
4685 caused a defer, replacing the real cause of the deferral.
4687 16. If an IP address had so many PTR records that the DNS lookup buffer
4688 was not large enough to hold them, Exim could crash while trying to process
4689 the truncated data. It now detects and logs this case.
4691 17. Further to 4.21/58, another change has been made: if (and only if) the
4692 first line of a message (the first header line) ends with CRLF, a bare LF
4693 in a subsequent header line has a space inserted after it, so as not to
4694 terminate the header.
4696 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4697 unnecessarily, used atoi() instead of strtol(), and didn't check the
4698 termination when getting file sizes from file names by regex.
4700 19. Completely re-implemented the support for maildirsize files, in the light
4701 of a number of problems with the previous contributed implementation
4702 (4.30/29). In particular:
4704 . If the quota is zero, the maildirsize file is maintained, but no quota is
4707 . If the maildir directory does not exist, it is created before any attempt
4708 to write a maildirsize file.
4710 . The quota value in the file is just a cache; if the quota is changed in
4711 the transport, the new value overrides.
4713 . A regular expression is available for excluding directories from the
4716 20. The autoreply transport checks the characters in options that define the
4717 message's headers; it allows continued headers, but it was checking with
4718 isspace() after an embedded newline instead of explicitly looking for a
4721 21. If all the "regular" hosts to which an address was routed had passed their
4722 expiry times, and had not reached their retry times, the address was
4723 bounced, even if fallback hosts were defined. Now Exim should go on to try
4726 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4727 equivalent code in the SMTP transport. Some hosts send humungous responses
4728 to HELO/EHLO, more than 1024 it seems.
4730 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4731 gives compiler warnings in some environments. I've now done it "properly",
4734 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4735 (because of problems with the built-in one) was declared to return uschar *
4736 instead of char *, causing compiler failure.
4738 25. Fixed a file descriptor leak when processing alias/forward files.
4740 26. Fixed a minor format string issue in dbfn.c.
4742 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4744 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4745 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4748 29. When a sender address is verified, it is cached, to save repeating the test
4749 when there is more than one recipient in a message. However, when the
4750 verification involves a callout, it is possible for different callout
4751 options to be set for different recipients. It is too complicated to keep
4752 track of this in the cache, so now Exim always runs a verification when a
4753 callout is required, relying on the callout cache for the optimization.
4754 The overhead is duplication of the address routing, but this should not be
4757 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4758 to be verified with callout=postmaster, and the main callout worked but the
4759 postmaster check failed, the verification correctly failed. However, if a
4760 subsequent RCPT command asked for sender verification *without* the
4761 postmaster check, incorrect caching caused this verification also to fail,
4764 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4765 it was not caching the DNS options (qualify_single, search_parents) that
4766 were used when the lookup failed. A subsequent lookup with different
4767 options therefore always gave the same answer, though there were cases
4768 where it should not have. (Example: a "domains = !$mx_any" option on a
4769 dnslookup router: the "domains" option is always processed without any
4770 widening, but the router might have qualify_single set.) Now Exim uses the
4771 cached value only when the same options are set.
4773 32. Added John Jetmore's "exipick" utility to the distribution.
4775 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4776 than a timeout (e.g. a certificate is required, and is not provided), an
4777 Exim server now closes the connection immediately. Previously it waited for
4778 the client to close - but if the client is SSL, it seems that they each
4779 wait for each other, leading to a delay before one of them times out.
4781 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4782 maintained 0.8.x compatibility because I don't think many are using it, and
4783 it is clearly obsolete.
4785 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4788 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4789 client certificate was expired. A simple patch fixes this, though I don't
4790 understand the full logic of why the verify callback is called multiple
4793 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4794 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4795 which causes problems with some clients (such as the Certicom SSL Plus
4796 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4797 disables the coutermeasure allowing Eudora to connect."
4799 38. Exim was not checking that a write() to a log file succeeded. This could
4800 lead to Bad Things if a log got too big, in particular if it hit a file
4801 size limit. Exim now panics and dies if it cannot write to a log file, just
4802 as it does if it cannot open a log file.
4804 39. Modified OS/Makefile-Linux so that it now contains
4806 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4808 The two -D definitions ensure that Exim is compiled with large file
4809 support, which makes it possible to handle log files that are bigger than
4812 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4813 instance) a domain was checked against a named list that involved a lookup,
4814 causing $domain_data to be set, then another domain was checked against the
4815 same list, then the first domain was re-checked, the value of $domain_data
4816 after the final check could be wrong. In particular, if the second check
4817 failed, it could be set empty. This bug probably also applied to
4820 41. The strip_trailing_dot option was not being applied to the address given
4821 with the -f command-line option.
4823 42. The code for reading a message's header from the spool was incrementing
4824 $received_count, but never initializing it. This meant that the value was
4825 incorrect (doubled) while delivering a message in the same process in which
4826 it was received. In the most common configuration of Exim, this never
4827 happens - a fresh exec is done - but it can happen when
4828 deliver_drop_privilege is set.
4830 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4831 it now includes up to 150 characters of the unexpected data in the log
4834 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4835 and building data strings. The size of both of these buffers was 10 000
4836 bytes - far larger than anybody would *ever* want, thought I. Needless to
4837 say, somebody hit the limit. I have increased the maximum line length to
4838 20 000 and the maximum data length of concatenated lines to 100 000. I have
4839 also fixed two bugs, because there was no checking on these buffers. Tsk,
4840 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4841 buffer is too small.
4843 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4844 lsearch lookups. Now it does.
4846 46. When parsing a route_list item in a manualroute router, a fixed-length
4847 buffer was used for the list of hosts. I made this 1024 bytes long,
4848 thinking that nobody would ever have a list of hosts that long. Wrong.
4849 Somebody had a whole pile of complicated expansion conditions, and the
4850 string was silently truncated, leading to an expansion error. It turns out
4851 that it is easier to change to an unlimited length (owing to other changes
4852 that have happened since this code was originally written) than to build
4853 structure for giving a limitation error. The length of the item that
4854 expands into the list of hosts is now unlimited.
4856 47. The lsearch lookup could not handle data where the length of text line was
4857 more than 4095 characters. Such lines were truncated, leading to shortened
4858 data being returned. It should now handle lines of any length.
4860 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4861 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4864 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4865 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4866 no longer happens in comment lines. A list of replacements is now placed
4867 at the head of all of the source files, except those whose only change is
4868 to replace PERL_COMMAND in the very first #! line.
4870 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4871 messages on the queue, with a bottom-up merge sort, using code contributed
4872 by Michael Haardt. This should make operations like -bp somewhat faster on
4873 large queues. It won't affect queue runners, except when queue_run_in_order
4876 51. Installed eximstats 1.31 in the distribution.
4878 52. Added support for SRV lookups to the dnslookup router.
4880 53. If an ACL referred to $message_body or $message_body_end, the value was not
4881 reset for any messages that followed in the same SMTP session.
4883 54. The store-handling optimization for building very long strings was not
4884 differentiating between the different store pools. I don't think this
4885 actually made any difference in practice, but I've tidied it.
4887 55. While running the routers to verify a sender address, $sender_address
4888 was still set to the sender address. This is wrong, because when routing to
4889 send a bounce to the sender, it would be empty. Therefore, I have changed
4890 it so that, while verifying a sender address, $sender_address is set to <>.
4891 (There is no change to what happens when verifying a recipient address.)
4893 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4894 target A or AAAA records (if not already returned) without resetting the
4895 qualify_single or search_parents options of the DNS resolver. These are
4896 inappropriate in this case because the targets of MX and SRV records must
4897 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4898 target that, when qualified, matched something in the local domain. These
4899 two options are now turned off when doing these lookups.
4901 57. It seems that at least some releases of Reiserfs (which does not have the
4902 concept of a fixed number of inodes) returns zero and not -1 for the
4903 number of available inodes. This interacted badly with check_spool_inodes,
4904 which assumed that -1 was the "no such thing" setting. What I have done is
4905 to check that the total number of inodes is greater than zero before doing
4906 the test of how many are available.
4908 58. When a "warn" ACL statement has a log_message modifier, the message is
4909 remembered, and not repeated. This is to avoid a lot of repetition when a
4910 message has many recipients that cause the same warning to be written.
4911 Howewer, Exim was preserving the list of already written lines for an
4912 entire SMTP session, which doesn't seem right. The memory is now reset if a
4913 new message is started.
4915 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4916 debugging output unless log_rewrite was also set.
4918 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4919 of (int)(handle) when we know that handle contains (void *)(-1).
4921 61. The Exim daemon panic-logs an error return when it closes the incoming
4922 connection. However "connection reset by peer" seems to be common, and
4923 isn't really an error worthy of noting specially, so that particular error
4926 62. When Exim is trying to find all the local interfaces, it used to panic and
4927 die if the ioctl to get the interface flags failed. However, it seems that
4928 on at least one OS (Solaris 9) it is possible to have an interface that is
4929 included in the list of interfaces, but for which you get a failure error
4930 for this call. This happens when the interface is not "plumbed" into a
4931 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4932 failure of the "get flags" call assumes that the interface is down.
4934 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4935 makes life easier for people who are doing arithmetic on fields extracted
4936 from dates, where you often get leading zeros that should not be
4937 interpreted as octal.
4939 64. Added qualify_domain to the redirect router, to override the global
4942 65. If a pathologically long header line contained very many addresses (the
4943 report of this problem mentioned 10 000) and each of them was rewritten,
4944 Exim could use up a very large amount of memory. (It kept on making new
4945 copies of the header line as it rewrote, and never released the old ones.)
4946 At the expense of a bit more processing, the header rewriting function has
4947 been changed so that it no longer eats memory in this way.
4949 66. The generation of the Received: header has been moved from the time that a
4950 message starts to be received, to the time that it finishes. The timestamp
4951 in the Received: header should now be very close to that of the <= log
4952 line. There are two side-effects of this change:
4954 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4955 logged header lines no longer include the local Received: line, because
4956 it has not yet been created. The same applies to a copy of the message
4957 that is returned to a non-SMTP sender when a message is rejected.
4959 (b) When a filter file is tested using -bf, no additional Received: header
4960 is added to the test message. After some thought, I decided that this
4963 This change does not affect the value of $received_for. It is still set
4964 after address rewriting, but before local_scan() is called.
4966 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4968 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4969 gave an unhelpful panic error message, and a defer error. I have managed to
4970 change this behaviour so that it now rejects any supplied certificate,
4971 which seems right, as the list of acceptable certificates is empty.
4973 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4974 gave an unhelpful defer error. I have not managed to make this reject any
4975 supplied certificates, but the error message it gives is "no certificate
4976 supplied", which is not helpful.
4978 70. exigrep's output now also includes lines that are not associated with any
4979 message, but which match the given pattern. Implemented by a patch from
4980 Martin Sluka, which also tidied up the Perl a bit.
4982 71. Recipient callout verification, like sender verification, was using <> in
4983 the MAIL FROM command. This isn't really the right thing, since the actual
4984 sender may affect whether the remote host accepts the recipient or not. I
4985 have changed it to use the actual sender in the callout; this means that
4986 the cache record is now keyed on a recipient/sender pair, not just the
4987 recipient address. There doesn't seem to be a real danger of callout loops,
4988 since a callout by the remote host to check the sender would use <>.
4989 [SEE ABOVE: changed after hitting problems.]
4991 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4992 temporary errors. However, in the case of such a code being given after
4993 the end of a data transmission (i.e. after ".") Exim was failing to write
4994 a retry record for the message. (Yes, there was some broken host that was
4995 actually sending 8xx at this point.)
4997 73. An unknown lookup type in a host list could cause Exim to panic-die when
4998 the list was checked. (An example that provoked this was putting <; in the
4999 middle of a list instead of at the start.) If this happened during a DATA
5000 ACL check, a -D file could be left lying around. This kind of configuration
5001 error no longer causes Exim to die; instead it causes a defer errror. The
5002 incident is still logged to the main and panic logs.
5004 74. Buglet left over from Exim 3 conversion. The message "too many messages
5005 in one connection" was written to the rejectlog but not the mainlog, except
5006 when address rewriting (yes!) was being logged.
5008 75. Added write_rejectlog option.
5010 76. When a system filter was run not as root (that is, when system_filter_user
5011 was set), the values of the $n variables were not being returned to the
5012 main process; thus, they were not subsequently available in the $sn
5015 77. Added +return_path_on_delivery log selector.
5017 78. A connection timeout was being treated differently from recipients deferred
5018 when testing hosts_max_try with a message that was older than the host's
5019 retry timeout. (The host should not be counted, thus allowing all hosts to
5020 be tried at least once before bouncing.) This may have been the cause of an
5021 occasionally reported bug whereby a message would remain on the queue
5022 longer than the retry timeout, but would be bounced if a delivery was
5023 forced. I say "may" because I never totally pinned down the problem;
5024 setting up timeout/retry tests is difficult. See also the next item.
5026 79. The ultimate address timeout was not being applied to errors that involved
5027 a combination of host plus message (for example, a timeout on a MAIL
5028 command). When an address resolved to a number of possible hosts, and they
5029 were not all tried for each delivery (e.g. because of hosts_max_try), a
5030 message could remain on the queue longer than the retry timeout.
5032 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5035 81. Fixed an obscure SMTP outgoing bug which required at least the following
5036 conditions: (a) there was another message waiting for the same server;
5037 (b) the server returned 5xx to all RCPT commands in the first message so
5038 that the message was not completed; (c) the server dropped the connection
5039 or gave a negative response to the RSET that Exim sends to abort the
5040 transaction. The observed case was a dropped connection after DATA that had
5041 been sent in pipelining mode. That is, the server had advertised PIPELINING
5042 but was not implementing it correctly. The effect of the bug was incorrect
5043 behaviour, such as trying another host, and this could lead to a crash.
5049 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5050 and daemon.c were passed as pointers to ints; they should have been
5051 pointers to socklen_t variables (which are typically unsigned ints).
5053 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5056 3. Fixed a really odd bug that affected only the testing scheme; patching a
5057 certain fixed string in the binary changed the value of another string that
5058 happened to be identical to the end of the original first string.
5060 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5061 name", it returns that address as the IP address. On some operating
5062 systems (e.g. Solaris), it also passes back the IP address string as the
5063 "host name". However, on others (e.g. Linux), it passes back an empty
5064 string. Exim wasn't checking for this, and was changing the host name to an
5065 empty string, assuming it had been canonicized.
5067 5. Although rare, it is permitted to have more than one PTR record for a given
5068 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5069 all the names associated with an address, because they do in Solaris.
5070 However, it seems that they do not in Linux for data that comes from the
5071 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5072 I found this out when I moved to a new Linux workstation and tried to run
5073 the Exim test suite.
5075 To get round this problem I have changed the code so that it now does its
5076 own call to the DNS to look up PTR records when searching for a host name.
5077 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5078 addresses that are only in /etc/hosts are still found.
5080 This behaviour is, however, controlled by an option called host_lookup_
5081 order, which defaults to "bydns:byaddr". If people want to use the other
5082 order, or indeed, just use one or the other means of lookup, they can
5083 specify it in this variable.
5085 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5086 some operating systems, this comes back from gethostbyaddr() as an empty
5087 string, and this is what Exim used to test for. However, it seems that in
5088 other systems, "." is yielded. Exim now tests for this case too.
5090 7. The values of check_spool_space and check_log_space are now held internally
5091 as a number of kilobytes instead of an absolute number of bytes. If a
5092 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5093 kilobyte. This means that much larger values can be stored.
5095 8. Exim monitor: an attempt to get the action menu when not actually pointing
5096 at a message produces an empty menu entitled "No message selected". This
5097 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5098 no entries in it ("Shell widget menu has zero width and/or height"). So I
5099 have added a single, blank menu entry in this case.
5101 9. Added ${quote_local_part.
5103 10. MIME decoding is now applied to the contents of Subject: header lines when
5106 11. Now that a reference to $sender_host_address automatically causes a reverse
5107 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5108 host lookup before query-style lookups in lists that might use this
5109 variable. This has therefore been abolished, and the "net-" prefix is no
5110 longer necessary for query-style lookups.
5112 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5113 have been a typo for LFLAGS, so it has been changed.
5115 13. The install script calls Exim with "-C /dev/null" in order to find the
5116 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5117 to be output. Howeve, since Exim outputs its version number before the
5118 error, it didn't break the script. It just looked ugly. I fixed this by
5119 always allowing "-C /dev/null" if the caller is root.
5121 14. Ignore overlarge ACL variable number when reading spool file - insurance
5122 against a later release with more variables having written the file.
5124 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5125 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5126 this, as well as the form without the "IPv6" on the front.
5128 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5129 OS/Makefile-Darwin file.
5131 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5132 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5134 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5135 variable. However, some OS use ints for inodes. I've added cast to long int
5136 to get rid of the compiler warning.
5138 19. I had forgotten to lock out "/../" in configuration file names when
5139 ALT_CONFIG_PREFIX was set.
5141 20. Routers used for verification do not need to specify transports. However,
5142 if such a router generated a host list, and callout was configured, Exim
5143 crashed, because it could not find a port number from the (non-existent)
5144 transport. It now assumes port 25 in this circumstance.
5146 21. Added the -t option to exigrep.
5148 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5149 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5150 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5151 them is set, LOOKUP_LSEARCH is forced.
5153 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5155 24. Added sender and host information to the "rejected by local_scan()" log
5156 line; previously there was no indication of these.
5158 25. Added .include_if_exists.
5160 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5161 Linux. It turns out that not all file systems support this. Apparently some
5162 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5163 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5164 not supported on the file descriptor, is now ignored when Exim is trying to
5165 sync a directory. This applies only to Linux.
5167 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5169 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5170 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5173 29. Added support for maildirsize files from supplied patch (modified a bit).
5175 30. The use of :fail: followed by an empty string could lead Exim to respond to
5176 sender verification failures with (e.g.):
5178 550 Verification failed for <xxx>
5179 550 Sender verify failed
5181 where the first response line was missing the '-' that indicates it is not
5182 the final line of the response.
5184 31. The loop for finding the name of the user that called Exim had a hardwired
5185 limit of 10; it now uses the value of finduser_retries, which is used for
5186 all other user lookups.
5188 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5191 33. Exim was neglecting to zero errno before one call of strtol() when
5192 expanding a string and expecting an integer value. On some systems this
5193 resulted in spurious "integer overflow" errors. Also, it was casting the
5194 result into an int without checking.
5196 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5197 did not work. The code looks as if it has *never* worked, though it appears
5198 to have been documented since at least releast 1.62. I have made it work.
5200 35. The "timeout_DNS" error in retry rules, also documented since at least
5201 1.62, also never worked. As it isn't clear exactly what this means, and
5202 clearly it isn't a major issue, I have abolished the feature by treating it
5203 as "timeout", and writing a warning to the main and panic logs.
5205 36. The display of retry rules for -brt wasn't always showing the error code
5208 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5209 timeout_connect_A, timeout_connect_MX.
5211 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5212 to the empty sender.
5214 39. The daemon was not analysing the content of -oX till after it had closed
5215 stderr and disconnected from the controlling terminal. This meant that any
5216 syntax errors were only noted on the panic log, and the return code from
5217 the command was 0. By re-arranging the code a little, I've made the
5218 decoding happen first, so such errors now appear on stderr, and the return
5219 code is 1. However, the actual setting up of the sockets still happens in
5220 the disconnected process, so errors there are still only recorded on the
5223 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5224 connections (as happens on some IP stacks) was logged at start up time as
5225 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5226 it from "IPv6 and IPv4", which means that two separate sockets are being
5229 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5230 says whether AF_INET or AF_INET6 was passed as an argument.
5232 42. Exiwhat output was messed up when time zones were included in log
5235 43. Exiwhat now gives more information about the daemon's listening ports,
5236 and whether -tls-on-connect was used.
5238 44. The "port" option of the smtp transport is now expanded.
5240 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5241 silently ignored. Now an error message is written to the main and panic
5244 46. There's a new ACL modifier called "logwrite" which writes to a log file
5245 as soon as it is encountered.
5247 47. Added $local_user_uid and $local_user_gid at routing time.
5249 48. Exim crashed when trying to verify a sender address that was being
5252 49. Exim was recognizing only a space character after ".include". It now also
5253 recognizes a tab character.
5255 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5256 extracting the relevant information from the specification. The man page no
5257 longer contains scrambled data for the -d option, and I've added a section
5258 at the front about calling Exim under different names.
5260 51. Added "extra_headers" argument to the "mail" command in filter files.
5262 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5265 53. Installed eximstats 1.29.
5267 54. Added transport_filter_timeout as a generic transport option.
5269 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5270 Cc: header lines. This was required by RFC 822, but it not required by RFC
5273 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5274 incoming messages that did not have them. Now it does so only if the
5275 message originates locally, that is, if there is no associated remote host
5276 address. When Resent- header lines are present, this applies to the Resent-
5277 lines rather than the non-Resent- lines.
5279 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5280 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5282 58. Messages for configuration errors now include the name of the main
5283 configuration file - useful now that there may be more than one file in a
5284 list (.included file names were always shown).
5286 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5287 for those rare installations that do not start the daemon as root or run it
5288 setuid root. I've cut out the call to initgroups() if the daemon is not
5291 60. The Exim user and group can now be bound into the binary as text strings
5292 that are looked up at the start of Exim's processing.
5294 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5296 62. Added $mailstore_basename variable.
5298 63. Installed patch to sieve.c from Michael Haardt.
5300 64. When Exim failed to open the panic log after failing to open the main log,
5301 the original message it was trying to log was written to stderr and debug
5302 output, but if they were not available (the usual case in production), it
5303 was lost. Now it is written to syslog before the two lines that record the
5304 failures to open the logs.
5306 65. Users' Exim filters run in subprocesses under the user's uid. It is
5307 possible for a "deliver" command or an alias in a "personal" command to
5308 provoke an address rewrite. If logging of address rewriting is configured,
5309 this fails because the process is not running as root or exim. There may be
5310 a better way of dealing with this, but for the moment (because 4.30 needs
5311 to be released), I have disabled address rewrite logging when running a
5312 filter in a non-root, non-exim process.
5318 1. The buildconfig auxiliary program wasn't quoting the value set for
5319 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5320 not defined. This bug was masked in 4.22 by the effect that was fixed in
5323 2. Some messages that were rejected after a message id was allocated were
5324 shown as "incomplete" by exigrep. It no longer does this for messages that
5325 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5327 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5328 have allow_domain_literals set, the ID did not get logged in the <= line.
5329 Domain literals are now always recognized in Message-ID: header lines.
5331 4. The first argument for a ${extract expansion item is the key name or field
5332 number. Leading and trailing spaces in this item were not being ignored,
5333 causing some misleading effects.
5335 5. When deliver_drop_privilege was set, single queue runner processes started
5336 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5337 same command in the process it spins off) were not dropping privilege.
5339 6. When the daemon running as "exim" started a queue runner, it always
5340 re-executed Exim in the spun-off process. This is a waste of effort when
5341 deliver_drop_privilege is set. The new process now just calls the
5342 queue-runner function directly.
5348 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5351 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5352 ignored. Though the use of -oP was forcing the writing of a pid file, it
5353 was always written to the default place.
5355 3. If the message "no IP address found for host xxxx" is generated during
5356 incoming verification, it is now followed by identification of the incoming
5357 connection (so you can more easily find what provoked it).
5359 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5361 5. Added some features to "harden" Exim a bit more against certain attacks:
5363 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5364 be put in Local/Makefile. This is like the never_users runtime option,
5365 but it cannot be overridden. The default setting is "root".
5367 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5368 prefix string with which any file named in a -C command line option
5371 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5372 is retained for -C and -D only if the caller of Exim is root. Without
5373 it, the exim user may also use -C and -D and retain privilege.
5375 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5376 command line option is disabled.
5378 6. Macro names set by the -D option must start with an upper case letter, just
5379 like macro names defined in the configuration file.
5381 7. Added "dereference=" facility to LDAP.
5383 8. Two instances of the typo "uknown" in the source files are fixed.
5385 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5386 the Configure-Makefile script screwed up while processing it.
5388 10. Incorporated PCRE 4.4.
5390 11. The SMTP synchronization check was not operating right at the start of an
5391 SMTP session. For example, it could not catch a HELO sent before the client
5392 waited for the greeting. There is now a check for outstanding input at the
5393 point when the greeting is written. Because of the duplex, asynchronous
5394 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5395 way, but not yet received, when the check is performed.
5397 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5398 on TCP/IP sockets, because this apparently causes some broken clients to
5401 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5402 unchanged) from the Cygwin maintainer.
5404 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5405 was supported instead of testing for mbx. Effectively a typo.
5407 15. The spa authenticator server code was not checking that the input it
5408 received was valid base64.
5410 16. The debug output line for the "set" modifier in ACLs was not showing the
5411 name of the variable that was being set.
5413 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5415 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5416 lookup on its own. Something else was needed to trigger the lookup. For
5417 example, a match in host_lookup or the need for a host name in a host list.
5418 Now, if $sender_host_name is referenced and the host name has not yet been
5419 looked up, a lookup is performed. If the lookup fails, the variable remains
5420 empty, and $host_lookup_failed is set to "1".
5422 19. Added "eqi" as a case-independent comparison operator.
5424 20. The saslauthd authentication condition could segfault if neither service
5425 nor realm was specified.
5427 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5428 error message that was logged was misleading, and incoming SMTP
5429 connections were dropped. The message is now more accurate, and temporary
5430 errors are given to SMTP connections.
5432 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5433 (except RSET) with a 421 error, until QUIT is received. However, it was
5434 failing to send a response to QUIT.
5436 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5437 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5438 ACL is now run after the helo_try_verify_hosts code.
5440 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5441 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5442 case-independent, so other case variants are also recognized). Apparently
5443 some systems use these upper case variants.
5445 25. If more than two messages were waiting for the same host, and a transport
5446 filter was specified for the transport, Exim sent two messages over the
5447 same TCP/IP connection, and then failed with "socket operation on non-
5448 socket" when it tried to send the third.
5450 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5452 27. The extern definition of crypt16() in expand.c was not being excluded when
5453 the OS had its own crypt16() function.
5455 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5456 as a preferred synonym for return_size_limit, both as an option and as an
5459 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5461 30. Changed the default configuration ACL to relax the local part checking rule
5462 for addresses that are not in any local domains. For these addresses,
5463 slashes and pipe symbols are allowed within local parts, but the sequence
5464 /../ is explicitly forbidden.
5466 31. SPA server authentication was not clearing the challenge buffer before
5469 32. log_message in a "warn" ACL statement was writing to the reject log as
5470 well as to the main log, which contradicts the documentation and doesn't
5471 seem right (because no rejection is happening). So I have stopped it.
5473 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5474 However, I am unable to do any testing of this.
5476 34. Fixed an infelicity in the appendfile transport. When checking directories
5477 for a mailbox, to see if any needed to be created, it was accidentally
5478 using path names with one or more superfluous leading slashes; tracing
5479 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5481 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5482 contents are added to the log line that is written for every discarded
5483 recipient. (Previously a log_message setting was ignored.)
5485 36. The ${quote: operator now quotes the string if it is empty.
5487 37. The install script runs exim in order to find its version number. If for
5488 some reason other than non-existence or emptiness, which it checks, it
5489 could not run './exim', it was installing it with an empty version number,
5490 i.e. as "exim-". This error state is now caught, and the installation is
5493 38. An argument was missing from the function that creates an error message
5494 when Exim fails to connect to the socket for saslauthd authentication.
5495 This could cause Exim to crash, or give a corrupted message.
5497 39. Added isip, isip4, and isip6 to ${if conditions.
5499 40. The ACL variables $acl_xx are now saved with the message, and can be
5500 accessed later in routers, transports, and filters.
5502 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5503 strings in the file are not string-expanded.
5505 42. If a MAIL command specified a SIZE value that was too large to fit into an
5506 int variable, the check against message_size_limit failed. Such values are
5507 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5508 day this will have to be increased, but I don't think I want to be around
5509 when emails are that large.
5516 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5517 iconv() is not standard in FreeBSD.
5519 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5520 IPv6 enabled. The observed symptom was a segmentation fault on return from
5521 the function os_common_find_running_interfaces() in src/os.c.
5523 3. In the check_special_case() function in daemon.c I had used "errno" as an
5524 argument name, which causes warnings on some systems. This was basically a
5525 typo, since it was named "eno" in the comments!
5527 4. The code that waits for the clock to tick (at a resolution of some fraction
5528 of a second) so as to ensure message-id uniqueness was always waiting for
5529 at least one whole tick, when it could have waited for less. [This is
5530 almost certainly not relevant at current processor speeds, where it is
5531 unlikely to ever wait at all. But we try to future-proof.]
5533 5. The function that sleeps for a time interval that includes fractions of a
5534 second contained a race. It did not block SIGALRM between setting the
5535 timer, and suspending (a couple of lines later). If the interval was short
5536 and the sigsuspend() was delayed until after it had expired, the suspension
5537 never ended. On busy systems this could lead to processes getting stuck for
5540 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5541 process, before it forks any delivery processes. The open lookup caching
5542 mechanism meant that the open file or database connection was passed into
5543 the delivery process. The problem was that delivery processes always tidy
5544 up cached lookup data. This could cause a problem for the next delivery
5545 process started by the queue runner, because the external queue runner
5546 process does not know about the closure. So the next delivery process
5547 still has data in the lookup cache. In the case of a file lookup, there was
5548 no problem because closing a file descriptor in a subprocess doesn't affect
5549 the parent. However, if the lookup was caching a connection to a database,
5550 the connection was closed, and the second delivery process was likely to
5551 see errors such as "PGSQL: query failed: server closed the connection
5552 unexpectedly". The problem has been fixed by closing all cached lookups
5553 in a queue runner before running a delivery process.
5555 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5556 seem to have the "const" qualifier which it has on other OS. I've
5559 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5560 *of the same type* (client or server) with the same public name that an
5561 error should be diagnosed.
5563 9. When Exim looked up a host name for an IP address, but failed to find the
5564 original IP address when looking up the host name (a safety check), it
5565 output the message "<ip address> does not match any IP for NULL", which was
5566 confusing, to say the least. The bug was that the host name should have
5567 appeared instead of "NULL".
5569 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5570 user that is built into the binary, and the -C or -D options is used, root
5571 privilege is dropped before the configuration file is read. In addition,
5572 logging is switched to stderr instead of the normal log files. If the
5573 configuration then re-defines the Exim user, the unprivileged environment
5574 is probably not what is expected, so Exim logs a panic warning message (but
5577 However, if deliver_drop_privilege is set, the unprivileged state may well
5578 be exactly what is intended, so the warning has been cut out in that case,
5579 and Exim is allowed to try to write to its normal log files.
5585 1. smtp_return_error_details was not giving details for temporary sender
5586 or receiver verification errors.
5588 2. Diagnose a configuration error if two authenticators have the same public
5591 3. Exim used not to create the message log file for a message until the first
5592 delivery attempt. This could be confusing when incoming messages were held
5593 for policy or load reasons. The message log file is now created at the time
5594 the message is received, and an initial "Received" line is written to it.
5596 4. The automatically generated man page for command line options had a minor
5597 bug that caused no ill effects; however, a more serious problem was that
5598 the procedure for building the man page automatically didn't always
5599 operate. Consequently, release 4.20 contains an out-of-date version. This
5600 shouldn't happen again.
5602 5. When building Exim with embedded Perl support, the script that builds the
5603 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5604 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5606 6. The freeze_tell option was not being used for messages that were frozen on
5607 arrival, either by an ACL or by local_scan().
5609 7. Added the smtp_incomplete_transaction log selector.
5611 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5612 was accepting AUTH without a new EHLO.
5614 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5615 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5618 10. Logging of TCP/IP connections (when configured) now happens in the main
5619 daemon process instead of the child process, so that the TCP/IP connection
5620 count is more accurate (but it can never be perfect).
5622 11. The use of "drop" in a nested ACL was not being handled correctly in the
5623 outer ACL. Now, if condition failure induced by the nested "drop" causes
5624 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5625 or "require"), the connection is dropped.
5627 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5628 that yield "discard" can now be used with an "accept" or a "discard" verb,
5629 but an error is generated for any others (because I can't see a useful way
5630 to define what should happen).
5632 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5633 are now processed in the same way as lines in the Exim configuration file.
5634 In particular, continuation lines are supported.
5636 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5638 15. Added -ti meaning -t -i.
5640 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5641 domains, and warn by logging if others are found.
5643 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5644 returned from GIFCONF when Exim is trying to find the list of interfaces on
5645 a host. The code in os.c has been modified to copy each ifreq to an aligned
5646 structure in all cases.
5648 Also, in some cases, the returned ifreq's were being copied to a 'struct
5649 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5650 means the last couple of bytes of an IPv6 address could be chopped if the
5651 ifreq contained only a normal sockaddr (14 bytes storage).
5653 18. Named domain lists were not supported in the hosts_treat_as_local option.
5654 An entry such as +xxxx was not recognized, and was treated as a literal
5657 19. Ensure that header lines added by a DATA ACL are included in the reject log
5658 if the ACL subsequently rejects the message.
5660 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5661 MD5 (which is deprecated).
5663 21. When testing a filter file using -bf, Exim was writing a message when it
5664 took the sender from a "From " line in the message, but it was not doing so
5665 when it took $return_path from a Return-Path: header line. It now does.
5667 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5668 with a valid header line field name (a series of printing characters
5669 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5671 23. Changed "disc" in the source to "disk" to conform to the documentation and
5672 the book and for uniformity.
5674 24. Ignore Sendmail's -Ooption=value command line item.
5676 25. When execve() failed while trying to run a command in a pipe transport,
5677 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5678 could be confused with a return value of 69 from the command itself. This
5679 has been changed to 127, the value the shell returns if it is asked to run
5680 a non-existent command. The wording for the related log line suggests a
5681 non-existent command as the problem.
5683 26. If received_header_text expands to an empty string, do not add a Received:
5684 header line to the message. (Well, it adds a token one on the spool, but
5685 marks it "old" so that it doesn't get used or transmitted.)
5687 27. Installed eximstats 1.28 (addition of -nt option).
5689 28. There was no check for failure on the call to getsockname() in the daemon
5690 code. This can fail if there is a shortage of resources on the system, with
5691 ENOMEM, for example. A temporary error is now given on failure.
5693 29. Contrary to the C standard, it seems that in some environments, the
5694 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5695 program. Exim now does this explicitly; it affects the formatting of
5696 timestamps using strftime().
5698 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5699 complaints. I've now initialized all the variables, to avoid this.
5701 32. Header lines added by a system filter were not being "seen" during
5702 transport-time rewrites.
5704 33. The info_callback() function passed to OpenSSL is set up with type void
5705 (*)(SSL *, int, int), as described somewhere. However, when calling the
5706 function (actually a macro) that sets it up, the type void(*)() is
5707 expected. I've put in a cast to prevent warnings from picky compilers.
5709 34. If a DNS black list lookup found a CNAME record, but there were no A
5710 records associated with the domain it pointed at, Exim crashed.
5712 35. If a DNS black list lookup returned more than one A record, Exim ignored
5713 all but the first. It now scans all returned addresses if a particular IP
5714 value is being sought. In this situation, the contents of the
5715 $dnslist_value variable are a list of all the addresses, separated by a
5718 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5719 to accept a host name and all its aliases if the forward lookup for any of
5720 them yielded the IP address of the incoming connection. Now it accepts only
5721 those names whose forward lookup yields the correct IP address. Any other
5722 names are discarded. This closes a loophole whereby a rogue DNS
5723 administrator could create reverse DNS records to break through a
5724 wildcarded host restriction in an ACL.
5726 37. If a user filter or a system filter that ran in a subprocess used any of
5727 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5728 the wrong values were passed to the pipe command ($thisaddress had the
5729 value of $0, $0 had the value of $1, etc). This bug was introduced by
5730 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5732 38. Improved the line breaking for long SMTP error messages from ACLs.
5733 Previously, if there was no break point between 40 and 75 characters, Exim
5734 left the rest of the message alone. Two changes have been made: (a) I've
5735 reduced the minimum length to 35 characters; (b) if it can't find a break
5736 point between 35 and 75 characters, it looks ahead and uses the first one
5737 that it finds. This may give the occasional overlong line, but at least the
5738 remaining text gets split now.
5740 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5741 file descriptors might be low, and that setting 1000 would always raise it.
5742 It turns out that in some environments, the limit is already over 1000 and
5743 that lowering it causes trouble. So now Exim takes care not to decrease it.
5745 40. When delivering a message, the value of $return_path is set to $sender_
5746 address at the start of routing (routers may change the value). By an
5747 oversight, this default was not being set up when an address was tested by
5748 -bt or -bv, which affected the outcome if any router or filter referred to
5751 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5752 log only when "message" or "log_message" are set. However, if one of the
5753 conditions was an address verification, or a call to a nested ACL, the
5754 messages generated by the underlying test were being passed through. This
5755 no longer happens. The underlying message is available in $acl_verify_
5756 message for both "message" and "log_message" expansions, so it can be
5757 passed through if needed.
5759 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5760 new expansion $bh_ to give the encoded byte string without charset
5761 translation. Translation happens only if iconv() is available; HAVE_ICONV
5762 indicates this at build time. HEADERS_CHARSET gives the charset to
5763 translate to; headers_charset can change it in the configuration, and
5764 "headers charset" can change it in an individual filter file.
5766 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5767 that creates RFC 2047 encoded "words" labels them as that charset instead
5768 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5769 expansion operator; (ii) when Exim creates a From: line for a local
5770 message; (iii) when a header line is rewritten to include a "phrase" part.
5772 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5773 buggy, causing it to skip the first lines of messages whose message ID
5774 ended in 'D'. This would not have bitten before Exim release 4.14, because
5775 message IDs were unlikely to end in 'D' before then. The effect was to have
5776 incorrect size information for certain domains.
5778 45. #include "config.h" was missing at the start of the crypt16.c module. This
5779 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5782 46. If there was a timeout during a "random" callout check, Exim treated it as
5783 a failure of the random address, and carried on sending RSET and the real
5784 address. If the delay was just some slowness somewhere, the response to the
5785 original RCPT would be taken as a response to RSET and so on, causing
5786 mayhem of various kinds.
5788 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5789 when I implemented it. It didn't allow for the fact that some option values
5790 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5791 the right test for positive values.
5793 48. Domain names in DNS records are case-independent. Exim always looks them up
5794 in lower case. Some resolvers return domain names in exactly the case they
5795 appear in the zone file, that is, they may contain uppercase letters. Not
5796 all resolvers do this - some return always lower case. Exim was treating a
5797 change of case by a resolver as a change of domain, similar to a widening
5798 of a domain abbreviation. This triggered its re-routing code and so it was
5799 trying to route what was effectively the same domain again. This normally
5800 caused routing to fail (because the router wouldn't handle the domain
5801 twice). Now Exim checks for this case specially, and just changes the
5802 casing of the domain that it ultimately uses when it transmits the message
5805 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5808 50. If a filter generated a file delivery with a non-absolute name (possible if
5809 no home directory exists for the router), the forbid_file option was not
5812 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5813 the existing equality matching.
5815 52. Exim was using ints instead of ino_t variables in some places where it was
5816 dealing with inode numbers.
5818 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5819 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5820 called TMPDIR, and if it finds it is different, it changes its value.
5822 54. The smtp_printf() function is now made available to local_scan() so
5823 additional output lines can be written before returning. There is also an
5824 smtp_fflush() function to enable the detection of a dropped connection.
5825 The variables smtp_input and smtp_batched_input are exported to
5828 55. Changed the default runtime configuration: the message "Unknown user"
5829 has been removed from the ACL, and instead placed on the localuser router,
5830 using the cannot_route_message feature. This means that any verification
5831 failures that generate their own messages won't get overridden. Similarly,
5832 the "Unrouteable address" message that was in the ACL for unverifiable
5833 relay addresses has also been removed.
5835 56. Added hosts_avoid_esmtp to the smtp transport.
5837 57. The exicyclog script was not checking for the esoteric option
5838 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5839 will work only if exicyclog is run under the appropriate euid.
5841 58. Following a discussion on the list, the rules by which Exim recognises line
5842 endings on incoming messages have been changed. The -dropcr and drop_cr
5843 options are now no-ops, retained only for backwards compatibility. The
5844 following line terminators are recognized: LF CRLF CR. However, special
5845 processing applies to CR:
5847 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5848 nor a local message in the state where . is a terminator.
5850 (ii) If a bare CR is encountered in a header line, an extra space is added
5851 after the line terminator so as not to end the header. The reasoning
5852 behind this is that bare CRs in header lines are most likely either
5853 to be mistakes, or people trying to play silly games.
5855 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5856 was being incorrectly given as 18 bytes larger than it should have been.
5857 This is a VOB (very old bug).
5859 60. This may never have affected anything current, but just in case it has:
5860 When the local host is found other than at the start of a list of hosts,
5861 the local host, those with the same MX, and any that follow, are discarded.
5862 When the list in question was part of a longer list of hosts, the following
5863 hosts (not currently being processed) were also being discarded. This no
5864 longer happens. I'm not sure if this situation could ever has previously
5867 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5870 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5871 header. This is something that is recommended in a new Internet Draft, and
5872 is something that is documented as being done by Sendmail. There are two
5873 possible values. For messages generated by the autoreply transport, Exim
5876 Auto-Submitted: auto-replied
5878 whereas for all other generated messages (e.g. bounces) it adds
5880 Auto-Submitted: auto-generated
5882 63. The "personal" condition in filters now includes a test for the
5883 Auto-Submitted: header. If it contains the string "auto-" the message it
5884 not considered personal.
5886 64. Added rcpt_include_affixes as a generic transport option.
5888 65. Added queue_only_override (default true).
5890 66. Added the syslog_duplication option.
5892 67. If what should have been the first header line of a message consisted of
5893 a space followed by a colon, Exim was mis-interpreting it as a header line.
5894 It isn't of course - it is syntactically invalid and should therefore be
5895 treated as the start of the message body. The misbehaviour could have
5896 caused a number of strange effects, including loss of data in subsequent
5897 header lines, and spool format errors.
5899 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5900 client host had authenticated. This control can now be exercised by an ACL
5901 for more flexibility.
5903 69. By default, callouts do not happen when testing with -bh. There is now a
5904 variant, -bhc, which does actually run the callout code, including
5905 consulting and updating the callout cache.
5907 70. Added support for saslauthd authentication, courtesy of Alexander
5910 71. If statvfs() failed on the spool or log directories while checking their
5911 size for availability, Exim confusingly gave the error "space shortage".
5912 Furthermore, in debugging mode it crashed with a floating point exception.
5913 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5914 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5915 serious problem, Exim now writes to the main and panic logs when this
5916 happens, with details of the failure. It then refuses to accept the
5917 incoming message, giving the message "spool directory problem" or "log
5918 directory problem" with a 421 code for SMTP messages.
5920 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5921 0, 1, and 2 exist, because some OS complain for execs without them (see
5922 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5923 descriptors. However, the code omitted to check that the open succeeded,
5924 causing mysterious errors if for some reason the permissions on /dev/null
5925 got screwed. Now Exim writes a message to the main and panic logs, and
5926 bombs out if it can't open /dev/null.
5928 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5929 interact so that it is all more flexible. It is supposed to remain
5930 backwards compatible. Also added extra_local_interfaces.
5932 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5933 to bomb out with an assertion failure - to the client this appears as a
5934 connection drop. This problem occurs in the part of the code that was taken
5935 from the Samba project. Fortunately, the assertion is in a very simple
5936 function, so I have fixed this by reproducing the function inline in the
5937 one place where it is called, and arranging for authentication to fail
5938 instead of killing the process with assert().
5940 75. The SPA client code was not working when the server requested OEM rather
5941 than Unicode encoding.
5943 76. Added code to make require_files with a specific uid setting more usable in
5944 the case where statting the file as root fails - usually a non-root-mounted
5945 NFS file system. When this happens and the failure is EACCES, Exim now
5946 forks a subprocess and does the per-uid checking as the relevant uid.
5948 77. Added process_log_path.
5950 78. If log_file_path was not explicitly set, a setting of check_log_space or
5951 check_log_inodes was ignored.
5953 79. If a space check for the spool or log partitions fails, the incident is now
5954 logged. Of course, in the latter case the data may get lost...
5956 80. Added the %p formatting code to string_format() so that it can be used to
5957 print addresses in debug_print(). Adjusted all the address printing in the
5958 debugging in store.c to use %p rather than %d.
5960 81. There was a concern that a line of code in smtp_in.c could overflow a
5961 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5962 initially expressed, the concern was not well-founded, because trailing
5963 spaces are removed early. However, if the trailing spaces were followed by
5964 a NULL, they did not get removed, so the overflow was possible. Two fixes
5967 (a) I re-wrote the offending code in a cleaner fashion.
5968 (b) If an incoming SMTP command contains a NULL character, it is rejected
5971 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5972 runs initgroups(), so that if the Exim user is in any additional groups,
5973 they will be used during message reception.
5979 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob