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. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router and the smtp transport (applying to the forward
91 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
92 of ldap servers used for a specific lookup. Patch provided by Heiko
95 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 New variable $lookup_dnssec_authenticated for observability.
98 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
99 Patch submitted by Lars Timman.
101 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
103 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
104 Requires trusted mode and valid format message id, aborts otherwise.
105 Patch contributed by Heiko Schlichting.
107 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
108 certextract with support for various fields. Bug 1358.
110 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
111 is requested by default, modifiable by smtp transport option
114 JH/22 Expansion operators ${md5:string} and ${sha1::string} can now
115 operate on certificate variables to give certificate fingerprints
116 Also new ${sha256:cert_variable}.
122 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
124 PP/02 Make -n do something, by making it not do something.
125 When combined with -bP, the name of an option is not output.
127 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
130 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
131 $sender_host_name and config options to manage this, and basic check
134 PP/05 DSCP support for outbound connections and control modifier for inbound.
136 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
137 (Only plugin which currently uses this is kerberos4, which nobody should
138 be using, but we should make it available and other future plugins might
139 conceivably use it, even though it would break NAT; stuff *should* be
140 using channel bindings instead).
142 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
143 name; added for Sendmail compatibility; requires admin caller.
144 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
145 just ignore it); requires trusted caller.
146 Also parse but ignore: -Ac -Am -X<logfile>
149 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
151 TL/02 Add +smtp_confirmation as a default logging option.
153 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
154 Patch by Magnus Holmgren from 2007-02-20.
156 TL/04 Bugzilla 1281 - Spec typo.
157 Bugzilla 1283 - Spec typo.
158 Bugzilla 1290 - Spec grammar fixes.
160 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
162 TL/06 Add Experimental DMARC support using libopendmarc libraries.
164 TL/07 Fix an out of order global option causing a segfault. Reported to dev
165 mailing list by by Dmitry Isaikin.
167 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
169 JH/02 Support "G" suffix to numbers in ${if comparisons.
171 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
173 NM/01 Bugzilla 1197 - Spec typo
174 Bugzilla 1196 - Spec examples corrections
176 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
178 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
179 gnutls_enable_pkcs11, but renamed to more accurately indicate its
182 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
183 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
185 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
186 "acl {{name}{arg}...}", and optional args on acl condition
189 JH/05 Permit multiple router/transport headers_add/remove lines.
191 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
193 JH/07 Avoid using a waiting database for a single-message-only transport.
194 Performance patch from Paul Fisher. Bugzilla 1262.
196 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
199 JH/09 Add $headers_added variable, with content from use of ACL modifier
200 add_header (but not yet added to the message). Bugzilla 199.
202 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
203 Pulled from Bugzilla 817 by Wolfgang Breyha.
205 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
207 (nb: this is the same fix as in Exim 4.80.1)
209 JH/11 Add A= logging on delivery lines, and a client_set_id option on
212 JH/12 Add optional authenticated_sender logging to A= and a log_selector
215 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
217 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
218 advertise SMTP AUTH mechanism to us, instead of a generic
219 protocol violation error. Also, make Exim more robust to bad
220 data from the Dovecot auth socket.
222 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
224 When a queue runner is handling a message, Exim first routes the
225 recipient addresses, during which it prunes them based on the retry
226 hints database. After that it attempts to deliver the message to
227 any remaining recipients. It then updates the hints database using
230 So if a recipient address works intermittently, it can get repeatedly
231 deferred at routing time. The retry hints record remains fresh so the
232 address never reaches the final cutoff time.
234 This is a fairly common occurrence when a user is bumping up against
235 their storage quota. Exim had some logic in its local delivery code
236 to deal with this. However it did not apply to per-recipient defers
237 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
239 This change adds a proper retry rule check during routing so that the
240 final cutoff time is checked against the message's age. We only do
241 this check if there is an address retry record and there is not a
242 domain retry record; this implies that previous attempts to handle
243 the address had the retry_use_local_parts option turned on. We use
244 this as an approximation for the destination being like a local
245 delivery, as in LMTP.
247 I suspect this new check makes the old local delivery cutoff check
248 redundant, but I have not verified this so I left the code in place.
250 TF/02 Correct gecos expansion when From: is a prefix of the username.
252 Test 0254 submits a message to Exim with the header
256 When I ran the test suite under the user fanf2, Exim expanded
257 the header to contain my full name, whereas it should have added
258 a Resent-Sender: header. It erroneously treats any prefix of the
259 username as equal to the username.
261 This change corrects that bug.
263 GF/01 DCC debug and logging tidyup
264 Error conditions log to paniclog rather than rejectlog.
265 Debug lines prefixed by "DCC: " to remove any ambiguity.
267 TF/03 Avoid unnecessary rebuilds of lookup-related code.
269 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
270 Bug spotted by Jeremy Harris; was flawed since initial commit.
271 Would have resulted in OCSP responses post-SNI triggering an Exim
272 NULL dereference and crash.
274 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
276 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
277 Bug detection, analysis and fix by Samuel Thibault.
278 Bugzilla 1331, Debian bug #698092.
280 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
282 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
283 Server implementation by Todd Lyons, client by JH.
284 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
285 config variable "prdr_enable" controls whether the server
286 advertises the facility. If the client requests PRDR a new
287 acl_data_smtp_prdr ACL is called once for each recipient, after
288 the body content is received and before the acl_smtp_data ACL.
289 The client is controlled by bolth of: a hosts_try_prdr option
290 on the smtp transport, and the server advertisement.
291 Default client logging of deliveries and rejections involving
292 PRDR are flagged with the string "PRDR".
294 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
295 fclose(). Diagnosis by Todd Lyons.
297 PP/17 Update configure.default to handle IPv6 localhost better.
298 Patch by Alain Williams (plus minor tweaks).
301 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
302 This is now consistent with GnuTLS, and is now documented: the
303 previous undocumented portable approach to treating the option as
304 unset was to force an expansion failure. That still works, and
305 an empty string is now equivalent.
307 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
308 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
309 not performing validation itself.
311 PP/20 Added force_command boolean option to pipe transport.
312 Patch from Nick Koston, of cPanel Inc.
314 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
317 TF/04 Added udpsend ACL modifer and hexquote expansion operator
319 PP/21 Fix eximon continuous updating with timestamped log-files.
320 Broken in a format-string cleanup in 4.80, missed when I repaired the
321 other false fix of the same issue.
322 Report and fix from Heiko Schlichting.
325 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
326 Report from Prashanth Katuri.
328 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
329 It's SecureTransport, so affects any MacOS clients which use the
330 system-integrated TLS libraries, including email clients.
332 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
333 using a MIME ACL for non-SMTP local injection.
334 Report and assistance in diagnosis by Warren Baker.
336 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
338 JH/16 Fix comparisons for 64b. Bugzilla 1385.
340 TL/09 Add expansion variable $authenticated_fail_id to keep track of
341 last id that failed so it may be referenced in subsequent ACL's.
343 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
346 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
347 ldap library initialization, allowing self-signed CA's to be
348 used. Also properly sets require_cert option later in code by
349 using NULL (global ldap config) instead of ldap handle (per
350 session). Bug diagnosis and testing by alxgomz.
352 TL/12 Enhanced documentation in the ratelimit.pl script provided in
353 the src/util/ subdirectory.
355 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
356 renamed to Transport Post Delivery Action by Jeremy Harris, as
359 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
360 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
361 redis_servers = needs to be configured which will be used by the redis
362 lookup. Patch from Warren Baker, of The Packet Hub.
364 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
366 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
367 hostname or reverse DNS when processing a host list. Used suggestions
368 from multiple comments on this bug.
370 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
372 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
373 Missed a few lines, added it to make the runtest require no keyboard
376 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
377 contains upper case chars. Make router use caseful_local_part.
379 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
380 support when GnuTLS has been built with p11-kit.
386 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
388 This, or similar/improved, will also be change PP/11 of 4.82.
394 PP/01 Handle short writes when writing local log-files.
395 In practice, only affects FreeBSD (8 onwards).
396 Bugzilla 1053, with thanks to Dmitry Isaikin.
398 NM/01 Bugzilla 949 - Documentation tweak
400 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
403 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
405 PP/02 Implemented gsasl authenticator.
407 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
409 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
410 `pkg-config foo` for cflags/libs.
412 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
413 with rest of GSASL and with heimdal_gssapi.
415 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
416 `pkg-config foo` for cflags/libs for the TLS implementation.
418 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
419 properties get this fed in as external SSF. A number of robustness
420 and debugging improvements to the cyrus_sasl authenticator.
422 PP/08 cyrus_sasl server now expands the server_realm option.
424 PP/09 Bugzilla 1214 - Log authentication information in reject log.
425 Patch by Jeremy Harris.
427 PP/10 Added dbmjz lookup type.
429 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
431 PP/12 MAIL args handles TAB as well as SP, for better interop with
432 non-compliant senders.
433 Analysis and variant patch by Todd Lyons.
435 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
436 Bug report from Lars Müller <lars@samba.org> (via SUSE),
437 Patch from Dirk Mueller <dmueller@suse.com>
439 PP/13 tls_peerdn now print-escaped for spool files.
440 Observed some $tls_peerdn in wild which contained \n, which resulted
441 in spool file corruption.
443 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
444 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
445 or write after TLS renegotiation, which otherwise led to messages
448 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
449 as a tracking header (ie: a signed header comes before the signature).
450 Patch from Wolfgang Breyha.
452 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
453 comma-sep list; embedded commas doubled.
455 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
457 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
459 Report and patch from Dmitry Banschikov.
461 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
462 Removed SSL_clear() after SSL_new() which led to protocol negotiation
463 failures. We appear to now support TLS1.1+ with Exim.
465 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
466 lets Exim select keys and certificates based upon TLS SNI from client.
467 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
468 before an outbound SMTP session. New log_selector, +tls_sni.
470 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
471 NULL dereference. Report and patch from Alun Jones.
473 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
474 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
475 Not seeing resolver debug output on NetBSD, but suspect this is a
476 resolver implementation change.
478 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
479 Left warnings. Added "eximon gdb" invocation mode.
481 PP/21 Defaulting "accept_8bitmime" to true, not false.
483 PP/22 Added -bw for inetd wait mode support.
485 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
486 locate the relevant includes and libraries. Made this the default.
488 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
489 Bugzilla 1246, report and most of solution from Tomasz Kusy.
491 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
492 This may cause build issues on older platforms.
494 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
495 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
496 gnutls_require_mac & gnutls_require_protocols (no longer supported).
497 Added SNI support via GnuTLS too.
498 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
500 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
502 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
503 Applied second patch from Janne, implementing suggestion to default
504 multiple-strings-in-record handling to match SPF spec.
506 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
508 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
509 read-only, out of scope).
510 Patch from Wolfgang Breyha, report from Stuart Northfield.
512 PP/29 Fix three issues highlighted by clang analyser static analysis.
513 Only crash-plausible issue would require the Cambridge-specific
514 iplookup router and a misconfiguration.
515 Report from Marcin Mirosław.
517 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
519 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
520 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
521 As part of this, removing so much warning spew let me fix some minor
522 real issues in debug logging.
524 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
525 assignment on my part. Fixed.
527 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
528 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
529 Janne Snabb (who went above and beyond: thank you).
531 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
532 string otherwise requires a connection and a bunch more work and it's
533 relatively easy to get wrong. Should also expose TLS library linkage
536 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
537 64-bit ${eval} (JH/03).
539 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
540 GNU libc to support some of the 64-bit stuff, should not lead to
541 conflicts. Defined before os.h is pulled in, so if a given platform
542 needs to override this, it can.
544 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
545 protection layer was required, which is not implemented.
546 Bugzilla 1254, patch from Wolfgang Breyha.
548 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
549 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
550 tls_dhparam take prime identifiers. Also unbreak combination of
551 OpenSSL+DH_params+TLSSNI.
553 PP/39 Disable SSLv2 by default in OpenSSL support.
559 PP/01 Solaris build fix for Oracle's LDAP libraries.
560 Bugzilla 1109, patch from Stephen Usher.
562 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
564 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
567 TF/02 Fix a couple more cases where we did not log the error message
568 when unlink() failed. See also change 4.74-TF/03.
570 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
571 lock up or crash if it happened to be inside a call to libc when it
572 got a SIGUSR1 from exiwhat.
574 The SIGUSR1 handler appends the current process status to the process
575 log which is later printed by exiwhat. It used to use the general
576 purpose logging code to do this, but several functions it calls are
577 not safe for signals.
579 The new output code in the SIGUSR1 handler is specific to the process
580 log, and simple enough that it's easy to inspect for signal safety.
581 Removing some special cases also simplifies the general logging code.
582 Removing the spurious timestamps from the process log simplifies
585 TF/04 Improved ratelimit ACL condition.
587 The /noupdate option has been deprecated in favour of /readonly which
588 has clearer semantics. The /leaky, /strict, and /readonly update modes
589 are mutually exclusive. The update mode is no longer included in the
590 database key; it just determines when the database is updated. (This
591 means that when you upgrde Exim will forget old rate measurements.)
593 Exim now checks that the per_* options are used with an update mode that
594 makes sense for the current ACL. For example, when Exim is processing a
595 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
596 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
597 must specify per_mail/readonly. If you omit the update mode it defaults to
598 /leaky where that makes sense (as before) or /readonly where required.
600 The /noupdate option is now undocumented but still supported for
601 backwards compatibility. It is equivalent to /readonly except that in
602 ACLs where /readonly is required you may specify /leaky/noupdate or
603 /strict/noupdate which are treated the same as /readonly.
605 A useful new feature is the /count= option. This is a generalization
606 of the per_byte option, so that you can measure the throughput of other
607 aggregate values. For example, the per_byte option is now equivalent
608 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
610 The per_rcpt option has been generalized using the /count= mechanism
611 (though it's more complicated than the per_byte equivalence). When it is
612 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
613 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
614 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
615 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
616 Note that using per_rcpt with a non-readonly update mode in more than
617 one ACL will cause the recipients to be double-counted. (The per_mail
618 and per_byte options don't have this problem.)
620 The handling of very low rates has changed slightly. If the computed rate
621 is less than the event's count (usually one) then this event is the first
622 after a long gap. In this case the rate is set to the same as this event's
623 count, so that the first message of a spam run is counted properly.
625 The major new feature is a mechanism for counting the rate of unique
626 events. The new per_addr option counts the number of different
627 recipients that someone has sent messages to in the last time period. It
628 behaves like per_rcpt if all the recipient addresses are different, but
629 duplicate recipient addresses do not increase the measured rate. Like
630 the /count= option this is a general mechanism, so the per_addr option
631 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
632 example, measure the rate that a client uses different sender addresses
633 with the options per_mail/unique=$sender_address. There are further
634 details in the main documentation.
636 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
638 TF/06 Removed a few PCRE remnants.
640 TF/07 Automatically extract Exim's version number from tags in the git
641 repository when doing development or release builds.
643 PP/02 Raise smtp_cmd_buffer_size to 16kB.
644 Bugzilla 879. Patch from Paul Fisher.
646 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
647 Heavily based on revision 40f9a89a from Simon Arlott's tree.
650 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
652 PP/05 Variable $av_failed, true if the AV scanner deferred.
653 Bugzilla 1078. Patch from John Horne.
655 PP/06 Stop make process more reliably on build failure.
656 Bugzilla 1087. Patch from Heiko Schlittermann.
658 PP/07 Make maildir_use_size_file an _expandable_ boolean.
659 Bugzilla 1089. Patch from Heiko Schlittermann.
661 PP/08 Handle ${run} returning more data than OS pipe buffer size.
662 Bugzilla 1131. Patch from Holger Weiß.
664 PP/09 Handle IPv6 addresses with SPF.
665 Bugzilla 860. Patch from Wolfgang Breyha.
667 PP/10 GnuTLS: support TLS 1.2 & 1.1.
669 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
672 PP/11 match_* no longer expand right-hand-side by default.
673 New compile-time build option, EXPAND_LISTMATCH_RHS.
674 New expansion conditions, "inlist", "inlisti".
676 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
678 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
680 PP/14 fix log_write() format string regression from TF/03.
681 Bugzilla 1152. Patch from Dmitry Isaikin.
687 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
689 PP/02 Harmonised TLS library version reporting; only show if debugging.
690 Layout now matches that introduced for other libraries in 4.74 PP/03.
692 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
694 PP/04 New "dns_use_edns0" global option.
696 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
699 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
700 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
702 TK/01 Updated PolarSSL code to 0.14.2.
703 Bugzilla 1097. Patch from Andreas Metzler.
705 PP/07 Catch divide-by-zero in ${eval:...}.
708 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
711 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
712 format-string attack -- SECURITY: remote arbitrary code execution.
714 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
715 time unintentionally subject to list matching rules, letting the header
716 cause arbitrary Exim lookups (of items which can occur in lists, *not*
717 arbitrary string expansion). This allowed for information disclosure.
719 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
720 INT_MIN/-1 -- value coerced to INT_MAX.
726 NM/01 Workround for PCRE version dependency in version reporting
729 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
730 This fixes portability to compilers other than gcc, notably
731 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
733 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
734 makefiles for portability to HP-UX and POSIX correctness.
736 PP/01 Permit LOOKUP_foo enabling on the make command-line.
737 Also via indented variable definition in the Makefile.
738 (Debugging by Oliver Heesakkers).
740 PP/02 Restore caching of spamd results with expanded spamd_address.
741 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
743 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
744 Improves build reliability. Fix from: Frank Elsner
746 NM/02 Fix wide character breakage in the rfc2047 coding
747 Fixes bug 1064. Patch from Andrey N. Oktyabrski
749 NM/03 Allow underscore in dnslist lookups
750 Fixes bug 1026. Patch from Graeme Fowler
752 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
753 Code patches from Adam Ciarcinski of NetBSD.
755 NM/04 Fixed exiqgrep to cope with mailq missing size issue
758 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
759 is logged, to avoid truncation. Patch from John Horne.
761 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
762 Patch from Jakob Hirsch.
764 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
765 SQL string expansion failure details.
766 Patch from Andrey Oktyabrski.
768 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
769 Patch from Simon Arlott.
771 PP/09 New lookups functionality failed to compile on old gcc which rejects
772 extern declarations in function scope.
773 Patch from Oliver Fleischmann
775 PP/10 Use sig_atomic_t for flags set from signal handlers.
776 Check getgroups() return and improve debugging.
777 Fixed developed for diagnosis in bug 927 (which turned out to be
780 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
781 Patch from Mark Zealey.
783 PP/12 Bugzilla 1056: Improved spamd server selection.
784 Patch from Mark Zealey.
786 PP/13 Bugzilla 1086: Deal with maildir quota file races.
787 Based on patch from Heiko Schlittermann.
789 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
790 Patch from Uwe Doering, sign-off by Michael Haardt.
792 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
793 variable declaration deep within a block. Bug and patch from
796 PP/15 lookups-Makefile IRIX compatibilty coercion.
798 PP/16 Make DISABLE_DKIM build knob functional.
800 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
801 Patch by Simon Arlott
803 TF/03 Fix valgrind.h portability to C89 compilers that do not support
804 variable argument macros. Our copy now differs from upstream.
810 TF/01 Failure to get a lock on a hints database can have serious
811 consequences so log it to the panic log.
813 TF/02 Log LMTP confirmation messages in the same way as SMTP,
814 controlled using the smtp_confirmation log selector.
816 TF/03 Include the error message when we fail to unlink a spool file.
818 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
819 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
820 for maintaining out-of-tree patches for some time.
822 PP/01 Bugzilla 139: Documentation and portability issues.
823 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
824 Handle per-OS dynamic-module compilation flags.
826 PP/02 Let /dev/null have normal permissions.
827 The 4.73 fixes were a little too stringent and complained about the
828 permissions on /dev/null. Exempt it from some checks.
829 Reported by Andreas M. Kirchwitz.
831 PP/03 Report version information for many libraries, including
832 Exim version information for dynamically loaded libraries. Created
833 version.h, now support a version extension string for distributors
834 who patch heavily. Dynamic module ABI change.
836 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
837 privilege escalation vulnerability whereby the Exim run-time user
838 can cause root to append content of the attacker's choosing to
841 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
844 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
845 If dropping privileges for untrusted macros, we disabled normal logging
846 on the basis that it would fail; for the Exim run-time user, this is not
847 the case, and it resulted in successful deliveries going unlogged.
848 Fixed. Reported by Andreas Metzler.
854 PP/01 Date: & Message-Id: revert to normally being appended to a message,
855 only prepend for the Resent-* case. Fixes regression introduced in
856 Exim 4.70 by NM/22 for Bugzilla 607.
858 PP/02 Include check_rfc2047_length in configure.default because we're seeing
859 increasing numbers of administrators be bitten by this.
861 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
863 PP/03 Bugzilla 994: added openssl_options main configuration option.
865 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
867 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
869 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
871 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
872 without a peer certificate, leading to a segfault because of an
873 assumption that peers always have certificates. Be a little more
874 paranoid. Problem reported by Martin Tscholak.
876 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
877 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
878 NB: ClamAV planning to remove STREAM in "middle of 2010".
879 CL also introduces -bmalware, various -d+acl logging additions and
880 more caution in buffer sizes.
882 PP/09 Implemented reverse_ip expansion operator.
884 PP/10 Bugzilla 937: provide a "debug" ACL control.
886 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
888 PP/12 Bugzilla 973: Implement --version.
890 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
892 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
894 PP/15 Bugzilla 816: support multiple condition rules on Routers.
896 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
897 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
898 ignore trailing whitespace.
900 JJ/02 prevent non-panic DKIM error from being sent to paniclog
902 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
905 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
906 Notification from Dr Andrew Aitchison.
908 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
909 ExtendedDetectionInfo response format.
910 Notification from John Horne.
912 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
915 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
916 XSL and documented dependency on system catalogs, with examples of how
919 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
922 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
923 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
924 configuration file which is writeable by the Exim user or group.
926 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
927 of configuration files to cover files specified with the -C option if
928 they are going to be used with root privileges, not just the default
931 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
932 option (effectively making it always true).
934 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
935 files to be used while preserving root privileges.
937 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
938 that rogue child processes cannot use them.
940 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
941 run-time user, instead of root.
943 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
944 Exim run-time user without dropping privileges.
946 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
947 result string, instead of calling string_vformat() twice with the same
950 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
951 for other users. Others should always drop root privileges if they use
952 -C on the command line, even for a whitelisted configure file.
954 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
956 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
962 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
963 $data_path, and $header_path variables; fixed documentation bugs and
966 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
967 exipick to access non-standard spools, including the "frozen" queue
970 NM/01 Bugzilla 965: Support mysql stored procedures.
971 Patch from Alain Williams
973 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
975 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
976 Patch from Andreas Metzler
978 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
979 Patch from Kirill Miazine
981 NM/05 Bugzilla 671: Added umask to procmail example.
983 JJ/03 installed exipick 20100323.0, fixing doc bug
985 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
986 directory. Notification and patch from Dan Rosenberg.
988 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
990 TK/02 Improve log output when DKIM signing operation fails.
992 MH/01 Treat the transport option dkim_domain as a colon separated
993 list, not as a single string, and sign the message with each element,
994 omitting multiple occurences of the same signer.
996 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
997 Bugzilla 985, 986. Patch by Simon Arlott
999 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1000 Patch by Simon Arlott
1002 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1003 MBX locking. Notification from Dan Rosenberg.
1009 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1011 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1013 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1015 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1017 NM/04 Bugzilla 915: Flag broken perl installation during build.
1023 TK/01 Added patch by Johannes Berg that expands the main option
1024 "spamd_address" if it starts with a dollar sign.
1026 TK/02 Write list of recipients to X-Envelope-Sender header when building
1027 the mbox-format spool file for content scanning (suggested by Jakob
1030 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1031 (http://www.dcc-servers.net/) support via dccifd. Activated by
1032 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1034 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1035 by Mark Daniel Reidel <mr@df.eu>.
1037 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1038 When building exim an external PCRE library is now needed -
1039 PCRE is a system library on the majority of modern systems.
1040 See entry on PCRE_LIBS in EDITME file.
1042 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1043 conversation. Added nologin parameter to request.
1044 Patch contributed by Kirill Miazine.
1046 TF/01 Do not log submission mode rewrites if they do not change the address.
1048 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1050 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1051 log files in place. Contributed by Roberto Lima.
1053 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1055 TF/03 Bugzilla 615: When checking the local_parts router precondition
1056 after a local_part_suffix or local_part_prefix option, Exim now
1057 does not use the address's named list lookup cache, since this
1058 contains cached lookups for the whole local part.
1060 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1061 Robert Millan. Documentation is in experimental-spec.txt.
1063 TF/04 Bugzilla 668: Fix parallel build (make -j).
1065 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1067 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1068 Patch provided by Jan Srzednicki.
1070 TF/05 Leading white space used to be stripped from $spam_report which
1071 wrecked the formatting. Now it is preserved.
1073 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1074 that they are available at delivery time.
1076 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1078 TF/08 TLS error reporting now respects the incoming_interface and
1079 incoming_port log selectors.
1081 TF/09 Produce a more useful error message if an SMTP transport's hosts
1082 setting expands to an empty string.
1084 NM/06 Bugzilla 744: EXPN did not work under TLS.
1085 Patch provided by Phil Pennock.
1087 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1088 Patch provided by Richard Godbee.
1090 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1091 acl_smtp_notquit, added index entry.
1093 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1094 Patch provided by Eugene Bujak.
1096 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1097 accept(). Patch provided by Maxim Dounin.
1099 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1100 Patch provided by Phil Pennock.
1102 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1104 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1105 Patch provided by Brad "anomie" Jorsch.
1107 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1108 Patch provided by Dean Brooks.
1110 TK/05 Add native DKIM support (does not depend on external libraries).
1112 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1113 Patch provided by Graeme Fowler.
1115 NM/16 Bugzilla 851: Documentation example syntax fix.
1117 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1119 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1122 NM/19 Bugzilla 745: TLS version reporting.
1123 Patch provided by Phil Pennock.
1125 NM/20 Bugzilla 167: bool: condition support.
1126 Patch provided by Phil Pennock.
1128 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1129 clients. Patch provided by Phil Pennock.
1131 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1132 Patch provided by Brad "anomie" Jorsch.
1134 NM/23 Bugzilla 687: Fix misparses in eximstats.
1135 Patch provided by Heiko Schlittermann.
1137 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1138 Patch provided by Heiko Schlittermann.
1140 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1141 plus update to original patch.
1143 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1145 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1146 Patch provided by David Brownlee.
1148 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1150 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1152 NM/30 Bugzilla 888: TLS documentation bugfixes.
1154 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1156 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1157 Unlike the original bugzilla I have changed all shell scripts in src tree.
1159 NM/33 Bugzilla 898: Transport filter timeout fix.
1160 Patch by Todd Rinaldo.
1162 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1163 Patch by Serge Demonchaux.
1165 NM/35 Bugzilla 39: Base64 decode bug fixes.
1166 Patch by Jakob Hirsch.
1168 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1170 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1172 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1174 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1180 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1181 ALT-N's libdkim that I have put here:
1182 http://duncanthrax.net/exim-experimental/
1184 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1185 were called 'true' and it seems that C99 defines that as a reserved
1186 keyword to be used with 'bool' variable types. That means you could
1187 not include C99-style headers which use bools without triggering
1188 build errors in sieve.c.
1190 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1191 as mailq or other aliases. Changed the --help handling significantly
1192 to do whats expected. exim_usage() emits usage/help information.
1194 SC/01 Added the -bylocaldomain option to eximstats.
1196 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1198 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1200 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1206 PH/01 Another patch from the Sieve maintainer.
1208 PH/02 When an IPv6 address is converted to a string for single-key lookup
1209 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1210 dots are used instead of colons so that keys in lsearch files need not
1211 contain colons. This was done some time before quoting was made available
1212 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1213 (notated using the quote facility) so as to distinguish them from IPv4
1214 keys. This meant that lookups for IP addresses in host lists did not work
1215 for iplsearch lookups.
1217 This has been fixed by arranging for IPv6 addresses to be expressed with
1218 colons if the lookup type is iplsearch. This is not incompatible, because
1219 previously such lookups could never work.
1221 The situation is now rather anomolous, since one *can* have colons in
1222 ordinary lsearch keys. However, making the change in all cases is
1223 incompatible and would probably break a number of configurations.
1225 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1228 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1229 conversion specification without a maximum field width, thereby enabling
1230 a rogue spamd server to cause a buffer overflow. While nobody in their
1231 right mind would setup Exim to query an untrusted spamd server, an
1232 attacker that gains access to a server running spamd could potentially
1233 exploit this vulnerability to run arbitrary code as the Exim user.
1235 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1236 $primary_hostname instead of what libspf2 thinks the hosts name is.
1238 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1239 a directory entry by the name of the lookup key. Previously, if a
1240 symlink pointed to a non-existing file or a file in a directory that
1241 Exim lacked permissions to read, a lookup for a key matching that
1242 symlink would fail. Now it is enough that a matching directory entry
1243 exists, symlink or not. (Bugzilla 503.)
1245 PH/03 The body_linecount and body_zerocount variables are now exported in the
1248 PH/04 Added the $dnslist_matched variable.
1250 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1251 This means they are set thereafter only if the connection becomes
1254 PH/06 Added the client_condition to authenticators so that some can be skipped
1255 by clients under certain conditions.
1257 PH/07 The error message for a badly-placed control=no_multiline_responses left
1258 "_responses" off the end of the name.
1260 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1262 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1263 (without spaces) instead of just copying the configuration text.
1265 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1267 PH/11 Added $max_received_linelength.
1269 PH/12 Added +ignore_defer and +include_defer to host lists.
1271 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1272 way in which PCRE > 7.0 is built.
1274 PH/14 Implemented queue_only_load_latch.
1276 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1277 MAIL command. The effect was to mangle the value on 64-bit systems.
1279 PH/16 Another patch from the Sieve maintainer.
1281 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1283 PH/18 If a system quota error occurred while trying to create the file for
1284 a maildir delivery, the message "Mailbox is full" was not appended to the
1285 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1286 only to a quota excession during the actual writing of the file.
1288 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1289 characters?) which causes problems in log lines. The DN values are now
1290 passed through string_printing() before being added to log lines.
1292 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1293 and InterBase are left for another time.)
1295 PH/21 Added message_body_newlines option.
1297 PH/22 Guard against possible overflow in moan_check_errorcopy().
1299 PH/23 POSIX allows open() to be a macro; guard against that.
1301 PH/24 If the recipient of an error message contained an @ in the local part
1302 (suitably quoted, of course), incorrect values were put in $domain and
1303 $local_part during the evaluation of errors_copy.
1309 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1310 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1313 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1314 issue a MAIL command.
1316 PH/02 In an ACL statement such as
1318 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1320 if a client was not listed at all, or was listed with a value other than
1321 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1322 the condition was not true (as it should be), so access was not denied.
1323 The bug was that the ! inversion was incorrectly passed on to the second
1324 item. This has been fixed.
1326 PH/03 Added additional dnslists conditions == and =& which are different from
1327 = and & when the dns lookup returns more than one IP address.
1329 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1330 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1332 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1333 FSYNC, which compiles an option called disable_fsync that allows for
1334 bypassing fsync(). The documentation is heavily laced with warnings.
1336 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1338 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1339 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1340 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1341 including adding "make clean"; (3) Added -fPIC when compiling the test
1342 dynamically loaded module, to get rid of a warning.
1344 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1345 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1346 = 0s. The bug is otherwise harmless.
1348 PH/07 There was a bug in the dovecot authenticator such that the value of
1349 $auth1 could be overwritten, and so not correctly preserved, after a
1350 successful authentication. This usually meant that the value preserved by
1351 the server_setid option was incorrect.
1353 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1355 PH/09 Installed PCRE release 7.0.
1357 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1358 run for batched SMTP input. It is now run at the start of every message
1359 in the batch. While fixing this I discovered that the process information
1360 (output by running exiwhat) was not always getting set for -bs and -bS
1361 input. This is fixed, and it now also says "batched" for BSMTP.
1363 PH/11 Added control=no_pipelining.
1365 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1366 patch, slightly modified), and move the expansion of helo_data till after
1367 the connection is made in the smtp transport (so it can use these
1370 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1372 PH/14 Added log_selector = +pid.
1374 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1376 PH/16 Add ${if forany and ${if forall.
1378 PH/17 Added dsn_from option to vary the From: line in DSNs.
1380 PH/18 Flush SMTP output before performing a callout, unless control =
1381 no_callout_flush is set.
1383 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1384 was true (the default) a successful delivery failed to delete the retry
1385 item, thus causing premature timeout of the address. The bug is now
1388 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1390 PH/21 Long custom messages for fakedefer and fakereject are now split up
1391 into multiline reponses in the same way that messages for "deny" and
1392 other ACL rejections are.
1394 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1395 with slight modification.
1397 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1398 draft, changing the syntax and factoring some duplicate code".
1400 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1401 for deliveries of the second and subsequent messages over the same SMTP
1404 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1405 ${reduce, with only minor "tidies".
1407 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1409 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1410 expansion side effects.
1412 PH/27 When a message times out after an over-quota error from an Exim-imposed
1413 quota, the bounce message says "mailbox is full". This message was not
1414 being given when it was a system quota that was exceeded. It now should
1417 MH/03 Made $recipients available in local_scan(). local_scan() already has
1418 better access to the recipient list through recipients_list[], but
1419 $recipients can be useful in postmaster-provided expansion strings.
1421 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1422 in the case of a MAIL command with additional options following the
1423 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1424 were accidentally chopped off.
1426 PH/29 SMTP synchronization checks are implemented when a command is read -
1427 there is a check that no more input is waiting when there shouldn't be
1428 any. However, for some commands, a delay in an ACL can mean that it is
1429 some time before the response is written. In this time, more input might
1430 arrive, invalidly. So now there are extra checks after an ACL has run for
1431 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1432 pipelining has not been advertised.
1434 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1436 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1437 returned in the SMTP response when smtp_return_error_details was false.
1438 This has been fixed.
1440 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1441 instead of the C I/O that was originally supplied, because problems were
1442 reported on Solaris.
1444 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1445 Exim which did not show up earlier: it was assuming that a call to
1446 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1447 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1448 was a macro that became an assignment, so it seemed to work. This has
1449 changed to a proper function call with a void return, hence the compile
1450 error. Exim's code has been fixed.
1452 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1455 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1457 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1459 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1460 because it turns out to be too big or there is a timeout) there is no
1461 "Completed" line in the log. When some messages of this type were
1462 selected by exigrep, they were listed as "not completed". Others were
1463 picked up by some special patterns. I have improved the selection
1464 criteria to be more general.
1466 PH/38 The host_find_failed option in the manualroute router can now be set
1467 to "ignore", to completely ignore a host whose IP address cannot be
1468 found. If all hosts are ignored, the behaviour is controlled by the new
1469 host_all_ignored option.
1471 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1472 homing or because of multiple MX records with /mx) generated more than
1473 one IP address, and the following item turned out to be the local host,
1474 all the secondary addresses of the first item were incorrectly removed
1475 from the list, along with the local host and any following hosts (which
1476 is what is supposed to happen).
1478 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1479 whoever called Exim into the -H file. In the case of the daemon it was
1480 behaving confusingly. When first started, it used values for whoever
1481 started the daemon, but after a SIGHUP it used the Exim user (because it
1482 calls itself on a restart). I have changed the code so that it now always
1485 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1486 message are rejected with the same error (e.g. no authentication or bad
1487 sender address), and a DATA command is nevertheless sent (as can happen
1488 with PIPELINING or a stupid MUA), the error message that was given to the
1489 RCPT commands is included in the rejection of the DATA command. This is
1490 intended to be helpful for MUAs that show only the final error to their
1493 PH/42 Another patch from the Sieve maintainer.
1495 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1496 Eximstats - Fixed some broken HTML links and added missing column headers
1498 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1499 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1501 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1503 PH/43 Yet another patch from the Sieve maintainer.
1505 PH/44 I found a way to check for a TCP/IP connection going away before sending
1506 the response to the final '.' that terminates a message, but only in the
1507 case where the client has not sent further data following the '.'
1508 (unfortunately, this is allowed). However, in many cases there won't be
1509 any further data because there won't be any more messages to send. A call
1510 to select() can be used: if it shows that the input is "ready", there is
1511 either input waiting, or the socket has been closed. An attempt to read
1512 the next input character can distinguish the two cases. Previously, Exim
1513 would have sent an OK response which the client would never have see.
1514 This could lead to message repetition. This fix should cure that, at
1515 least in a lot of common cases.
1517 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1518 advertised in response to EHLO.
1524 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1525 fixed by 4.65/MH/01 (is this a record?) are fixed:
1527 (i) An empty string was always treated as zero by the numeric comparison
1528 operators. This behaviour has been restored.
1530 (ii) It is documented that the numeric comparison operators always treat
1531 their arguments as decimal numbers. This was broken in that numbers
1532 starting with 0 were being interpreted as octal.
1534 While fixing these problems I realized that there was another issue that
1535 hadn't been noticed. Values of message_size_limit (both the global option
1536 and the transport option) were treated as octal if they started with 0.
1537 The documentation was vague. These values are now always treated as
1538 decimal, and I will make that clear in the documentation.
1544 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1545 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1548 MH/01 Don't check that the operands of numeric comparison operators are
1549 integers when their expansion is in "skipping" mode (fixes bug
1550 introduced by 4.64-PH/07).
1552 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1553 child addresses, Exim now panics and dies. Previously, because the count
1554 is held in a short int, deliveries were likely to be lost. As such a
1555 large number of recipients for a single message is ridiculous
1556 (performance will be very, very poor), I have chosen to impose a limit
1557 rather than extend the field.
1563 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1564 leftover -K file (the existence of which was triggered by #402).
1565 While we were at it, introduced process PID as part of the -K
1566 filename. This should rule out race conditions when creating
1569 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1570 processing considerably. Previous code took too long for large mails,
1571 triggering a timeout which in turn triggers #401.
1573 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1574 in the DK code in transports.c. sendfile() is not really portable,
1575 hence the _LINUX specificness.
1577 TF/01 In the add_headers option to the mail command in an Exim filter,
1578 there was a bug that Exim would claim a syntax error in any
1579 header after the first one which had an odd number of characters
1582 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1583 callout verification, Exim cached a "reject" for the entire domain. This
1584 is correct for most verifications, but it is not correct for a recipient
1585 verification with use_sender or use_postmaster set, because in that case
1586 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1587 case of MAIL FROM:<> rejection from other early rejections (e.g.
1588 rejection of HELO). When verifying a recipient using a non-null MAIL
1589 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1590 Whatever the result of the callout, the value of the domain cache is
1591 left unchanged (for any other kind of callout, getting as far as trying
1592 RCPT means that the domain itself is ok).
1594 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1597 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1598 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1599 session, the daemon ignores ECONNECT errors and logs others; it now
1600 ignores EPIPE as well.
1602 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1603 (quoted-printable decoding).
1605 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1606 later the small subsequent patch to fix an introduced bug.
1608 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1610 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1612 PH/08 An error is now given if message_size_limit is specified negative.
1614 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1615 to be given (somewhat) arbitrary names.
1617 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1620 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1621 miscellaneous code fixes
1623 PH/10 Added the log_reject_target ACL modifier to specify where to log
1626 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1627 hostname. This is wrong, because it relates to the incoming message (and
1628 probably the interface on which it is arriving) and not to the outgoing
1629 callout (which could be using a different interface). This has been
1630 changed to use the value of the helo_data option from the smtp transport
1631 instead - this is what is used when a message is actually being sent. If
1632 there is no remote transport (possible with a router that sets up host
1633 addresses), $smtp_active_hostname is used.
1635 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1636 tweaks were necessary in order to get it to work (see also 21 below):
1637 (a) The code assumed that strncpy() returns a negative number on buffer
1638 overflow, which isn't the case. Replaced with Exim's string_format()
1640 (b) There were several signed/unsigned issues. I just did the minimum
1641 hacking in of casts. There is scope for a larger refactoring.
1642 (c) The code used strcasecmp() which is not a standard C function.
1643 Replaced with Exim's strcmpic() function.
1644 (d) The code set only $1; it now sets $auth1 as well.
1645 (e) A simple test gave the error "authentication client didn't specify
1646 service in request". It would seem that Dovecot has changed its
1647 interface. Fortunately there's a specification; I followed it and
1648 changed what the client sends and it appears to be working now.
1650 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1653 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1654 address A is aliased to B and C, where B exists and C does not. Without
1655 -v the output is "A verified" because verification stops after a
1656 successful redirection if more than one address is generated. However,
1657 with -v the child addresses are also verified. Exim was outputting "A
1658 failed to verify" and then showing the successful verification for C,
1659 with its parentage. It now outputs "B failed to verify", showing B's
1660 parentage before showing the successful verification of C.
1662 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1663 look up a TXT record in a specific list after matching in a combined
1666 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1667 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1668 they consult the DNS. I had assumed they would set it the way they
1669 wanted; and indeed my experiments on Linux seem to show that in some
1670 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1671 To be on the safe side, however, I have now made the interface to
1672 host_find_byname() similar to host_find_bydns(), with an argument
1673 containing the DNS resolver options. The host_find_byname() function now
1674 sets these options at its start, just as host_find_bydns() does. The smtp
1675 transport options dns_qualify_single and dns_search_parents are passed to
1676 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1677 of host_find_byname() use the default settings of RES_DEFNAMES
1678 (qualify_single) but not RES_DNSRCH (search_parents).
1680 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1681 spool_read_header() do less string testing, by means of a preliminary
1682 switch on the second character of optional "-foo" lines. (This is
1683 overdue, caused by the large number of possibilities that now exist.
1684 Originally there were few.) While I was there, I also converted the
1685 str(n)cmp tests so they don't re-test the leading "-" and the first
1686 character, in the hope this might squeeze out yet more improvement.
1688 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1689 flag allowing group syntax was set by the header_syntax check but not
1690 turned off, possible causing trouble later; (2) The flag was not being
1691 set at all for the header_verify test, causing "group"-style headers to
1692 be rejected. I have now set it in this case, and also caused header_
1693 verify to ignore an empty address taken from a group. While doing this, I
1694 came across some other cases where the code for allowing group syntax
1695 while scanning a header line wasn't quite right (mostly, not resetting
1696 the flag correctly in the right place). These bugs could have caused
1697 trouble for malformed header lines. I hope it is now all correct.
1699 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1700 with the "reply" argument non-NULL. The code, however (which originally
1701 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1702 but it didn't always do it. This confused somebody who was copying the
1703 code for some other use. I have removed all the tests.
1705 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1706 feature that was used to support insecure browsers during the U.S. crypto
1707 embargo. It requires special client support, and Exim is probably the
1708 only MTA that supported it -- and would never use it because real RSA is
1709 always available. This code has been removed, because it had the bad
1710 effect of slowing Exim down by computing (never used) parameters for the
1711 RSA_EXPORT functionality.
1713 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1714 authenticator to fail if there's a tab character in the incoming data
1715 (there should never be unless someone is messing about, as it's supposed
1716 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1717 if the connection is using TLS or if the remote IP is the same as the
1718 local IP, and the "valid-client-cert option" if a client certificate has
1721 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1722 authenticators. This can be used for authorization after authentication
1723 succeeds. (In the case of plaintext, it servers for both authentication
1726 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1727 if any retry times were supplied.
1729 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1730 connection, where there is no client IP address to check. In this
1731 situation, the verify now always succeeds.
1733 PH/25 Applied John Jetmore's -Mset patch.
1735 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1737 PH/27 In a string expansion for a processed (not raw) header when multiple
1738 headers of the same name were present, leading whitespace was being
1739 removed from all of them, but trailing whitespace was being removed only
1740 from the last one. Now trailing whitespace is removed from each header
1741 before concatenation. Completely empty headers in a concatenation (as
1742 before) are ignored.
1744 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1745 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1747 PH/29 [Removed. This was a change that I later backed out, and forgot to
1748 correct the ChangeLog entry (that I had efficiently created) before
1749 committing the later change.]
1751 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1752 address errors (4xx response to RCPT) over the same connection as other
1753 messages routed to the same hosts. Such deliveries are always "forced",
1754 so retry times are not inspected. This resulted in far too many retries
1755 for the affected addresses. The effect occurred only when there were more
1756 hosts than the hosts_max_try setting in the smtp transport when it had
1757 the 4xx errors. Those hosts that it had tried were not added to the list
1758 of hosts for which the message was waiting, so if all were tried, there
1759 was no problem. Two fixes have been applied:
1761 (i) If there are any address or message errors in an SMTP delivery, none
1762 of the hosts (tried or untried) are now added to the list of hosts
1763 for which the message is waiting, so the message should not be a
1764 candidate for sending over the same connection that was used for a
1765 successful delivery of some other message. This seems entirely
1766 reasonable: after all the message is NOT "waiting for some host".
1767 This is so "obvious" that I'm not sure why it wasn't done
1768 previously. Hope I haven't missed anything, but it can't do any
1769 harm, as the worst effect is to miss an optimization.
1771 (ii) If, despite (i), such a delivery is accidentally attempted, the
1772 routing retry time is respected, so at least it doesn't keep
1773 hammering the server.
1775 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1776 in ${readsocket because some servers need this prod.
1778 PH/32 Added some extra debug output when updating a wait-xxx database.
1780 PH/33 The hint "could be header name not terminated by colon", which has been
1781 given for certain expansion errors for a long time, was not being given
1782 for the ${if def:h_colon_omitted{... case.
1784 PH/34 The spec says: "With one important exception, whenever a domain list is
1785 being scanned, $domain contains the subject domain." There was at least
1786 one case where this was not true.
1788 PH/35 The error "getsockname() failed: connection reset by peer" was being
1789 written to the panic log as well as the main log, but it isn't really
1790 panic-worthy as it just means the connection died rather early on. I have
1791 removed the panic log writing for the ECONNRESET error when getsockname()
1794 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1795 runs only) independently of the message's sender address. This meant
1796 that, if the 4xx error was in fact related to the sender, a different
1797 message to the same recipient with a different sender could confuse
1798 things. In particualar, this can happen when sending to a greylisting
1799 server, but other circumstances could also provoke similar problems.
1800 I have changed the default so that the retry time for these errors is now
1801 based a combination of the sender and recipient addresses. This change
1802 can be overridden by setting address_retry_include_sender=false in the
1805 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1806 remote server are returned as part of bounce messages. This was not
1807 happening for LMTP over a pipe (the lmtp transport), but now it is the
1808 same for both kinds of LMTP.
1810 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1811 in header lines that were in fact CNAMEs. This is no longer the case.
1813 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1814 and queue runs started by the daemon processed all messages. This has
1815 been fixed so that -R and -S can now usefully be given with -q<time>.
1817 PH/40 Import PCRE release 6.7 (fixes some bugs).
1819 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1821 PH/42 Give an error if -q is specified more than once.
1823 PH/43 Renamed the variables $interface_address and $interface_port as
1824 $received_ip_address and $received_port, to make it clear that these
1825 values apply to message reception, and not to the outgoing interface when
1826 a message is delivered. (The old names remain recognized, of course.)
1828 PH/44 There was no timeout on the connect() call when using a Unix domain
1829 socket in the ${readsocket expansion. There now is.
1831 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1832 be meaningful with "accept".
1834 SC/01 Eximstats V1.43
1835 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1837 SC/02 Eximstats V1.44
1838 Use a glob alias rather than an array ref in the generated
1839 parser. This improves both readability and performance.
1841 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1842 Collect SpamAssassin and rejection statistics.
1843 Don't display local sender or destination tables unless
1844 there is data to show.
1845 Added average volumes into the top table text output.
1847 SC/04 Eximstats V1.46
1848 Collect data on the number of addresses (recipients)
1849 as well as the number of messages.
1851 SC/05 Eximstats V1.47
1852 Added 'Message too big' to the list of mail rejection
1853 reasons (thanks to Marco Gaiarin).
1855 SC/06 Eximstats V1.48
1856 Mainlog lines which have GMT offsets and are too short to
1857 have a flag are now skipped.
1859 SC/07 Eximstats V1.49 (Alain Williams)
1860 Added the -emptyok flag.
1862 SC/08 Eximstats V1.50
1863 Fixes for obtaining the IP address from reject messages.
1865 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1866 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1867 whitesspace changes from 4.64-PH/27
1869 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1872 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1873 are found, allow negative numbers in numeric criteria)
1875 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1877 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1878 to match changes made in 4.64-PH/43
1880 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1882 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1883 "domain" is actually an IP address, and force a failure. This locks out
1884 those revolvers/nameservers that support "A-for-A" lookups, in
1885 contravention of the specifications.
1887 PH/48 When a host name was looked up from an IP address, and the subsequent
1888 forward lookup of the name timed out, the host name was left in
1889 $sender_host_name, contrary to the specification.
1891 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1892 restricted to single-key lookups, Exim was not diagnosing an error if
1893 * or *@ was used with a query-style lookup.
1895 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1897 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1898 long ago, but noone interested enough thought of it. Let's just say that
1899 the "1.1" means that there are some new functions that weren't there at
1900 some point in the past.
1902 PH/51 Error processing for expansion failure of helo_data from an smtp
1903 transport during callout processing was broken.
1905 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1906 tested/used via the -bh/-bhc/-bs options.
1908 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1909 bug, fixed in subsequent PCRE releases).
1911 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1912 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1914 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1920 SC/01 Use a glob alias rather than an array ref in eximstats generated
1921 parser. This improves both readability and performance.
1923 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1924 Don't display local sender or destination tables in eximstats unless
1925 there is data to show.
1926 Added average volumes into the eximstats top table text output.
1928 SC/03 Collect data on the number of addresses (recipients) as well
1929 as the number of messages in eximstats.
1931 TF/01 Correct an error in the documentation for the redirect router. Exim
1932 does (usually) call initgroups() when daemonizing.
1934 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1935 with consistent privilege compared to when running as a daemon.
1937 TF/03 Note in the spec that $authenticated_id is not set for local
1938 submissions from trusted users.
1940 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1941 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1943 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1944 by adding some example configuration directives to the default
1945 configuration file. A little bit of work is required to uncomment the
1946 directives and define how usernames and passwords are checked, but
1947 there is now a framework to start from.
1949 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1950 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1951 without this. I don't know how relevant this is to other LDAP libraries.
1953 PH/02 Add the verb name to the "unknown ACL verb" error.
1955 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1957 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1959 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1960 directory not expanded when it should be if an expanded home directory
1961 was set for the address (which is overridden by the transport).
1963 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1966 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1967 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1968 because it is too late at that time, and has no effect.
1970 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1971 security issue with \' (bugzilla #107). I could not use the
1972 PQescapeStringConn() function, because it needs a PGconn value as one of
1975 PH/08 When testing addresses using -bt, indicate those final addresses that
1976 are duplicates that would not cause an additional delivery. At least one
1977 person was confused, thinking that -bt output corresponded to deliveries.
1978 (Suppressing duplicates isn't a good idea as you lose the information
1979 about possibly different redirections that led to the duplicates.)
1981 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1982 systems where poll() doesn't work, in particular OS X.
1984 PH/10 Added more information to debugging output for retry time not reached.
1986 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1987 operations in malware.c.
1989 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1992 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1993 syslog_duplication set false, log lines that would normally be written
1994 both the the main log and to the reject log were not written to syslog at
1997 PH/14 In the default configuration, change the use of "message" in ACL warn
1998 statements to "add_header".
2000 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2001 not followed by a command (e.g. "seen endif").
2003 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2004 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2007 PH/17 Added extra conditions to the default value of delay_warning_condition
2011 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2012 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2013 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2016 The Auto-Submitted: and various List- headers are standardised, whereas I
2017 don't think Precedence: ever was.
2019 PH/18 Refactored debugging code in route_finduser() to show more information,
2020 in particular, the error code if getpwnam() issues one.
2022 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2023 This is apparently needed in addition to the PH/07 change above to avoid
2024 any possible encoding problems.
2026 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2027 but not after initializing Perl.
2029 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2030 output them only if debugging. By default they are written stderr,
2031 apparently, which is not desirable.
2033 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2036 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2039 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2041 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2042 authenticated or an ident call has been made. Suppress the default
2043 values for $authenticated_id and $authenticated_sender (but permit -oMai
2044 and -oMas) when testing with -bh.
2046 PH/24 Re-jigged the order of the tests in the default configuration so that the
2047 tests for valid domains and recipients precede the DNS black list and CSA
2048 tests, on the grounds that those ones are more expensive.
2050 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2051 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2052 This bug exists in every version of Exim that I still have, right back to
2055 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2056 However, an attempt to turn on case-sensitivity in a regex key by
2057 including (?-i) didn't work because the subject string was already
2058 lowercased, and the effects were non-intuitive. It turns out that a
2059 one-line patch can be used to allow (?-i) to work as expected.
2065 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2066 other effects) broke the use of negated acl sub-conditions.
2068 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2071 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2072 "Deny" causes Exim to reject the incoming connection with a 554 error.
2073 Unfortunately, if there is a major crisis, such as a disk failure,
2074 tcp-wrappers gives "deny", whereas what one would like would be some
2075 kind of temporary error. A kludge has been added to help with this.
2076 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2077 554 error is used if errno is still zero or contains ENOENT (which occurs
2078 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2081 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2083 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2084 errors. Otherwise a message that provokes a temporary error (when other
2085 messages do not) can cause a whole host to time out.
2087 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2088 addresses were routed directly to files or pipes from a redirect router.
2089 File deliveries just didn't batch; pipe deliveries might have suffered
2092 PH/06 A failure to get a lock for a hints database would erroneously always say
2093 "Failed to get write lock", even when it was really a read lock.
2095 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2096 of 0600. This has been changed to use the value of the lockfile_mode
2097 option (which defaults to 0600).
2099 PH/08 Applied small patch from the Sieve maintainer.
2101 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2102 folder from quota calculations, a direct delivery into this folder messed
2103 up the contents of the maildirsize file. This was because the regex was
2104 used only to exclude .Trash (or whatever) when the size of the mailbox
2105 was calculated. There was no check that a delivery was happening into an
2106 excluded directory. This bug has been fixed by ignoring all quota
2107 processing for deliveries into excluded directories.
2109 PH/10 Added the maildirfolder_create_regex option to appendfile.
2115 PH/01 The code for finding all the local interface addresses on a FreeBSD
2116 system running IPv6 was broken. This may well have applied to all BSD
2117 systems, as well as to others that have similar system calls. The broken
2118 code found IPv4 interfaces correctly, but gave incorrect values for the
2119 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2120 that it would not match correctly against @[] and not recognize the IPv6
2123 PH/02 The ipliteral router was not recognizing addresses of the form user@
2124 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2126 PH/03 Added disable_ipv6.
2128 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2129 decoded form, because it is most often used to construct To: headers
2130 lines in autoreplies, and the decoded form may well be syntactically
2131 invalid. However, $reply_address has leading white space removed, and all
2132 newlines turned into spaces so that the autoreply transport does not
2135 PH/05 If group was specified without a user on a router, and no group or user
2136 was specified on a transport, the group from the router was ignored.
2138 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2139 for visible compile-time settings that can be used to change these
2140 numbers, for those that want even more. Backwards compatibility with old
2141 spool files has been maintained. However, going back to a previous Exim
2142 release will lost any variables that are in spool files.
2144 PH/07 Two small changes when running in the test harness: increase delay when
2145 passing a TCP/IP connection to a new process, in case the original
2146 process has to generate a bounce, and remove special handling of
2147 127.0.0.2 (sic), which is no longer necessary.
2149 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2150 be the same on different OS.
2152 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2155 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2156 whether --show-vars was specified or not
2158 JJ/02 exipick: Added support for new ACL variable spool format introduced
2161 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2162 syntactically invalid From: or Reply-to: line, and a filter used this to
2163 generate an autoreply, and therefore failed to obtain an address for the
2164 autoreply, Exim could try to deliver to a non-existent relative file
2165 name, causing unrelated and misleading errors. What now happens is that
2166 it logs this as a hard delivery error, but does not attempt to create a
2169 PH/11 The exinext utility has a -C option for testing purposes, but although
2170 the given file was scanned by exinext itself; it wasn't being passed on
2171 when Exim was called.
2173 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2174 an end-of-file indication when reading a command response.
2176 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2177 compiled. In many other places in Exim, IPv6 addresses are always
2178 recognized, so I have changed this. It also means that IPv4 domain
2179 literals of the form [IPV4:n.n.n.n] are now always recognized.
2181 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2182 used if the router is not running as root, for example, when verifying at
2183 ACL time, or when using -bh. The debugging output from this situation was
2184 non-existent - all you got was a failure to exec. I have made two
2187 (a) Failures to set uid/gid, the current directory, or a process leader
2188 in a subprocess such as that created by queryprogram now generate
2189 suitable debugging ouput when -d is set.
2191 (b) The queryprogram router detects when it is not running as root,
2192 outputs suitable debugging information if -d is set, and then runs
2193 the subprocess without attempting to change uid/gid.
2195 PH/15 Minor change to Makefile for building test_host (undocumented testing
2198 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2199 additional section of a DNS packet that returns MX or SRV records.
2200 Instead, it always explicitly searches for A/AAAA records. This avoids
2201 major problems that occur when a DNS server includes only records of one
2202 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2203 fixed another bug: if SRV records were looked up and the corresponding
2204 address records were *not* found in the additional section, the port
2205 values from the SRV records were lost.
2207 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2208 using the correct key (the original address) when searching the retry
2209 rules in order to find which one to use for generating the retry hint.
2211 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2212 adding the default one. Similarly, if it contains a Reply-To: header, the
2213 errors_reply_to option, if set, is not used.
2215 PH/19 When calculating a retry time, Exim used to measure the "time since
2216 failure" by looking at the "first failed" field in the retry record. Now
2217 it does not use this if it is later than than the arrival time of the
2218 message. Instead it uses the arrival time. This makes for better
2219 behaviour in cases where some deliveries succeed, thus re-setting the
2220 "first failed" field. An example is a quota failure for a huge message
2221 when small messages continue to be delivered. Without this change, the
2222 "time since failure" will always be short, possible causing more frequent
2223 delivery attempts for the huge message than are intended.
2224 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2226 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2227 $1, $2, $3) because the numerical variables can be reset during some
2228 expansion items (e.g. "match"), thereby losing the authentication data.
2230 PH/21 Make -bV show the size of off_t variables so that the test suite can
2231 decide whether to run tests for quotas > 2G.
2233 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2234 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2235 filecount value is greater than 2G or if a quota value is greater than 2G
2236 on a system where the size of off_t is not greater than 4, a panic error
2239 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2240 never match. The debug and -bh output now contains an explicit error
2241 message indicating a malformed IPv4 address or mask.
2243 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2244 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2245 PH/23 above applies.
2247 PH/25 Do not write to syslog when running in the test harness. The only
2248 occasion when this arises is a failure to open the main or panic logs
2249 (for which there is an explicit test).
2251 PH/26 Added the /no_tell option to "control=freeze".
2253 PH/27 If a host name lookup failed very early in a connection, for example, if
2254 the IP address matched host_lookup and the reverse lookup yielded a name
2255 that did not have a forward lookup, an error message of the form "no IP
2256 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2257 could be logged. Now it outputs the IP address instead of "NULL".
2259 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2260 allows the sender and the authenticated sender to be set when
2261 submitting a message from within Exim. Since child_open_exim() is
2262 documented for local_scan(), the new function should be too.
2264 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2265 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2266 results in an empty string is now treated as unset.
2268 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2270 PH/31 Added sender_verify_fail logging option.
2272 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2273 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2274 tidied the source and removed it altogether.
2276 PH/33 When a queue run was abandoned because the load average was too high, a
2277 log line was always written; now it is written only if the queue_run log
2278 selector is set. In addition, the log line for abandonment now contains
2279 information about the queue run such as the pid. This is always present
2280 in "start" and "stop" lines but was omitted from the "abandon" line.
2282 PH/34 Omit spaces between a header name and the colon in the error message that
2283 is given when verify = headers_syntax fails (if there are lots of them,
2284 the message gets confusing).
2286 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2287 names, as there are now some PTR records that contain slashes. This check
2288 is only to protect against broken name servers that fall over on strange
2289 characters, so the fact that it applies to all lookups doesn't matter.
2291 PH/36 Now that the new test suite is complete, we can remove some of the
2292 special code in Exim that was needed for the old test suite. For example,
2293 sorting DNS records because real resolvers return them in an arbitrary
2294 order. The new test suite's fake resolver always returns records in the
2297 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2298 bounces) except when queue_only is set, to avoid logging races between
2299 the different processes.
2301 PH/38 Panic-die if .include specifies a non-absolute path.
2303 PH/39 A tweak to the "H" retry rule from its user.
2305 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2306 a label. They prevented compilation on older perls.
2308 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2309 a warning to be raised on newish perls.
2311 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2312 on queue. Changes to match documented behaviour of showing count of
2313 messages matching specified criteria.
2315 PH/40 Changed the default ident timeout from 30s to 5s.
2317 PH/41 Added support for the use of login_cap features, on those BSD systems
2318 that have them, for controlling the resources used by pipe deliveries.
2320 PH/42 The content-scanning code uses fopen() to create files in which to put
2321 message data. Previously it was not paying any attention to the mode of
2322 the files. Exim runs with umask(0) because the rest of the code creates
2323 files with open(), and sets the required mode explicitly. Thus, these
2324 files were ending up world-writeable. This was not a big issue, because,
2325 being within the spool directory, they were not world-accessible. I have
2326 created a function called modefopen, which takes an additional mode
2327 argument. It sets umask(777), creates the file, chmods it to the required
2328 mode, then resets the umask. All the relevant calls to fopen() in the
2329 content scanning code have been changed to use this function.
2331 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2332 to 24 hours. This avoids potential overflow problems when processing G
2333 and H retry rules. I suspect nobody ever tinkers with this value.
2335 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2337 PH/45 When the plaintext authenticator is running as a client, the server's
2338 challenges are checked to ensure they are valid base64 strings. By
2339 default, the authentication attempt is cancelled if an invalid string is
2340 received. Setting client_ignore_invalid_base64 true ignores these errors.
2341 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2342 they are received. Thus, the responses can be made to depend on the
2343 challenges. If an invalid string is ignored, an empty string is placed in
2346 PH/46 Messages that are created by the autoreply transport now contains a
2347 References: header, in accordance with RFCs 2822 and 3834.
2349 PH/47 Added authenticated_sender_force to the smtp transport.
2351 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2353 PH/49 Installed latest patch from the Sieve maintainer.
2355 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2356 was also set, the appendfile transport was unnecessarily scanning a
2357 directory of message files (e.g. for maildir delivery) to find the count
2358 of files (along with the size), even though it did not need this
2359 information. It now does the scan only if it needs to find either the
2360 size of the count of files.
2362 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2364 PH/52 Two bugs concerned with error handling when the smtp transport is
2367 (i) Exim was not creating retry information for temporary errors given
2368 for individual recipients after the DATA command when the smtp transport
2369 was used in LMTP mode. This meant that they could be retried too
2370 frequently, and not timed out correctly.
2372 (ii) Exim was setting the flag that allows error details to be returned
2373 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2374 recipients that were returned after the DATA command.
2376 PH/53 This is related to PH/52, but is more general: for any failing address,
2377 when detailed error information was permitted to be returned to the
2378 sender, but the error was temporary, then after the final timeout, only
2379 "retry timeout exceeded" was returned. Now it returns the full error as
2380 well as "retry timeout exceeded".
2382 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2383 do this, and (what is worse) MTAs that accept it.
2385 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2386 will now be deprecated.
2388 PH/56 New os.c-cygwin from the Cygwin maintainer.
2390 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2391 formats (previously only available in exim formats via -bpr, -bpru,
2392 and -bpra. Now also available in native and exiqgrep formats)
2394 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2395 with very large, slow to parse queues
2397 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2399 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2401 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2402 responses to authentication challenges, though it was showing the
2403 challenges; (ii) I've removed the CR characters from the debug output for
2406 PH/58 Allow for the insertion of a newline as well as a space when a string
2407 is turned into more than one encoded-word during RFC 2047 encoding. The
2408 Sieve code now uses this.
2410 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2411 data_4xx, lost_connection, tls_required.
2413 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2414 message was being sent as an SMTP response.
2416 PH/61 Add -l and -k options to exicyclog.
2418 PH/62 When verifying, if an address was redirected to one new address, so that
2419 verification continued, and the new address failed or deferred after
2420 having set something in $address_data, the value of $address_data was not
2421 passed back to the ACL. This was different to the case when no
2422 redirection occurred. The value is now passed back in both cases.
2424 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2425 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2426 use login_cap.h, so on its own it isn't the distinguishing feature. The
2427 new name refers directly to the setclassresources() function.
2429 PH/65 Added configuration files for NetBSD3.
2431 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2433 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2434 is preferred over IPv4.
2436 PH/68 The bounce_return_message and bounce_return_body options were not being
2437 honoured for bounces generated during the reception of non-SMTP messages.
2438 In particular, this applied to messages rejected by the ACL. This bug has
2439 been fixed. However, if bounce_return_message is true and bounce_return_
2440 body is false, the headers that are returned for a non-SMTP message
2441 include only those that have been read before the error was detected.
2442 (In the case of an ACL rejection, they have all been read.)
2444 PH/69 The HTML version of the specification is now built in a directory called
2445 spec_html instead of spec.html, because the latter looks like a path with
2446 a MIME-type, and this confuses some software.
2448 PH/70 Catch two compiler warnings in sieve.c.
2450 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2451 function verify_get_ident() calls ip_connect() to connect a socket, but
2452 if the "connect()" function timed out, ip_connect() used to close the
2453 socket. However, verify_get_ident() also closes the socket later, and in
2454 between Exim writes to the log, which may get opened at this point. When
2455 the socket was closed in ip_connect(), the log could get the same file
2456 descriptor number as the socket. This naturally causes chaos. The fix is
2457 not to close the socket in ip_connect(); the socket should be closed by
2458 the function that creates it. There was only one place in the code where
2459 this was missing, in the iplookup router, which I don't think anybody now
2460 uses, but I've fixed it anyway.
2462 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2463 well as to direct DNS lookups. Otherwise the handling of names in host
2464 lists is inconsistent and therefore confusing.
2470 PH/01 Two changes to the default runtime configuration:
2472 (1) Move the checks for relay_from_hosts and authenticated clients from
2473 after to before the (commented out) DNS black list checks.
2475 (2) Add control=submission to the relay_from_hosts and authenticated
2476 clients checks, on the grounds that messages accepted by these
2477 statements are most likely to be submissions.
2479 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2481 (1) Generate an error if the third argument for the ${prvs expansion is
2484 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2487 (3) Reset the variables that are obtained from the first argument of
2488 ${prvscheck and used in the second argument before leaving the code,
2489 because their memory is reclaimed, so using them afterwards may do
2492 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2493 one (it's much easier than Tom thought :-).
2495 (5) Because of (4), we can now allow for the use of $prvscheck_result
2496 inside the third argument.
2498 PH/03 For some reason, the default setting of PATH when running a command from
2499 a pipe transport was just "/usr/bin". I have changed it to
2502 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2503 anything to be listed in the output from -bV.
2505 PH/05 When a filter generated an autoreply, the entire To: header line was
2506 quoted in the delivery log line, like this:
2508 => >A.N.Other <ano@some.domain> <original@ddress> ...
2510 This has been changed so that it extracts the operative address. There
2511 may be more than one such address. If so, they are comma-separated, like
2514 => >ano@some.domain,ona@other.domain <original@ddress> ...
2516 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2517 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2518 not being looked up in the rDNS to get a host name, Exim was showing the
2519 IP address twice in Received: lines, even though the IP addresses were
2520 identical. For example:
2522 Received: from [1.2.3.4] (helo=[1.2.3.4])
2524 However, if the real host name was known, it was omitting the HELO data
2525 if it matched the actual IP address. This has been tidied up so that it
2526 doesn't show the same IP address twice.
2528 PH/07 When both +timestamp and +memory debugging was on, the value given by
2529 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2530 called by the debug printing, thereby overwriting the timestamp buffer.
2531 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2533 PH/08 When the original message was included in an autoreply transport, it
2534 always said "this is a copy of the message, including all the headers",
2535 even if body_only or headers_only was set. It now gives an appropriate
2538 PH/09 Applied a patch from the Sieve maintainer which:
2540 o fixes some comments
2541 o adds the (disabled) notify extension core
2542 o adds some debug output for the result of if/elsif tests
2543 o points to the current vacation draft in the documentation
2544 and documents the missing references header update
2548 o fixes a bug in processing the envelope test (when testing
2549 multiple envelope elements, the last element determinted the
2552 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2553 Electronic Mail") by including:
2555 Auto-submitted: auto-generated
2557 in the messages that it generates (bounce messages and others, such as
2558 warnings). In the case of bounce messages for non-SMTP mesages, there was
2559 also a typo: it was using "Auto_submitted" (underscore instead of
2560 hyphen). Since every message generated by Exim is necessarily in response
2561 to another message, thes have all been changed to:
2563 Auto-Submitted: auto-replied
2565 in accordance with these statements in the RFC:
2567 The auto-replied keyword:
2569 - SHOULD be used on messages sent in direct response to another
2570 message by an automatic process,
2572 - MUST NOT be used on manually-generated messages,
2574 - MAY be used on Delivery Status Notifications (DSNs) and Message
2575 Disposition Notifications (MDNs),
2577 - MUST NOT be used on messages generated by automatic or periodic
2578 processes, except for messages which are automatic responses to
2581 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2582 to the default Received: header definition.
2584 PH/12 Added log selector acl_warn_skipped (default on).
2586 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2587 variables because (a) they are in the wrong storage pool and (b) even if
2588 they were copied, it wouldn't work properly because of the caching.
2590 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2591 checking when decoding. Apparently there are clients that generate
2592 overlong encoded strings. Why am I not surprised?
2594 PH/15 If the first argument of "${if match_address" was not empty, but did not
2595 contain an "@" character, Exim crashed. Now it writes a panic log message
2596 and treats the condition as false.
2598 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2600 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2601 extension "envelope-auth". The code is finished and in agreement with
2602 other implementations, but there is no documentation so far and in fact,
2603 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2604 not changing the active code.
2606 Print executed "if" and "elsif" statements when debugging is used. This
2607 helps a great deal to understand what a filter does.
2609 Document more things not specified clearly in RFC3028. I had all this
2610 sorted out, when out of a sudden new issues came to my mind. Oops."
2612 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2615 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2616 canonical form (as documented). However, after a host name lookup from
2617 the IP address, check_host() was doing a simple string comparison with
2618 addresses acquired from the DNS when checking that the found name did
2619 have the original IP as one of its addresses. Since any found IPv6
2620 addresses are likely to be in abbreviated form, the comparison could
2621 fail. Luckily, there already exists a function for doing the comparison
2622 by converting both addresses to binary, so now that is used instead of
2623 the text comparison.
2625 PH/20 There was another similar case to PH/19, when a complete host name was
2626 given in a host list; looking up its IP address could give an abbreviated
2627 form, whereas the current host's name might or might not be abbreviated.
2628 The same fix has been applied.
2634 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2635 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2638 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2639 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2641 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2643 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2644 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2645 Also, for "Auto-Submitted", treat anything other than "no" as
2646 non-personal, in accordance with RFC 3834. (Previously it treated
2647 anything starting "auto-" as non-personal.)
2649 TF/01 The control=submission/name=... option had a problem with syntax
2650 errors if the name included a slash character. The /name= option
2651 now slurps the rest of the string, so it can include any characters
2652 but it must come last in the list of options (after /sender_retain
2655 PH/05 Some modifications to the interface to the fake nameserver for the new
2663 TK/01 Added the "success_on_redirect" address verification option. See
2664 NewStuff for rationale and an example.
2666 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2668 PH/02 Patch to exigrep to allow it to work on syslog lines.
2670 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2671 fread() to skip over the body file's header line, because in Cygwin the
2672 header line is locked and is inaccessible.
2674 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2675 co-exist for some time) to make it clear that it is the Exim ID that is
2676 referenced, not the Message-ID: header line.
2678 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2679 string_format() function, because snprintf() does not exist on all
2682 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2683 ${stat: expansion item.
2685 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2686 protocol synchronization error", to keep the pedants happy.
2688 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2689 well as for IRIX systems, when gcc is being used. See the host.c source
2692 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2694 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2697 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2698 transport and to the smtp transport in LMTP mode.
2700 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2702 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2703 run a filter in a subprocess. This could lead to confusion in subsequent
2704 lookups in the parent process. There should also be a search_tidyup() at
2705 the end of the subprocess.
2707 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2708 only if the host matched helo_try_verify_hosts, which caused the
2709 verification to occur when the EHLO/HELO command was issued. The ACL just
2710 tested the remembered result. Now, if a previous verification attempt has
2711 not happened, "verify = helo" does it there and then.
2713 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2715 TK/03 Fix log output including CR from clamd.
2717 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2718 exist provoked a memory error which could cause a segfault.
2720 PH/15 Installed PCRE 6.2
2722 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2724 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2725 of the problem. Specifically, suggested +O2 rather than +O1 for the
2728 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2730 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2731 record, the port was not correctly logged when the outgoing_port log
2732 selector was set (it logged the transort's default port).
2734 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2735 fallback_hosts, and "hosts" in the smtp transport.
2737 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2738 host errors such as "Connection refused".
2740 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2741 authentication with radiusclient 0.4.9:
2743 - Error returned from rc_read_config was caught wrongly
2744 - Username/password not passed on to radius server due to wrong length.
2746 The presumption is that some radiusclient API changes for 4.51/PH/17
2747 were not taken care of correctly. The code is still untested by me (my
2748 Linux distribution still has 0.3.2 of radiusclient), but it was
2749 contributed by a Radius user.
2751 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2752 expanding the "port" option of the smtp transport.
2754 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2755 while reading a MIME header. Thanks to Tom Hughes for a patch.
2757 PH/24 Include config.h inside local_scan.h so that configuration settings are
2760 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2761 that in an ACL for RCPT (for example), you can examine exactly what was
2764 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2765 commands, but it was not correctly comparing the address with the actual
2766 client host address. Thus, it would show the EHLO address in Received:
2767 header lines when this was not necessary.
2769 PH/27 Added the % operator to ${eval:}.
2771 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2772 it should be ignoring failures (because with -C, for example, it has lost
2773 privilege). It wasn't ignoring creation failures other than "already
2776 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2779 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2780 because an input line was too long, either on its own, or by virtue of
2781 too many continuations, the temporary file was not being removed, and the
2782 return code was incorrect.
2784 PH/31 Missing "BOOL" in function definition in filtertest.c.
2786 PH/32 Applied Sieve patches from the maintainer.
2788 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2790 PH/33 Added "verify = not_blind".
2792 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2793 Local/Makefile (with some defaults set). These are used in built scripts
2794 such as exicyclog, but they have never been used in the exim_install
2795 script (though there are many overriding facilities there). I have
2796 arranged that the exim_install script now takes note of these two
2799 PH/35 Installed configuration files for Dragonfly.
2801 PH/36 When a locally submitted message by a trusted user did not contain a
2802 From: header, and the sender address was obtained from -f or from an SMTP
2803 MAIL command, and the trusted user did not use -F to supply a sender
2804 name, $originator_name was incorrectly used when constructing a From:
2805 header. Furthermore, $originator_name was used for submission mode
2806 messages from external hosts without From: headers in a similar way,
2807 which is clearly wrong.
2809 PH/37 Added control=suppress_local_fixups.
2811 PH/38 When log_selector = +received_sender was set, and the addition of the
2812 sender made the log line's construction buffer exactly full, or one byte
2813 less than full, an overflow happened when the terminating "\n" was
2816 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2817 when the result of a list match is failure because a DNS lookup failed.
2819 PH/40 RM_COMMAND is now used in the building process.
2821 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2822 the "build-* directories that it finds.
2824 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2825 address in a domain literal was a prefix of an interface address.
2827 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2828 when verifying a sender address, unless rewrite_headers is false.
2830 PH/44 Wrote a long comment about why errors_to addresses are verified as
2831 recipients, not senders.
2833 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2834 the ratelimit ACL was added.
2836 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2838 PH/46 Added extra information about PostgreSQL errors to the error string.
2840 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2841 avoiding the need to install special zones in a real server. This is
2842 backwards compatible; if it can't find the fake resolver, it drops back.
2843 Thus, both old and new test suites can be run.
2845 TF/02 Added util/ratelimit.pl
2847 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2848 clock is set back in time.
2850 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2851 Candler <B.Candler@pobox.com>.
2853 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2854 OK for addresses that are the result of redirections.
2856 PH/48 A number of further additions for the benefit of the new test suite,
2857 including a fake gethostbyname() that interfaces to the fake DNS resolver
2860 TF/06 The fix for widen_domains has also been applied to qualify_single and
2861 search_parents which are the other dnslookup options that can cause
2864 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2867 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2869 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2870 service" if it encountered an MX record with an empty target hostname.
2871 The message is now "an MX or SRV record indicated no SMTP service".
2873 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2874 if the DNS of the sending site is misconfigured. This is quite a
2875 common situation. This change restores the behaviour of treating a
2876 helo verification defer as a failure.
2878 PH/51 If self=fail was set on a router, the bounce message did not include the
2879 actual error message.
2885 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2887 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2888 command itself ended in error, the underlying message about the transport
2889 filter timeout was being overwritten with the pipe command error. Now the
2890 underlying error message should be appended to the second error message.
2892 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2894 PH/02 Reduce the amount of output that "make" produces by default. Full output
2895 can still be requested.
2897 PH/03 The warning log line about a condition test deferring for a "warn" verb
2898 was being output only once per connection, rather than after each
2899 occurrence (because it was using the same function as for successful
2900 "warn" verbs). This seems wrong, so I have changed it.
2902 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2903 it should not have, which might have caused a crash in the right
2904 circumstances, but probably never did.
2906 PH/04 Installed a modified version of Tony Finch's patch to make submission
2907 mode fix the return path as well as the Sender: header line, and to
2908 add a /name= option so that you can make the user's friendly name appear
2911 TF/03 Added the control = fakedefer ACL modifier.
2913 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2914 Mark Lowes for thorough testing.
2916 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2918 TK/03 Merged latest SRS patch from Miles Wilton.
2920 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2921 with the definition in sysexits.h (which is #included earlier).
2922 Fortunately, Exim does not actually use EX_OK. The code used to try to
2923 preserve the sysexits.h value, by assumimg that macro definitions were
2924 scanned for macro replacements. I have been disabused of this notion,
2925 so now the code just undefines EX_OK before #including unistd.h.
2927 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2928 in the smtp transport. When a block could not be written in a single
2929 write() function, the timeout was being re-applied to each part-write.
2930 This seems wrong - if the receiver was accepting one byte at a time it
2931 would take for ever. The timeout is now adjusted when this happens. It
2932 doesn't have to be particularly precise.
2934 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2935 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2937 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2938 without a domain if the check to <postmaster@domain> fails.
2940 SC/01 Eximstats: added -xls and the ability to specify output files
2941 (patch written by Frank Heydlauf).
2943 SC/02 Eximstats: use FileHandles for outputing results.
2945 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2947 SC/04 Eximstats: fixed display of large numbers with -nvr option
2949 SC/05 Eximstats: fixed merging of reports with empty tables.
2951 SC/06 Eximstats: added the -include_original_destination flag
2953 SC/07 Eximstats: removed tabs and trailing whitespace.
2955 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2957 TK/06 MBOX spool code: Add real "From " MBOX separator line
2958 so the .eml file is really in mbox format (even though
2959 most programs do not really care). Patch from Alex Miller.
2961 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2962 The latter is generated from $received_to and is only set if the
2963 message has one envelope recipient. SA can use these headers,
2964 obviously out-of-the-box. Patch from Alex Miller.
2966 PH/08 The ${def test on a variable was returning false if the variable's
2967 value was "0", contrary to what the specification has always said!
2968 The result should be true unless the variable is empty.
2970 PH/09 The syntax error of a character other than { following "${if
2971 def:variable_name" (after optional whitespace) was not being diagnosed.
2972 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2973 accidental colon was present, for example, could give incorrect results.
2975 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2976 result is used (not including appendfile, where other changes are about
2979 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2980 This involved changing a lot of size variables from int to off_t. It
2981 should work with maildirs and everything.
2983 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2984 spamd dying while we are connected to it.
2986 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2989 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2990 function for BDB 4.3.
2992 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2994 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2995 the sources for PCRE have been re-arranged and more files are now
2998 PH/15 The code I had for printing potentially long long variables in PH/11
2999 above was not the best (it lost precision). The length of off_t variables
3000 is now inspected at build time, and an appropriate printing format (%ld
3001 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3002 to be "long long int" or "long int". This is needed for the internal
3003 formatting function string_vformat().
3005 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3006 the configuration file to be ":syslog", then the script "guesses" where
3007 the logs files are, rather than using the compiled in default. In our
3008 case the guess is not the same as the compiled default, so the script
3009 suddenly stopped working when I started to use syslog. The patch checks
3010 to see if log_file_path is "". If so, it attempts to read it from exim
3011 with no configuration file to get the compiled in version, before it
3012 falls back to the previous guessing code."
3014 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3015 implementing BATV in an Exim configuration. See NewStuff for the gory
3018 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3019 Makefile that are specific to HP-UX.
3021 PH/18 If the "use_postmaster" option was set for a recipient callout together
3022 with the "random" option, the postmaster address was used as the MAIL
3023 FROM address for the random test, but not for the subsequent recipient
3024 test. It is now used for both.
3026 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3027 patch removes a few documentation additions to RFC 3028, because the
3028 latest draft now contains them. It adds the new en;ascii-case comparator
3029 and a new error check for 8bit text in MIME parts. Comparator and
3030 require names are now matched exactly. I enabled the subaddress
3031 extension, but it is not well tested yet (read: it works for me)."
3033 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3034 rework some of the code of TK/09 above to avoid the hardwired use of
3035 "%lld" and "long long". Replaced the call to snprintf() with a call to
3038 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3039 records point to non-existent hosts", "retry timeout exceeded", and
3040 "retry time not reached for any host after a long failure period".
3042 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3043 experimental DomainKeys support:
3045 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3046 (2) On an error such as an illegally used "control", the wrong name for
3047 the control was given.
3049 These problems did NOT occur unless DomainKeys support was compiled.
3051 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3053 PH/24 Added ${if match_ip condition.
3055 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3056 databases so that it will be absolutely obvious if a crash occurs in the
3057 DB library. This is a regular occurrence (often caused by mis-matched
3060 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3061 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3062 picked up on a user's system that detects such things. There doesn't seem
3063 to be a gcc warning option for this - only an attribute that has to be
3064 put on the function's prototype. It seems that in Fedora Core 4 they have
3065 set this on a number of new functions. No doubt there will be more in due
3068 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3069 specify a transport. However, if an address that was verified by such a
3070 router was the subject of a callout, Exim crashed because it tried to
3071 read the rcpt_include_affixes from the non-existent transport. Now it
3072 just assumes that the setting of that option is false. This bug was
3073 introduced by 4.51/PH/31.
3075 PH/28 Changed -d+all to exclude +memory, because that information is very
3076 rarely of interest, but it makes the output a lot bigger. People tend to
3077 do -d+all out of habit.
3079 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3080 code in os-type was giving problems when libc.so lives in lib64, like on
3083 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3084 aren't the modern standard, and it seems that some systems' include files
3085 don't always have them. Exim was already checking for some of the newer
3086 ones like T_AAAA, and defining it itself. I've added checks for all the
3087 record types that Exim uses.
3089 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3090 not automatically generating a new one, as it is supposed to. This
3091 prevented TLS from working. If the file did exist, but contained invalid
3092 data, a new version was generated, as expected. It was only the case of a
3093 non-existent file that was broken.
3095 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3096 with a change in libdomainkeys > 0.64.
3098 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3099 from DNS. If the selector record carries the flag, it now has
3100 precedence over the domain-wide flag.
3102 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3104 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3105 the use of a port name that isn't defined in /etc/services) occurred, the
3106 message was deferred as in a normal delivery, and thus remained on the
3107 spool, instead of being failed because of the mua_wrapper setting. This
3108 is now fixed, and I tidied up some of the mua_wrapper messages at the
3111 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3112 the messages in a hash of arrays rather than using individual hashes.
3113 This is a bit cleaner and results in dramatic memory savings, albeit
3114 at a slight CPU cost.
3116 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3117 as requested by Marc Sherman.
3119 SC/10 Eximstats: added histograms for user specified patterns as requested
3122 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3124 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3125 fopen() in the content-scanning modules that did not already have it.
3131 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3132 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3134 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3136 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3138 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3139 to test Sieve filters that use "vacation".
3141 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3142 that changes the way the GnuTLS parameters are stored in the cache file.
3143 The new format can be generated externally. For backward compatibility,
3144 if the data in the cache doesn't make sense, Exim assumes it has read an
3145 old-format file, and it generates new data and writes a new file. This
3146 means that you can't go back to an older release without removing the
3149 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3150 work if there are any delivery delays because "one_time" forces the
3151 parent to be marked "delivered", so its unseen clone is never tried
3152 again. For this reason, Exim now forbids the simultaneous setting of
3155 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3156 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3157 entry if you want to know the details. The fix, however, neglected to
3158 consider the case when local delivery batching is involved. The test for
3159 "previously delivered" was not happening when checking to see if an
3160 address could be batched with a previous (undelivered) one; under
3161 certain circumstances this could lead to multiple deliveries to the same
3164 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3165 in its include files, and this causes problems building Exim.
3167 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3168 header_syntax) but Exim was just ignoring anything given after a slash.
3169 In particular, this caused confusion with an attempt to use "verify =
3170 reverse_host_lookup/defer_ok". An error is now given when options are
3171 supplied for verify items that do not have them. (Maybe reverse_host_
3172 lookup should have a defer_ok option, but that's a different point.)
3174 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3175 defined by RFC 821) to 2048, because there were problems with some AUTH
3176 commands, and RFC 1869 says the size should be increased for extended
3177 SMTP commands that take arguments.
3179 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3182 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3183 "unknown" error; now it says that the functionality isn't in the binary.
3185 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3186 an address' error message when a string expansion fails (syntax or
3187 whatever). Otherwise the password may appear in the log. Following change
3188 PH/42 below, there is no longer a chance of it appearing in a bounce
3191 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3193 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3194 of its addresses was ever tried. (Bugzilla bug #2.)
3196 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3197 the result incorrectly in the debug output. (It correctly added a newline
3198 to what was transported.)
3200 TF/01 Added $received_time.
3202 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3203 commented out examples of how to interface to a virus scanner and to
3204 SpamAssassin. Also added commented examples of av_scanner and
3205 spamd_address settings.
3207 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3208 and controls are allowed in which ACLs. There were a couple of minor
3209 errors. Some of the entries in the conditions table (which is a table of
3210 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3211 negation of where the condition IS allowed.
3213 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3215 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3216 header file does not have a version number, so I've had to invent a new
3217 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3218 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3219 radiusclient), but it was contributed by a Radius user.
3221 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3222 files or directories, for OpenSSL.
3224 PH/19 When an Exim process that is running as root has to create an Exim log
3225 file, it does so in a subprocess that runs as exim:exim so as to get the
3226 ownership right at creation (otherwise, other Exim processes might see
3227 the file with the wrong ownership). There was no test for failure of this
3228 fork() call, which would lead to the process getting stuck as it waited
3229 for a non-existent subprocess. Forks do occasionally fail when resources
3230 run out. I reviewed all the other calls to fork(); they all seem to check
3233 PH/20 When checking for unexpected SMTP input at connect time (before writing
3234 the banner), Exim was not dealing correctly with a non-positive return
3235 from the read() function. If the client had disconnected by this time,
3236 the result was a log entry for a synchronization error with an empty
3237 string after "input=" when read() returned zero. If read() returned -1
3238 (an event I could not check), uninitialized data bytes were printed.
3239 There were reports of junk text (parts of files, etc) appearing after
3242 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3244 PH/22 Added support for macro redefinition, and (re)definition in between
3245 driver and ACL definitions.
3247 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3248 forgetting to use the resulting value; it was using the unexpanded value.
3250 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3251 hadn't been configured. The fix is from Juergen Kreileder, who
3252 understands it better than I do:
3254 "Here's what I see happening with three configured cyrus_sasl
3255 authenticators configured (plain, login, cram-md5):
3257 On startup auth_cyrus_sasl_init() gets called for each of these.
3258 This means three calls to sasl_listmech() without a specified mech_list.
3259 => SASL tests which mechs of all available mechs actually work
3260 => three warnings about OTP not working
3261 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3263 With the patch, sasl_listmech() also gets called three times. But now
3264 SASL's mech_list option is set to the server_mech specified in the the
3265 authenticator. Or in other words, the answer from sasl_listmech()
3266 gets limited to just the mech you're testing for (which is different
3268 => the return list contains just 'plain' or 'login', 'cram-md5' or
3269 nothing depending on the value of ob->server_mech.
3271 I've just tested the patch: Authentication still works fine,
3272 unavailable mechs specified in the exim configuration are still
3273 caught, and the auth.log warnings about OTP are gone."
3275 PH/25 When debugging is enabled, the contents of the command line are added
3276 to the debugging output, even when log_selector=+arguments is not
3279 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3280 answer is "GNU", and only if the return is "GNU/something" is the answer
3283 PH/27 $acl_verify_message is now set immediately after the failure of a
3284 verification in an ACL, and so is available in subsequent modifiers. In
3285 particular, the message can be preserved by coding like this:
3287 warn !verify = sender
3288 set acl_m0 = $acl_verify_message
3290 Previously, $acl_verify_message was set only while expanding "message"
3291 and "log_message" when a very denied access.
3293 PH/28 Modified OS/os.c-Linux with
3295 -#ifndef OS_LOAD_AVERAGE
3296 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3298 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3299 nomenclature these days.)
3301 PH/29 Installed patch from the Sieve maintainer that adds the options
3302 sieve_useraddress and sieve_subaddress to the redirect router.
3304 PH/30 In these circumstances:
3305 . Two addresses routed to the same list of hosts;
3306 . First host does not offer TLS;
3307 . First host accepts first address;
3308 . First host gives temporary error to second address;
3309 . Second host offers TLS and a TLS session is established;
3310 . Second host accepts second address.
3311 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3312 and peerdn, if requested) that were in fact used only for the second
3315 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3316 attention to any local part prefix or suffix that was matched by the
3317 router that accepted the address. It now behaves in the same way as it
3318 does for delivery: the affixes are removed from the local part unless
3319 rcpt_include_affixes is set on the transport.
3321 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3322 timeout during the DATA phase of an incoming message.
3324 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3325 applied a patch sanctioned by the Sieve maintainer.
3327 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3328 the uid or gid is negative. A case of a negative gid caused this to be
3329 noticed. The fix allows for either to be negative.
3331 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3332 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3335 PH/36 The change PH/12 above was broken. Fixed it.
3337 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3338 the grounds that routing the same address twice would always produce the
3339 same answer. This might have been true once, but it is certainly no
3340 longer true now. Routing a child address may depend on the previous
3341 routing that produced that child. Some complicated redirection strategies
3342 went wrong when messages had multiple recipients, and made Exim's
3343 behaviour dependent on the order in which the addresses were given.
3345 I have moved the duplicate checking until after the routing is complete.
3346 Exim scans the addresses that are assigned to local and remote
3347 transports, and removes any duplicates. This means that more work will be
3348 done, as duplicates will always all be routed, but duplicates are
3349 presumably rare, so I don't expect this is of any significance.
3351 For deliveries to pipes, files, and autoreplies, the duplicate checking
3352 still happens during the routing process, since they are not going to be
3355 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3356 It corrects a timeout issue with spamd. This is Ian's comment: "The
3357 background is that sometimes spamd either never reads data from a
3358 connection it has accepted, or it never writes response data. The exiscan
3359 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3360 blindly assumes that writes won't block so it may never time out."
3362 PH/39 Allow G after quota size as well as K and M.
3364 PH/40 The value set for $authenticated_id in an authenticator may not contain
3365 binary zeroes or newlines because the value is written to log lines and
3366 to spool files. There was no check on this. Now the value is run through
3367 the string_printing() function so that such characters are converted to
3368 printable escape sequences.
3370 PH/41 $message_linecount is a new variable that contains the total number of
3371 lines in the message. Compare $body_linecount, which is the count for the
3374 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3375 bounce and delay warning messages, except in certain special cases, which
3378 (a) An SMTP error message from a remote host;
3379 (b) A message specified in a :fail: redirection;
3380 (c) A message specified in a "fail" command in a system filter;
3381 (d) A message specified in a FAIL return from the queryprogram router;
3382 (e) A message specified by the cannot_route_message router option.
3384 In these cases only, Exim does include the error details in bounce and
3385 warning messages. There are also a few cases where bland messages such
3386 as "unrouteable address" or "local delivery error" are given.
3388 PH/43 $value is now also set for the "else" part of a ${run expansion.
3390 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3391 being worked on, but at least Exim now implements the latest version to
3394 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3395 process to complete was treated as a delivery failure, a timeout while
3396 writing the message to the pipe was logged, but erroneously treated as a
3397 successful delivery. Such timeouts include transport filter timeouts. For
3398 consistency with the overall process timeout, these timeouts are now
3399 treated as errors, giving rise to delivery failures by default. However,
3400 there is now a new Boolean option for the pipe transport called
3401 timeout_defer, which, if set TRUE, converts the failures into defers for
3402 both kinds of timeout. A transport filter timeout is now identified in
3405 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3406 systems where "make" and "gmake" are different, calling "gmake" at top
3407 level broke things. I've arranged for the value of $(MAKE) to be passed
3408 from the Makefile to this script so that it can call the same version of
3412 A note about Exim versions 4.44 and 4.50
3413 ----------------------------------------
3415 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3416 changes of various kinds. As a consequence, a big documentation update was
3417 needed. This delayed the release for rather longer than seemed good, especially
3418 in the light of a couple of (minor) security issues. Therefore, the changes
3419 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3420 release. So 4.44 and 4.50 are in effect two different branches that both start
3423 I have left the 4.50 change log unchanged; it contains all the changes since
3424 4.43. The change log for 4.44 is below; many of its items are identical to
3425 those for 4.50. This seems to be the most sensible way to preserve the
3426 historical information.
3432 1. Minor wording change to the doc/README.SIEVE file.
3434 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3435 computation of the current number of files was incorrect.
3437 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3438 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3441 4. Give more explanation in the error message when the command for a transport
3442 filter fails to execute.
3444 5. There are several places where Exim runs a non-Exim command in a
3445 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3446 was being done only for the command run by the queryprogram router. It is
3447 now done for all such subprocesses. The other cases are: ${run, transport
3448 filters, and the commands run by the lmtp and pipe transports.
3450 6. Added CONFIGURE_GROUP build-time option.
3452 7. Some older OS have a limit of 256 on the maximum number of file
3453 descriptors. Exim was using setrlimit() to set 1000 as a large value
3454 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3455 systems. I've change it so that if it can't get 1000, it tries for 256.
3457 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3458 was an oversight, and furthermore, ever since the addition of extra
3459 controls (e.g. 4.43/32), the checks on when to allow different forms of
3460 "control" were broken. There should now be diagnostics for all cases when a
3461 control that does not make sense is encountered.
3463 9. Added the /retain_sender option to "control=submission".
3465 10. $recipients is now available in the predata ACL (oversight).
3467 11. Tidy the search cache before the fork to do a delivery from a message
3468 received from the command line. Otherwise the child will trigger a lookup
3469 failure and thereby defer the delivery if it tries to use (for example) a
3470 cached ldap connection that the parent has called unbind on.
3472 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3473 of $address_data from the recipient verification was clobbered by the
3474 sender verification.
3476 13. The value of address_data from a sender verification is now available in
3477 $sender_address_data in subsequent conditions in the ACL statement.
3479 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3481 15. Added a new option "connect=<time>" to callout options, to set a different
3484 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3485 was its contents. (It was OK if the option was not defined at all.)
3487 17. A "Completed" log line is now written for messages that are removed from
3488 the spool by the -Mrm option.
3490 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3491 information about exactly what failed.
3493 19. Added -dd to debug only the daemon process.
3495 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3496 handles timeouts, both on the server side and network timeouts. Renamed the
3497 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3499 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3500 It is now set to "smtps".
3502 22. $host_address is now set to the target address during the checking of
3503 ignore_target_hosts.
3505 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3506 being passed; this would have caused $sender_host_name to have been used if
3507 matching the list had actually called for a host name (not very likely,
3508 since this list is usually IP addresses). A host name is now passed as
3511 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3512 code that specifies a non-restarting handler (typically sigaction() in
3513 modern systems) in an attempt to fix a rare and obscure crash bug.
3515 25. Narrowed the window for a race in the daemon that could cause it to ignore
3516 SIGCHLD signals. This is not a major problem, because they are used only to
3517 wake it up if nothing else does.
3519 26. A malformed maildirsize file could cause Exim to calculate negative values
3520 for the mailbox size or file count. Odd effects could occur as a result.
3521 The maildirsize information is now recalculated if the size or filecount
3524 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3525 support for a long time. Removed HAVE_SYS_VFS_H.
3527 28. Installed the latest version of exipick from John Jetmore.
3529 29. In an address list, if the pattern was not a regular expression, an empty
3530 subject address (from a bounce message) matched only if the pattern was an
3531 empty string. Non-empty patterns were not even tested. This was the wrong
3532 because it is perfectly reasonable to use an empty address as part of a
3533 database query. An empty address is now tested by patterns that are
3534 lookups. However, all the other forms of pattern expect the subject to
3535 contain a local part and a domain, and therefore, for them, an empty
3536 address still always fails if the pattern is not itself empty.
3538 30. Exim went into a mad DNS loop when attempting to do a callout where the
3539 host was specified on an smtp transport, and looking it up yielded more
3540 than one IP address.
3542 31. Re-factored the code for checking spool and log partition space into a
3543 function that finds that data and another that does the check. The former
3544 is then used to implement four new variables: $spool_space, $log_space,
3545 $spool_inodes, and $log_inodes.
3547 32. The RFC2047 encoding function was originally intended for short strings
3548 such as real names; it was not keeping to the 75-character limit for
3549 encoded words that the RFC imposes. It now respects the limit, and
3550 generates multiple encoded words if necessary. To be on the safe side, I
3551 have increased the buffer size for the ${rfc2047: expansion operator from
3554 33. It is now permitted to omit both strings after an "if" condition; if the
3555 condition is true, the result is "true". As before, when the second string
3556 is omitted, a false condition yields an empty string. This makes it less
3557 cumbersome to write custom ACL and router conditions.
3559 34. Failure to deliver a bounce message always caused it to be frozen, even if
3560 there was an errors_to setting on the router. The errors_to setting is now
3563 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3564 canonical form (fully expanded) before being placed in
3565 $sender_host_address.
3567 36. The table in the code that translates DNS record types into text (T_A to
3568 "A" for instance) was missing entries for NS and CNAME. It is just possible
3569 that this could have caused confusion if both these types were looked up
3570 for the same domain, because the text type is used as part of Exim's
3571 per-process caching. But the chance of anyone hitting this buglet seems
3574 37. The dnsdb lookup has been extended in a number of ways.
3576 (1) There is a new type, "zns", which walks up the domain tree until it
3577 finds some nameserver records. It should be used with care.
3579 (2) There is a new type, "mxh", which is like "mx" except that it returns
3580 just the host names, not the priorities.
3582 (3) It is now possible to give a list of domains (or IP addresses) to be
3583 looked up. The behaviour when one of the lookups defers can be
3584 controlled by a keyword.
3586 (4) It is now possible to specify the separator character for use when
3587 multiple records are returned.
3589 38. The dnslists ACL condition has been extended: it is now possible to supply
3590 a list of IP addresses and/or domains to be looked up in a particular DNS
3593 39. Added log_selector=+queue_time_overall.
3595 40. When running the queue in the test harness, wait just a tad after forking a
3596 delivery process, to get repeatability of debugging output.
3598 41. Include certificate and key file names in error message when GnuTLS fails
3599 to set them up, because the GnuTLS error message doesn't include the name
3600 of the failing file when there is a problem reading it.
3602 42. Allow both -bf and -bF in the same test run.
3604 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3606 44. The "Exiscan patch" is now merged into the mainline Exim source.
3608 45. Sometimes the final signoff response after QUIT could fail to get
3609 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3610 < 0 before doing a fflush(). This bug looks as though it goes back to the
3611 introduction of TLS in release 3.20, but "sometimes" must have been rare
3612 because the tests only now provoked it.
3614 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3615 (this can affect the format of dates).
3617 47. exim_tidydb, when checking for the continued existence of a message for
3618 which it has found a message-specific retry record, was not finding
3619 messages that were in split spool directories. Consequently, it was
3620 deleting retry records that should have stayed in existence.
3622 48. Steve fixed some bugs in eximstats.
3624 49. The SPA authentication driver was not abandoning authentication and moving
3625 on to the next authenticator when an expansion was forced to fail,
3626 contradicting the general specification for all authenticators. Instead it
3627 was generating a temporary error. It now behaves as specified.
3629 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3630 (the order specifies the preference for clients). The order is now AES256,
3631 AES128, 3DES, ARCFOUR128.
3633 51. Small patch to Sieve code - explicitly set From: when generating an
3636 52. Exim crashed if a remote delivery caused a very long error message to be
3637 recorded - for instance if somebody sent an entire SpamAssassin report back
3638 as a large number of 550 error lines. This bug was coincidentally fixed by
3639 increasing the size of one of Exim's internal buffers (big_buffer) that
3640 happened as part of the Exiscan merge. However, to be on the safe side, I
3641 have made the code more robust (and fixed the comments that describe what
3644 53. Now that there can be additional text after "Completed" in log lines (if
3645 the queue_time_overall log selector is set), a one-byte patch to exigrep
3646 was needed to allow it to recognize "Completed" as not the last thing in
3649 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3650 patch that reportedly fixes this has been added. I am not expert enough to
3651 create a test for it. This is what the patch creator wrote:
3653 "I found a little strange behaviour of ldap code when working with
3654 Windows 2003 AD Domain, where users was placed in more than one
3655 Organization Units. When I tried to give exim partial DN, the exit code
3656 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3657 But simultaneously result of request was absolutely normal ldap result,
3658 so I produce this patch..."
3660 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3661 REFERENCE, so I have modified the code to exclude the patch when that macro
3664 55. Some experimental protocols are using DNS PTR records for new purposes. The
3665 keys for these records are domain names, not reversed IP addresses. The
3666 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3667 leaves it alone. Component reversal etc. now happens only for IP addresses.
3670 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3672 57. Double the size of the debug message buffer (to 2048) so that more of very
3673 long debug lines gets shown.
3675 58. The exicyclog utility now does better if the number of log files to keep
3676 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3678 59. Two changes related to the smtp_active_hostname option:
3680 (1) $smtp_active_hostname is now available as a variable.
3681 (2) The default for smtp_banner uses $smtp_active_hostname instead
3682 of $primary_hostname.
3684 60. The host_aton() function is supposed to be passed a string that is known
3685 to be a valid IP address. However, in the case of IPv6 addresses, it was
3686 not checking this. This is a hostage to fortune. Exim now panics and dies
3687 if the condition is not met. A case was found where this could be provoked
3688 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3689 components; fortuitously, this particular loophole had already been fixed
3690 by change 4.50/55 above.
3692 If there are any other similar loopholes, the new check in host_aton()
3693 itself should stop them being exploited. The report I received stated that
3694 data on the command line could provoke the exploit when Exim was running as
3695 exim, but did not say which command line option was involved. All I could
3696 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3697 running as the user.
3700 61. There was a buffer overflow vulnerability in the SPA authentication code
3701 (which came originally from the Samba project). I have added a test to the
3702 spa_base64_to_bits() function which I hope fixes it.
3705 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3706 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3708 63. The daemon start-up calls getloadavg() while still root for those OS that
3709 need the first call to be done as root, but it missed one case: when
3710 deliver_queue_load_max is set with deliver_drop_privilege. This is
3711 necessary for the benefit of the queue runner, because there is no re-exec
3712 when deliver_drop_privilege is set.
3714 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3715 This has been fixed.
3717 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3718 was in use, was not putting the data itself into the right store pool;
3719 consequently, it could be overwritten for a subsequent message in the same
3720 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3723 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3725 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3726 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3727 the function were treating the return as a boolean value, which happened to
3728 work because 0=false and not-0=true, but is not correct code.
3730 68. The host_aton() function was not handling scoped IPv6 addresses (those
3731 with, for example, "%eth0" on the end) correctly.
3733 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3734 negated items (that is, ~something) in unsigned ints. Some compilers
3735 apparently mutter when there is no cast.
3737 70. If an address verification called from an ACL failed, and did not produce a
3738 user-specific message (i.e. there was only a "system" message), nothing was
3739 put in $acl_verify_message. In this situation, it now puts the system
3742 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3743 session; change 4.31/43 added the unwanted input to the log line - except
3744 that it did not do this in the start of session case. It now does.
3746 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3747 This is wrong and can cause the other end to generate a synchronization
3748 error if it is another Exim or anything else that does the synchronization
3749 check. A QUIT command is no longer sent after a timeout.
3751 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3752 during host lookups.
3754 74. The defer_ok option of callout verification was not working if it was used
3755 when verifying addresses in header lines, that is, for this case:
3757 verify = header_sender/callout=defer_ok
3759 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3760 those file descriptors could be used for SMTP connections. If anything
3761 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3762 could be sent to the SMTP client, causing chaos. The daemon now opens
3763 stdin, stdout, and stderr to /dev/null when it puts itself into the
3766 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3767 log by default. The user can override this with suitable Perl magic.
3769 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3770 the log message when discard triggers, was not working for the DATA ACL or
3771 for the non-SMTP ACL.
3773 78. Error message wording change in sieve.c.
3775 79. If smtp_accept_max_per_host was set, the number of connections could be
3776 restricted to fewer than expected, because the daemon was trying to set up
3777 a new connection before checking whether the processes handling previous
3778 connections had finished. The check for completed processes is now done
3779 earlier. On busy systems, this bug wouldn't be noticed because something
3780 else would have woken the daemon, and it would have reaped the completed
3783 80. If a message was submitted locally by a user whose login name contained one
3784 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3785 It caused a spool format error. I have fixed the spool reading code. A
3786 related problem was that the "from" clause in the Received: line became
3787 illegal because of the space(s). It is now covered by ${quote_local_part.
3789 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3792 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3794 83. Patch from Sieve maintainer to fix unterminated string problem in
3795 "vacation" handling.
3797 84. Some minor changes to the Linux configuration files to help with other
3798 OS variants using glibc.
3800 85. One more patch for Sieve to update vacation handling to latest spec.
3803 ----------------------------------------------------
3804 See the note above about the 4.44 and 4.50 releases.
3805 ----------------------------------------------------
3811 1. Change 4.43/35 introduced a bug that caused file counts to be
3812 incorrectly computed when quota_filecount was set in an appendfile
3815 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3816 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3819 3. Give more explanation in the error message when the command for a transport
3820 filter fails to execute.
3822 4. There are several places where Exim runs a non-Exim command in a
3823 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3824 was being done only for the command run by the queryprogram router. It is
3825 now done for all such subprocesses. The other cases are: ${run, transport
3826 filters, and the commands run by the lmtp and pipe transports.
3828 5. Some older OS have a limit of 256 on the maximum number of file
3829 descriptors. Exim was using setrlimit() to set 1000 as a large value
3830 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3831 systems. I've change it so that if it can't get 1000, it tries for 256.
3833 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3834 was an oversight, and furthermore, ever since the addition of extra
3835 controls (e.g. 4.43/32), the checks on when to allow different forms of
3836 "control" were broken. There should now be diagnostics for all cases when a
3837 control that does not make sense is encountered.
3839 7. $recipients is now available in the predata ACL (oversight).
3841 8. Tidy the search cache before the fork to do a delivery from a message
3842 received from the command line. Otherwise the child will trigger a lookup
3843 failure and thereby defer the delivery if it tries to use (for example) a
3844 cached ldap connection that the parent has called unbind on.
3846 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3847 of $address_data from the recipient verification was clobbered by the
3848 sender verification.
3850 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3851 was its contents. (It was OK if the option was not defined at all.)
3853 11. A "Completed" log line is now written for messages that are removed from
3854 the spool by the -Mrm option.
3856 12. $host_address is now set to the target address during the checking of
3857 ignore_target_hosts.
3859 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3860 being passed; this would have caused $sender_host_name to have been used if
3861 matching the list had actually called for a host name (not very likely,
3862 since this list is usually IP addresses). A host name is now passed as
3865 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3866 code that specifies a non-restarting handler (typically sigaction() in
3867 modern systems) in an attempt to fix a rare and obscure crash bug.
3869 15. Narrowed the window for a race in the daemon that could cause it to ignore
3870 SIGCHLD signals. This is not a major problem, because they are used only to
3871 wake it up if nothing else does.
3873 16. A malformed maildirsize file could cause Exim to calculate negative values
3874 for the mailbox size or file count. Odd effects could occur as a result.
3875 The maildirsize information is now recalculated if the size or filecount
3878 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3879 support for a long time. Removed HAVE_SYS_VFS_H.
3881 18. Updated exipick to current release from John Jetmore.
3883 19. Allow an empty sender to be matched against a lookup in an address list.
3884 Previously the only cases considered were a regular expression, or an
3887 20. Exim went into a mad DNS lookup loop when doing a callout where the
3888 host was specified on the transport, if the DNS lookup yielded more than
3891 21. The RFC2047 encoding function was originally intended for short strings
3892 such as real names; it was not keeping to the 75-character limit for
3893 encoded words that the RFC imposes. It now respects the limit, and
3894 generates multiple encoded words if necessary. To be on the safe side, I
3895 have increased the buffer size for the ${rfc2047: expansion operator from
3898 22. Failure to deliver a bounce message always caused it to be frozen, even if
3899 there was an errors_to setting on the router. The errors_to setting is now
3902 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3903 canonical form (fully expanded) before being placed in
3904 $sender_host_address.
3906 24. Updated eximstats to version 1.33
3908 25. Include certificate and key file names in error message when GnuTLS fails
3909 to set them up, because the GnuTLS error message doesn't include the name
3910 of the failing file when there is a problem reading it.
3912 26. Expand error message when OpenSSL has problems setting up cert/key files.
3915 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3916 (this can affect the format of dates).
3918 28. exim_tidydb, when checking for the continued existence of a message for
3919 which it has found a message-specific retry record, was not finding
3920 messages that were in split spool directories. Consequently, it was
3921 deleting retry records that should have stayed in existence.
3923 29. eximstats updated to version 1.35
3924 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3925 1.35 - bugfix such that pie charts by volume are generated correctly
3927 30. The SPA authentication driver was not abandoning authentication and moving
3928 on to the next authenticator when an expansion was forced to fail,
3929 contradicting the general specification for all authenticators. Instead it
3930 was generating a temporary error. It now behaves as specified.
3932 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3933 (the order specifies the preference for clients). The order is now AES256,
3934 AES128, 3DES, ARCFOUR128.
3936 31. Small patch to Sieve code - explicitly set From: when generating an
3939 32. Exim crashed if a remote delivery caused a very long error message to be
3940 recorded - for instance if somebody sent an entire SpamAssassin report back
3941 as a large number of 550 error lines. This bug was coincidentally fixed by
3942 increasing the size of one of Exim's internal buffers (big_buffer) that
3943 happened as part of the Exiscan merge. However, to be on the safe side, I
3944 have made the code more robust (and fixed the comments that describe what
3947 33. Some experimental protocols are using DNS PTR records for new purposes. The
3948 keys for these records are domain names, not reversed IP addresses. The
3949 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3950 leaves it alone. Component reversal etc. now happens only for IP addresses.
3953 34. The host_aton() function is supposed to be passed a string that is known
3954 to be a valid IP address. However, in the case of IPv6 addresses, it was
3955 not checking this. This is a hostage to fortune. Exim now panics and dies
3956 if the condition is not met. A case was found where this could be provoked
3957 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3958 components; fortuitously, this particular loophole had already been fixed
3959 by change 4.50/55 or 4.44/33 above.
3961 If there are any other similar loopholes, the new check in host_aton()
3962 itself should stop them being exploited. The report I received stated that
3963 data on the command line could provoke the exploit when Exim was running as
3964 exim, but did not say which command line option was involved. All I could
3965 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3966 running as the user.
3969 35. There was a buffer overflow vulnerability in the SPA authentication code
3970 (which came originally from the Samba project). I have added a test to the
3971 spa_base64_to_bits() function which I hope fixes it.
3974 36. The daemon start-up calls getloadavg() while still root for those OS that
3975 need the first call to be done as root, but it missed one case: when
3976 deliver_queue_load_max is set with deliver_drop_privilege. This is
3977 necessary for the benefit of the queue runner, because there is no re-exec
3978 when deliver_drop_privilege is set.
3980 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3981 was in use, was not putting the data itself into the right store pool;
3982 consequently, it could be overwritten for a subsequent message in the same
3983 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3986 38. Sometimes the final signoff response after QUIT could fail to get
3987 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3988 < 0 before doing a fflush(). This bug looks as though it goes back to the
3989 introduction of TLS in release 3.20, but "sometimes" must have been rare
3990 because the tests only now provoked it.
3996 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3997 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3998 Now it returns an integer. A number of calls were still expecting a T/F
3999 return. Fortuitously, in all cases, the tests worked in OK situations,
4000 which is the norm. However, things would have gone wrong on any write
4001 failures on the smtp file descriptor. This function is used when sending
4002 messages over SMTP and also when doing verify callouts.
4004 2. When Exim is called to do synchronous delivery of a locally submitted
4005 message (the -odf or -odi options), it no longer closes stderr before doing
4008 3. Implemented the mua_wrapper option.
4010 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4012 5. Implemented the functions header_remove(), header_testname(),
4013 header_add_at_position(), and receive_remove_recipient(), and exported them
4016 6. If an ACL "warn" statement specified the addition of headers, Exim already
4017 inserted X-ACL-Warn: at the start if there was no header name. However, it
4018 was not making this test for the second and subsequent header lines if
4019 there were newlines in the string. This meant that an invalid header could
4020 be inserted if Exim was badly configured.
4022 7. Allow an ACL "warn" statement to add header lines at the start or after all
4023 the Received: headers, as well as at the end.
4025 8. Added the rcpt_4xx retry error code.
4027 9. Added postmaster_mailfrom=xxx to callout verification option.
4029 10. Added mailfrom=xxxx to the callout verification option, for verify=
4032 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4033 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4035 12. Inserted some casts to stop certain compilers warning when using pointer
4036 differences as field lengths or precisions in printf-type calls (mostly
4037 affecting debugging statements).
4039 13. Added optional readline() support for -be (dynamically loaded).
4041 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4042 same clock tick as a message's arrival, so that its received time was the
4043 same as the "first fail" time on the retry record, and that message
4044 remained on the queue past the ultimate address timeout, every queue runner
4045 would try a delivery (because it was past the ultimate address timeout) but
4046 after another failure, the ultimate address timeout, which should have then
4047 bounced the address, did not kick in. This was a "< instead of <=" error;
4048 in most cases the first failure would have been in the next clock tick
4049 after the received time, and all would be well.
4051 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4052 being recognized when the domain list was tested by the match_domain
4053 condition in an expansion string.
4055 16. Added the ${str2b64: operator.
4057 17. Exim was always calling setrlimit() to set a large limit for the number of
4058 processes, without checking whether the existing limit was already
4059 adequate. (It did check for the limit on file descriptors.) Furthermore,
4060 errors from getrlimit() and setrlimit() were being ignored. Now they are
4061 logged to the main and panic logs, but Exim does carry on, to try to do its
4062 job under whatever limits there are.
4064 18. Imported PCRE 5.0.
4066 19. Trivial typo in log message " temporarily refused connection" (the leading
4069 20. If the log selector return_path_on_delivery was set and an address was
4070 redirected to /dev/null, the delivery process crashed because it assumed
4071 that a return path would always be set for a "successful" delivery. In this
4072 case, the whole delivery is bypassed as an optimization, and therefore no
4075 21. Internal re-arrangement: the function for sending a challenge and reading
4076 a response while authentication was assuming a zero-terminated challenge
4077 string. It's now changed to take a pointer and a length, to allow for
4078 binary data in such strings.
4080 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4082 23. Exim was not respecting finduser_retries when seeking the login of the
4083 uid under which it was called; it was always trying 10 times. (The default
4084 setting of finduser_retries is zero.) Also, it was sleeping after the final
4085 failure, which is pointless.
4087 24. Implemented tls_on_connect_ports.
4089 25. Implemented acl_smtp_predata.
4091 26. If the domain in control=submission is set empty, Exim assumes that the
4092 authenticated id is a complete email address when it generates From: or
4093 Sender: header lines.
4095 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4096 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4097 chown and chgrp in /bin and hostname in /usr/bin.
4099 28. Exim was keeping the "process log" file open after each use, just as it
4100 does for the main log. This opens the possibility of it remaining open for
4101 long periods when the USR1 signal hits a daemon. Occasional processlog
4102 errors were reported, that could have been caused by this. Anyway, it seems
4103 much more sensible not to leave this file open at all, so that is what now
4106 29. The long-running daemon process does not normally write to the log once it
4107 has entered its main loop, and it closes the log before doing so. This is
4108 so that log files can straightforwardly be renamed and moved. However,
4109 there are a couple of unusual error situations where the daemon does write
4110 log entries, and I had neglected to close the log afterwards.
4112 30. The text of an SMTP error response that was received during a remote
4113 delivery was being truncated at 512 bytes. This is too short for some of
4114 the long messages that one sometimes sees. I've increased the limit to
4117 31. It is now possible to make retry rules that apply only when a message has a
4118 specific sender, in particular, an empty sender.
4120 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4121 it possible to be selective about when SMTP synchronization is enforced.
4123 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4125 32. Implemented hosts_connection_nolog.
4127 33. Added an ACL for QUIT.
4129 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4132 35. Added mailbox_size and mailbox_filecount to appendfile.
4134 36. Added control = no_multiline_responses to ACLs.
4136 37. There was a bug in the logic of the code that waits for the clock to tick
4137 in the case where the clock went backwards by a substantial amount such
4138 that the microsecond fraction of "now" was more than the microsecond
4139 fraction of "then" (but the whole seconds number was less).
4141 38. Added support for the libradius Radius client library this is found on
4142 FreeBSD (previously only the radiusclient library was supported).
4148 1. When certain lookups returned multiple values in the form name=value, the
4149 quoting of the values was not always being done properly. Specifically:
4150 (a) If the value started with a double quote, but contained no whitespace,
4152 (b) If the value contained whitespace other than a space character (i.e.
4153 tabs or newlines or carriage returns) it was not quoted.
4154 This fix has been applied to the mysql and pgsql lookups by writing a
4155 separate quoting function and calling it from the lookup code. The fix
4156 should probably also be applied to nisplus, ibase and oracle lookups, but
4157 since I cannot test any of those, I have not disturbed their existing code.
4159 2. A hit in the callout cache for a specific address caused a log line with no
4160 reason for rejecting RCPT. Now it says "Previous (cached) callout
4161 verification failure".
4163 3. There was an off-by-one bug in the queryprogram router. An over-long
4164 return line was truncated at 256 instead of 255 characters, thereby
4165 overflowing its buffer with the terminating zero. As well as fixing this, I
4166 have increased the buffer size to 1024 (and made a note to document this).
4168 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4169 when Exim is waiting for an SMTP response from a remote server, Exim
4170 restarts its select() call on the socket, thereby resetting its timeout.
4171 This is not a problem when such interrupts are rare. Somebody set up a cron
4172 job to run exiwhat every 2 minutes, which is less than the normal select()
4173 timeout (5 or 10 minutes). This meant that the select() timeout never
4174 kicked in because it was always reset. I have fixed this by comparing the
4175 time when an interrupt arrives with the time at the start of the first call
4176 to select(). If more time than the timeout has elapsed, the interrupt is
4177 treated as a timeout.
4179 5. Some internal re-factoring in preparation for the addition of Sieve
4180 extensions (by MH). In particular, the "personal" test is moved to a
4181 separate function, and given an option for scanning Cc: and Bcc: (which is
4182 not set for Exim filters).
4184 6. When Exim created an email address using the login of the caller as the
4185 local part (e.g. when creating a From: or Sender: header line), it was not
4186 quoting the local part when it contained special characters such as @.
4188 7. Installed new OpenBSD configuration files.
4190 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4191 try to make them clearer.
4193 9. Callout options, other than the timeout value, were being ignored when
4194 verifying sender addresses in header lines. For example, when using
4196 verify = header_sender/callout=no_cache
4198 the cache was (incorrectly) being used.
4200 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4201 only the Cygwin environment.
4203 11. When return_path_on_delivery was set as a log selector, if different remote
4204 addresses in the same message used different return paths and parallel
4205 remote delivery occurred, the wrong values would sometimes be logged.
4206 (Whenever a remote delivery process finished, the return path value from
4207 the most recently started remote delivery process was logged.)
4209 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4210 lines when AUTH and/or TLS are in use. This is the "received protocol"
4211 field. Exim used to use "asmtp" for authenticated SMTP, without any
4212 indication (in the protocol name) for TLS use. Now it follows the RFC and
4213 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4214 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4215 names appear in log lines as well as in Received: header lines.
4217 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4218 extensions, and comparison tests, and to fix some bugs.
4220 14. Changes to the "personal" filter test:
4222 (1) The test was buggy in that it was just doing the equivalent of
4223 "contains" tests on header lines. For example, if a user's address was
4224 anne@some.where, the "personal" test would incorrectly be true for
4226 To: susanne@some.where
4228 This test is now done by extracting each address from the header in turn,
4229 and checking the entire address. Other tests that are part of "personal"
4230 are now done using regular expressions (for example, to check local parts
4231 of addresses in From: header lines).
4233 (2) The list of non-personal local parts in From: addresses has been
4234 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4235 taken from the Sieve specification recommendations.
4237 (3) If the message contains any header line starting with "List-" it is
4238 treated as non-personal.
4240 (4) The test for "circular" in the Subject: header line has been removed
4241 because it now seems ill-conceived.
4243 15. Minor typos in src/EDITME comments corrected.
4245 16. Installed latest exipick from John Jetmore.
4247 17. If headers_add on a router specified a text string that was too long for
4248 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4249 of string_sprintf() is now avoided.
4251 18. $message_body_size was not set (it was always zero) when running the DATA
4252 ACL and the local_scan() function.
4254 19. For the "mail" command in an Exim filter, no default was being set for
4255 the once_repeat time, causing a random time value to be used if "once" was
4256 specified. (If the value happened to be <= 0, no repeat happened.) The
4257 default is now 0s, meaning "never repeat". The "vacation" command was OK
4258 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4259 (I found it when inspecting the code).
4261 20. There is now an overall timeout for performing a callout verification. It
4262 defaults to 4 times the callout timeout, which applies to individual SMTP
4263 commands during the callout. The overall timeout applies when there is more
4264 than one host that can be tried. The timeout is checked before trying the
4265 next host. This prevents very long delays if there are a large number of
4266 hosts and all are timing out (e.g. when the network connections are timing
4267 out). The value of the overall timeout can be changed by specifying an
4268 additional sub-option for "callout", called "maxwait". For example:
4270 verify = sender/callout=5s,maxwait=20s
4272 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4273 to the end before writing, but this should make it even safer).
4275 22. Exim was forgetting that it had advertised PIPELINING for the second and
4276 subsequent messages on an SMTP connection. It was also not resetting its
4277 memory on STARTTLS and an internal HELO.
4279 23. When Exim logs an SMTP synchronization error within a session, it now
4280 records whether PIPELINING has been advertised or not.
4282 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4283 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4284 rather than long int.
4286 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4288 26. Added the never_mail option to autoreply.
4294 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4295 crash if the getsockname() call failed; this can happen if a connection is
4296 closed very soon after it is established. The problem was simply in the
4297 order in which certain operations were done, causing Exim to try to write
4298 to the SMTP stream before it had set up the file descriptor. The bug has
4299 been fixed by making things happen in the correct order.
4305 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4306 before the connection was closed, thus losing the rejection response.
4308 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4309 some early Solaris releases, but causes trouble in current releases where
4310 socklen_t is defined.
4312 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4315 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4318 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4319 that releasing the top of it at the end releases what was used for sub-
4320 expansions (unless the block got too big). However, discard this block if
4321 the first thing is a variable or header, so that we can use its block when
4322 it is dynamic (useful for very large $message_headers, for example).
4324 6. Lookups now cache *every* query, not just the most recent. A new, separate
4325 store pool is used for this. It can be recovered when all lookup caches are
4326 flushed. Lookups now release memory at the end of their result strings.
4327 This has involved some general refactoring of the lookup sources.
4329 7. Some code has been added to the store_xxx() functions to reduce the amount
4330 of flapping under certain conditions.
4332 8. log_incoming_interface used to affect only the <= reception log lines. Now
4333 it causes the local interface and port to be added to several more SMTP log
4334 lines, for example "SMTP connection from", and rejection lines.
4336 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4338 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4340 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4341 could be overwritten at the end of the current message (or the start of a
4342 new message if it was set in a HELO ACL). The value is now preserved for
4343 the duration of the SMTP connection.
4345 12. If a transport had a headers_rewrite setting, and a matching header line
4346 contained an unqualified address, that address was qualified, even if it
4347 did not match any rewriting rules. The underlying bug was that the values
4348 of the flags that permit the existence of unqualified sender and recipient
4349 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4350 non-local messages, and by -bnq for local messages) were not being
4351 preserved with the message after it was received.
4353 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4354 "next input=" as part of the text comprising the host identity instead of
4355 the correct text. The code was using the same buffer for two different
4356 strings. However, depending on which order the printing function evaluated
4357 its arguments, the bug did not always show up. Under Linux, for example, my
4358 test suite worked just fine.
4360 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4361 doesn't work with some older versions of Perl. It has been changed to "my",
4362 which in any case is probably the better facility to use.
4364 15. A really picky compiler found some instances of statements for creating
4365 error messages that either had too many or two few arguments for the format
4368 16. The size of the buffer for calls to the DNS resolver has been increased
4369 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4370 for addresses that have a lot of PTR records. This alleviates a problem; it
4371 does not fully solve it.
4373 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4374 buffer now truncates the list and logs the incident, which is the same
4375 action as happens when Exim is looking up a host name and its aliases.
4376 Previously in this situation something unpredictable would happen;
4377 sometimes it was "internal error: store_reset failed".
4379 18. If a server dropped the connection unexpectedly when an Exim client was
4380 using GnuTLS and trying to read a response, the client delivery process
4381 crashed while trying to generate an error log message.
4383 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4384 string, for example:
4386 warn message = H1: something\nH2: something
4388 the text was added as a single header line from Exim's point of view
4389 though it ended up OK in the delivered message. However, searching for the
4390 second and subsequent header lines using $h_h2: did not work. This has been
4391 fixed. Similarly, if a system filter added multiple headers in this way,
4392 the routers could not see them.
4394 20. Expanded the error message when iplsearch is called with an invalid key to
4395 suggest using net-iplsearch in a host list.
4397 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4398 ACLs are no longer actually imposed (and a message to that effect is
4401 22. If a "gecos" field in a passwd entry contained escaped characters, in
4402 particular, if it contained a \" sequence, Exim got it wrong when building
4403 a From: or a Sender: header from that name. A second bug also caused
4404 incorrect handling when an unquoted " was present following a character
4405 that needed quoting.
4407 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4408 was not being matched caselessly.
4410 24. Arranged for all hyphens in the exim.8 source to be escaped with
4413 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4414 properly. Recipient callout cache records were still being keyed to include
4415 the sender, even when use_sender was set false. This led to far more
4416 callouts that were necessary. The sender is no longer included in the key
4417 when use_sender is false.
4419 26. Added "control = submission" modifier to ACLs.
4421 27. Added the ${base62d: operator to decode base 62 numbers.
4423 28. dnsdb lookups can now access SRV records.
4425 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4426 the configuration file.
4428 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4429 (-v) mode. This makes the output for a verbose queue run more intelligible.
4431 31. Added a use_postmaster feature to recipient callouts.
4433 32. Added the $body_zerocount variable, containing the number of binary zero
4434 bytes in the message body.
4436 33. The time of last modification of the "new" subdirectory is now used as the
4437 "mailbox time last read" when there is a quota error for a maildir
4440 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4442 35. Added +ignore_unknown as a special item in host lists.
4444 36. Code for decoding IPv6 addresses in host lists is now included, even if
4445 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4446 address was recognized as an IP address, but was then not correctly decoded
4447 into binary, causing unexpected and incorrect effects when compared with
4454 1. Very minor rewording of debugging text in manualroute to say "list of
4455 hosts" instead of "hostlist".
4457 2. If verify=header_syntax was set, and a header line with an unqualified
4458 address (no domain) and a large number of spaces between the end of the
4459 name and the colon was received, the reception process suffered a buffer
4460 overflow, and (when I tested it) crashed. This was caused by some obsolete
4461 code that should have been removed. The fix is to remove it!
4463 3. When running in the test harness, delay a bit after writing a bounce
4464 message to get a bit more predictability in the log output.
4466 4. Added a call to search_tidyup() just before forking a reception process. In
4467 theory, someone could use a lookup in the expansion of smtp_accept_max_
4468 per_host which, without the tidyup, could leave open a database connection.
4470 5. Added the variables $recipient_data and $sender_data which get set from a
4471 lookup success in an ACL "recipients" or "senders" condition, or a router
4472 "senders" option, similar to $domain_data and $local_part_data.
4474 6. Moved the writing of debug_print from before to after the "senders" test
4477 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4478 problems for message scanning, either using a data ACL, or using
4479 local_scan() because the Received: header was not generated till after they
4480 were called (in order to set the time as the time of reception completion).
4481 I have revised the way this works. The header is now generated after the
4482 body is received, but before the ACL or local_scan() are called. After they
4483 are run, the timestamp in the header is updated.
4489 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4490 before starting a queue runner without re-exec. This happened only when
4491 deliver_drop_privilege was set or when the Exim user was set to root. The
4492 effect of the bug was that timeouts during subsequent deliveries caused
4493 crashes instead of being properly handled. The handler is now left at its
4494 default (and expected) setting.
4496 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4497 message, again when deliver_drop_privilege is set or Exim is run as root.
4498 The bug described in (1) was not present in this case, but the tidying up
4499 of the other signals was missing. I have made the two cases consistent.
4501 3. The ignore_target_hosts setting on a manualroute router was being ignored
4502 for hosts that were looked up using the /MX notation.
4504 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4507 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4508 operated on the sender address. After changing the $sender_address to <>
4509 for the sender address verify, Exim was re-instated it as the original
4510 (before rewriting) address, but remembering that it had rewritten it, so it
4511 wasn't rewriting it again. This bug also had the effect of breaking the
4512 sender address verification caching when the sender address was rewritten.
4514 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4515 This has been changed so that if the ip literal address matches
4516 ignore_target_hosts, the router declines.
4518 7. Added expansion conditions match_domain, match_address, and match_local_
4519 part (NOT match_host).
4521 8. The placeholder for the Received: header didn't have a length field set.
4523 9. Added code to Exim itself and to exim_lock to test for a specific race
4524 condition that could lead to file corruption when using MBX delivery. The
4525 issue is with the lockfile that is created in /tmp. If this file is removed
4526 after a process has opened it but before that process has acquired a lock,
4527 there is the potential for a second process to recreate the file and also
4528 acquire a lock. This could lead to two Exim processes writing to the file
4529 at the same time. The added code performs the same test as UW imapd; it
4530 checks after acquiring the lock that its file descriptor still refers to
4531 the same named file.
4533 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4534 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4537 11. Added the smtp_active_hostname option. If used, this will typically be made
4538 to depend on the incoming interface address. Because $interface_address is
4539 not set up until the daemon has forked a reception process, error responses
4540 that can happen earlier (such as "too many connections") no longer contain
4543 12. If an expansion in a condition on a "warn" statement fails because a lookup
4544 defers, the "warn" statement is abandoned, and the next ACL statement is
4545 processed. Previously this caused the whole ACL to be aborted.
4547 13. Added the iplsearch lookup type.
4549 14. Added ident_timeout as a log selector.
4551 15. Added tls_certificate_verified as a log selector.
4553 16. Added a global option tls_require_ciphers (compare the smtp transport
4554 option of the same name). This controls incoming TLS connections.
4556 17. I finally figured out how to make tls_require_ciphers do a similar thing
4557 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4558 before starting the TLS session.
4560 18. Tabs are now shown as \t in -bP output.
4562 19. If the log selector return_path_on_delivery was set, Exim crashed when
4563 bouncing a message because it had too many Received: header lines.
4565 20. If two routers both had headers_remove settings, and the first one included
4566 a superfluous trailing colon, the final name in the first list and the
4567 first name in the second list were incorrectly joined into one item (with a
4568 colon in the middle).
4574 1. Added -C and -D options to the exinext utility, mainly to make it easier
4575 to include in the automated testing, but these could be helpful when
4576 multiple configurations are in use.
4578 2. The exinext utility was not formatting the output nicely when there was
4579 an alternate port involved in the retry record key, nor when there was a
4580 message id as well (for retries that were specific to a specific message
4581 and a specific host). It was also confused by IPv6 addresses, because of
4582 the additional colons they contain. I have fixed the IPv4 problem, and
4583 patched it up to do a reasonable job for IPv6.
4585 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4586 delivery, the log line now contains "pipelined" if PIPELINING was used.
4588 4. An SMTP transport process used to panic and die if the bind() call to set
4589 an explicit outgoing interface failed. This has been changed; it is now
4590 treated in the same way as a connect() failure.
4592 5. A reference to $sender_host_name in the part of a conditional expansion
4593 that was being skipped was still causing a DNS lookup. This no longer
4596 6. The def: expansion condition was not recognizing references to header lines
4597 that used bh_ and bheader_.
4599 7. Added the _cache feature to named lists.
4601 8. The code for checking quota_filecount in the appendfile transport was
4602 allowing one more file than it should have been.
4604 9. For compatibility with Sendmail, the command line option
4612 and sets the incoming protocol and host name (for trusted callers). The
4613 host name and its colon can be omitted when only the protocol is to be set.
4614 Note the Exim already has two private options, -pd and -ps, that refer to
4615 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4616 "s", but I don't think that's a major issue.
4618 10. A number of refactoring changes to the code, none of which should affect
4621 (a) The number of logging options was getting close to filling up the
4622 32-bit word that was used as a bit map. I have split them into two classes:
4623 those that are passed in the argument to log_write(), and those that are
4624 only ever tested independently outside of that function. These are now in
4625 separate 32-bit words, so there is plenty of room for expansion again.
4626 There is no change in the user interface or the logging behaviour.
4628 (b) When building, for example, log lines, the code previously used a
4629 macro that called string_cat() twice, in order to add two strings. This is
4630 not really sufficiently general. Furthermore, there was one instance where
4631 it was actually wrong because one of the argument was used twice, and in
4632 one call a function was used. (As it happened, calling the function twice
4633 did not affect the overall behaviour.) The macro has been replaced by a
4634 function that can join an arbitrary number of extra strings onto a growing
4637 (c) The code for expansion conditions now uses a table and a binary chop
4638 instead of a serial search (which was left over from when there were very
4639 few conditions). Also, it now recognizes conditions like "pam" even when
4640 the relevant support is not compiled in: a suitably worded error message is
4641 given if an attempt is made to use such a condition.
4643 11. Added ${time_interval:xxxxx}.
4645 12. A bug was causing one of the ddress fields not to be passed back correctly
4646 from remote delivery subprocesses. The field in question was not being
4647 subsequently used, so this caused to problems in practice.
4649 13. Added new log selectors queue_time and deliver_time.
4651 14. Might have fixed a bug in maildirsizefile handling that threw up
4652 "unexpected character" debug warnings, and recalculated the data
4653 unnecessarily. In any case, I expanded the warning message to give more
4656 15. Added the message "Restricted characters in address" to the statements in
4657 the default ACL that block characters like @ and % in local parts.
4659 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4660 Three changes have been made:
4662 (a) There was a serious bug; a negative response to MAIL caused the whole
4663 recipient domain to be cached as invalid, thereby blocking all messages
4664 to all local parts at the same domain, from all senders. This bug has
4665 been fixed. The domain is no longer cached after a negative response to
4666 MAIL if the sender used is not empty.
4668 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4671 (c) A new callout option, "use_sender" has been added for people who want
4672 the modified behaviour.
4678 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4681 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4682 indeed breaks things for older releases.
4684 3. Added additional logging to the case where there is a problem reading data
4685 from a filter that is running in a subprocess using a pipe, in order to
4686 try to track down a specific problem.
4688 4. Testing facility fudge: when running in the test harness and attempting
4689 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4690 getting "No route to host". Convert this to a timeout.
4692 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4695 6. Some OS don't have socklen_t but use size_t instead. This affects the
4696 fifth argument of getsockopt() amongst other things. This is now
4697 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4698 can be set for individual OS. I have set it for SunOS5, OSF1, and
4699 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4700 some earlier ones do not.
4702 7. Change 4.30/15 was not doing the test caselessly.
4704 8. The standard form for an IPv6 address literal was being rejected by address
4705 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4706 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4707 this, as well as the form without the "IPv6" on the front (but only when
4708 address literals are enabled, of course).
4710 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4712 10. Exim crashed if a message with an empty sender address specified by -f
4713 encountered a router with an errors_to setting. This could be provoked only
4714 by a command such as
4718 where an empty string was supplied; "<>" did not hit this bug.
4720 11. Installed PCRE release 4.5.
4722 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4723 remained set. It is now erased.
4725 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4726 times from message ids (which are base 36 rather than the normal 62).
4728 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4729 were being counted as actual protocol errors, and logged if the log
4730 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4731 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4732 and DATA following a set of rejected RCPTs do not count as protocol errors.
4733 In other words, Exim assumes they were pipelined, though this may not
4734 actually be the case. Of course, in all cases the client gets an
4735 appropriate error code.
4737 15. If a lookup fails in an ACL condition, a message about the failure may
4738 be available; it is used if testing the ACL cannot continue, because most
4739 such messages specify what the cause of the deferral is. However, some
4740 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4741 that caused an old message to be retained and used if a later statement
4742 caused a defer, replacing the real cause of the deferral.
4744 16. If an IP address had so many PTR records that the DNS lookup buffer
4745 was not large enough to hold them, Exim could crash while trying to process
4746 the truncated data. It now detects and logs this case.
4748 17. Further to 4.21/58, another change has been made: if (and only if) the
4749 first line of a message (the first header line) ends with CRLF, a bare LF
4750 in a subsequent header line has a space inserted after it, so as not to
4751 terminate the header.
4753 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4754 unnecessarily, used atoi() instead of strtol(), and didn't check the
4755 termination when getting file sizes from file names by regex.
4757 19. Completely re-implemented the support for maildirsize files, in the light
4758 of a number of problems with the previous contributed implementation
4759 (4.30/29). In particular:
4761 . If the quota is zero, the maildirsize file is maintained, but no quota is
4764 . If the maildir directory does not exist, it is created before any attempt
4765 to write a maildirsize file.
4767 . The quota value in the file is just a cache; if the quota is changed in
4768 the transport, the new value overrides.
4770 . A regular expression is available for excluding directories from the
4773 20. The autoreply transport checks the characters in options that define the
4774 message's headers; it allows continued headers, but it was checking with
4775 isspace() after an embedded newline instead of explicitly looking for a
4778 21. If all the "regular" hosts to which an address was routed had passed their
4779 expiry times, and had not reached their retry times, the address was
4780 bounced, even if fallback hosts were defined. Now Exim should go on to try
4783 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4784 equivalent code in the SMTP transport. Some hosts send humungous responses
4785 to HELO/EHLO, more than 1024 it seems.
4787 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4788 gives compiler warnings in some environments. I've now done it "properly",
4791 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4792 (because of problems with the built-in one) was declared to return uschar *
4793 instead of char *, causing compiler failure.
4795 25. Fixed a file descriptor leak when processing alias/forward files.
4797 26. Fixed a minor format string issue in dbfn.c.
4799 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4801 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4802 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4805 29. When a sender address is verified, it is cached, to save repeating the test
4806 when there is more than one recipient in a message. However, when the
4807 verification involves a callout, it is possible for different callout
4808 options to be set for different recipients. It is too complicated to keep
4809 track of this in the cache, so now Exim always runs a verification when a
4810 callout is required, relying on the callout cache for the optimization.
4811 The overhead is duplication of the address routing, but this should not be
4814 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4815 to be verified with callout=postmaster, and the main callout worked but the
4816 postmaster check failed, the verification correctly failed. However, if a
4817 subsequent RCPT command asked for sender verification *without* the
4818 postmaster check, incorrect caching caused this verification also to fail,
4821 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4822 it was not caching the DNS options (qualify_single, search_parents) that
4823 were used when the lookup failed. A subsequent lookup with different
4824 options therefore always gave the same answer, though there were cases
4825 where it should not have. (Example: a "domains = !$mx_any" option on a
4826 dnslookup router: the "domains" option is always processed without any
4827 widening, but the router might have qualify_single set.) Now Exim uses the
4828 cached value only when the same options are set.
4830 32. Added John Jetmore's "exipick" utility to the distribution.
4832 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4833 than a timeout (e.g. a certificate is required, and is not provided), an
4834 Exim server now closes the connection immediately. Previously it waited for
4835 the client to close - but if the client is SSL, it seems that they each
4836 wait for each other, leading to a delay before one of them times out.
4838 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4839 maintained 0.8.x compatibility because I don't think many are using it, and
4840 it is clearly obsolete.
4842 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4845 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4846 client certificate was expired. A simple patch fixes this, though I don't
4847 understand the full logic of why the verify callback is called multiple
4850 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4851 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4852 which causes problems with some clients (such as the Certicom SSL Plus
4853 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4854 disables the coutermeasure allowing Eudora to connect."
4856 38. Exim was not checking that a write() to a log file succeeded. This could
4857 lead to Bad Things if a log got too big, in particular if it hit a file
4858 size limit. Exim now panics and dies if it cannot write to a log file, just
4859 as it does if it cannot open a log file.
4861 39. Modified OS/Makefile-Linux so that it now contains
4863 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4865 The two -D definitions ensure that Exim is compiled with large file
4866 support, which makes it possible to handle log files that are bigger than
4869 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4870 instance) a domain was checked against a named list that involved a lookup,
4871 causing $domain_data to be set, then another domain was checked against the
4872 same list, then the first domain was re-checked, the value of $domain_data
4873 after the final check could be wrong. In particular, if the second check
4874 failed, it could be set empty. This bug probably also applied to
4877 41. The strip_trailing_dot option was not being applied to the address given
4878 with the -f command-line option.
4880 42. The code for reading a message's header from the spool was incrementing
4881 $received_count, but never initializing it. This meant that the value was
4882 incorrect (doubled) while delivering a message in the same process in which
4883 it was received. In the most common configuration of Exim, this never
4884 happens - a fresh exec is done - but it can happen when
4885 deliver_drop_privilege is set.
4887 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4888 it now includes up to 150 characters of the unexpected data in the log
4891 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4892 and building data strings. The size of both of these buffers was 10 000
4893 bytes - far larger than anybody would *ever* want, thought I. Needless to
4894 say, somebody hit the limit. I have increased the maximum line length to
4895 20 000 and the maximum data length of concatenated lines to 100 000. I have
4896 also fixed two bugs, because there was no checking on these buffers. Tsk,
4897 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4898 buffer is too small.
4900 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4901 lsearch lookups. Now it does.
4903 46. When parsing a route_list item in a manualroute router, a fixed-length
4904 buffer was used for the list of hosts. I made this 1024 bytes long,
4905 thinking that nobody would ever have a list of hosts that long. Wrong.
4906 Somebody had a whole pile of complicated expansion conditions, and the
4907 string was silently truncated, leading to an expansion error. It turns out
4908 that it is easier to change to an unlimited length (owing to other changes
4909 that have happened since this code was originally written) than to build
4910 structure for giving a limitation error. The length of the item that
4911 expands into the list of hosts is now unlimited.
4913 47. The lsearch lookup could not handle data where the length of text line was
4914 more than 4095 characters. Such lines were truncated, leading to shortened
4915 data being returned. It should now handle lines of any length.
4917 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4918 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4921 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4922 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4923 no longer happens in comment lines. A list of replacements is now placed
4924 at the head of all of the source files, except those whose only change is
4925 to replace PERL_COMMAND in the very first #! line.
4927 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4928 messages on the queue, with a bottom-up merge sort, using code contributed
4929 by Michael Haardt. This should make operations like -bp somewhat faster on
4930 large queues. It won't affect queue runners, except when queue_run_in_order
4933 51. Installed eximstats 1.31 in the distribution.
4935 52. Added support for SRV lookups to the dnslookup router.
4937 53. If an ACL referred to $message_body or $message_body_end, the value was not
4938 reset for any messages that followed in the same SMTP session.
4940 54. The store-handling optimization for building very long strings was not
4941 differentiating between the different store pools. I don't think this
4942 actually made any difference in practice, but I've tidied it.
4944 55. While running the routers to verify a sender address, $sender_address
4945 was still set to the sender address. This is wrong, because when routing to
4946 send a bounce to the sender, it would be empty. Therefore, I have changed
4947 it so that, while verifying a sender address, $sender_address is set to <>.
4948 (There is no change to what happens when verifying a recipient address.)
4950 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4951 target A or AAAA records (if not already returned) without resetting the
4952 qualify_single or search_parents options of the DNS resolver. These are
4953 inappropriate in this case because the targets of MX and SRV records must
4954 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4955 target that, when qualified, matched something in the local domain. These
4956 two options are now turned off when doing these lookups.
4958 57. It seems that at least some releases of Reiserfs (which does not have the
4959 concept of a fixed number of inodes) returns zero and not -1 for the
4960 number of available inodes. This interacted badly with check_spool_inodes,
4961 which assumed that -1 was the "no such thing" setting. What I have done is
4962 to check that the total number of inodes is greater than zero before doing
4963 the test of how many are available.
4965 58. When a "warn" ACL statement has a log_message modifier, the message is
4966 remembered, and not repeated. This is to avoid a lot of repetition when a
4967 message has many recipients that cause the same warning to be written.
4968 Howewer, Exim was preserving the list of already written lines for an
4969 entire SMTP session, which doesn't seem right. The memory is now reset if a
4970 new message is started.
4972 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4973 debugging output unless log_rewrite was also set.
4975 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4976 of (int)(handle) when we know that handle contains (void *)(-1).
4978 61. The Exim daemon panic-logs an error return when it closes the incoming
4979 connection. However "connection reset by peer" seems to be common, and
4980 isn't really an error worthy of noting specially, so that particular error
4983 62. When Exim is trying to find all the local interfaces, it used to panic and
4984 die if the ioctl to get the interface flags failed. However, it seems that
4985 on at least one OS (Solaris 9) it is possible to have an interface that is
4986 included in the list of interfaces, but for which you get a failure error
4987 for this call. This happens when the interface is not "plumbed" into a
4988 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4989 failure of the "get flags" call assumes that the interface is down.
4991 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4992 makes life easier for people who are doing arithmetic on fields extracted
4993 from dates, where you often get leading zeros that should not be
4994 interpreted as octal.
4996 64. Added qualify_domain to the redirect router, to override the global
4999 65. If a pathologically long header line contained very many addresses (the
5000 report of this problem mentioned 10 000) and each of them was rewritten,
5001 Exim could use up a very large amount of memory. (It kept on making new
5002 copies of the header line as it rewrote, and never released the old ones.)
5003 At the expense of a bit more processing, the header rewriting function has
5004 been changed so that it no longer eats memory in this way.
5006 66. The generation of the Received: header has been moved from the time that a
5007 message starts to be received, to the time that it finishes. The timestamp
5008 in the Received: header should now be very close to that of the <= log
5009 line. There are two side-effects of this change:
5011 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5012 logged header lines no longer include the local Received: line, because
5013 it has not yet been created. The same applies to a copy of the message
5014 that is returned to a non-SMTP sender when a message is rejected.
5016 (b) When a filter file is tested using -bf, no additional Received: header
5017 is added to the test message. After some thought, I decided that this
5020 This change does not affect the value of $received_for. It is still set
5021 after address rewriting, but before local_scan() is called.
5023 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5025 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5026 gave an unhelpful panic error message, and a defer error. I have managed to
5027 change this behaviour so that it now rejects any supplied certificate,
5028 which seems right, as the list of acceptable certificates is empty.
5030 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5031 gave an unhelpful defer error. I have not managed to make this reject any
5032 supplied certificates, but the error message it gives is "no certificate
5033 supplied", which is not helpful.
5035 70. exigrep's output now also includes lines that are not associated with any
5036 message, but which match the given pattern. Implemented by a patch from
5037 Martin Sluka, which also tidied up the Perl a bit.
5039 71. Recipient callout verification, like sender verification, was using <> in
5040 the MAIL FROM command. This isn't really the right thing, since the actual
5041 sender may affect whether the remote host accepts the recipient or not. I
5042 have changed it to use the actual sender in the callout; this means that
5043 the cache record is now keyed on a recipient/sender pair, not just the
5044 recipient address. There doesn't seem to be a real danger of callout loops,
5045 since a callout by the remote host to check the sender would use <>.
5046 [SEE ABOVE: changed after hitting problems.]
5048 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5049 temporary errors. However, in the case of such a code being given after
5050 the end of a data transmission (i.e. after ".") Exim was failing to write
5051 a retry record for the message. (Yes, there was some broken host that was
5052 actually sending 8xx at this point.)
5054 73. An unknown lookup type in a host list could cause Exim to panic-die when
5055 the list was checked. (An example that provoked this was putting <; in the
5056 middle of a list instead of at the start.) If this happened during a DATA
5057 ACL check, a -D file could be left lying around. This kind of configuration
5058 error no longer causes Exim to die; instead it causes a defer errror. The
5059 incident is still logged to the main and panic logs.
5061 74. Buglet left over from Exim 3 conversion. The message "too many messages
5062 in one connection" was written to the rejectlog but not the mainlog, except
5063 when address rewriting (yes!) was being logged.
5065 75. Added write_rejectlog option.
5067 76. When a system filter was run not as root (that is, when system_filter_user
5068 was set), the values of the $n variables were not being returned to the
5069 main process; thus, they were not subsequently available in the $sn
5072 77. Added +return_path_on_delivery log selector.
5074 78. A connection timeout was being treated differently from recipients deferred
5075 when testing hosts_max_try with a message that was older than the host's
5076 retry timeout. (The host should not be counted, thus allowing all hosts to
5077 be tried at least once before bouncing.) This may have been the cause of an
5078 occasionally reported bug whereby a message would remain on the queue
5079 longer than the retry timeout, but would be bounced if a delivery was
5080 forced. I say "may" because I never totally pinned down the problem;
5081 setting up timeout/retry tests is difficult. See also the next item.
5083 79. The ultimate address timeout was not being applied to errors that involved
5084 a combination of host plus message (for example, a timeout on a MAIL
5085 command). When an address resolved to a number of possible hosts, and they
5086 were not all tried for each delivery (e.g. because of hosts_max_try), a
5087 message could remain on the queue longer than the retry timeout.
5089 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5092 81. Fixed an obscure SMTP outgoing bug which required at least the following
5093 conditions: (a) there was another message waiting for the same server;
5094 (b) the server returned 5xx to all RCPT commands in the first message so
5095 that the message was not completed; (c) the server dropped the connection
5096 or gave a negative response to the RSET that Exim sends to abort the
5097 transaction. The observed case was a dropped connection after DATA that had
5098 been sent in pipelining mode. That is, the server had advertised PIPELINING
5099 but was not implementing it correctly. The effect of the bug was incorrect
5100 behaviour, such as trying another host, and this could lead to a crash.
5106 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5107 and daemon.c were passed as pointers to ints; they should have been
5108 pointers to socklen_t variables (which are typically unsigned ints).
5110 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5113 3. Fixed a really odd bug that affected only the testing scheme; patching a
5114 certain fixed string in the binary changed the value of another string that
5115 happened to be identical to the end of the original first string.
5117 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5118 name", it returns that address as the IP address. On some operating
5119 systems (e.g. Solaris), it also passes back the IP address string as the
5120 "host name". However, on others (e.g. Linux), it passes back an empty
5121 string. Exim wasn't checking for this, and was changing the host name to an
5122 empty string, assuming it had been canonicized.
5124 5. Although rare, it is permitted to have more than one PTR record for a given
5125 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5126 all the names associated with an address, because they do in Solaris.
5127 However, it seems that they do not in Linux for data that comes from the
5128 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5129 I found this out when I moved to a new Linux workstation and tried to run
5130 the Exim test suite.
5132 To get round this problem I have changed the code so that it now does its
5133 own call to the DNS to look up PTR records when searching for a host name.
5134 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5135 addresses that are only in /etc/hosts are still found.
5137 This behaviour is, however, controlled by an option called host_lookup_
5138 order, which defaults to "bydns:byaddr". If people want to use the other
5139 order, or indeed, just use one or the other means of lookup, they can
5140 specify it in this variable.
5142 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5143 some operating systems, this comes back from gethostbyaddr() as an empty
5144 string, and this is what Exim used to test for. However, it seems that in
5145 other systems, "." is yielded. Exim now tests for this case too.
5147 7. The values of check_spool_space and check_log_space are now held internally
5148 as a number of kilobytes instead of an absolute number of bytes. If a
5149 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5150 kilobyte. This means that much larger values can be stored.
5152 8. Exim monitor: an attempt to get the action menu when not actually pointing
5153 at a message produces an empty menu entitled "No message selected". This
5154 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5155 no entries in it ("Shell widget menu has zero width and/or height"). So I
5156 have added a single, blank menu entry in this case.
5158 9. Added ${quote_local_part.
5160 10. MIME decoding is now applied to the contents of Subject: header lines when
5163 11. Now that a reference to $sender_host_address automatically causes a reverse
5164 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5165 host lookup before query-style lookups in lists that might use this
5166 variable. This has therefore been abolished, and the "net-" prefix is no
5167 longer necessary for query-style lookups.
5169 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5170 have been a typo for LFLAGS, so it has been changed.
5172 13. The install script calls Exim with "-C /dev/null" in order to find the
5173 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5174 to be output. Howeve, since Exim outputs its version number before the
5175 error, it didn't break the script. It just looked ugly. I fixed this by
5176 always allowing "-C /dev/null" if the caller is root.
5178 14. Ignore overlarge ACL variable number when reading spool file - insurance
5179 against a later release with more variables having written the file.
5181 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5182 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5183 this, as well as the form without the "IPv6" on the front.
5185 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5186 OS/Makefile-Darwin file.
5188 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5189 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5191 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5192 variable. However, some OS use ints for inodes. I've added cast to long int
5193 to get rid of the compiler warning.
5195 19. I had forgotten to lock out "/../" in configuration file names when
5196 ALT_CONFIG_PREFIX was set.
5198 20. Routers used for verification do not need to specify transports. However,
5199 if such a router generated a host list, and callout was configured, Exim
5200 crashed, because it could not find a port number from the (non-existent)
5201 transport. It now assumes port 25 in this circumstance.
5203 21. Added the -t option to exigrep.
5205 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5206 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5207 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5208 them is set, LOOKUP_LSEARCH is forced.
5210 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5212 24. Added sender and host information to the "rejected by local_scan()" log
5213 line; previously there was no indication of these.
5215 25. Added .include_if_exists.
5217 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5218 Linux. It turns out that not all file systems support this. Apparently some
5219 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5220 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5221 not supported on the file descriptor, is now ignored when Exim is trying to
5222 sync a directory. This applies only to Linux.
5224 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5226 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5227 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5230 29. Added support for maildirsize files from supplied patch (modified a bit).
5232 30. The use of :fail: followed by an empty string could lead Exim to respond to
5233 sender verification failures with (e.g.):
5235 550 Verification failed for <xxx>
5236 550 Sender verify failed
5238 where the first response line was missing the '-' that indicates it is not
5239 the final line of the response.
5241 31. The loop for finding the name of the user that called Exim had a hardwired
5242 limit of 10; it now uses the value of finduser_retries, which is used for
5243 all other user lookups.
5245 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5248 33. Exim was neglecting to zero errno before one call of strtol() when
5249 expanding a string and expecting an integer value. On some systems this
5250 resulted in spurious "integer overflow" errors. Also, it was casting the
5251 result into an int without checking.
5253 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5254 did not work. The code looks as if it has *never* worked, though it appears
5255 to have been documented since at least releast 1.62. I have made it work.
5257 35. The "timeout_DNS" error in retry rules, also documented since at least
5258 1.62, also never worked. As it isn't clear exactly what this means, and
5259 clearly it isn't a major issue, I have abolished the feature by treating it
5260 as "timeout", and writing a warning to the main and panic logs.
5262 36. The display of retry rules for -brt wasn't always showing the error code
5265 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5266 timeout_connect_A, timeout_connect_MX.
5268 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5269 to the empty sender.
5271 39. The daemon was not analysing the content of -oX till after it had closed
5272 stderr and disconnected from the controlling terminal. This meant that any
5273 syntax errors were only noted on the panic log, and the return code from
5274 the command was 0. By re-arranging the code a little, I've made the
5275 decoding happen first, so such errors now appear on stderr, and the return
5276 code is 1. However, the actual setting up of the sockets still happens in
5277 the disconnected process, so errors there are still only recorded on the
5280 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5281 connections (as happens on some IP stacks) was logged at start up time as
5282 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5283 it from "IPv6 and IPv4", which means that two separate sockets are being
5286 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5287 says whether AF_INET or AF_INET6 was passed as an argument.
5289 42. Exiwhat output was messed up when time zones were included in log
5292 43. Exiwhat now gives more information about the daemon's listening ports,
5293 and whether -tls-on-connect was used.
5295 44. The "port" option of the smtp transport is now expanded.
5297 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5298 silently ignored. Now an error message is written to the main and panic
5301 46. There's a new ACL modifier called "logwrite" which writes to a log file
5302 as soon as it is encountered.
5304 47. Added $local_user_uid and $local_user_gid at routing time.
5306 48. Exim crashed when trying to verify a sender address that was being
5309 49. Exim was recognizing only a space character after ".include". It now also
5310 recognizes a tab character.
5312 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5313 extracting the relevant information from the specification. The man page no
5314 longer contains scrambled data for the -d option, and I've added a section
5315 at the front about calling Exim under different names.
5317 51. Added "extra_headers" argument to the "mail" command in filter files.
5319 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5322 53. Installed eximstats 1.29.
5324 54. Added transport_filter_timeout as a generic transport option.
5326 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5327 Cc: header lines. This was required by RFC 822, but it not required by RFC
5330 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5331 incoming messages that did not have them. Now it does so only if the
5332 message originates locally, that is, if there is no associated remote host
5333 address. When Resent- header lines are present, this applies to the Resent-
5334 lines rather than the non-Resent- lines.
5336 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5337 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5339 58. Messages for configuration errors now include the name of the main
5340 configuration file - useful now that there may be more than one file in a
5341 list (.included file names were always shown).
5343 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5344 for those rare installations that do not start the daemon as root or run it
5345 setuid root. I've cut out the call to initgroups() if the daemon is not
5348 60. The Exim user and group can now be bound into the binary as text strings
5349 that are looked up at the start of Exim's processing.
5351 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5353 62. Added $mailstore_basename variable.
5355 63. Installed patch to sieve.c from Michael Haardt.
5357 64. When Exim failed to open the panic log after failing to open the main log,
5358 the original message it was trying to log was written to stderr and debug
5359 output, but if they were not available (the usual case in production), it
5360 was lost. Now it is written to syslog before the two lines that record the
5361 failures to open the logs.
5363 65. Users' Exim filters run in subprocesses under the user's uid. It is
5364 possible for a "deliver" command or an alias in a "personal" command to
5365 provoke an address rewrite. If logging of address rewriting is configured,
5366 this fails because the process is not running as root or exim. There may be
5367 a better way of dealing with this, but for the moment (because 4.30 needs
5368 to be released), I have disabled address rewrite logging when running a
5369 filter in a non-root, non-exim process.
5375 1. The buildconfig auxiliary program wasn't quoting the value set for
5376 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5377 not defined. This bug was masked in 4.22 by the effect that was fixed in
5380 2. Some messages that were rejected after a message id was allocated were
5381 shown as "incomplete" by exigrep. It no longer does this for messages that
5382 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5384 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5385 have allow_domain_literals set, the ID did not get logged in the <= line.
5386 Domain literals are now always recognized in Message-ID: header lines.
5388 4. The first argument for a ${extract expansion item is the key name or field
5389 number. Leading and trailing spaces in this item were not being ignored,
5390 causing some misleading effects.
5392 5. When deliver_drop_privilege was set, single queue runner processes started
5393 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5394 same command in the process it spins off) were not dropping privilege.
5396 6. When the daemon running as "exim" started a queue runner, it always
5397 re-executed Exim in the spun-off process. This is a waste of effort when
5398 deliver_drop_privilege is set. The new process now just calls the
5399 queue-runner function directly.
5405 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5408 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5409 ignored. Though the use of -oP was forcing the writing of a pid file, it
5410 was always written to the default place.
5412 3. If the message "no IP address found for host xxxx" is generated during
5413 incoming verification, it is now followed by identification of the incoming
5414 connection (so you can more easily find what provoked it).
5416 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5418 5. Added some features to "harden" Exim a bit more against certain attacks:
5420 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5421 be put in Local/Makefile. This is like the never_users runtime option,
5422 but it cannot be overridden. The default setting is "root".
5424 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5425 prefix string with which any file named in a -C command line option
5428 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5429 is retained for -C and -D only if the caller of Exim is root. Without
5430 it, the exim user may also use -C and -D and retain privilege.
5432 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5433 command line option is disabled.
5435 6. Macro names set by the -D option must start with an upper case letter, just
5436 like macro names defined in the configuration file.
5438 7. Added "dereference=" facility to LDAP.
5440 8. Two instances of the typo "uknown" in the source files are fixed.
5442 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5443 the Configure-Makefile script screwed up while processing it.
5445 10. Incorporated PCRE 4.4.
5447 11. The SMTP synchronization check was not operating right at the start of an
5448 SMTP session. For example, it could not catch a HELO sent before the client
5449 waited for the greeting. There is now a check for outstanding input at the
5450 point when the greeting is written. Because of the duplex, asynchronous
5451 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5452 way, but not yet received, when the check is performed.
5454 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5455 on TCP/IP sockets, because this apparently causes some broken clients to
5458 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5459 unchanged) from the Cygwin maintainer.
5461 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5462 was supported instead of testing for mbx. Effectively a typo.
5464 15. The spa authenticator server code was not checking that the input it
5465 received was valid base64.
5467 16. The debug output line for the "set" modifier in ACLs was not showing the
5468 name of the variable that was being set.
5470 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5472 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5473 lookup on its own. Something else was needed to trigger the lookup. For
5474 example, a match in host_lookup or the need for a host name in a host list.
5475 Now, if $sender_host_name is referenced and the host name has not yet been
5476 looked up, a lookup is performed. If the lookup fails, the variable remains
5477 empty, and $host_lookup_failed is set to "1".
5479 19. Added "eqi" as a case-independent comparison operator.
5481 20. The saslauthd authentication condition could segfault if neither service
5482 nor realm was specified.
5484 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5485 error message that was logged was misleading, and incoming SMTP
5486 connections were dropped. The message is now more accurate, and temporary
5487 errors are given to SMTP connections.
5489 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5490 (except RSET) with a 421 error, until QUIT is received. However, it was
5491 failing to send a response to QUIT.
5493 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5494 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5495 ACL is now run after the helo_try_verify_hosts code.
5497 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5498 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5499 case-independent, so other case variants are also recognized). Apparently
5500 some systems use these upper case variants.
5502 25. If more than two messages were waiting for the same host, and a transport
5503 filter was specified for the transport, Exim sent two messages over the
5504 same TCP/IP connection, and then failed with "socket operation on non-
5505 socket" when it tried to send the third.
5507 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5509 27. The extern definition of crypt16() in expand.c was not being excluded when
5510 the OS had its own crypt16() function.
5512 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5513 as a preferred synonym for return_size_limit, both as an option and as an
5516 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5518 30. Changed the default configuration ACL to relax the local part checking rule
5519 for addresses that are not in any local domains. For these addresses,
5520 slashes and pipe symbols are allowed within local parts, but the sequence
5521 /../ is explicitly forbidden.
5523 31. SPA server authentication was not clearing the challenge buffer before
5526 32. log_message in a "warn" ACL statement was writing to the reject log as
5527 well as to the main log, which contradicts the documentation and doesn't
5528 seem right (because no rejection is happening). So I have stopped it.
5530 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5531 However, I am unable to do any testing of this.
5533 34. Fixed an infelicity in the appendfile transport. When checking directories
5534 for a mailbox, to see if any needed to be created, it was accidentally
5535 using path names with one or more superfluous leading slashes; tracing
5536 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5538 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5539 contents are added to the log line that is written for every discarded
5540 recipient. (Previously a log_message setting was ignored.)
5542 36. The ${quote: operator now quotes the string if it is empty.
5544 37. The install script runs exim in order to find its version number. If for
5545 some reason other than non-existence or emptiness, which it checks, it
5546 could not run './exim', it was installing it with an empty version number,
5547 i.e. as "exim-". This error state is now caught, and the installation is
5550 38. An argument was missing from the function that creates an error message
5551 when Exim fails to connect to the socket for saslauthd authentication.
5552 This could cause Exim to crash, or give a corrupted message.
5554 39. Added isip, isip4, and isip6 to ${if conditions.
5556 40. The ACL variables $acl_xx are now saved with the message, and can be
5557 accessed later in routers, transports, and filters.
5559 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5560 strings in the file are not string-expanded.
5562 42. If a MAIL command specified a SIZE value that was too large to fit into an
5563 int variable, the check against message_size_limit failed. Such values are
5564 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5565 day this will have to be increased, but I don't think I want to be around
5566 when emails are that large.
5573 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5574 iconv() is not standard in FreeBSD.
5576 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5577 IPv6 enabled. The observed symptom was a segmentation fault on return from
5578 the function os_common_find_running_interfaces() in src/os.c.
5580 3. In the check_special_case() function in daemon.c I had used "errno" as an
5581 argument name, which causes warnings on some systems. This was basically a
5582 typo, since it was named "eno" in the comments!
5584 4. The code that waits for the clock to tick (at a resolution of some fraction
5585 of a second) so as to ensure message-id uniqueness was always waiting for
5586 at least one whole tick, when it could have waited for less. [This is
5587 almost certainly not relevant at current processor speeds, where it is
5588 unlikely to ever wait at all. But we try to future-proof.]
5590 5. The function that sleeps for a time interval that includes fractions of a
5591 second contained a race. It did not block SIGALRM between setting the
5592 timer, and suspending (a couple of lines later). If the interval was short
5593 and the sigsuspend() was delayed until after it had expired, the suspension
5594 never ended. On busy systems this could lead to processes getting stuck for
5597 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5598 process, before it forks any delivery processes. The open lookup caching
5599 mechanism meant that the open file or database connection was passed into
5600 the delivery process. The problem was that delivery processes always tidy
5601 up cached lookup data. This could cause a problem for the next delivery
5602 process started by the queue runner, because the external queue runner
5603 process does not know about the closure. So the next delivery process
5604 still has data in the lookup cache. In the case of a file lookup, there was
5605 no problem because closing a file descriptor in a subprocess doesn't affect
5606 the parent. However, if the lookup was caching a connection to a database,
5607 the connection was closed, and the second delivery process was likely to
5608 see errors such as "PGSQL: query failed: server closed the connection
5609 unexpectedly". The problem has been fixed by closing all cached lookups
5610 in a queue runner before running a delivery process.
5612 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5613 seem to have the "const" qualifier which it has on other OS. I've
5616 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5617 *of the same type* (client or server) with the same public name that an
5618 error should be diagnosed.
5620 9. When Exim looked up a host name for an IP address, but failed to find the
5621 original IP address when looking up the host name (a safety check), it
5622 output the message "<ip address> does not match any IP for NULL", which was
5623 confusing, to say the least. The bug was that the host name should have
5624 appeared instead of "NULL".
5626 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5627 user that is built into the binary, and the -C or -D options is used, root
5628 privilege is dropped before the configuration file is read. In addition,
5629 logging is switched to stderr instead of the normal log files. If the
5630 configuration then re-defines the Exim user, the unprivileged environment
5631 is probably not what is expected, so Exim logs a panic warning message (but
5634 However, if deliver_drop_privilege is set, the unprivileged state may well
5635 be exactly what is intended, so the warning has been cut out in that case,
5636 and Exim is allowed to try to write to its normal log files.
5642 1. smtp_return_error_details was not giving details for temporary sender
5643 or receiver verification errors.
5645 2. Diagnose a configuration error if two authenticators have the same public
5648 3. Exim used not to create the message log file for a message until the first
5649 delivery attempt. This could be confusing when incoming messages were held
5650 for policy or load reasons. The message log file is now created at the time
5651 the message is received, and an initial "Received" line is written to it.
5653 4. The automatically generated man page for command line options had a minor
5654 bug that caused no ill effects; however, a more serious problem was that
5655 the procedure for building the man page automatically didn't always
5656 operate. Consequently, release 4.20 contains an out-of-date version. This
5657 shouldn't happen again.
5659 5. When building Exim with embedded Perl support, the script that builds the
5660 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5661 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5663 6. The freeze_tell option was not being used for messages that were frozen on
5664 arrival, either by an ACL or by local_scan().
5666 7. Added the smtp_incomplete_transaction log selector.
5668 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5669 was accepting AUTH without a new EHLO.
5671 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5672 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5675 10. Logging of TCP/IP connections (when configured) now happens in the main
5676 daemon process instead of the child process, so that the TCP/IP connection
5677 count is more accurate (but it can never be perfect).
5679 11. The use of "drop" in a nested ACL was not being handled correctly in the
5680 outer ACL. Now, if condition failure induced by the nested "drop" causes
5681 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5682 or "require"), the connection is dropped.
5684 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5685 that yield "discard" can now be used with an "accept" or a "discard" verb,
5686 but an error is generated for any others (because I can't see a useful way
5687 to define what should happen).
5689 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5690 are now processed in the same way as lines in the Exim configuration file.
5691 In particular, continuation lines are supported.
5693 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5695 15. Added -ti meaning -t -i.
5697 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5698 domains, and warn by logging if others are found.
5700 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5701 returned from GIFCONF when Exim is trying to find the list of interfaces on
5702 a host. The code in os.c has been modified to copy each ifreq to an aligned
5703 structure in all cases.
5705 Also, in some cases, the returned ifreq's were being copied to a 'struct
5706 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5707 means the last couple of bytes of an IPv6 address could be chopped if the
5708 ifreq contained only a normal sockaddr (14 bytes storage).
5710 18. Named domain lists were not supported in the hosts_treat_as_local option.
5711 An entry such as +xxxx was not recognized, and was treated as a literal
5714 19. Ensure that header lines added by a DATA ACL are included in the reject log
5715 if the ACL subsequently rejects the message.
5717 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5718 MD5 (which is deprecated).
5720 21. When testing a filter file using -bf, Exim was writing a message when it
5721 took the sender from a "From " line in the message, but it was not doing so
5722 when it took $return_path from a Return-Path: header line. It now does.
5724 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5725 with a valid header line field name (a series of printing characters
5726 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5728 23. Changed "disc" in the source to "disk" to conform to the documentation and
5729 the book and for uniformity.
5731 24. Ignore Sendmail's -Ooption=value command line item.
5733 25. When execve() failed while trying to run a command in a pipe transport,
5734 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5735 could be confused with a return value of 69 from the command itself. This
5736 has been changed to 127, the value the shell returns if it is asked to run
5737 a non-existent command. The wording for the related log line suggests a
5738 non-existent command as the problem.
5740 26. If received_header_text expands to an empty string, do not add a Received:
5741 header line to the message. (Well, it adds a token one on the spool, but
5742 marks it "old" so that it doesn't get used or transmitted.)
5744 27. Installed eximstats 1.28 (addition of -nt option).
5746 28. There was no check for failure on the call to getsockname() in the daemon
5747 code. This can fail if there is a shortage of resources on the system, with
5748 ENOMEM, for example. A temporary error is now given on failure.
5750 29. Contrary to the C standard, it seems that in some environments, the
5751 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5752 program. Exim now does this explicitly; it affects the formatting of
5753 timestamps using strftime().
5755 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5756 complaints. I've now initialized all the variables, to avoid this.
5758 32. Header lines added by a system filter were not being "seen" during
5759 transport-time rewrites.
5761 33. The info_callback() function passed to OpenSSL is set up with type void
5762 (*)(SSL *, int, int), as described somewhere. However, when calling the
5763 function (actually a macro) that sets it up, the type void(*)() is
5764 expected. I've put in a cast to prevent warnings from picky compilers.
5766 34. If a DNS black list lookup found a CNAME record, but there were no A
5767 records associated with the domain it pointed at, Exim crashed.
5769 35. If a DNS black list lookup returned more than one A record, Exim ignored
5770 all but the first. It now scans all returned addresses if a particular IP
5771 value is being sought. In this situation, the contents of the
5772 $dnslist_value variable are a list of all the addresses, separated by a
5775 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5776 to accept a host name and all its aliases if the forward lookup for any of
5777 them yielded the IP address of the incoming connection. Now it accepts only
5778 those names whose forward lookup yields the correct IP address. Any other
5779 names are discarded. This closes a loophole whereby a rogue DNS
5780 administrator could create reverse DNS records to break through a
5781 wildcarded host restriction in an ACL.
5783 37. If a user filter or a system filter that ran in a subprocess used any of
5784 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5785 the wrong values were passed to the pipe command ($thisaddress had the
5786 value of $0, $0 had the value of $1, etc). This bug was introduced by
5787 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5789 38. Improved the line breaking for long SMTP error messages from ACLs.
5790 Previously, if there was no break point between 40 and 75 characters, Exim
5791 left the rest of the message alone. Two changes have been made: (a) I've
5792 reduced the minimum length to 35 characters; (b) if it can't find a break
5793 point between 35 and 75 characters, it looks ahead and uses the first one
5794 that it finds. This may give the occasional overlong line, but at least the
5795 remaining text gets split now.
5797 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5798 file descriptors might be low, and that setting 1000 would always raise it.
5799 It turns out that in some environments, the limit is already over 1000 and
5800 that lowering it causes trouble. So now Exim takes care not to decrease it.
5802 40. When delivering a message, the value of $return_path is set to $sender_
5803 address at the start of routing (routers may change the value). By an
5804 oversight, this default was not being set up when an address was tested by
5805 -bt or -bv, which affected the outcome if any router or filter referred to
5808 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5809 log only when "message" or "log_message" are set. However, if one of the
5810 conditions was an address verification, or a call to a nested ACL, the
5811 messages generated by the underlying test were being passed through. This
5812 no longer happens. The underlying message is available in $acl_verify_
5813 message for both "message" and "log_message" expansions, so it can be
5814 passed through if needed.
5816 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5817 new expansion $bh_ to give the encoded byte string without charset
5818 translation. Translation happens only if iconv() is available; HAVE_ICONV
5819 indicates this at build time. HEADERS_CHARSET gives the charset to
5820 translate to; headers_charset can change it in the configuration, and
5821 "headers charset" can change it in an individual filter file.
5823 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5824 that creates RFC 2047 encoded "words" labels them as that charset instead
5825 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5826 expansion operator; (ii) when Exim creates a From: line for a local
5827 message; (iii) when a header line is rewritten to include a "phrase" part.
5829 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5830 buggy, causing it to skip the first lines of messages whose message ID
5831 ended in 'D'. This would not have bitten before Exim release 4.14, because
5832 message IDs were unlikely to end in 'D' before then. The effect was to have
5833 incorrect size information for certain domains.
5835 45. #include "config.h" was missing at the start of the crypt16.c module. This
5836 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5839 46. If there was a timeout during a "random" callout check, Exim treated it as
5840 a failure of the random address, and carried on sending RSET and the real
5841 address. If the delay was just some slowness somewhere, the response to the
5842 original RCPT would be taken as a response to RSET and so on, causing
5843 mayhem of various kinds.
5845 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5846 when I implemented it. It didn't allow for the fact that some option values
5847 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5848 the right test for positive values.
5850 48. Domain names in DNS records are case-independent. Exim always looks them up
5851 in lower case. Some resolvers return domain names in exactly the case they
5852 appear in the zone file, that is, they may contain uppercase letters. Not
5853 all resolvers do this - some return always lower case. Exim was treating a
5854 change of case by a resolver as a change of domain, similar to a widening
5855 of a domain abbreviation. This triggered its re-routing code and so it was
5856 trying to route what was effectively the same domain again. This normally
5857 caused routing to fail (because the router wouldn't handle the domain
5858 twice). Now Exim checks for this case specially, and just changes the
5859 casing of the domain that it ultimately uses when it transmits the message
5862 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5865 50. If a filter generated a file delivery with a non-absolute name (possible if
5866 no home directory exists for the router), the forbid_file option was not
5869 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5870 the existing equality matching.
5872 52. Exim was using ints instead of ino_t variables in some places where it was
5873 dealing with inode numbers.
5875 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5876 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5877 called TMPDIR, and if it finds it is different, it changes its value.
5879 54. The smtp_printf() function is now made available to local_scan() so
5880 additional output lines can be written before returning. There is also an
5881 smtp_fflush() function to enable the detection of a dropped connection.
5882 The variables smtp_input and smtp_batched_input are exported to
5885 55. Changed the default runtime configuration: the message "Unknown user"
5886 has been removed from the ACL, and instead placed on the localuser router,
5887 using the cannot_route_message feature. This means that any verification
5888 failures that generate their own messages won't get overridden. Similarly,
5889 the "Unrouteable address" message that was in the ACL for unverifiable
5890 relay addresses has also been removed.
5892 56. Added hosts_avoid_esmtp to the smtp transport.
5894 57. The exicyclog script was not checking for the esoteric option
5895 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5896 will work only if exicyclog is run under the appropriate euid.
5898 58. Following a discussion on the list, the rules by which Exim recognises line
5899 endings on incoming messages have been changed. The -dropcr and drop_cr
5900 options are now no-ops, retained only for backwards compatibility. The
5901 following line terminators are recognized: LF CRLF CR. However, special
5902 processing applies to CR:
5904 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5905 nor a local message in the state where . is a terminator.
5907 (ii) If a bare CR is encountered in a header line, an extra space is added
5908 after the line terminator so as not to end the header. The reasoning
5909 behind this is that bare CRs in header lines are most likely either
5910 to be mistakes, or people trying to play silly games.
5912 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5913 was being incorrectly given as 18 bytes larger than it should have been.
5914 This is a VOB (very old bug).
5916 60. This may never have affected anything current, but just in case it has:
5917 When the local host is found other than at the start of a list of hosts,
5918 the local host, those with the same MX, and any that follow, are discarded.
5919 When the list in question was part of a longer list of hosts, the following
5920 hosts (not currently being processed) were also being discarded. This no
5921 longer happens. I'm not sure if this situation could ever has previously
5924 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5927 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5928 header. This is something that is recommended in a new Internet Draft, and
5929 is something that is documented as being done by Sendmail. There are two
5930 possible values. For messages generated by the autoreply transport, Exim
5933 Auto-Submitted: auto-replied
5935 whereas for all other generated messages (e.g. bounces) it adds
5937 Auto-Submitted: auto-generated
5939 63. The "personal" condition in filters now includes a test for the
5940 Auto-Submitted: header. If it contains the string "auto-" the message it
5941 not considered personal.
5943 64. Added rcpt_include_affixes as a generic transport option.
5945 65. Added queue_only_override (default true).
5947 66. Added the syslog_duplication option.
5949 67. If what should have been the first header line of a message consisted of
5950 a space followed by a colon, Exim was mis-interpreting it as a header line.
5951 It isn't of course - it is syntactically invalid and should therefore be
5952 treated as the start of the message body. The misbehaviour could have
5953 caused a number of strange effects, including loss of data in subsequent
5954 header lines, and spool format errors.
5956 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5957 client host had authenticated. This control can now be exercised by an ACL
5958 for more flexibility.
5960 69. By default, callouts do not happen when testing with -bh. There is now a
5961 variant, -bhc, which does actually run the callout code, including
5962 consulting and updating the callout cache.
5964 70. Added support for saslauthd authentication, courtesy of Alexander
5967 71. If statvfs() failed on the spool or log directories while checking their
5968 size for availability, Exim confusingly gave the error "space shortage".
5969 Furthermore, in debugging mode it crashed with a floating point exception.
5970 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5971 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5972 serious problem, Exim now writes to the main and panic logs when this
5973 happens, with details of the failure. It then refuses to accept the
5974 incoming message, giving the message "spool directory problem" or "log
5975 directory problem" with a 421 code for SMTP messages.
5977 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5978 0, 1, and 2 exist, because some OS complain for execs without them (see
5979 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5980 descriptors. However, the code omitted to check that the open succeeded,
5981 causing mysterious errors if for some reason the permissions on /dev/null
5982 got screwed. Now Exim writes a message to the main and panic logs, and
5983 bombs out if it can't open /dev/null.
5985 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5986 interact so that it is all more flexible. It is supposed to remain
5987 backwards compatible. Also added extra_local_interfaces.
5989 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5990 to bomb out with an assertion failure - to the client this appears as a
5991 connection drop. This problem occurs in the part of the code that was taken
5992 from the Samba project. Fortunately, the assertion is in a very simple
5993 function, so I have fixed this by reproducing the function inline in the
5994 one place where it is called, and arranging for authentication to fail
5995 instead of killing the process with assert().
5997 75. The SPA client code was not working when the server requested OEM rather
5998 than Unicode encoding.
6000 76. Added code to make require_files with a specific uid setting more usable in
6001 the case where statting the file as root fails - usually a non-root-mounted
6002 NFS file system. When this happens and the failure is EACCES, Exim now
6003 forks a subprocess and does the per-uid checking as the relevant uid.
6005 77. Added process_log_path.
6007 78. If log_file_path was not explicitly set, a setting of check_log_space or
6008 check_log_inodes was ignored.
6010 79. If a space check for the spool or log partitions fails, the incident is now
6011 logged. Of course, in the latter case the data may get lost...
6013 80. Added the %p formatting code to string_format() so that it can be used to
6014 print addresses in debug_print(). Adjusted all the address printing in the
6015 debugging in store.c to use %p rather than %d.
6017 81. There was a concern that a line of code in smtp_in.c could overflow a
6018 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6019 initially expressed, the concern was not well-founded, because trailing
6020 spaces are removed early. However, if the trailing spaces were followed by
6021 a NULL, they did not get removed, so the overflow was possible. Two fixes
6024 (a) I re-wrote the offending code in a cleaner fashion.
6025 (b) If an incoming SMTP command contains a NULL character, it is rejected
6028 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6029 runs initgroups(), so that if the Exim user is in any additional groups,
6030 they will be used during message reception.
6036 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob