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 (applying to the forward lookup).
90 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
91 of ldap servers used for a specific lookup. Patch provided by Heiko
94 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
97 Patch submitted by Lars Timman.
103 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
105 PP/02 Make -n do something, by making it not do something.
106 When combined with -bP, the name of an option is not output.
108 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
111 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
112 $sender_host_name and config options to manage this, and basic check
115 PP/05 DSCP support for outbound connections and control modifier for inbound.
117 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
118 (Only plugin which currently uses this is kerberos4, which nobody should
119 be using, but we should make it available and other future plugins might
120 conceivably use it, even though it would break NAT; stuff *should* be
121 using channel bindings instead).
123 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
124 name; added for Sendmail compatibility; requires admin caller.
125 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
126 just ignore it); requires trusted caller.
127 Also parse but ignore: -Ac -Am -X<logfile>
130 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
132 TL/02 Add +smtp_confirmation as a default logging option.
134 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
135 Patch by Magnus Holmgren from 2007-02-20.
137 TL/04 Bugzilla 1281 - Spec typo.
138 Bugzilla 1283 - Spec typo.
139 Bugzilla 1290 - Spec grammar fixes.
141 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
143 TL/06 Add Experimental DMARC support using libopendmarc libraries.
145 TL/07 Fix an out of order global option causing a segfault. Reported to dev
146 mailing list by by Dmitry Isaikin.
148 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
150 JH/02 Support "G" suffix to numbers in ${if comparisons.
152 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
154 NM/01 Bugzilla 1197 - Spec typo
155 Bugzilla 1196 - Spec examples corrections
157 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
159 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
160 gnutls_enable_pkcs11, but renamed to more accurately indicate its
163 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
164 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
166 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
167 "acl {{name}{arg}...}", and optional args on acl condition
170 JH/05 Permit multiple router/transport headers_add/remove lines.
172 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
174 JH/07 Avoid using a waiting database for a single-message-only transport.
175 Performance patch from Paul Fisher. Bugzilla 1262.
177 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
180 JH/09 Add $headers_added variable, with content from use of ACL modifier
181 add_header (but not yet added to the message). Bugzilla 199.
183 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
184 Pulled from Bugzilla 817 by Wolfgang Breyha.
186 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
188 (nb: this is the same fix as in Exim 4.80.1)
190 JH/11 Add A= logging on delivery lines, and a client_set_id option on
193 JH/12 Add optional authenticated_sender logging to A= and a log_selector
196 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
198 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
199 advertise SMTP AUTH mechanism to us, instead of a generic
200 protocol violation error. Also, make Exim more robust to bad
201 data from the Dovecot auth socket.
203 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
205 When a queue runner is handling a message, Exim first routes the
206 recipient addresses, during which it prunes them based on the retry
207 hints database. After that it attempts to deliver the message to
208 any remaining recipients. It then updates the hints database using
211 So if a recipient address works intermittently, it can get repeatedly
212 deferred at routing time. The retry hints record remains fresh so the
213 address never reaches the final cutoff time.
215 This is a fairly common occurrence when a user is bumping up against
216 their storage quota. Exim had some logic in its local delivery code
217 to deal with this. However it did not apply to per-recipient defers
218 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
220 This change adds a proper retry rule check during routing so that the
221 final cutoff time is checked against the message's age. We only do
222 this check if there is an address retry record and there is not a
223 domain retry record; this implies that previous attempts to handle
224 the address had the retry_use_local_parts option turned on. We use
225 this as an approximation for the destination being like a local
226 delivery, as in LMTP.
228 I suspect this new check makes the old local delivery cutoff check
229 redundant, but I have not verified this so I left the code in place.
231 TF/02 Correct gecos expansion when From: is a prefix of the username.
233 Test 0254 submits a message to Exim with the header
237 When I ran the test suite under the user fanf2, Exim expanded
238 the header to contain my full name, whereas it should have added
239 a Resent-Sender: header. It erroneously treats any prefix of the
240 username as equal to the username.
242 This change corrects that bug.
244 GF/01 DCC debug and logging tidyup
245 Error conditions log to paniclog rather than rejectlog.
246 Debug lines prefixed by "DCC: " to remove any ambiguity.
248 TF/03 Avoid unnecessary rebuilds of lookup-related code.
250 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
251 Bug spotted by Jeremy Harris; was flawed since initial commit.
252 Would have resulted in OCSP responses post-SNI triggering an Exim
253 NULL dereference and crash.
255 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
257 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
258 Bug detection, analysis and fix by Samuel Thibault.
259 Bugzilla 1331, Debian bug #698092.
261 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
263 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
264 Server implementation by Todd Lyons, client by JH.
265 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
266 config variable "prdr_enable" controls whether the server
267 advertises the facility. If the client requests PRDR a new
268 acl_data_smtp_prdr ACL is called once for each recipient, after
269 the body content is received and before the acl_smtp_data ACL.
270 The client is controlled by bolth of: a hosts_try_prdr option
271 on the smtp transport, and the server advertisement.
272 Default client logging of deliveries and rejections involving
273 PRDR are flagged with the string "PRDR".
275 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
276 fclose(). Diagnosis by Todd Lyons.
278 PP/17 Update configure.default to handle IPv6 localhost better.
279 Patch by Alain Williams (plus minor tweaks).
282 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
283 This is now consistent with GnuTLS, and is now documented: the
284 previous undocumented portable approach to treating the option as
285 unset was to force an expansion failure. That still works, and
286 an empty string is now equivalent.
288 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
289 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
290 not performing validation itself.
292 PP/20 Added force_command boolean option to pipe transport.
293 Patch from Nick Koston, of cPanel Inc.
295 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
298 TF/04 Added udpsend ACL modifer and hexquote expansion operator
300 PP/21 Fix eximon continuous updating with timestamped log-files.
301 Broken in a format-string cleanup in 4.80, missed when I repaired the
302 other false fix of the same issue.
303 Report and fix from Heiko Schlichting.
306 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
307 Report from Prashanth Katuri.
309 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
310 It's SecureTransport, so affects any MacOS clients which use the
311 system-integrated TLS libraries, including email clients.
313 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
314 using a MIME ACL for non-SMTP local injection.
315 Report and assistance in diagnosis by Warren Baker.
317 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
319 JH/16 Fix comparisons for 64b. Bugzilla 1385.
321 TL/09 Add expansion variable $authenticated_fail_id to keep track of
322 last id that failed so it may be referenced in subsequent ACL's.
324 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
327 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
328 ldap library initialization, allowing self-signed CA's to be
329 used. Also properly sets require_cert option later in code by
330 using NULL (global ldap config) instead of ldap handle (per
331 session). Bug diagnosis and testing by alxgomz.
333 TL/12 Enhanced documentation in the ratelimit.pl script provided in
334 the src/util/ subdirectory.
336 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
337 renamed to Transport Post Delivery Action by Jeremy Harris, as
340 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
341 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
342 redis_servers = needs to be configured which will be used by the redis
343 lookup. Patch from Warren Baker, of The Packet Hub.
345 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
347 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
348 hostname or reverse DNS when processing a host list. Used suggestions
349 from multiple comments on this bug.
351 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
353 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
354 Missed a few lines, added it to make the runtest require no keyboard
357 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
358 contains upper case chars. Make router use caseful_local_part.
360 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
361 support when GnuTLS has been built with p11-kit.
367 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
369 This, or similar/improved, will also be change PP/11 of 4.82.
375 PP/01 Handle short writes when writing local log-files.
376 In practice, only affects FreeBSD (8 onwards).
377 Bugzilla 1053, with thanks to Dmitry Isaikin.
379 NM/01 Bugzilla 949 - Documentation tweak
381 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
384 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
386 PP/02 Implemented gsasl authenticator.
388 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
390 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
391 `pkg-config foo` for cflags/libs.
393 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
394 with rest of GSASL and with heimdal_gssapi.
396 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
397 `pkg-config foo` for cflags/libs for the TLS implementation.
399 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
400 properties get this fed in as external SSF. A number of robustness
401 and debugging improvements to the cyrus_sasl authenticator.
403 PP/08 cyrus_sasl server now expands the server_realm option.
405 PP/09 Bugzilla 1214 - Log authentication information in reject log.
406 Patch by Jeremy Harris.
408 PP/10 Added dbmjz lookup type.
410 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
412 PP/12 MAIL args handles TAB as well as SP, for better interop with
413 non-compliant senders.
414 Analysis and variant patch by Todd Lyons.
416 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
417 Bug report from Lars Müller <lars@samba.org> (via SUSE),
418 Patch from Dirk Mueller <dmueller@suse.com>
420 PP/13 tls_peerdn now print-escaped for spool files.
421 Observed some $tls_peerdn in wild which contained \n, which resulted
422 in spool file corruption.
424 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
425 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
426 or write after TLS renegotiation, which otherwise led to messages
429 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
430 as a tracking header (ie: a signed header comes before the signature).
431 Patch from Wolfgang Breyha.
433 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
434 comma-sep list; embedded commas doubled.
436 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
438 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
440 Report and patch from Dmitry Banschikov.
442 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
443 Removed SSL_clear() after SSL_new() which led to protocol negotiation
444 failures. We appear to now support TLS1.1+ with Exim.
446 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
447 lets Exim select keys and certificates based upon TLS SNI from client.
448 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
449 before an outbound SMTP session. New log_selector, +tls_sni.
451 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
452 NULL dereference. Report and patch from Alun Jones.
454 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
455 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
456 Not seeing resolver debug output on NetBSD, but suspect this is a
457 resolver implementation change.
459 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
460 Left warnings. Added "eximon gdb" invocation mode.
462 PP/21 Defaulting "accept_8bitmime" to true, not false.
464 PP/22 Added -bw for inetd wait mode support.
466 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
467 locate the relevant includes and libraries. Made this the default.
469 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
470 Bugzilla 1246, report and most of solution from Tomasz Kusy.
472 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
473 This may cause build issues on older platforms.
475 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
476 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
477 gnutls_require_mac & gnutls_require_protocols (no longer supported).
478 Added SNI support via GnuTLS too.
479 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
481 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
483 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
484 Applied second patch from Janne, implementing suggestion to default
485 multiple-strings-in-record handling to match SPF spec.
487 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
489 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
490 read-only, out of scope).
491 Patch from Wolfgang Breyha, report from Stuart Northfield.
493 PP/29 Fix three issues highlighted by clang analyser static analysis.
494 Only crash-plausible issue would require the Cambridge-specific
495 iplookup router and a misconfiguration.
496 Report from Marcin Mirosław.
498 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
500 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
501 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
502 As part of this, removing so much warning spew let me fix some minor
503 real issues in debug logging.
505 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
506 assignment on my part. Fixed.
508 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
509 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
510 Janne Snabb (who went above and beyond: thank you).
512 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
513 string otherwise requires a connection and a bunch more work and it's
514 relatively easy to get wrong. Should also expose TLS library linkage
517 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
518 64-bit ${eval} (JH/03).
520 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
521 GNU libc to support some of the 64-bit stuff, should not lead to
522 conflicts. Defined before os.h is pulled in, so if a given platform
523 needs to override this, it can.
525 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
526 protection layer was required, which is not implemented.
527 Bugzilla 1254, patch from Wolfgang Breyha.
529 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
530 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
531 tls_dhparam take prime identifiers. Also unbreak combination of
532 OpenSSL+DH_params+TLSSNI.
534 PP/39 Disable SSLv2 by default in OpenSSL support.
540 PP/01 Solaris build fix for Oracle's LDAP libraries.
541 Bugzilla 1109, patch from Stephen Usher.
543 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
545 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
548 TF/02 Fix a couple more cases where we did not log the error message
549 when unlink() failed. See also change 4.74-TF/03.
551 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
552 lock up or crash if it happened to be inside a call to libc when it
553 got a SIGUSR1 from exiwhat.
555 The SIGUSR1 handler appends the current process status to the process
556 log which is later printed by exiwhat. It used to use the general
557 purpose logging code to do this, but several functions it calls are
558 not safe for signals.
560 The new output code in the SIGUSR1 handler is specific to the process
561 log, and simple enough that it's easy to inspect for signal safety.
562 Removing some special cases also simplifies the general logging code.
563 Removing the spurious timestamps from the process log simplifies
566 TF/04 Improved ratelimit ACL condition.
568 The /noupdate option has been deprecated in favour of /readonly which
569 has clearer semantics. The /leaky, /strict, and /readonly update modes
570 are mutually exclusive. The update mode is no longer included in the
571 database key; it just determines when the database is updated. (This
572 means that when you upgrde Exim will forget old rate measurements.)
574 Exim now checks that the per_* options are used with an update mode that
575 makes sense for the current ACL. For example, when Exim is processing a
576 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
577 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
578 must specify per_mail/readonly. If you omit the update mode it defaults to
579 /leaky where that makes sense (as before) or /readonly where required.
581 The /noupdate option is now undocumented but still supported for
582 backwards compatibility. It is equivalent to /readonly except that in
583 ACLs where /readonly is required you may specify /leaky/noupdate or
584 /strict/noupdate which are treated the same as /readonly.
586 A useful new feature is the /count= option. This is a generalization
587 of the per_byte option, so that you can measure the throughput of other
588 aggregate values. For example, the per_byte option is now equivalent
589 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
591 The per_rcpt option has been generalized using the /count= mechanism
592 (though it's more complicated than the per_byte equivalence). When it is
593 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
594 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
595 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
596 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
597 Note that using per_rcpt with a non-readonly update mode in more than
598 one ACL will cause the recipients to be double-counted. (The per_mail
599 and per_byte options don't have this problem.)
601 The handling of very low rates has changed slightly. If the computed rate
602 is less than the event's count (usually one) then this event is the first
603 after a long gap. In this case the rate is set to the same as this event's
604 count, so that the first message of a spam run is counted properly.
606 The major new feature is a mechanism for counting the rate of unique
607 events. The new per_addr option counts the number of different
608 recipients that someone has sent messages to in the last time period. It
609 behaves like per_rcpt if all the recipient addresses are different, but
610 duplicate recipient addresses do not increase the measured rate. Like
611 the /count= option this is a general mechanism, so the per_addr option
612 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
613 example, measure the rate that a client uses different sender addresses
614 with the options per_mail/unique=$sender_address. There are further
615 details in the main documentation.
617 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
619 TF/06 Removed a few PCRE remnants.
621 TF/07 Automatically extract Exim's version number from tags in the git
622 repository when doing development or release builds.
624 PP/02 Raise smtp_cmd_buffer_size to 16kB.
625 Bugzilla 879. Patch from Paul Fisher.
627 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
628 Heavily based on revision 40f9a89a from Simon Arlott's tree.
631 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
633 PP/05 Variable $av_failed, true if the AV scanner deferred.
634 Bugzilla 1078. Patch from John Horne.
636 PP/06 Stop make process more reliably on build failure.
637 Bugzilla 1087. Patch from Heiko Schlittermann.
639 PP/07 Make maildir_use_size_file an _expandable_ boolean.
640 Bugzilla 1089. Patch from Heiko Schlittermann.
642 PP/08 Handle ${run} returning more data than OS pipe buffer size.
643 Bugzilla 1131. Patch from Holger Weiß.
645 PP/09 Handle IPv6 addresses with SPF.
646 Bugzilla 860. Patch from Wolfgang Breyha.
648 PP/10 GnuTLS: support TLS 1.2 & 1.1.
650 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
653 PP/11 match_* no longer expand right-hand-side by default.
654 New compile-time build option, EXPAND_LISTMATCH_RHS.
655 New expansion conditions, "inlist", "inlisti".
657 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
659 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
661 PP/14 fix log_write() format string regression from TF/03.
662 Bugzilla 1152. Patch from Dmitry Isaikin.
668 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
670 PP/02 Harmonised TLS library version reporting; only show if debugging.
671 Layout now matches that introduced for other libraries in 4.74 PP/03.
673 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
675 PP/04 New "dns_use_edns0" global option.
677 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
680 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
681 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
683 TK/01 Updated PolarSSL code to 0.14.2.
684 Bugzilla 1097. Patch from Andreas Metzler.
686 PP/07 Catch divide-by-zero in ${eval:...}.
689 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
692 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
693 format-string attack -- SECURITY: remote arbitrary code execution.
695 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
696 time unintentionally subject to list matching rules, letting the header
697 cause arbitrary Exim lookups (of items which can occur in lists, *not*
698 arbitrary string expansion). This allowed for information disclosure.
700 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
701 INT_MIN/-1 -- value coerced to INT_MAX.
707 NM/01 Workround for PCRE version dependency in version reporting
710 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
711 This fixes portability to compilers other than gcc, notably
712 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
714 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
715 makefiles for portability to HP-UX and POSIX correctness.
717 PP/01 Permit LOOKUP_foo enabling on the make command-line.
718 Also via indented variable definition in the Makefile.
719 (Debugging by Oliver Heesakkers).
721 PP/02 Restore caching of spamd results with expanded spamd_address.
722 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
724 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
725 Improves build reliability. Fix from: Frank Elsner
727 NM/02 Fix wide character breakage in the rfc2047 coding
728 Fixes bug 1064. Patch from Andrey N. Oktyabrski
730 NM/03 Allow underscore in dnslist lookups
731 Fixes bug 1026. Patch from Graeme Fowler
733 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
734 Code patches from Adam Ciarcinski of NetBSD.
736 NM/04 Fixed exiqgrep to cope with mailq missing size issue
739 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
740 is logged, to avoid truncation. Patch from John Horne.
742 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
743 Patch from Jakob Hirsch.
745 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
746 SQL string expansion failure details.
747 Patch from Andrey Oktyabrski.
749 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
750 Patch from Simon Arlott.
752 PP/09 New lookups functionality failed to compile on old gcc which rejects
753 extern declarations in function scope.
754 Patch from Oliver Fleischmann
756 PP/10 Use sig_atomic_t for flags set from signal handlers.
757 Check getgroups() return and improve debugging.
758 Fixed developed for diagnosis in bug 927 (which turned out to be
761 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
762 Patch from Mark Zealey.
764 PP/12 Bugzilla 1056: Improved spamd server selection.
765 Patch from Mark Zealey.
767 PP/13 Bugzilla 1086: Deal with maildir quota file races.
768 Based on patch from Heiko Schlittermann.
770 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
771 Patch from Uwe Doering, sign-off by Michael Haardt.
773 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
774 variable declaration deep within a block. Bug and patch from
777 PP/15 lookups-Makefile IRIX compatibilty coercion.
779 PP/16 Make DISABLE_DKIM build knob functional.
781 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
782 Patch by Simon Arlott
784 TF/03 Fix valgrind.h portability to C89 compilers that do not support
785 variable argument macros. Our copy now differs from upstream.
791 TF/01 Failure to get a lock on a hints database can have serious
792 consequences so log it to the panic log.
794 TF/02 Log LMTP confirmation messages in the same way as SMTP,
795 controlled using the smtp_confirmation log selector.
797 TF/03 Include the error message when we fail to unlink a spool file.
799 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
800 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
801 for maintaining out-of-tree patches for some time.
803 PP/01 Bugzilla 139: Documentation and portability issues.
804 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
805 Handle per-OS dynamic-module compilation flags.
807 PP/02 Let /dev/null have normal permissions.
808 The 4.73 fixes were a little too stringent and complained about the
809 permissions on /dev/null. Exempt it from some checks.
810 Reported by Andreas M. Kirchwitz.
812 PP/03 Report version information for many libraries, including
813 Exim version information for dynamically loaded libraries. Created
814 version.h, now support a version extension string for distributors
815 who patch heavily. Dynamic module ABI change.
817 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
818 privilege escalation vulnerability whereby the Exim run-time user
819 can cause root to append content of the attacker's choosing to
822 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
825 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
826 If dropping privileges for untrusted macros, we disabled normal logging
827 on the basis that it would fail; for the Exim run-time user, this is not
828 the case, and it resulted in successful deliveries going unlogged.
829 Fixed. Reported by Andreas Metzler.
835 PP/01 Date: & Message-Id: revert to normally being appended to a message,
836 only prepend for the Resent-* case. Fixes regression introduced in
837 Exim 4.70 by NM/22 for Bugzilla 607.
839 PP/02 Include check_rfc2047_length in configure.default because we're seeing
840 increasing numbers of administrators be bitten by this.
842 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
844 PP/03 Bugzilla 994: added openssl_options main configuration option.
846 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
848 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
850 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
852 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
853 without a peer certificate, leading to a segfault because of an
854 assumption that peers always have certificates. Be a little more
855 paranoid. Problem reported by Martin Tscholak.
857 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
858 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
859 NB: ClamAV planning to remove STREAM in "middle of 2010".
860 CL also introduces -bmalware, various -d+acl logging additions and
861 more caution in buffer sizes.
863 PP/09 Implemented reverse_ip expansion operator.
865 PP/10 Bugzilla 937: provide a "debug" ACL control.
867 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
869 PP/12 Bugzilla 973: Implement --version.
871 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
873 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
875 PP/15 Bugzilla 816: support multiple condition rules on Routers.
877 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
878 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
879 ignore trailing whitespace.
881 JJ/02 prevent non-panic DKIM error from being sent to paniclog
883 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
886 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
887 Notification from Dr Andrew Aitchison.
889 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
890 ExtendedDetectionInfo response format.
891 Notification from John Horne.
893 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
896 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
897 XSL and documented dependency on system catalogs, with examples of how
900 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
903 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
904 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
905 configuration file which is writeable by the Exim user or group.
907 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
908 of configuration files to cover files specified with the -C option if
909 they are going to be used with root privileges, not just the default
912 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
913 option (effectively making it always true).
915 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
916 files to be used while preserving root privileges.
918 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
919 that rogue child processes cannot use them.
921 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
922 run-time user, instead of root.
924 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
925 Exim run-time user without dropping privileges.
927 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
928 result string, instead of calling string_vformat() twice with the same
931 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
932 for other users. Others should always drop root privileges if they use
933 -C on the command line, even for a whitelisted configure file.
935 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
937 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
943 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
944 $data_path, and $header_path variables; fixed documentation bugs and
947 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
948 exipick to access non-standard spools, including the "frozen" queue
951 NM/01 Bugzilla 965: Support mysql stored procedures.
952 Patch from Alain Williams
954 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
956 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
957 Patch from Andreas Metzler
959 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
960 Patch from Kirill Miazine
962 NM/05 Bugzilla 671: Added umask to procmail example.
964 JJ/03 installed exipick 20100323.0, fixing doc bug
966 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
967 directory. Notification and patch from Dan Rosenberg.
969 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
971 TK/02 Improve log output when DKIM signing operation fails.
973 MH/01 Treat the transport option dkim_domain as a colon separated
974 list, not as a single string, and sign the message with each element,
975 omitting multiple occurences of the same signer.
977 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
978 Bugzilla 985, 986. Patch by Simon Arlott
980 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
981 Patch by Simon Arlott
983 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
984 MBX locking. Notification from Dan Rosenberg.
990 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
992 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
994 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
996 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
998 NM/04 Bugzilla 915: Flag broken perl installation during build.
1004 TK/01 Added patch by Johannes Berg that expands the main option
1005 "spamd_address" if it starts with a dollar sign.
1007 TK/02 Write list of recipients to X-Envelope-Sender header when building
1008 the mbox-format spool file for content scanning (suggested by Jakob
1011 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1012 (http://www.dcc-servers.net/) support via dccifd. Activated by
1013 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1015 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1016 by Mark Daniel Reidel <mr@df.eu>.
1018 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1019 When building exim an external PCRE library is now needed -
1020 PCRE is a system library on the majority of modern systems.
1021 See entry on PCRE_LIBS in EDITME file.
1023 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1024 conversation. Added nologin parameter to request.
1025 Patch contributed by Kirill Miazine.
1027 TF/01 Do not log submission mode rewrites if they do not change the address.
1029 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1031 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1032 log files in place. Contributed by Roberto Lima.
1034 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1036 TF/03 Bugzilla 615: When checking the local_parts router precondition
1037 after a local_part_suffix or local_part_prefix option, Exim now
1038 does not use the address's named list lookup cache, since this
1039 contains cached lookups for the whole local part.
1041 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1042 Robert Millan. Documentation is in experimental-spec.txt.
1044 TF/04 Bugzilla 668: Fix parallel build (make -j).
1046 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1048 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1049 Patch provided by Jan Srzednicki.
1051 TF/05 Leading white space used to be stripped from $spam_report which
1052 wrecked the formatting. Now it is preserved.
1054 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1055 that they are available at delivery time.
1057 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1059 TF/08 TLS error reporting now respects the incoming_interface and
1060 incoming_port log selectors.
1062 TF/09 Produce a more useful error message if an SMTP transport's hosts
1063 setting expands to an empty string.
1065 NM/06 Bugzilla 744: EXPN did not work under TLS.
1066 Patch provided by Phil Pennock.
1068 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1069 Patch provided by Richard Godbee.
1071 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1072 acl_smtp_notquit, added index entry.
1074 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1075 Patch provided by Eugene Bujak.
1077 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1078 accept(). Patch provided by Maxim Dounin.
1080 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1081 Patch provided by Phil Pennock.
1083 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1085 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1086 Patch provided by Brad "anomie" Jorsch.
1088 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1089 Patch provided by Dean Brooks.
1091 TK/05 Add native DKIM support (does not depend on external libraries).
1093 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1094 Patch provided by Graeme Fowler.
1096 NM/16 Bugzilla 851: Documentation example syntax fix.
1098 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1100 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1103 NM/19 Bugzilla 745: TLS version reporting.
1104 Patch provided by Phil Pennock.
1106 NM/20 Bugzilla 167: bool: condition support.
1107 Patch provided by Phil Pennock.
1109 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1110 clients. Patch provided by Phil Pennock.
1112 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1113 Patch provided by Brad "anomie" Jorsch.
1115 NM/23 Bugzilla 687: Fix misparses in eximstats.
1116 Patch provided by Heiko Schlittermann.
1118 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1119 Patch provided by Heiko Schlittermann.
1121 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1122 plus update to original patch.
1124 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1126 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1127 Patch provided by David Brownlee.
1129 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1131 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1133 NM/30 Bugzilla 888: TLS documentation bugfixes.
1135 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1137 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1138 Unlike the original bugzilla I have changed all shell scripts in src tree.
1140 NM/33 Bugzilla 898: Transport filter timeout fix.
1141 Patch by Todd Rinaldo.
1143 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1144 Patch by Serge Demonchaux.
1146 NM/35 Bugzilla 39: Base64 decode bug fixes.
1147 Patch by Jakob Hirsch.
1149 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1151 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1153 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1155 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1161 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1162 ALT-N's libdkim that I have put here:
1163 http://duncanthrax.net/exim-experimental/
1165 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1166 were called 'true' and it seems that C99 defines that as a reserved
1167 keyword to be used with 'bool' variable types. That means you could
1168 not include C99-style headers which use bools without triggering
1169 build errors in sieve.c.
1171 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1172 as mailq or other aliases. Changed the --help handling significantly
1173 to do whats expected. exim_usage() emits usage/help information.
1175 SC/01 Added the -bylocaldomain option to eximstats.
1177 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1179 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1181 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1187 PH/01 Another patch from the Sieve maintainer.
1189 PH/02 When an IPv6 address is converted to a string for single-key lookup
1190 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1191 dots are used instead of colons so that keys in lsearch files need not
1192 contain colons. This was done some time before quoting was made available
1193 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1194 (notated using the quote facility) so as to distinguish them from IPv4
1195 keys. This meant that lookups for IP addresses in host lists did not work
1196 for iplsearch lookups.
1198 This has been fixed by arranging for IPv6 addresses to be expressed with
1199 colons if the lookup type is iplsearch. This is not incompatible, because
1200 previously such lookups could never work.
1202 The situation is now rather anomolous, since one *can* have colons in
1203 ordinary lsearch keys. However, making the change in all cases is
1204 incompatible and would probably break a number of configurations.
1206 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1209 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1210 conversion specification without a maximum field width, thereby enabling
1211 a rogue spamd server to cause a buffer overflow. While nobody in their
1212 right mind would setup Exim to query an untrusted spamd server, an
1213 attacker that gains access to a server running spamd could potentially
1214 exploit this vulnerability to run arbitrary code as the Exim user.
1216 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1217 $primary_hostname instead of what libspf2 thinks the hosts name is.
1219 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1220 a directory entry by the name of the lookup key. Previously, if a
1221 symlink pointed to a non-existing file or a file in a directory that
1222 Exim lacked permissions to read, a lookup for a key matching that
1223 symlink would fail. Now it is enough that a matching directory entry
1224 exists, symlink or not. (Bugzilla 503.)
1226 PH/03 The body_linecount and body_zerocount variables are now exported in the
1229 PH/04 Added the $dnslist_matched variable.
1231 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1232 This means they are set thereafter only if the connection becomes
1235 PH/06 Added the client_condition to authenticators so that some can be skipped
1236 by clients under certain conditions.
1238 PH/07 The error message for a badly-placed control=no_multiline_responses left
1239 "_responses" off the end of the name.
1241 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1243 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1244 (without spaces) instead of just copying the configuration text.
1246 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1248 PH/11 Added $max_received_linelength.
1250 PH/12 Added +ignore_defer and +include_defer to host lists.
1252 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1253 way in which PCRE > 7.0 is built.
1255 PH/14 Implemented queue_only_load_latch.
1257 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1258 MAIL command. The effect was to mangle the value on 64-bit systems.
1260 PH/16 Another patch from the Sieve maintainer.
1262 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1264 PH/18 If a system quota error occurred while trying to create the file for
1265 a maildir delivery, the message "Mailbox is full" was not appended to the
1266 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1267 only to a quota excession during the actual writing of the file.
1269 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1270 characters?) which causes problems in log lines. The DN values are now
1271 passed through string_printing() before being added to log lines.
1273 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1274 and InterBase are left for another time.)
1276 PH/21 Added message_body_newlines option.
1278 PH/22 Guard against possible overflow in moan_check_errorcopy().
1280 PH/23 POSIX allows open() to be a macro; guard against that.
1282 PH/24 If the recipient of an error message contained an @ in the local part
1283 (suitably quoted, of course), incorrect values were put in $domain and
1284 $local_part during the evaluation of errors_copy.
1290 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1291 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1294 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1295 issue a MAIL command.
1297 PH/02 In an ACL statement such as
1299 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1301 if a client was not listed at all, or was listed with a value other than
1302 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1303 the condition was not true (as it should be), so access was not denied.
1304 The bug was that the ! inversion was incorrectly passed on to the second
1305 item. This has been fixed.
1307 PH/03 Added additional dnslists conditions == and =& which are different from
1308 = and & when the dns lookup returns more than one IP address.
1310 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1311 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1313 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1314 FSYNC, which compiles an option called disable_fsync that allows for
1315 bypassing fsync(). The documentation is heavily laced with warnings.
1317 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1319 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1320 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1321 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1322 including adding "make clean"; (3) Added -fPIC when compiling the test
1323 dynamically loaded module, to get rid of a warning.
1325 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1326 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1327 = 0s. The bug is otherwise harmless.
1329 PH/07 There was a bug in the dovecot authenticator such that the value of
1330 $auth1 could be overwritten, and so not correctly preserved, after a
1331 successful authentication. This usually meant that the value preserved by
1332 the server_setid option was incorrect.
1334 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1336 PH/09 Installed PCRE release 7.0.
1338 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1339 run for batched SMTP input. It is now run at the start of every message
1340 in the batch. While fixing this I discovered that the process information
1341 (output by running exiwhat) was not always getting set for -bs and -bS
1342 input. This is fixed, and it now also says "batched" for BSMTP.
1344 PH/11 Added control=no_pipelining.
1346 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1347 patch, slightly modified), and move the expansion of helo_data till after
1348 the connection is made in the smtp transport (so it can use these
1351 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1353 PH/14 Added log_selector = +pid.
1355 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1357 PH/16 Add ${if forany and ${if forall.
1359 PH/17 Added dsn_from option to vary the From: line in DSNs.
1361 PH/18 Flush SMTP output before performing a callout, unless control =
1362 no_callout_flush is set.
1364 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1365 was true (the default) a successful delivery failed to delete the retry
1366 item, thus causing premature timeout of the address. The bug is now
1369 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1371 PH/21 Long custom messages for fakedefer and fakereject are now split up
1372 into multiline reponses in the same way that messages for "deny" and
1373 other ACL rejections are.
1375 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1376 with slight modification.
1378 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1379 draft, changing the syntax and factoring some duplicate code".
1381 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1382 for deliveries of the second and subsequent messages over the same SMTP
1385 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1386 ${reduce, with only minor "tidies".
1388 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1390 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1391 expansion side effects.
1393 PH/27 When a message times out after an over-quota error from an Exim-imposed
1394 quota, the bounce message says "mailbox is full". This message was not
1395 being given when it was a system quota that was exceeded. It now should
1398 MH/03 Made $recipients available in local_scan(). local_scan() already has
1399 better access to the recipient list through recipients_list[], but
1400 $recipients can be useful in postmaster-provided expansion strings.
1402 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1403 in the case of a MAIL command with additional options following the
1404 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1405 were accidentally chopped off.
1407 PH/29 SMTP synchronization checks are implemented when a command is read -
1408 there is a check that no more input is waiting when there shouldn't be
1409 any. However, for some commands, a delay in an ACL can mean that it is
1410 some time before the response is written. In this time, more input might
1411 arrive, invalidly. So now there are extra checks after an ACL has run for
1412 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1413 pipelining has not been advertised.
1415 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1417 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1418 returned in the SMTP response when smtp_return_error_details was false.
1419 This has been fixed.
1421 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1422 instead of the C I/O that was originally supplied, because problems were
1423 reported on Solaris.
1425 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1426 Exim which did not show up earlier: it was assuming that a call to
1427 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1428 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1429 was a macro that became an assignment, so it seemed to work. This has
1430 changed to a proper function call with a void return, hence the compile
1431 error. Exim's code has been fixed.
1433 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1436 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1438 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1440 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1441 because it turns out to be too big or there is a timeout) there is no
1442 "Completed" line in the log. When some messages of this type were
1443 selected by exigrep, they were listed as "not completed". Others were
1444 picked up by some special patterns. I have improved the selection
1445 criteria to be more general.
1447 PH/38 The host_find_failed option in the manualroute router can now be set
1448 to "ignore", to completely ignore a host whose IP address cannot be
1449 found. If all hosts are ignored, the behaviour is controlled by the new
1450 host_all_ignored option.
1452 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1453 homing or because of multiple MX records with /mx) generated more than
1454 one IP address, and the following item turned out to be the local host,
1455 all the secondary addresses of the first item were incorrectly removed
1456 from the list, along with the local host and any following hosts (which
1457 is what is supposed to happen).
1459 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1460 whoever called Exim into the -H file. In the case of the daemon it was
1461 behaving confusingly. When first started, it used values for whoever
1462 started the daemon, but after a SIGHUP it used the Exim user (because it
1463 calls itself on a restart). I have changed the code so that it now always
1466 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1467 message are rejected with the same error (e.g. no authentication or bad
1468 sender address), and a DATA command is nevertheless sent (as can happen
1469 with PIPELINING or a stupid MUA), the error message that was given to the
1470 RCPT commands is included in the rejection of the DATA command. This is
1471 intended to be helpful for MUAs that show only the final error to their
1474 PH/42 Another patch from the Sieve maintainer.
1476 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1477 Eximstats - Fixed some broken HTML links and added missing column headers
1479 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1480 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1482 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1484 PH/43 Yet another patch from the Sieve maintainer.
1486 PH/44 I found a way to check for a TCP/IP connection going away before sending
1487 the response to the final '.' that terminates a message, but only in the
1488 case where the client has not sent further data following the '.'
1489 (unfortunately, this is allowed). However, in many cases there won't be
1490 any further data because there won't be any more messages to send. A call
1491 to select() can be used: if it shows that the input is "ready", there is
1492 either input waiting, or the socket has been closed. An attempt to read
1493 the next input character can distinguish the two cases. Previously, Exim
1494 would have sent an OK response which the client would never have see.
1495 This could lead to message repetition. This fix should cure that, at
1496 least in a lot of common cases.
1498 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1499 advertised in response to EHLO.
1505 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1506 fixed by 4.65/MH/01 (is this a record?) are fixed:
1508 (i) An empty string was always treated as zero by the numeric comparison
1509 operators. This behaviour has been restored.
1511 (ii) It is documented that the numeric comparison operators always treat
1512 their arguments as decimal numbers. This was broken in that numbers
1513 starting with 0 were being interpreted as octal.
1515 While fixing these problems I realized that there was another issue that
1516 hadn't been noticed. Values of message_size_limit (both the global option
1517 and the transport option) were treated as octal if they started with 0.
1518 The documentation was vague. These values are now always treated as
1519 decimal, and I will make that clear in the documentation.
1525 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1526 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1529 MH/01 Don't check that the operands of numeric comparison operators are
1530 integers when their expansion is in "skipping" mode (fixes bug
1531 introduced by 4.64-PH/07).
1533 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1534 child addresses, Exim now panics and dies. Previously, because the count
1535 is held in a short int, deliveries were likely to be lost. As such a
1536 large number of recipients for a single message is ridiculous
1537 (performance will be very, very poor), I have chosen to impose a limit
1538 rather than extend the field.
1544 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1545 leftover -K file (the existence of which was triggered by #402).
1546 While we were at it, introduced process PID as part of the -K
1547 filename. This should rule out race conditions when creating
1550 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1551 processing considerably. Previous code took too long for large mails,
1552 triggering a timeout which in turn triggers #401.
1554 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1555 in the DK code in transports.c. sendfile() is not really portable,
1556 hence the _LINUX specificness.
1558 TF/01 In the add_headers option to the mail command in an Exim filter,
1559 there was a bug that Exim would claim a syntax error in any
1560 header after the first one which had an odd number of characters
1563 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1564 callout verification, Exim cached a "reject" for the entire domain. This
1565 is correct for most verifications, but it is not correct for a recipient
1566 verification with use_sender or use_postmaster set, because in that case
1567 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1568 case of MAIL FROM:<> rejection from other early rejections (e.g.
1569 rejection of HELO). When verifying a recipient using a non-null MAIL
1570 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1571 Whatever the result of the callout, the value of the domain cache is
1572 left unchanged (for any other kind of callout, getting as far as trying
1573 RCPT means that the domain itself is ok).
1575 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1578 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1579 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1580 session, the daemon ignores ECONNECT errors and logs others; it now
1581 ignores EPIPE as well.
1583 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1584 (quoted-printable decoding).
1586 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1587 later the small subsequent patch to fix an introduced bug.
1589 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1591 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1593 PH/08 An error is now given if message_size_limit is specified negative.
1595 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1596 to be given (somewhat) arbitrary names.
1598 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1601 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1602 miscellaneous code fixes
1604 PH/10 Added the log_reject_target ACL modifier to specify where to log
1607 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1608 hostname. This is wrong, because it relates to the incoming message (and
1609 probably the interface on which it is arriving) and not to the outgoing
1610 callout (which could be using a different interface). This has been
1611 changed to use the value of the helo_data option from the smtp transport
1612 instead - this is what is used when a message is actually being sent. If
1613 there is no remote transport (possible with a router that sets up host
1614 addresses), $smtp_active_hostname is used.
1616 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1617 tweaks were necessary in order to get it to work (see also 21 below):
1618 (a) The code assumed that strncpy() returns a negative number on buffer
1619 overflow, which isn't the case. Replaced with Exim's string_format()
1621 (b) There were several signed/unsigned issues. I just did the minimum
1622 hacking in of casts. There is scope for a larger refactoring.
1623 (c) The code used strcasecmp() which is not a standard C function.
1624 Replaced with Exim's strcmpic() function.
1625 (d) The code set only $1; it now sets $auth1 as well.
1626 (e) A simple test gave the error "authentication client didn't specify
1627 service in request". It would seem that Dovecot has changed its
1628 interface. Fortunately there's a specification; I followed it and
1629 changed what the client sends and it appears to be working now.
1631 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1634 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1635 address A is aliased to B and C, where B exists and C does not. Without
1636 -v the output is "A verified" because verification stops after a
1637 successful redirection if more than one address is generated. However,
1638 with -v the child addresses are also verified. Exim was outputting "A
1639 failed to verify" and then showing the successful verification for C,
1640 with its parentage. It now outputs "B failed to verify", showing B's
1641 parentage before showing the successful verification of C.
1643 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1644 look up a TXT record in a specific list after matching in a combined
1647 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1648 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1649 they consult the DNS. I had assumed they would set it the way they
1650 wanted; and indeed my experiments on Linux seem to show that in some
1651 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1652 To be on the safe side, however, I have now made the interface to
1653 host_find_byname() similar to host_find_bydns(), with an argument
1654 containing the DNS resolver options. The host_find_byname() function now
1655 sets these options at its start, just as host_find_bydns() does. The smtp
1656 transport options dns_qualify_single and dns_search_parents are passed to
1657 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1658 of host_find_byname() use the default settings of RES_DEFNAMES
1659 (qualify_single) but not RES_DNSRCH (search_parents).
1661 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1662 spool_read_header() do less string testing, by means of a preliminary
1663 switch on the second character of optional "-foo" lines. (This is
1664 overdue, caused by the large number of possibilities that now exist.
1665 Originally there were few.) While I was there, I also converted the
1666 str(n)cmp tests so they don't re-test the leading "-" and the first
1667 character, in the hope this might squeeze out yet more improvement.
1669 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1670 flag allowing group syntax was set by the header_syntax check but not
1671 turned off, possible causing trouble later; (2) The flag was not being
1672 set at all for the header_verify test, causing "group"-style headers to
1673 be rejected. I have now set it in this case, and also caused header_
1674 verify to ignore an empty address taken from a group. While doing this, I
1675 came across some other cases where the code for allowing group syntax
1676 while scanning a header line wasn't quite right (mostly, not resetting
1677 the flag correctly in the right place). These bugs could have caused
1678 trouble for malformed header lines. I hope it is now all correct.
1680 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1681 with the "reply" argument non-NULL. The code, however (which originally
1682 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1683 but it didn't always do it. This confused somebody who was copying the
1684 code for some other use. I have removed all the tests.
1686 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1687 feature that was used to support insecure browsers during the U.S. crypto
1688 embargo. It requires special client support, and Exim is probably the
1689 only MTA that supported it -- and would never use it because real RSA is
1690 always available. This code has been removed, because it had the bad
1691 effect of slowing Exim down by computing (never used) parameters for the
1692 RSA_EXPORT functionality.
1694 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1695 authenticator to fail if there's a tab character in the incoming data
1696 (there should never be unless someone is messing about, as it's supposed
1697 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1698 if the connection is using TLS or if the remote IP is the same as the
1699 local IP, and the "valid-client-cert option" if a client certificate has
1702 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1703 authenticators. This can be used for authorization after authentication
1704 succeeds. (In the case of plaintext, it servers for both authentication
1707 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1708 if any retry times were supplied.
1710 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1711 connection, where there is no client IP address to check. In this
1712 situation, the verify now always succeeds.
1714 PH/25 Applied John Jetmore's -Mset patch.
1716 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1718 PH/27 In a string expansion for a processed (not raw) header when multiple
1719 headers of the same name were present, leading whitespace was being
1720 removed from all of them, but trailing whitespace was being removed only
1721 from the last one. Now trailing whitespace is removed from each header
1722 before concatenation. Completely empty headers in a concatenation (as
1723 before) are ignored.
1725 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1726 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1728 PH/29 [Removed. This was a change that I later backed out, and forgot to
1729 correct the ChangeLog entry (that I had efficiently created) before
1730 committing the later change.]
1732 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1733 address errors (4xx response to RCPT) over the same connection as other
1734 messages routed to the same hosts. Such deliveries are always "forced",
1735 so retry times are not inspected. This resulted in far too many retries
1736 for the affected addresses. The effect occurred only when there were more
1737 hosts than the hosts_max_try setting in the smtp transport when it had
1738 the 4xx errors. Those hosts that it had tried were not added to the list
1739 of hosts for which the message was waiting, so if all were tried, there
1740 was no problem. Two fixes have been applied:
1742 (i) If there are any address or message errors in an SMTP delivery, none
1743 of the hosts (tried or untried) are now added to the list of hosts
1744 for which the message is waiting, so the message should not be a
1745 candidate for sending over the same connection that was used for a
1746 successful delivery of some other message. This seems entirely
1747 reasonable: after all the message is NOT "waiting for some host".
1748 This is so "obvious" that I'm not sure why it wasn't done
1749 previously. Hope I haven't missed anything, but it can't do any
1750 harm, as the worst effect is to miss an optimization.
1752 (ii) If, despite (i), such a delivery is accidentally attempted, the
1753 routing retry time is respected, so at least it doesn't keep
1754 hammering the server.
1756 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1757 in ${readsocket because some servers need this prod.
1759 PH/32 Added some extra debug output when updating a wait-xxx database.
1761 PH/33 The hint "could be header name not terminated by colon", which has been
1762 given for certain expansion errors for a long time, was not being given
1763 for the ${if def:h_colon_omitted{... case.
1765 PH/34 The spec says: "With one important exception, whenever a domain list is
1766 being scanned, $domain contains the subject domain." There was at least
1767 one case where this was not true.
1769 PH/35 The error "getsockname() failed: connection reset by peer" was being
1770 written to the panic log as well as the main log, but it isn't really
1771 panic-worthy as it just means the connection died rather early on. I have
1772 removed the panic log writing for the ECONNRESET error when getsockname()
1775 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1776 runs only) independently of the message's sender address. This meant
1777 that, if the 4xx error was in fact related to the sender, a different
1778 message to the same recipient with a different sender could confuse
1779 things. In particualar, this can happen when sending to a greylisting
1780 server, but other circumstances could also provoke similar problems.
1781 I have changed the default so that the retry time for these errors is now
1782 based a combination of the sender and recipient addresses. This change
1783 can be overridden by setting address_retry_include_sender=false in the
1786 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1787 remote server are returned as part of bounce messages. This was not
1788 happening for LMTP over a pipe (the lmtp transport), but now it is the
1789 same for both kinds of LMTP.
1791 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1792 in header lines that were in fact CNAMEs. This is no longer the case.
1794 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1795 and queue runs started by the daemon processed all messages. This has
1796 been fixed so that -R and -S can now usefully be given with -q<time>.
1798 PH/40 Import PCRE release 6.7 (fixes some bugs).
1800 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1802 PH/42 Give an error if -q is specified more than once.
1804 PH/43 Renamed the variables $interface_address and $interface_port as
1805 $received_ip_address and $received_port, to make it clear that these
1806 values apply to message reception, and not to the outgoing interface when
1807 a message is delivered. (The old names remain recognized, of course.)
1809 PH/44 There was no timeout on the connect() call when using a Unix domain
1810 socket in the ${readsocket expansion. There now is.
1812 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1813 be meaningful with "accept".
1815 SC/01 Eximstats V1.43
1816 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1818 SC/02 Eximstats V1.44
1819 Use a glob alias rather than an array ref in the generated
1820 parser. This improves both readability and performance.
1822 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1823 Collect SpamAssassin and rejection statistics.
1824 Don't display local sender or destination tables unless
1825 there is data to show.
1826 Added average volumes into the top table text output.
1828 SC/04 Eximstats V1.46
1829 Collect data on the number of addresses (recipients)
1830 as well as the number of messages.
1832 SC/05 Eximstats V1.47
1833 Added 'Message too big' to the list of mail rejection
1834 reasons (thanks to Marco Gaiarin).
1836 SC/06 Eximstats V1.48
1837 Mainlog lines which have GMT offsets and are too short to
1838 have a flag are now skipped.
1840 SC/07 Eximstats V1.49 (Alain Williams)
1841 Added the -emptyok flag.
1843 SC/08 Eximstats V1.50
1844 Fixes for obtaining the IP address from reject messages.
1846 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1847 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1848 whitesspace changes from 4.64-PH/27
1850 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1853 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1854 are found, allow negative numbers in numeric criteria)
1856 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1858 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1859 to match changes made in 4.64-PH/43
1861 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1863 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1864 "domain" is actually an IP address, and force a failure. This locks out
1865 those revolvers/nameservers that support "A-for-A" lookups, in
1866 contravention of the specifications.
1868 PH/48 When a host name was looked up from an IP address, and the subsequent
1869 forward lookup of the name timed out, the host name was left in
1870 $sender_host_name, contrary to the specification.
1872 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1873 restricted to single-key lookups, Exim was not diagnosing an error if
1874 * or *@ was used with a query-style lookup.
1876 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1878 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1879 long ago, but noone interested enough thought of it. Let's just say that
1880 the "1.1" means that there are some new functions that weren't there at
1881 some point in the past.
1883 PH/51 Error processing for expansion failure of helo_data from an smtp
1884 transport during callout processing was broken.
1886 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1887 tested/used via the -bh/-bhc/-bs options.
1889 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1890 bug, fixed in subsequent PCRE releases).
1892 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1893 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1895 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1901 SC/01 Use a glob alias rather than an array ref in eximstats generated
1902 parser. This improves both readability and performance.
1904 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1905 Don't display local sender or destination tables in eximstats unless
1906 there is data to show.
1907 Added average volumes into the eximstats top table text output.
1909 SC/03 Collect data on the number of addresses (recipients) as well
1910 as the number of messages in eximstats.
1912 TF/01 Correct an error in the documentation for the redirect router. Exim
1913 does (usually) call initgroups() when daemonizing.
1915 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1916 with consistent privilege compared to when running as a daemon.
1918 TF/03 Note in the spec that $authenticated_id is not set for local
1919 submissions from trusted users.
1921 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1922 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1924 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1925 by adding some example configuration directives to the default
1926 configuration file. A little bit of work is required to uncomment the
1927 directives and define how usernames and passwords are checked, but
1928 there is now a framework to start from.
1930 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1931 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1932 without this. I don't know how relevant this is to other LDAP libraries.
1934 PH/02 Add the verb name to the "unknown ACL verb" error.
1936 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1938 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1940 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1941 directory not expanded when it should be if an expanded home directory
1942 was set for the address (which is overridden by the transport).
1944 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1947 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1948 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1949 because it is too late at that time, and has no effect.
1951 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1952 security issue with \' (bugzilla #107). I could not use the
1953 PQescapeStringConn() function, because it needs a PGconn value as one of
1956 PH/08 When testing addresses using -bt, indicate those final addresses that
1957 are duplicates that would not cause an additional delivery. At least one
1958 person was confused, thinking that -bt output corresponded to deliveries.
1959 (Suppressing duplicates isn't a good idea as you lose the information
1960 about possibly different redirections that led to the duplicates.)
1962 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1963 systems where poll() doesn't work, in particular OS X.
1965 PH/10 Added more information to debugging output for retry time not reached.
1967 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1968 operations in malware.c.
1970 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1973 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1974 syslog_duplication set false, log lines that would normally be written
1975 both the the main log and to the reject log were not written to syslog at
1978 PH/14 In the default configuration, change the use of "message" in ACL warn
1979 statements to "add_header".
1981 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1982 not followed by a command (e.g. "seen endif").
1984 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1985 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1988 PH/17 Added extra conditions to the default value of delay_warning_condition
1992 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1993 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1994 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1997 The Auto-Submitted: and various List- headers are standardised, whereas I
1998 don't think Precedence: ever was.
2000 PH/18 Refactored debugging code in route_finduser() to show more information,
2001 in particular, the error code if getpwnam() issues one.
2003 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2004 This is apparently needed in addition to the PH/07 change above to avoid
2005 any possible encoding problems.
2007 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2008 but not after initializing Perl.
2010 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2011 output them only if debugging. By default they are written stderr,
2012 apparently, which is not desirable.
2014 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2017 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2020 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2022 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2023 authenticated or an ident call has been made. Suppress the default
2024 values for $authenticated_id and $authenticated_sender (but permit -oMai
2025 and -oMas) when testing with -bh.
2027 PH/24 Re-jigged the order of the tests in the default configuration so that the
2028 tests for valid domains and recipients precede the DNS black list and CSA
2029 tests, on the grounds that those ones are more expensive.
2031 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2032 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2033 This bug exists in every version of Exim that I still have, right back to
2036 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2037 However, an attempt to turn on case-sensitivity in a regex key by
2038 including (?-i) didn't work because the subject string was already
2039 lowercased, and the effects were non-intuitive. It turns out that a
2040 one-line patch can be used to allow (?-i) to work as expected.
2046 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2047 other effects) broke the use of negated acl sub-conditions.
2049 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2052 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2053 "Deny" causes Exim to reject the incoming connection with a 554 error.
2054 Unfortunately, if there is a major crisis, such as a disk failure,
2055 tcp-wrappers gives "deny", whereas what one would like would be some
2056 kind of temporary error. A kludge has been added to help with this.
2057 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2058 554 error is used if errno is still zero or contains ENOENT (which occurs
2059 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2062 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2064 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2065 errors. Otherwise a message that provokes a temporary error (when other
2066 messages do not) can cause a whole host to time out.
2068 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2069 addresses were routed directly to files or pipes from a redirect router.
2070 File deliveries just didn't batch; pipe deliveries might have suffered
2073 PH/06 A failure to get a lock for a hints database would erroneously always say
2074 "Failed to get write lock", even when it was really a read lock.
2076 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2077 of 0600. This has been changed to use the value of the lockfile_mode
2078 option (which defaults to 0600).
2080 PH/08 Applied small patch from the Sieve maintainer.
2082 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2083 folder from quota calculations, a direct delivery into this folder messed
2084 up the contents of the maildirsize file. This was because the regex was
2085 used only to exclude .Trash (or whatever) when the size of the mailbox
2086 was calculated. There was no check that a delivery was happening into an
2087 excluded directory. This bug has been fixed by ignoring all quota
2088 processing for deliveries into excluded directories.
2090 PH/10 Added the maildirfolder_create_regex option to appendfile.
2096 PH/01 The code for finding all the local interface addresses on a FreeBSD
2097 system running IPv6 was broken. This may well have applied to all BSD
2098 systems, as well as to others that have similar system calls. The broken
2099 code found IPv4 interfaces correctly, but gave incorrect values for the
2100 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2101 that it would not match correctly against @[] and not recognize the IPv6
2104 PH/02 The ipliteral router was not recognizing addresses of the form user@
2105 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2107 PH/03 Added disable_ipv6.
2109 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2110 decoded form, because it is most often used to construct To: headers
2111 lines in autoreplies, and the decoded form may well be syntactically
2112 invalid. However, $reply_address has leading white space removed, and all
2113 newlines turned into spaces so that the autoreply transport does not
2116 PH/05 If group was specified without a user on a router, and no group or user
2117 was specified on a transport, the group from the router was ignored.
2119 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2120 for visible compile-time settings that can be used to change these
2121 numbers, for those that want even more. Backwards compatibility with old
2122 spool files has been maintained. However, going back to a previous Exim
2123 release will lost any variables that are in spool files.
2125 PH/07 Two small changes when running in the test harness: increase delay when
2126 passing a TCP/IP connection to a new process, in case the original
2127 process has to generate a bounce, and remove special handling of
2128 127.0.0.2 (sic), which is no longer necessary.
2130 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2131 be the same on different OS.
2133 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2136 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2137 whether --show-vars was specified or not
2139 JJ/02 exipick: Added support for new ACL variable spool format introduced
2142 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2143 syntactically invalid From: or Reply-to: line, and a filter used this to
2144 generate an autoreply, and therefore failed to obtain an address for the
2145 autoreply, Exim could try to deliver to a non-existent relative file
2146 name, causing unrelated and misleading errors. What now happens is that
2147 it logs this as a hard delivery error, but does not attempt to create a
2150 PH/11 The exinext utility has a -C option for testing purposes, but although
2151 the given file was scanned by exinext itself; it wasn't being passed on
2152 when Exim was called.
2154 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2155 an end-of-file indication when reading a command response.
2157 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2158 compiled. In many other places in Exim, IPv6 addresses are always
2159 recognized, so I have changed this. It also means that IPv4 domain
2160 literals of the form [IPV4:n.n.n.n] are now always recognized.
2162 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2163 used if the router is not running as root, for example, when verifying at
2164 ACL time, or when using -bh. The debugging output from this situation was
2165 non-existent - all you got was a failure to exec. I have made two
2168 (a) Failures to set uid/gid, the current directory, or a process leader
2169 in a subprocess such as that created by queryprogram now generate
2170 suitable debugging ouput when -d is set.
2172 (b) The queryprogram router detects when it is not running as root,
2173 outputs suitable debugging information if -d is set, and then runs
2174 the subprocess without attempting to change uid/gid.
2176 PH/15 Minor change to Makefile for building test_host (undocumented testing
2179 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2180 additional section of a DNS packet that returns MX or SRV records.
2181 Instead, it always explicitly searches for A/AAAA records. This avoids
2182 major problems that occur when a DNS server includes only records of one
2183 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2184 fixed another bug: if SRV records were looked up and the corresponding
2185 address records were *not* found in the additional section, the port
2186 values from the SRV records were lost.
2188 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2189 using the correct key (the original address) when searching the retry
2190 rules in order to find which one to use for generating the retry hint.
2192 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2193 adding the default one. Similarly, if it contains a Reply-To: header, the
2194 errors_reply_to option, if set, is not used.
2196 PH/19 When calculating a retry time, Exim used to measure the "time since
2197 failure" by looking at the "first failed" field in the retry record. Now
2198 it does not use this if it is later than than the arrival time of the
2199 message. Instead it uses the arrival time. This makes for better
2200 behaviour in cases where some deliveries succeed, thus re-setting the
2201 "first failed" field. An example is a quota failure for a huge message
2202 when small messages continue to be delivered. Without this change, the
2203 "time since failure" will always be short, possible causing more frequent
2204 delivery attempts for the huge message than are intended.
2205 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2207 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2208 $1, $2, $3) because the numerical variables can be reset during some
2209 expansion items (e.g. "match"), thereby losing the authentication data.
2211 PH/21 Make -bV show the size of off_t variables so that the test suite can
2212 decide whether to run tests for quotas > 2G.
2214 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2215 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2216 filecount value is greater than 2G or if a quota value is greater than 2G
2217 on a system where the size of off_t is not greater than 4, a panic error
2220 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2221 never match. The debug and -bh output now contains an explicit error
2222 message indicating a malformed IPv4 address or mask.
2224 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2225 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2226 PH/23 above applies.
2228 PH/25 Do not write to syslog when running in the test harness. The only
2229 occasion when this arises is a failure to open the main or panic logs
2230 (for which there is an explicit test).
2232 PH/26 Added the /no_tell option to "control=freeze".
2234 PH/27 If a host name lookup failed very early in a connection, for example, if
2235 the IP address matched host_lookup and the reverse lookup yielded a name
2236 that did not have a forward lookup, an error message of the form "no IP
2237 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2238 could be logged. Now it outputs the IP address instead of "NULL".
2240 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2241 allows the sender and the authenticated sender to be set when
2242 submitting a message from within Exim. Since child_open_exim() is
2243 documented for local_scan(), the new function should be too.
2245 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2246 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2247 results in an empty string is now treated as unset.
2249 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2251 PH/31 Added sender_verify_fail logging option.
2253 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2254 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2255 tidied the source and removed it altogether.
2257 PH/33 When a queue run was abandoned because the load average was too high, a
2258 log line was always written; now it is written only if the queue_run log
2259 selector is set. In addition, the log line for abandonment now contains
2260 information about the queue run such as the pid. This is always present
2261 in "start" and "stop" lines but was omitted from the "abandon" line.
2263 PH/34 Omit spaces between a header name and the colon in the error message that
2264 is given when verify = headers_syntax fails (if there are lots of them,
2265 the message gets confusing).
2267 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2268 names, as there are now some PTR records that contain slashes. This check
2269 is only to protect against broken name servers that fall over on strange
2270 characters, so the fact that it applies to all lookups doesn't matter.
2272 PH/36 Now that the new test suite is complete, we can remove some of the
2273 special code in Exim that was needed for the old test suite. For example,
2274 sorting DNS records because real resolvers return them in an arbitrary
2275 order. The new test suite's fake resolver always returns records in the
2278 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2279 bounces) except when queue_only is set, to avoid logging races between
2280 the different processes.
2282 PH/38 Panic-die if .include specifies a non-absolute path.
2284 PH/39 A tweak to the "H" retry rule from its user.
2286 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2287 a label. They prevented compilation on older perls.
2289 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2290 a warning to be raised on newish perls.
2292 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2293 on queue. Changes to match documented behaviour of showing count of
2294 messages matching specified criteria.
2296 PH/40 Changed the default ident timeout from 30s to 5s.
2298 PH/41 Added support for the use of login_cap features, on those BSD systems
2299 that have them, for controlling the resources used by pipe deliveries.
2301 PH/42 The content-scanning code uses fopen() to create files in which to put
2302 message data. Previously it was not paying any attention to the mode of
2303 the files. Exim runs with umask(0) because the rest of the code creates
2304 files with open(), and sets the required mode explicitly. Thus, these
2305 files were ending up world-writeable. This was not a big issue, because,
2306 being within the spool directory, they were not world-accessible. I have
2307 created a function called modefopen, which takes an additional mode
2308 argument. It sets umask(777), creates the file, chmods it to the required
2309 mode, then resets the umask. All the relevant calls to fopen() in the
2310 content scanning code have been changed to use this function.
2312 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2313 to 24 hours. This avoids potential overflow problems when processing G
2314 and H retry rules. I suspect nobody ever tinkers with this value.
2316 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2318 PH/45 When the plaintext authenticator is running as a client, the server's
2319 challenges are checked to ensure they are valid base64 strings. By
2320 default, the authentication attempt is cancelled if an invalid string is
2321 received. Setting client_ignore_invalid_base64 true ignores these errors.
2322 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2323 they are received. Thus, the responses can be made to depend on the
2324 challenges. If an invalid string is ignored, an empty string is placed in
2327 PH/46 Messages that are created by the autoreply transport now contains a
2328 References: header, in accordance with RFCs 2822 and 3834.
2330 PH/47 Added authenticated_sender_force to the smtp transport.
2332 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2334 PH/49 Installed latest patch from the Sieve maintainer.
2336 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2337 was also set, the appendfile transport was unnecessarily scanning a
2338 directory of message files (e.g. for maildir delivery) to find the count
2339 of files (along with the size), even though it did not need this
2340 information. It now does the scan only if it needs to find either the
2341 size of the count of files.
2343 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2345 PH/52 Two bugs concerned with error handling when the smtp transport is
2348 (i) Exim was not creating retry information for temporary errors given
2349 for individual recipients after the DATA command when the smtp transport
2350 was used in LMTP mode. This meant that they could be retried too
2351 frequently, and not timed out correctly.
2353 (ii) Exim was setting the flag that allows error details to be returned
2354 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2355 recipients that were returned after the DATA command.
2357 PH/53 This is related to PH/52, but is more general: for any failing address,
2358 when detailed error information was permitted to be returned to the
2359 sender, but the error was temporary, then after the final timeout, only
2360 "retry timeout exceeded" was returned. Now it returns the full error as
2361 well as "retry timeout exceeded".
2363 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2364 do this, and (what is worse) MTAs that accept it.
2366 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2367 will now be deprecated.
2369 PH/56 New os.c-cygwin from the Cygwin maintainer.
2371 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2372 formats (previously only available in exim formats via -bpr, -bpru,
2373 and -bpra. Now also available in native and exiqgrep formats)
2375 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2376 with very large, slow to parse queues
2378 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2380 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2382 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2383 responses to authentication challenges, though it was showing the
2384 challenges; (ii) I've removed the CR characters from the debug output for
2387 PH/58 Allow for the insertion of a newline as well as a space when a string
2388 is turned into more than one encoded-word during RFC 2047 encoding. The
2389 Sieve code now uses this.
2391 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2392 data_4xx, lost_connection, tls_required.
2394 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2395 message was being sent as an SMTP response.
2397 PH/61 Add -l and -k options to exicyclog.
2399 PH/62 When verifying, if an address was redirected to one new address, so that
2400 verification continued, and the new address failed or deferred after
2401 having set something in $address_data, the value of $address_data was not
2402 passed back to the ACL. This was different to the case when no
2403 redirection occurred. The value is now passed back in both cases.
2405 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2406 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2407 use login_cap.h, so on its own it isn't the distinguishing feature. The
2408 new name refers directly to the setclassresources() function.
2410 PH/65 Added configuration files for NetBSD3.
2412 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2414 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2415 is preferred over IPv4.
2417 PH/68 The bounce_return_message and bounce_return_body options were not being
2418 honoured for bounces generated during the reception of non-SMTP messages.
2419 In particular, this applied to messages rejected by the ACL. This bug has
2420 been fixed. However, if bounce_return_message is true and bounce_return_
2421 body is false, the headers that are returned for a non-SMTP message
2422 include only those that have been read before the error was detected.
2423 (In the case of an ACL rejection, they have all been read.)
2425 PH/69 The HTML version of the specification is now built in a directory called
2426 spec_html instead of spec.html, because the latter looks like a path with
2427 a MIME-type, and this confuses some software.
2429 PH/70 Catch two compiler warnings in sieve.c.
2431 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2432 function verify_get_ident() calls ip_connect() to connect a socket, but
2433 if the "connect()" function timed out, ip_connect() used to close the
2434 socket. However, verify_get_ident() also closes the socket later, and in
2435 between Exim writes to the log, which may get opened at this point. When
2436 the socket was closed in ip_connect(), the log could get the same file
2437 descriptor number as the socket. This naturally causes chaos. The fix is
2438 not to close the socket in ip_connect(); the socket should be closed by
2439 the function that creates it. There was only one place in the code where
2440 this was missing, in the iplookup router, which I don't think anybody now
2441 uses, but I've fixed it anyway.
2443 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2444 well as to direct DNS lookups. Otherwise the handling of names in host
2445 lists is inconsistent and therefore confusing.
2451 PH/01 Two changes to the default runtime configuration:
2453 (1) Move the checks for relay_from_hosts and authenticated clients from
2454 after to before the (commented out) DNS black list checks.
2456 (2) Add control=submission to the relay_from_hosts and authenticated
2457 clients checks, on the grounds that messages accepted by these
2458 statements are most likely to be submissions.
2460 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2462 (1) Generate an error if the third argument for the ${prvs expansion is
2465 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2468 (3) Reset the variables that are obtained from the first argument of
2469 ${prvscheck and used in the second argument before leaving the code,
2470 because their memory is reclaimed, so using them afterwards may do
2473 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2474 one (it's much easier than Tom thought :-).
2476 (5) Because of (4), we can now allow for the use of $prvscheck_result
2477 inside the third argument.
2479 PH/03 For some reason, the default setting of PATH when running a command from
2480 a pipe transport was just "/usr/bin". I have changed it to
2483 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2484 anything to be listed in the output from -bV.
2486 PH/05 When a filter generated an autoreply, the entire To: header line was
2487 quoted in the delivery log line, like this:
2489 => >A.N.Other <ano@some.domain> <original@ddress> ...
2491 This has been changed so that it extracts the operative address. There
2492 may be more than one such address. If so, they are comma-separated, like
2495 => >ano@some.domain,ona@other.domain <original@ddress> ...
2497 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2498 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2499 not being looked up in the rDNS to get a host name, Exim was showing the
2500 IP address twice in Received: lines, even though the IP addresses were
2501 identical. For example:
2503 Received: from [1.2.3.4] (helo=[1.2.3.4])
2505 However, if the real host name was known, it was omitting the HELO data
2506 if it matched the actual IP address. This has been tidied up so that it
2507 doesn't show the same IP address twice.
2509 PH/07 When both +timestamp and +memory debugging was on, the value given by
2510 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2511 called by the debug printing, thereby overwriting the timestamp buffer.
2512 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2514 PH/08 When the original message was included in an autoreply transport, it
2515 always said "this is a copy of the message, including all the headers",
2516 even if body_only or headers_only was set. It now gives an appropriate
2519 PH/09 Applied a patch from the Sieve maintainer which:
2521 o fixes some comments
2522 o adds the (disabled) notify extension core
2523 o adds some debug output for the result of if/elsif tests
2524 o points to the current vacation draft in the documentation
2525 and documents the missing references header update
2529 o fixes a bug in processing the envelope test (when testing
2530 multiple envelope elements, the last element determinted the
2533 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2534 Electronic Mail") by including:
2536 Auto-submitted: auto-generated
2538 in the messages that it generates (bounce messages and others, such as
2539 warnings). In the case of bounce messages for non-SMTP mesages, there was
2540 also a typo: it was using "Auto_submitted" (underscore instead of
2541 hyphen). Since every message generated by Exim is necessarily in response
2542 to another message, thes have all been changed to:
2544 Auto-Submitted: auto-replied
2546 in accordance with these statements in the RFC:
2548 The auto-replied keyword:
2550 - SHOULD be used on messages sent in direct response to another
2551 message by an automatic process,
2553 - MUST NOT be used on manually-generated messages,
2555 - MAY be used on Delivery Status Notifications (DSNs) and Message
2556 Disposition Notifications (MDNs),
2558 - MUST NOT be used on messages generated by automatic or periodic
2559 processes, except for messages which are automatic responses to
2562 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2563 to the default Received: header definition.
2565 PH/12 Added log selector acl_warn_skipped (default on).
2567 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2568 variables because (a) they are in the wrong storage pool and (b) even if
2569 they were copied, it wouldn't work properly because of the caching.
2571 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2572 checking when decoding. Apparently there are clients that generate
2573 overlong encoded strings. Why am I not surprised?
2575 PH/15 If the first argument of "${if match_address" was not empty, but did not
2576 contain an "@" character, Exim crashed. Now it writes a panic log message
2577 and treats the condition as false.
2579 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2581 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2582 extension "envelope-auth". The code is finished and in agreement with
2583 other implementations, but there is no documentation so far and in fact,
2584 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2585 not changing the active code.
2587 Print executed "if" and "elsif" statements when debugging is used. This
2588 helps a great deal to understand what a filter does.
2590 Document more things not specified clearly in RFC3028. I had all this
2591 sorted out, when out of a sudden new issues came to my mind. Oops."
2593 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2596 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2597 canonical form (as documented). However, after a host name lookup from
2598 the IP address, check_host() was doing a simple string comparison with
2599 addresses acquired from the DNS when checking that the found name did
2600 have the original IP as one of its addresses. Since any found IPv6
2601 addresses are likely to be in abbreviated form, the comparison could
2602 fail. Luckily, there already exists a function for doing the comparison
2603 by converting both addresses to binary, so now that is used instead of
2604 the text comparison.
2606 PH/20 There was another similar case to PH/19, when a complete host name was
2607 given in a host list; looking up its IP address could give an abbreviated
2608 form, whereas the current host's name might or might not be abbreviated.
2609 The same fix has been applied.
2615 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2616 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2619 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2620 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2622 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2624 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2625 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2626 Also, for "Auto-Submitted", treat anything other than "no" as
2627 non-personal, in accordance with RFC 3834. (Previously it treated
2628 anything starting "auto-" as non-personal.)
2630 TF/01 The control=submission/name=... option had a problem with syntax
2631 errors if the name included a slash character. The /name= option
2632 now slurps the rest of the string, so it can include any characters
2633 but it must come last in the list of options (after /sender_retain
2636 PH/05 Some modifications to the interface to the fake nameserver for the new
2644 TK/01 Added the "success_on_redirect" address verification option. See
2645 NewStuff for rationale and an example.
2647 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2649 PH/02 Patch to exigrep to allow it to work on syslog lines.
2651 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2652 fread() to skip over the body file's header line, because in Cygwin the
2653 header line is locked and is inaccessible.
2655 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2656 co-exist for some time) to make it clear that it is the Exim ID that is
2657 referenced, not the Message-ID: header line.
2659 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2660 string_format() function, because snprintf() does not exist on all
2663 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2664 ${stat: expansion item.
2666 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2667 protocol synchronization error", to keep the pedants happy.
2669 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2670 well as for IRIX systems, when gcc is being used. See the host.c source
2673 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2675 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2678 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2679 transport and to the smtp transport in LMTP mode.
2681 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2683 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2684 run a filter in a subprocess. This could lead to confusion in subsequent
2685 lookups in the parent process. There should also be a search_tidyup() at
2686 the end of the subprocess.
2688 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2689 only if the host matched helo_try_verify_hosts, which caused the
2690 verification to occur when the EHLO/HELO command was issued. The ACL just
2691 tested the remembered result. Now, if a previous verification attempt has
2692 not happened, "verify = helo" does it there and then.
2694 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2696 TK/03 Fix log output including CR from clamd.
2698 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2699 exist provoked a memory error which could cause a segfault.
2701 PH/15 Installed PCRE 6.2
2703 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2705 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2706 of the problem. Specifically, suggested +O2 rather than +O1 for the
2709 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2711 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2712 record, the port was not correctly logged when the outgoing_port log
2713 selector was set (it logged the transort's default port).
2715 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2716 fallback_hosts, and "hosts" in the smtp transport.
2718 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2719 host errors such as "Connection refused".
2721 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2722 authentication with radiusclient 0.4.9:
2724 - Error returned from rc_read_config was caught wrongly
2725 - Username/password not passed on to radius server due to wrong length.
2727 The presumption is that some radiusclient API changes for 4.51/PH/17
2728 were not taken care of correctly. The code is still untested by me (my
2729 Linux distribution still has 0.3.2 of radiusclient), but it was
2730 contributed by a Radius user.
2732 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2733 expanding the "port" option of the smtp transport.
2735 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2736 while reading a MIME header. Thanks to Tom Hughes for a patch.
2738 PH/24 Include config.h inside local_scan.h so that configuration settings are
2741 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2742 that in an ACL for RCPT (for example), you can examine exactly what was
2745 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2746 commands, but it was not correctly comparing the address with the actual
2747 client host address. Thus, it would show the EHLO address in Received:
2748 header lines when this was not necessary.
2750 PH/27 Added the % operator to ${eval:}.
2752 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2753 it should be ignoring failures (because with -C, for example, it has lost
2754 privilege). It wasn't ignoring creation failures other than "already
2757 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2760 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2761 because an input line was too long, either on its own, or by virtue of
2762 too many continuations, the temporary file was not being removed, and the
2763 return code was incorrect.
2765 PH/31 Missing "BOOL" in function definition in filtertest.c.
2767 PH/32 Applied Sieve patches from the maintainer.
2769 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2771 PH/33 Added "verify = not_blind".
2773 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2774 Local/Makefile (with some defaults set). These are used in built scripts
2775 such as exicyclog, but they have never been used in the exim_install
2776 script (though there are many overriding facilities there). I have
2777 arranged that the exim_install script now takes note of these two
2780 PH/35 Installed configuration files for Dragonfly.
2782 PH/36 When a locally submitted message by a trusted user did not contain a
2783 From: header, and the sender address was obtained from -f or from an SMTP
2784 MAIL command, and the trusted user did not use -F to supply a sender
2785 name, $originator_name was incorrectly used when constructing a From:
2786 header. Furthermore, $originator_name was used for submission mode
2787 messages from external hosts without From: headers in a similar way,
2788 which is clearly wrong.
2790 PH/37 Added control=suppress_local_fixups.
2792 PH/38 When log_selector = +received_sender was set, and the addition of the
2793 sender made the log line's construction buffer exactly full, or one byte
2794 less than full, an overflow happened when the terminating "\n" was
2797 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2798 when the result of a list match is failure because a DNS lookup failed.
2800 PH/40 RM_COMMAND is now used in the building process.
2802 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2803 the "build-* directories that it finds.
2805 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2806 address in a domain literal was a prefix of an interface address.
2808 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2809 when verifying a sender address, unless rewrite_headers is false.
2811 PH/44 Wrote a long comment about why errors_to addresses are verified as
2812 recipients, not senders.
2814 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2815 the ratelimit ACL was added.
2817 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2819 PH/46 Added extra information about PostgreSQL errors to the error string.
2821 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2822 avoiding the need to install special zones in a real server. This is
2823 backwards compatible; if it can't find the fake resolver, it drops back.
2824 Thus, both old and new test suites can be run.
2826 TF/02 Added util/ratelimit.pl
2828 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2829 clock is set back in time.
2831 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2832 Candler <B.Candler@pobox.com>.
2834 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2835 OK for addresses that are the result of redirections.
2837 PH/48 A number of further additions for the benefit of the new test suite,
2838 including a fake gethostbyname() that interfaces to the fake DNS resolver
2841 TF/06 The fix for widen_domains has also been applied to qualify_single and
2842 search_parents which are the other dnslookup options that can cause
2845 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2848 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2850 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2851 service" if it encountered an MX record with an empty target hostname.
2852 The message is now "an MX or SRV record indicated no SMTP service".
2854 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2855 if the DNS of the sending site is misconfigured. This is quite a
2856 common situation. This change restores the behaviour of treating a
2857 helo verification defer as a failure.
2859 PH/51 If self=fail was set on a router, the bounce message did not include the
2860 actual error message.
2866 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2868 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2869 command itself ended in error, the underlying message about the transport
2870 filter timeout was being overwritten with the pipe command error. Now the
2871 underlying error message should be appended to the second error message.
2873 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2875 PH/02 Reduce the amount of output that "make" produces by default. Full output
2876 can still be requested.
2878 PH/03 The warning log line about a condition test deferring for a "warn" verb
2879 was being output only once per connection, rather than after each
2880 occurrence (because it was using the same function as for successful
2881 "warn" verbs). This seems wrong, so I have changed it.
2883 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2884 it should not have, which might have caused a crash in the right
2885 circumstances, but probably never did.
2887 PH/04 Installed a modified version of Tony Finch's patch to make submission
2888 mode fix the return path as well as the Sender: header line, and to
2889 add a /name= option so that you can make the user's friendly name appear
2892 TF/03 Added the control = fakedefer ACL modifier.
2894 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2895 Mark Lowes for thorough testing.
2897 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2899 TK/03 Merged latest SRS patch from Miles Wilton.
2901 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2902 with the definition in sysexits.h (which is #included earlier).
2903 Fortunately, Exim does not actually use EX_OK. The code used to try to
2904 preserve the sysexits.h value, by assumimg that macro definitions were
2905 scanned for macro replacements. I have been disabused of this notion,
2906 so now the code just undefines EX_OK before #including unistd.h.
2908 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2909 in the smtp transport. When a block could not be written in a single
2910 write() function, the timeout was being re-applied to each part-write.
2911 This seems wrong - if the receiver was accepting one byte at a time it
2912 would take for ever. The timeout is now adjusted when this happens. It
2913 doesn't have to be particularly precise.
2915 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2916 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2918 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2919 without a domain if the check to <postmaster@domain> fails.
2921 SC/01 Eximstats: added -xls and the ability to specify output files
2922 (patch written by Frank Heydlauf).
2924 SC/02 Eximstats: use FileHandles for outputing results.
2926 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2928 SC/04 Eximstats: fixed display of large numbers with -nvr option
2930 SC/05 Eximstats: fixed merging of reports with empty tables.
2932 SC/06 Eximstats: added the -include_original_destination flag
2934 SC/07 Eximstats: removed tabs and trailing whitespace.
2936 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2938 TK/06 MBOX spool code: Add real "From " MBOX separator line
2939 so the .eml file is really in mbox format (even though
2940 most programs do not really care). Patch from Alex Miller.
2942 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2943 The latter is generated from $received_to and is only set if the
2944 message has one envelope recipient. SA can use these headers,
2945 obviously out-of-the-box. Patch from Alex Miller.
2947 PH/08 The ${def test on a variable was returning false if the variable's
2948 value was "0", contrary to what the specification has always said!
2949 The result should be true unless the variable is empty.
2951 PH/09 The syntax error of a character other than { following "${if
2952 def:variable_name" (after optional whitespace) was not being diagnosed.
2953 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2954 accidental colon was present, for example, could give incorrect results.
2956 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2957 result is used (not including appendfile, where other changes are about
2960 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2961 This involved changing a lot of size variables from int to off_t. It
2962 should work with maildirs and everything.
2964 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2965 spamd dying while we are connected to it.
2967 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2970 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2971 function for BDB 4.3.
2973 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2975 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2976 the sources for PCRE have been re-arranged and more files are now
2979 PH/15 The code I had for printing potentially long long variables in PH/11
2980 above was not the best (it lost precision). The length of off_t variables
2981 is now inspected at build time, and an appropriate printing format (%ld
2982 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2983 to be "long long int" or "long int". This is needed for the internal
2984 formatting function string_vformat().
2986 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2987 the configuration file to be ":syslog", then the script "guesses" where
2988 the logs files are, rather than using the compiled in default. In our
2989 case the guess is not the same as the compiled default, so the script
2990 suddenly stopped working when I started to use syslog. The patch checks
2991 to see if log_file_path is "". If so, it attempts to read it from exim
2992 with no configuration file to get the compiled in version, before it
2993 falls back to the previous guessing code."
2995 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2996 implementing BATV in an Exim configuration. See NewStuff for the gory
2999 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3000 Makefile that are specific to HP-UX.
3002 PH/18 If the "use_postmaster" option was set for a recipient callout together
3003 with the "random" option, the postmaster address was used as the MAIL
3004 FROM address for the random test, but not for the subsequent recipient
3005 test. It is now used for both.
3007 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3008 patch removes a few documentation additions to RFC 3028, because the
3009 latest draft now contains them. It adds the new en;ascii-case comparator
3010 and a new error check for 8bit text in MIME parts. Comparator and
3011 require names are now matched exactly. I enabled the subaddress
3012 extension, but it is not well tested yet (read: it works for me)."
3014 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3015 rework some of the code of TK/09 above to avoid the hardwired use of
3016 "%lld" and "long long". Replaced the call to snprintf() with a call to
3019 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3020 records point to non-existent hosts", "retry timeout exceeded", and
3021 "retry time not reached for any host after a long failure period".
3023 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3024 experimental DomainKeys support:
3026 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3027 (2) On an error such as an illegally used "control", the wrong name for
3028 the control was given.
3030 These problems did NOT occur unless DomainKeys support was compiled.
3032 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3034 PH/24 Added ${if match_ip condition.
3036 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3037 databases so that it will be absolutely obvious if a crash occurs in the
3038 DB library. This is a regular occurrence (often caused by mis-matched
3041 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3042 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3043 picked up on a user's system that detects such things. There doesn't seem
3044 to be a gcc warning option for this - only an attribute that has to be
3045 put on the function's prototype. It seems that in Fedora Core 4 they have
3046 set this on a number of new functions. No doubt there will be more in due
3049 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3050 specify a transport. However, if an address that was verified by such a
3051 router was the subject of a callout, Exim crashed because it tried to
3052 read the rcpt_include_affixes from the non-existent transport. Now it
3053 just assumes that the setting of that option is false. This bug was
3054 introduced by 4.51/PH/31.
3056 PH/28 Changed -d+all to exclude +memory, because that information is very
3057 rarely of interest, but it makes the output a lot bigger. People tend to
3058 do -d+all out of habit.
3060 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3061 code in os-type was giving problems when libc.so lives in lib64, like on
3064 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3065 aren't the modern standard, and it seems that some systems' include files
3066 don't always have them. Exim was already checking for some of the newer
3067 ones like T_AAAA, and defining it itself. I've added checks for all the
3068 record types that Exim uses.
3070 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3071 not automatically generating a new one, as it is supposed to. This
3072 prevented TLS from working. If the file did exist, but contained invalid
3073 data, a new version was generated, as expected. It was only the case of a
3074 non-existent file that was broken.
3076 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3077 with a change in libdomainkeys > 0.64.
3079 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3080 from DNS. If the selector record carries the flag, it now has
3081 precedence over the domain-wide flag.
3083 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3085 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3086 the use of a port name that isn't defined in /etc/services) occurred, the
3087 message was deferred as in a normal delivery, and thus remained on the
3088 spool, instead of being failed because of the mua_wrapper setting. This
3089 is now fixed, and I tidied up some of the mua_wrapper messages at the
3092 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3093 the messages in a hash of arrays rather than using individual hashes.
3094 This is a bit cleaner and results in dramatic memory savings, albeit
3095 at a slight CPU cost.
3097 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3098 as requested by Marc Sherman.
3100 SC/10 Eximstats: added histograms for user specified patterns as requested
3103 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3105 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3106 fopen() in the content-scanning modules that did not already have it.
3112 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3113 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3115 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3117 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3119 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3120 to test Sieve filters that use "vacation".
3122 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3123 that changes the way the GnuTLS parameters are stored in the cache file.
3124 The new format can be generated externally. For backward compatibility,
3125 if the data in the cache doesn't make sense, Exim assumes it has read an
3126 old-format file, and it generates new data and writes a new file. This
3127 means that you can't go back to an older release without removing the
3130 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3131 work if there are any delivery delays because "one_time" forces the
3132 parent to be marked "delivered", so its unseen clone is never tried
3133 again. For this reason, Exim now forbids the simultaneous setting of
3136 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3137 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3138 entry if you want to know the details. The fix, however, neglected to
3139 consider the case when local delivery batching is involved. The test for
3140 "previously delivered" was not happening when checking to see if an
3141 address could be batched with a previous (undelivered) one; under
3142 certain circumstances this could lead to multiple deliveries to the same
3145 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3146 in its include files, and this causes problems building Exim.
3148 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3149 header_syntax) but Exim was just ignoring anything given after a slash.
3150 In particular, this caused confusion with an attempt to use "verify =
3151 reverse_host_lookup/defer_ok". An error is now given when options are
3152 supplied for verify items that do not have them. (Maybe reverse_host_
3153 lookup should have a defer_ok option, but that's a different point.)
3155 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3156 defined by RFC 821) to 2048, because there were problems with some AUTH
3157 commands, and RFC 1869 says the size should be increased for extended
3158 SMTP commands that take arguments.
3160 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3163 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3164 "unknown" error; now it says that the functionality isn't in the binary.
3166 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3167 an address' error message when a string expansion fails (syntax or
3168 whatever). Otherwise the password may appear in the log. Following change
3169 PH/42 below, there is no longer a chance of it appearing in a bounce
3172 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3174 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3175 of its addresses was ever tried. (Bugzilla bug #2.)
3177 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3178 the result incorrectly in the debug output. (It correctly added a newline
3179 to what was transported.)
3181 TF/01 Added $received_time.
3183 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3184 commented out examples of how to interface to a virus scanner and to
3185 SpamAssassin. Also added commented examples of av_scanner and
3186 spamd_address settings.
3188 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3189 and controls are allowed in which ACLs. There were a couple of minor
3190 errors. Some of the entries in the conditions table (which is a table of
3191 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3192 negation of where the condition IS allowed.
3194 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3196 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3197 header file does not have a version number, so I've had to invent a new
3198 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3199 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3200 radiusclient), but it was contributed by a Radius user.
3202 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3203 files or directories, for OpenSSL.
3205 PH/19 When an Exim process that is running as root has to create an Exim log
3206 file, it does so in a subprocess that runs as exim:exim so as to get the
3207 ownership right at creation (otherwise, other Exim processes might see
3208 the file with the wrong ownership). There was no test for failure of this
3209 fork() call, which would lead to the process getting stuck as it waited
3210 for a non-existent subprocess. Forks do occasionally fail when resources
3211 run out. I reviewed all the other calls to fork(); they all seem to check
3214 PH/20 When checking for unexpected SMTP input at connect time (before writing
3215 the banner), Exim was not dealing correctly with a non-positive return
3216 from the read() function. If the client had disconnected by this time,
3217 the result was a log entry for a synchronization error with an empty
3218 string after "input=" when read() returned zero. If read() returned -1
3219 (an event I could not check), uninitialized data bytes were printed.
3220 There were reports of junk text (parts of files, etc) appearing after
3223 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3225 PH/22 Added support for macro redefinition, and (re)definition in between
3226 driver and ACL definitions.
3228 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3229 forgetting to use the resulting value; it was using the unexpanded value.
3231 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3232 hadn't been configured. The fix is from Juergen Kreileder, who
3233 understands it better than I do:
3235 "Here's what I see happening with three configured cyrus_sasl
3236 authenticators configured (plain, login, cram-md5):
3238 On startup auth_cyrus_sasl_init() gets called for each of these.
3239 This means three calls to sasl_listmech() without a specified mech_list.
3240 => SASL tests which mechs of all available mechs actually work
3241 => three warnings about OTP not working
3242 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3244 With the patch, sasl_listmech() also gets called three times. But now
3245 SASL's mech_list option is set to the server_mech specified in the the
3246 authenticator. Or in other words, the answer from sasl_listmech()
3247 gets limited to just the mech you're testing for (which is different
3249 => the return list contains just 'plain' or 'login', 'cram-md5' or
3250 nothing depending on the value of ob->server_mech.
3252 I've just tested the patch: Authentication still works fine,
3253 unavailable mechs specified in the exim configuration are still
3254 caught, and the auth.log warnings about OTP are gone."
3256 PH/25 When debugging is enabled, the contents of the command line are added
3257 to the debugging output, even when log_selector=+arguments is not
3260 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3261 answer is "GNU", and only if the return is "GNU/something" is the answer
3264 PH/27 $acl_verify_message is now set immediately after the failure of a
3265 verification in an ACL, and so is available in subsequent modifiers. In
3266 particular, the message can be preserved by coding like this:
3268 warn !verify = sender
3269 set acl_m0 = $acl_verify_message
3271 Previously, $acl_verify_message was set only while expanding "message"
3272 and "log_message" when a very denied access.
3274 PH/28 Modified OS/os.c-Linux with
3276 -#ifndef OS_LOAD_AVERAGE
3277 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3279 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3280 nomenclature these days.)
3282 PH/29 Installed patch from the Sieve maintainer that adds the options
3283 sieve_useraddress and sieve_subaddress to the redirect router.
3285 PH/30 In these circumstances:
3286 . Two addresses routed to the same list of hosts;
3287 . First host does not offer TLS;
3288 . First host accepts first address;
3289 . First host gives temporary error to second address;
3290 . Second host offers TLS and a TLS session is established;
3291 . Second host accepts second address.
3292 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3293 and peerdn, if requested) that were in fact used only for the second
3296 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3297 attention to any local part prefix or suffix that was matched by the
3298 router that accepted the address. It now behaves in the same way as it
3299 does for delivery: the affixes are removed from the local part unless
3300 rcpt_include_affixes is set on the transport.
3302 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3303 timeout during the DATA phase of an incoming message.
3305 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3306 applied a patch sanctioned by the Sieve maintainer.
3308 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3309 the uid or gid is negative. A case of a negative gid caused this to be
3310 noticed. The fix allows for either to be negative.
3312 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3313 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3316 PH/36 The change PH/12 above was broken. Fixed it.
3318 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3319 the grounds that routing the same address twice would always produce the
3320 same answer. This might have been true once, but it is certainly no
3321 longer true now. Routing a child address may depend on the previous
3322 routing that produced that child. Some complicated redirection strategies
3323 went wrong when messages had multiple recipients, and made Exim's
3324 behaviour dependent on the order in which the addresses were given.
3326 I have moved the duplicate checking until after the routing is complete.
3327 Exim scans the addresses that are assigned to local and remote
3328 transports, and removes any duplicates. This means that more work will be
3329 done, as duplicates will always all be routed, but duplicates are
3330 presumably rare, so I don't expect this is of any significance.
3332 For deliveries to pipes, files, and autoreplies, the duplicate checking
3333 still happens during the routing process, since they are not going to be
3336 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3337 It corrects a timeout issue with spamd. This is Ian's comment: "The
3338 background is that sometimes spamd either never reads data from a
3339 connection it has accepted, or it never writes response data. The exiscan
3340 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3341 blindly assumes that writes won't block so it may never time out."
3343 PH/39 Allow G after quota size as well as K and M.
3345 PH/40 The value set for $authenticated_id in an authenticator may not contain
3346 binary zeroes or newlines because the value is written to log lines and
3347 to spool files. There was no check on this. Now the value is run through
3348 the string_printing() function so that such characters are converted to
3349 printable escape sequences.
3351 PH/41 $message_linecount is a new variable that contains the total number of
3352 lines in the message. Compare $body_linecount, which is the count for the
3355 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3356 bounce and delay warning messages, except in certain special cases, which
3359 (a) An SMTP error message from a remote host;
3360 (b) A message specified in a :fail: redirection;
3361 (c) A message specified in a "fail" command in a system filter;
3362 (d) A message specified in a FAIL return from the queryprogram router;
3363 (e) A message specified by the cannot_route_message router option.
3365 In these cases only, Exim does include the error details in bounce and
3366 warning messages. There are also a few cases where bland messages such
3367 as "unrouteable address" or "local delivery error" are given.
3369 PH/43 $value is now also set for the "else" part of a ${run expansion.
3371 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3372 being worked on, but at least Exim now implements the latest version to
3375 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3376 process to complete was treated as a delivery failure, a timeout while
3377 writing the message to the pipe was logged, but erroneously treated as a
3378 successful delivery. Such timeouts include transport filter timeouts. For
3379 consistency with the overall process timeout, these timeouts are now
3380 treated as errors, giving rise to delivery failures by default. However,
3381 there is now a new Boolean option for the pipe transport called
3382 timeout_defer, which, if set TRUE, converts the failures into defers for
3383 both kinds of timeout. A transport filter timeout is now identified in
3386 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3387 systems where "make" and "gmake" are different, calling "gmake" at top
3388 level broke things. I've arranged for the value of $(MAKE) to be passed
3389 from the Makefile to this script so that it can call the same version of
3393 A note about Exim versions 4.44 and 4.50
3394 ----------------------------------------
3396 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3397 changes of various kinds. As a consequence, a big documentation update was
3398 needed. This delayed the release for rather longer than seemed good, especially
3399 in the light of a couple of (minor) security issues. Therefore, the changes
3400 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3401 release. So 4.44 and 4.50 are in effect two different branches that both start
3404 I have left the 4.50 change log unchanged; it contains all the changes since
3405 4.43. The change log for 4.44 is below; many of its items are identical to
3406 those for 4.50. This seems to be the most sensible way to preserve the
3407 historical information.
3413 1. Minor wording change to the doc/README.SIEVE file.
3415 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3416 computation of the current number of files was incorrect.
3418 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3419 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3422 4. Give more explanation in the error message when the command for a transport
3423 filter fails to execute.
3425 5. There are several places where Exim runs a non-Exim command in a
3426 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3427 was being done only for the command run by the queryprogram router. It is
3428 now done for all such subprocesses. The other cases are: ${run, transport
3429 filters, and the commands run by the lmtp and pipe transports.
3431 6. Added CONFIGURE_GROUP build-time option.
3433 7. Some older OS have a limit of 256 on the maximum number of file
3434 descriptors. Exim was using setrlimit() to set 1000 as a large value
3435 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3436 systems. I've change it so that if it can't get 1000, it tries for 256.
3438 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3439 was an oversight, and furthermore, ever since the addition of extra
3440 controls (e.g. 4.43/32), the checks on when to allow different forms of
3441 "control" were broken. There should now be diagnostics for all cases when a
3442 control that does not make sense is encountered.
3444 9. Added the /retain_sender option to "control=submission".
3446 10. $recipients is now available in the predata ACL (oversight).
3448 11. Tidy the search cache before the fork to do a delivery from a message
3449 received from the command line. Otherwise the child will trigger a lookup
3450 failure and thereby defer the delivery if it tries to use (for example) a
3451 cached ldap connection that the parent has called unbind on.
3453 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3454 of $address_data from the recipient verification was clobbered by the
3455 sender verification.
3457 13. The value of address_data from a sender verification is now available in
3458 $sender_address_data in subsequent conditions in the ACL statement.
3460 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3462 15. Added a new option "connect=<time>" to callout options, to set a different
3465 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3466 was its contents. (It was OK if the option was not defined at all.)
3468 17. A "Completed" log line is now written for messages that are removed from
3469 the spool by the -Mrm option.
3471 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3472 information about exactly what failed.
3474 19. Added -dd to debug only the daemon process.
3476 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3477 handles timeouts, both on the server side and network timeouts. Renamed the
3478 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3480 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3481 It is now set to "smtps".
3483 22. $host_address is now set to the target address during the checking of
3484 ignore_target_hosts.
3486 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3487 being passed; this would have caused $sender_host_name to have been used if
3488 matching the list had actually called for a host name (not very likely,
3489 since this list is usually IP addresses). A host name is now passed as
3492 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3493 code that specifies a non-restarting handler (typically sigaction() in
3494 modern systems) in an attempt to fix a rare and obscure crash bug.
3496 25. Narrowed the window for a race in the daemon that could cause it to ignore
3497 SIGCHLD signals. This is not a major problem, because they are used only to
3498 wake it up if nothing else does.
3500 26. A malformed maildirsize file could cause Exim to calculate negative values
3501 for the mailbox size or file count. Odd effects could occur as a result.
3502 The maildirsize information is now recalculated if the size or filecount
3505 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3506 support for a long time. Removed HAVE_SYS_VFS_H.
3508 28. Installed the latest version of exipick from John Jetmore.
3510 29. In an address list, if the pattern was not a regular expression, an empty
3511 subject address (from a bounce message) matched only if the pattern was an
3512 empty string. Non-empty patterns were not even tested. This was the wrong
3513 because it is perfectly reasonable to use an empty address as part of a
3514 database query. An empty address is now tested by patterns that are
3515 lookups. However, all the other forms of pattern expect the subject to
3516 contain a local part and a domain, and therefore, for them, an empty
3517 address still always fails if the pattern is not itself empty.
3519 30. Exim went into a mad DNS loop when attempting to do a callout where the
3520 host was specified on an smtp transport, and looking it up yielded more
3521 than one IP address.
3523 31. Re-factored the code for checking spool and log partition space into a
3524 function that finds that data and another that does the check. The former
3525 is then used to implement four new variables: $spool_space, $log_space,
3526 $spool_inodes, and $log_inodes.
3528 32. The RFC2047 encoding function was originally intended for short strings
3529 such as real names; it was not keeping to the 75-character limit for
3530 encoded words that the RFC imposes. It now respects the limit, and
3531 generates multiple encoded words if necessary. To be on the safe side, I
3532 have increased the buffer size for the ${rfc2047: expansion operator from
3535 33. It is now permitted to omit both strings after an "if" condition; if the
3536 condition is true, the result is "true". As before, when the second string
3537 is omitted, a false condition yields an empty string. This makes it less
3538 cumbersome to write custom ACL and router conditions.
3540 34. Failure to deliver a bounce message always caused it to be frozen, even if
3541 there was an errors_to setting on the router. The errors_to setting is now
3544 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3545 canonical form (fully expanded) before being placed in
3546 $sender_host_address.
3548 36. The table in the code that translates DNS record types into text (T_A to
3549 "A" for instance) was missing entries for NS and CNAME. It is just possible
3550 that this could have caused confusion if both these types were looked up
3551 for the same domain, because the text type is used as part of Exim's
3552 per-process caching. But the chance of anyone hitting this buglet seems
3555 37. The dnsdb lookup has been extended in a number of ways.
3557 (1) There is a new type, "zns", which walks up the domain tree until it
3558 finds some nameserver records. It should be used with care.
3560 (2) There is a new type, "mxh", which is like "mx" except that it returns
3561 just the host names, not the priorities.
3563 (3) It is now possible to give a list of domains (or IP addresses) to be
3564 looked up. The behaviour when one of the lookups defers can be
3565 controlled by a keyword.
3567 (4) It is now possible to specify the separator character for use when
3568 multiple records are returned.
3570 38. The dnslists ACL condition has been extended: it is now possible to supply
3571 a list of IP addresses and/or domains to be looked up in a particular DNS
3574 39. Added log_selector=+queue_time_overall.
3576 40. When running the queue in the test harness, wait just a tad after forking a
3577 delivery process, to get repeatability of debugging output.
3579 41. Include certificate and key file names in error message when GnuTLS fails
3580 to set them up, because the GnuTLS error message doesn't include the name
3581 of the failing file when there is a problem reading it.
3583 42. Allow both -bf and -bF in the same test run.
3585 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3587 44. The "Exiscan patch" is now merged into the mainline Exim source.
3589 45. Sometimes the final signoff response after QUIT could fail to get
3590 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3591 < 0 before doing a fflush(). This bug looks as though it goes back to the
3592 introduction of TLS in release 3.20, but "sometimes" must have been rare
3593 because the tests only now provoked it.
3595 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3596 (this can affect the format of dates).
3598 47. exim_tidydb, when checking for the continued existence of a message for
3599 which it has found a message-specific retry record, was not finding
3600 messages that were in split spool directories. Consequently, it was
3601 deleting retry records that should have stayed in existence.
3603 48. Steve fixed some bugs in eximstats.
3605 49. The SPA authentication driver was not abandoning authentication and moving
3606 on to the next authenticator when an expansion was forced to fail,
3607 contradicting the general specification for all authenticators. Instead it
3608 was generating a temporary error. It now behaves as specified.
3610 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3611 (the order specifies the preference for clients). The order is now AES256,
3612 AES128, 3DES, ARCFOUR128.
3614 51. Small patch to Sieve code - explicitly set From: when generating an
3617 52. Exim crashed if a remote delivery caused a very long error message to be
3618 recorded - for instance if somebody sent an entire SpamAssassin report back
3619 as a large number of 550 error lines. This bug was coincidentally fixed by
3620 increasing the size of one of Exim's internal buffers (big_buffer) that
3621 happened as part of the Exiscan merge. However, to be on the safe side, I
3622 have made the code more robust (and fixed the comments that describe what
3625 53. Now that there can be additional text after "Completed" in log lines (if
3626 the queue_time_overall log selector is set), a one-byte patch to exigrep
3627 was needed to allow it to recognize "Completed" as not the last thing in
3630 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3631 patch that reportedly fixes this has been added. I am not expert enough to
3632 create a test for it. This is what the patch creator wrote:
3634 "I found a little strange behaviour of ldap code when working with
3635 Windows 2003 AD Domain, where users was placed in more than one
3636 Organization Units. When I tried to give exim partial DN, the exit code
3637 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3638 But simultaneously result of request was absolutely normal ldap result,
3639 so I produce this patch..."
3641 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3642 REFERENCE, so I have modified the code to exclude the patch when that macro
3645 55. Some experimental protocols are using DNS PTR records for new purposes. The
3646 keys for these records are domain names, not reversed IP addresses. The
3647 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3648 leaves it alone. Component reversal etc. now happens only for IP addresses.
3651 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3653 57. Double the size of the debug message buffer (to 2048) so that more of very
3654 long debug lines gets shown.
3656 58. The exicyclog utility now does better if the number of log files to keep
3657 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3659 59. Two changes related to the smtp_active_hostname option:
3661 (1) $smtp_active_hostname is now available as a variable.
3662 (2) The default for smtp_banner uses $smtp_active_hostname instead
3663 of $primary_hostname.
3665 60. The host_aton() function is supposed to be passed a string that is known
3666 to be a valid IP address. However, in the case of IPv6 addresses, it was
3667 not checking this. This is a hostage to fortune. Exim now panics and dies
3668 if the condition is not met. A case was found where this could be provoked
3669 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3670 components; fortuitously, this particular loophole had already been fixed
3671 by change 4.50/55 above.
3673 If there are any other similar loopholes, the new check in host_aton()
3674 itself should stop them being exploited. The report I received stated that
3675 data on the command line could provoke the exploit when Exim was running as
3676 exim, but did not say which command line option was involved. All I could
3677 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3678 running as the user.
3681 61. There was a buffer overflow vulnerability in the SPA authentication code
3682 (which came originally from the Samba project). I have added a test to the
3683 spa_base64_to_bits() function which I hope fixes it.
3686 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3687 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3689 63. The daemon start-up calls getloadavg() while still root for those OS that
3690 need the first call to be done as root, but it missed one case: when
3691 deliver_queue_load_max is set with deliver_drop_privilege. This is
3692 necessary for the benefit of the queue runner, because there is no re-exec
3693 when deliver_drop_privilege is set.
3695 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3696 This has been fixed.
3698 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3699 was in use, was not putting the data itself into the right store pool;
3700 consequently, it could be overwritten for a subsequent message in the same
3701 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3704 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3706 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3707 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3708 the function were treating the return as a boolean value, which happened to
3709 work because 0=false and not-0=true, but is not correct code.
3711 68. The host_aton() function was not handling scoped IPv6 addresses (those
3712 with, for example, "%eth0" on the end) correctly.
3714 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3715 negated items (that is, ~something) in unsigned ints. Some compilers
3716 apparently mutter when there is no cast.
3718 70. If an address verification called from an ACL failed, and did not produce a
3719 user-specific message (i.e. there was only a "system" message), nothing was
3720 put in $acl_verify_message. In this situation, it now puts the system
3723 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3724 session; change 4.31/43 added the unwanted input to the log line - except
3725 that it did not do this in the start of session case. It now does.
3727 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3728 This is wrong and can cause the other end to generate a synchronization
3729 error if it is another Exim or anything else that does the synchronization
3730 check. A QUIT command is no longer sent after a timeout.
3732 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3733 during host lookups.
3735 74. The defer_ok option of callout verification was not working if it was used
3736 when verifying addresses in header lines, that is, for this case:
3738 verify = header_sender/callout=defer_ok
3740 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3741 those file descriptors could be used for SMTP connections. If anything
3742 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3743 could be sent to the SMTP client, causing chaos. The daemon now opens
3744 stdin, stdout, and stderr to /dev/null when it puts itself into the
3747 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3748 log by default. The user can override this with suitable Perl magic.
3750 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3751 the log message when discard triggers, was not working for the DATA ACL or
3752 for the non-SMTP ACL.
3754 78. Error message wording change in sieve.c.
3756 79. If smtp_accept_max_per_host was set, the number of connections could be
3757 restricted to fewer than expected, because the daemon was trying to set up
3758 a new connection before checking whether the processes handling previous
3759 connections had finished. The check for completed processes is now done
3760 earlier. On busy systems, this bug wouldn't be noticed because something
3761 else would have woken the daemon, and it would have reaped the completed
3764 80. If a message was submitted locally by a user whose login name contained one
3765 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3766 It caused a spool format error. I have fixed the spool reading code. A
3767 related problem was that the "from" clause in the Received: line became
3768 illegal because of the space(s). It is now covered by ${quote_local_part.
3770 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3773 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3775 83. Patch from Sieve maintainer to fix unterminated string problem in
3776 "vacation" handling.
3778 84. Some minor changes to the Linux configuration files to help with other
3779 OS variants using glibc.
3781 85. One more patch for Sieve to update vacation handling to latest spec.
3784 ----------------------------------------------------
3785 See the note above about the 4.44 and 4.50 releases.
3786 ----------------------------------------------------
3792 1. Change 4.43/35 introduced a bug that caused file counts to be
3793 incorrectly computed when quota_filecount was set in an appendfile
3796 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3797 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3800 3. Give more explanation in the error message when the command for a transport
3801 filter fails to execute.
3803 4. There are several places where Exim runs a non-Exim command in a
3804 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3805 was being done only for the command run by the queryprogram router. It is
3806 now done for all such subprocesses. The other cases are: ${run, transport
3807 filters, and the commands run by the lmtp and pipe transports.
3809 5. Some older OS have a limit of 256 on the maximum number of file
3810 descriptors. Exim was using setrlimit() to set 1000 as a large value
3811 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3812 systems. I've change it so that if it can't get 1000, it tries for 256.
3814 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3815 was an oversight, and furthermore, ever since the addition of extra
3816 controls (e.g. 4.43/32), the checks on when to allow different forms of
3817 "control" were broken. There should now be diagnostics for all cases when a
3818 control that does not make sense is encountered.
3820 7. $recipients is now available in the predata ACL (oversight).
3822 8. Tidy the search cache before the fork to do a delivery from a message
3823 received from the command line. Otherwise the child will trigger a lookup
3824 failure and thereby defer the delivery if it tries to use (for example) a
3825 cached ldap connection that the parent has called unbind on.
3827 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3828 of $address_data from the recipient verification was clobbered by the
3829 sender verification.
3831 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3832 was its contents. (It was OK if the option was not defined at all.)
3834 11. A "Completed" log line is now written for messages that are removed from
3835 the spool by the -Mrm option.
3837 12. $host_address is now set to the target address during the checking of
3838 ignore_target_hosts.
3840 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3841 being passed; this would have caused $sender_host_name to have been used if
3842 matching the list had actually called for a host name (not very likely,
3843 since this list is usually IP addresses). A host name is now passed as
3846 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3847 code that specifies a non-restarting handler (typically sigaction() in
3848 modern systems) in an attempt to fix a rare and obscure crash bug.
3850 15. Narrowed the window for a race in the daemon that could cause it to ignore
3851 SIGCHLD signals. This is not a major problem, because they are used only to
3852 wake it up if nothing else does.
3854 16. A malformed maildirsize file could cause Exim to calculate negative values
3855 for the mailbox size or file count. Odd effects could occur as a result.
3856 The maildirsize information is now recalculated if the size or filecount
3859 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3860 support for a long time. Removed HAVE_SYS_VFS_H.
3862 18. Updated exipick to current release from John Jetmore.
3864 19. Allow an empty sender to be matched against a lookup in an address list.
3865 Previously the only cases considered were a regular expression, or an
3868 20. Exim went into a mad DNS lookup loop when doing a callout where the
3869 host was specified on the transport, if the DNS lookup yielded more than
3872 21. The RFC2047 encoding function was originally intended for short strings
3873 such as real names; it was not keeping to the 75-character limit for
3874 encoded words that the RFC imposes. It now respects the limit, and
3875 generates multiple encoded words if necessary. To be on the safe side, I
3876 have increased the buffer size for the ${rfc2047: expansion operator from
3879 22. Failure to deliver a bounce message always caused it to be frozen, even if
3880 there was an errors_to setting on the router. The errors_to setting is now
3883 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3884 canonical form (fully expanded) before being placed in
3885 $sender_host_address.
3887 24. Updated eximstats to version 1.33
3889 25. Include certificate and key file names in error message when GnuTLS fails
3890 to set them up, because the GnuTLS error message doesn't include the name
3891 of the failing file when there is a problem reading it.
3893 26. Expand error message when OpenSSL has problems setting up cert/key files.
3896 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3897 (this can affect the format of dates).
3899 28. exim_tidydb, when checking for the continued existence of a message for
3900 which it has found a message-specific retry record, was not finding
3901 messages that were in split spool directories. Consequently, it was
3902 deleting retry records that should have stayed in existence.
3904 29. eximstats updated to version 1.35
3905 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3906 1.35 - bugfix such that pie charts by volume are generated correctly
3908 30. The SPA authentication driver was not abandoning authentication and moving
3909 on to the next authenticator when an expansion was forced to fail,
3910 contradicting the general specification for all authenticators. Instead it
3911 was generating a temporary error. It now behaves as specified.
3913 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3914 (the order specifies the preference for clients). The order is now AES256,
3915 AES128, 3DES, ARCFOUR128.
3917 31. Small patch to Sieve code - explicitly set From: when generating an
3920 32. Exim crashed if a remote delivery caused a very long error message to be
3921 recorded - for instance if somebody sent an entire SpamAssassin report back
3922 as a large number of 550 error lines. This bug was coincidentally fixed by
3923 increasing the size of one of Exim's internal buffers (big_buffer) that
3924 happened as part of the Exiscan merge. However, to be on the safe side, I
3925 have made the code more robust (and fixed the comments that describe what
3928 33. Some experimental protocols are using DNS PTR records for new purposes. The
3929 keys for these records are domain names, not reversed IP addresses. The
3930 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3931 leaves it alone. Component reversal etc. now happens only for IP addresses.
3934 34. The host_aton() function is supposed to be passed a string that is known
3935 to be a valid IP address. However, in the case of IPv6 addresses, it was
3936 not checking this. This is a hostage to fortune. Exim now panics and dies
3937 if the condition is not met. A case was found where this could be provoked
3938 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3939 components; fortuitously, this particular loophole had already been fixed
3940 by change 4.50/55 or 4.44/33 above.
3942 If there are any other similar loopholes, the new check in host_aton()
3943 itself should stop them being exploited. The report I received stated that
3944 data on the command line could provoke the exploit when Exim was running as
3945 exim, but did not say which command line option was involved. All I could
3946 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3947 running as the user.
3950 35. There was a buffer overflow vulnerability in the SPA authentication code
3951 (which came originally from the Samba project). I have added a test to the
3952 spa_base64_to_bits() function which I hope fixes it.
3955 36. The daemon start-up calls getloadavg() while still root for those OS that
3956 need the first call to be done as root, but it missed one case: when
3957 deliver_queue_load_max is set with deliver_drop_privilege. This is
3958 necessary for the benefit of the queue runner, because there is no re-exec
3959 when deliver_drop_privilege is set.
3961 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3962 was in use, was not putting the data itself into the right store pool;
3963 consequently, it could be overwritten for a subsequent message in the same
3964 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3967 38. Sometimes the final signoff response after QUIT could fail to get
3968 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3969 < 0 before doing a fflush(). This bug looks as though it goes back to the
3970 introduction of TLS in release 3.20, but "sometimes" must have been rare
3971 because the tests only now provoked it.
3977 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3978 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3979 Now it returns an integer. A number of calls were still expecting a T/F
3980 return. Fortuitously, in all cases, the tests worked in OK situations,
3981 which is the norm. However, things would have gone wrong on any write
3982 failures on the smtp file descriptor. This function is used when sending
3983 messages over SMTP and also when doing verify callouts.
3985 2. When Exim is called to do synchronous delivery of a locally submitted
3986 message (the -odf or -odi options), it no longer closes stderr before doing
3989 3. Implemented the mua_wrapper option.
3991 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3993 5. Implemented the functions header_remove(), header_testname(),
3994 header_add_at_position(), and receive_remove_recipient(), and exported them
3997 6. If an ACL "warn" statement specified the addition of headers, Exim already
3998 inserted X-ACL-Warn: at the start if there was no header name. However, it
3999 was not making this test for the second and subsequent header lines if
4000 there were newlines in the string. This meant that an invalid header could
4001 be inserted if Exim was badly configured.
4003 7. Allow an ACL "warn" statement to add header lines at the start or after all
4004 the Received: headers, as well as at the end.
4006 8. Added the rcpt_4xx retry error code.
4008 9. Added postmaster_mailfrom=xxx to callout verification option.
4010 10. Added mailfrom=xxxx to the callout verification option, for verify=
4013 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4014 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4016 12. Inserted some casts to stop certain compilers warning when using pointer
4017 differences as field lengths or precisions in printf-type calls (mostly
4018 affecting debugging statements).
4020 13. Added optional readline() support for -be (dynamically loaded).
4022 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4023 same clock tick as a message's arrival, so that its received time was the
4024 same as the "first fail" time on the retry record, and that message
4025 remained on the queue past the ultimate address timeout, every queue runner
4026 would try a delivery (because it was past the ultimate address timeout) but
4027 after another failure, the ultimate address timeout, which should have then
4028 bounced the address, did not kick in. This was a "< instead of <=" error;
4029 in most cases the first failure would have been in the next clock tick
4030 after the received time, and all would be well.
4032 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4033 being recognized when the domain list was tested by the match_domain
4034 condition in an expansion string.
4036 16. Added the ${str2b64: operator.
4038 17. Exim was always calling setrlimit() to set a large limit for the number of
4039 processes, without checking whether the existing limit was already
4040 adequate. (It did check for the limit on file descriptors.) Furthermore,
4041 errors from getrlimit() and setrlimit() were being ignored. Now they are
4042 logged to the main and panic logs, but Exim does carry on, to try to do its
4043 job under whatever limits there are.
4045 18. Imported PCRE 5.0.
4047 19. Trivial typo in log message " temporarily refused connection" (the leading
4050 20. If the log selector return_path_on_delivery was set and an address was
4051 redirected to /dev/null, the delivery process crashed because it assumed
4052 that a return path would always be set for a "successful" delivery. In this
4053 case, the whole delivery is bypassed as an optimization, and therefore no
4056 21. Internal re-arrangement: the function for sending a challenge and reading
4057 a response while authentication was assuming a zero-terminated challenge
4058 string. It's now changed to take a pointer and a length, to allow for
4059 binary data in such strings.
4061 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4063 23. Exim was not respecting finduser_retries when seeking the login of the
4064 uid under which it was called; it was always trying 10 times. (The default
4065 setting of finduser_retries is zero.) Also, it was sleeping after the final
4066 failure, which is pointless.
4068 24. Implemented tls_on_connect_ports.
4070 25. Implemented acl_smtp_predata.
4072 26. If the domain in control=submission is set empty, Exim assumes that the
4073 authenticated id is a complete email address when it generates From: or
4074 Sender: header lines.
4076 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4077 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4078 chown and chgrp in /bin and hostname in /usr/bin.
4080 28. Exim was keeping the "process log" file open after each use, just as it
4081 does for the main log. This opens the possibility of it remaining open for
4082 long periods when the USR1 signal hits a daemon. Occasional processlog
4083 errors were reported, that could have been caused by this. Anyway, it seems
4084 much more sensible not to leave this file open at all, so that is what now
4087 29. The long-running daemon process does not normally write to the log once it
4088 has entered its main loop, and it closes the log before doing so. This is
4089 so that log files can straightforwardly be renamed and moved. However,
4090 there are a couple of unusual error situations where the daemon does write
4091 log entries, and I had neglected to close the log afterwards.
4093 30. The text of an SMTP error response that was received during a remote
4094 delivery was being truncated at 512 bytes. This is too short for some of
4095 the long messages that one sometimes sees. I've increased the limit to
4098 31. It is now possible to make retry rules that apply only when a message has a
4099 specific sender, in particular, an empty sender.
4101 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4102 it possible to be selective about when SMTP synchronization is enforced.
4104 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4106 32. Implemented hosts_connection_nolog.
4108 33. Added an ACL for QUIT.
4110 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4113 35. Added mailbox_size and mailbox_filecount to appendfile.
4115 36. Added control = no_multiline_responses to ACLs.
4117 37. There was a bug in the logic of the code that waits for the clock to tick
4118 in the case where the clock went backwards by a substantial amount such
4119 that the microsecond fraction of "now" was more than the microsecond
4120 fraction of "then" (but the whole seconds number was less).
4122 38. Added support for the libradius Radius client library this is found on
4123 FreeBSD (previously only the radiusclient library was supported).
4129 1. When certain lookups returned multiple values in the form name=value, the
4130 quoting of the values was not always being done properly. Specifically:
4131 (a) If the value started with a double quote, but contained no whitespace,
4133 (b) If the value contained whitespace other than a space character (i.e.
4134 tabs or newlines or carriage returns) it was not quoted.
4135 This fix has been applied to the mysql and pgsql lookups by writing a
4136 separate quoting function and calling it from the lookup code. The fix
4137 should probably also be applied to nisplus, ibase and oracle lookups, but
4138 since I cannot test any of those, I have not disturbed their existing code.
4140 2. A hit in the callout cache for a specific address caused a log line with no
4141 reason for rejecting RCPT. Now it says "Previous (cached) callout
4142 verification failure".
4144 3. There was an off-by-one bug in the queryprogram router. An over-long
4145 return line was truncated at 256 instead of 255 characters, thereby
4146 overflowing its buffer with the terminating zero. As well as fixing this, I
4147 have increased the buffer size to 1024 (and made a note to document this).
4149 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4150 when Exim is waiting for an SMTP response from a remote server, Exim
4151 restarts its select() call on the socket, thereby resetting its timeout.
4152 This is not a problem when such interrupts are rare. Somebody set up a cron
4153 job to run exiwhat every 2 minutes, which is less than the normal select()
4154 timeout (5 or 10 minutes). This meant that the select() timeout never
4155 kicked in because it was always reset. I have fixed this by comparing the
4156 time when an interrupt arrives with the time at the start of the first call
4157 to select(). If more time than the timeout has elapsed, the interrupt is
4158 treated as a timeout.
4160 5. Some internal re-factoring in preparation for the addition of Sieve
4161 extensions (by MH). In particular, the "personal" test is moved to a
4162 separate function, and given an option for scanning Cc: and Bcc: (which is
4163 not set for Exim filters).
4165 6. When Exim created an email address using the login of the caller as the
4166 local part (e.g. when creating a From: or Sender: header line), it was not
4167 quoting the local part when it contained special characters such as @.
4169 7. Installed new OpenBSD configuration files.
4171 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4172 try to make them clearer.
4174 9. Callout options, other than the timeout value, were being ignored when
4175 verifying sender addresses in header lines. For example, when using
4177 verify = header_sender/callout=no_cache
4179 the cache was (incorrectly) being used.
4181 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4182 only the Cygwin environment.
4184 11. When return_path_on_delivery was set as a log selector, if different remote
4185 addresses in the same message used different return paths and parallel
4186 remote delivery occurred, the wrong values would sometimes be logged.
4187 (Whenever a remote delivery process finished, the return path value from
4188 the most recently started remote delivery process was logged.)
4190 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4191 lines when AUTH and/or TLS are in use. This is the "received protocol"
4192 field. Exim used to use "asmtp" for authenticated SMTP, without any
4193 indication (in the protocol name) for TLS use. Now it follows the RFC and
4194 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4195 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4196 names appear in log lines as well as in Received: header lines.
4198 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4199 extensions, and comparison tests, and to fix some bugs.
4201 14. Changes to the "personal" filter test:
4203 (1) The test was buggy in that it was just doing the equivalent of
4204 "contains" tests on header lines. For example, if a user's address was
4205 anne@some.where, the "personal" test would incorrectly be true for
4207 To: susanne@some.where
4209 This test is now done by extracting each address from the header in turn,
4210 and checking the entire address. Other tests that are part of "personal"
4211 are now done using regular expressions (for example, to check local parts
4212 of addresses in From: header lines).
4214 (2) The list of non-personal local parts in From: addresses has been
4215 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4216 taken from the Sieve specification recommendations.
4218 (3) If the message contains any header line starting with "List-" it is
4219 treated as non-personal.
4221 (4) The test for "circular" in the Subject: header line has been removed
4222 because it now seems ill-conceived.
4224 15. Minor typos in src/EDITME comments corrected.
4226 16. Installed latest exipick from John Jetmore.
4228 17. If headers_add on a router specified a text string that was too long for
4229 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4230 of string_sprintf() is now avoided.
4232 18. $message_body_size was not set (it was always zero) when running the DATA
4233 ACL and the local_scan() function.
4235 19. For the "mail" command in an Exim filter, no default was being set for
4236 the once_repeat time, causing a random time value to be used if "once" was
4237 specified. (If the value happened to be <= 0, no repeat happened.) The
4238 default is now 0s, meaning "never repeat". The "vacation" command was OK
4239 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4240 (I found it when inspecting the code).
4242 20. There is now an overall timeout for performing a callout verification. It
4243 defaults to 4 times the callout timeout, which applies to individual SMTP
4244 commands during the callout. The overall timeout applies when there is more
4245 than one host that can be tried. The timeout is checked before trying the
4246 next host. This prevents very long delays if there are a large number of
4247 hosts and all are timing out (e.g. when the network connections are timing
4248 out). The value of the overall timeout can be changed by specifying an
4249 additional sub-option for "callout", called "maxwait". For example:
4251 verify = sender/callout=5s,maxwait=20s
4253 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4254 to the end before writing, but this should make it even safer).
4256 22. Exim was forgetting that it had advertised PIPELINING for the second and
4257 subsequent messages on an SMTP connection. It was also not resetting its
4258 memory on STARTTLS and an internal HELO.
4260 23. When Exim logs an SMTP synchronization error within a session, it now
4261 records whether PIPELINING has been advertised or not.
4263 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4264 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4265 rather than long int.
4267 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4269 26. Added the never_mail option to autoreply.
4275 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4276 crash if the getsockname() call failed; this can happen if a connection is
4277 closed very soon after it is established. The problem was simply in the
4278 order in which certain operations were done, causing Exim to try to write
4279 to the SMTP stream before it had set up the file descriptor. The bug has
4280 been fixed by making things happen in the correct order.
4286 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4287 before the connection was closed, thus losing the rejection response.
4289 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4290 some early Solaris releases, but causes trouble in current releases where
4291 socklen_t is defined.
4293 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4296 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4299 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4300 that releasing the top of it at the end releases what was used for sub-
4301 expansions (unless the block got too big). However, discard this block if
4302 the first thing is a variable or header, so that we can use its block when
4303 it is dynamic (useful for very large $message_headers, for example).
4305 6. Lookups now cache *every* query, not just the most recent. A new, separate
4306 store pool is used for this. It can be recovered when all lookup caches are
4307 flushed. Lookups now release memory at the end of their result strings.
4308 This has involved some general refactoring of the lookup sources.
4310 7. Some code has been added to the store_xxx() functions to reduce the amount
4311 of flapping under certain conditions.
4313 8. log_incoming_interface used to affect only the <= reception log lines. Now
4314 it causes the local interface and port to be added to several more SMTP log
4315 lines, for example "SMTP connection from", and rejection lines.
4317 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4319 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4321 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4322 could be overwritten at the end of the current message (or the start of a
4323 new message if it was set in a HELO ACL). The value is now preserved for
4324 the duration of the SMTP connection.
4326 12. If a transport had a headers_rewrite setting, and a matching header line
4327 contained an unqualified address, that address was qualified, even if it
4328 did not match any rewriting rules. The underlying bug was that the values
4329 of the flags that permit the existence of unqualified sender and recipient
4330 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4331 non-local messages, and by -bnq for local messages) were not being
4332 preserved with the message after it was received.
4334 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4335 "next input=" as part of the text comprising the host identity instead of
4336 the correct text. The code was using the same buffer for two different
4337 strings. However, depending on which order the printing function evaluated
4338 its arguments, the bug did not always show up. Under Linux, for example, my
4339 test suite worked just fine.
4341 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4342 doesn't work with some older versions of Perl. It has been changed to "my",
4343 which in any case is probably the better facility to use.
4345 15. A really picky compiler found some instances of statements for creating
4346 error messages that either had too many or two few arguments for the format
4349 16. The size of the buffer for calls to the DNS resolver has been increased
4350 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4351 for addresses that have a lot of PTR records. This alleviates a problem; it
4352 does not fully solve it.
4354 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4355 buffer now truncates the list and logs the incident, which is the same
4356 action as happens when Exim is looking up a host name and its aliases.
4357 Previously in this situation something unpredictable would happen;
4358 sometimes it was "internal error: store_reset failed".
4360 18. If a server dropped the connection unexpectedly when an Exim client was
4361 using GnuTLS and trying to read a response, the client delivery process
4362 crashed while trying to generate an error log message.
4364 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4365 string, for example:
4367 warn message = H1: something\nH2: something
4369 the text was added as a single header line from Exim's point of view
4370 though it ended up OK in the delivered message. However, searching for the
4371 second and subsequent header lines using $h_h2: did not work. This has been
4372 fixed. Similarly, if a system filter added multiple headers in this way,
4373 the routers could not see them.
4375 20. Expanded the error message when iplsearch is called with an invalid key to
4376 suggest using net-iplsearch in a host list.
4378 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4379 ACLs are no longer actually imposed (and a message to that effect is
4382 22. If a "gecos" field in a passwd entry contained escaped characters, in
4383 particular, if it contained a \" sequence, Exim got it wrong when building
4384 a From: or a Sender: header from that name. A second bug also caused
4385 incorrect handling when an unquoted " was present following a character
4386 that needed quoting.
4388 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4389 was not being matched caselessly.
4391 24. Arranged for all hyphens in the exim.8 source to be escaped with
4394 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4395 properly. Recipient callout cache records were still being keyed to include
4396 the sender, even when use_sender was set false. This led to far more
4397 callouts that were necessary. The sender is no longer included in the key
4398 when use_sender is false.
4400 26. Added "control = submission" modifier to ACLs.
4402 27. Added the ${base62d: operator to decode base 62 numbers.
4404 28. dnsdb lookups can now access SRV records.
4406 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4407 the configuration file.
4409 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4410 (-v) mode. This makes the output for a verbose queue run more intelligible.
4412 31. Added a use_postmaster feature to recipient callouts.
4414 32. Added the $body_zerocount variable, containing the number of binary zero
4415 bytes in the message body.
4417 33. The time of last modification of the "new" subdirectory is now used as the
4418 "mailbox time last read" when there is a quota error for a maildir
4421 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4423 35. Added +ignore_unknown as a special item in host lists.
4425 36. Code for decoding IPv6 addresses in host lists is now included, even if
4426 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4427 address was recognized as an IP address, but was then not correctly decoded
4428 into binary, causing unexpected and incorrect effects when compared with
4435 1. Very minor rewording of debugging text in manualroute to say "list of
4436 hosts" instead of "hostlist".
4438 2. If verify=header_syntax was set, and a header line with an unqualified
4439 address (no domain) and a large number of spaces between the end of the
4440 name and the colon was received, the reception process suffered a buffer
4441 overflow, and (when I tested it) crashed. This was caused by some obsolete
4442 code that should have been removed. The fix is to remove it!
4444 3. When running in the test harness, delay a bit after writing a bounce
4445 message to get a bit more predictability in the log output.
4447 4. Added a call to search_tidyup() just before forking a reception process. In
4448 theory, someone could use a lookup in the expansion of smtp_accept_max_
4449 per_host which, without the tidyup, could leave open a database connection.
4451 5. Added the variables $recipient_data and $sender_data which get set from a
4452 lookup success in an ACL "recipients" or "senders" condition, or a router
4453 "senders" option, similar to $domain_data and $local_part_data.
4455 6. Moved the writing of debug_print from before to after the "senders" test
4458 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4459 problems for message scanning, either using a data ACL, or using
4460 local_scan() because the Received: header was not generated till after they
4461 were called (in order to set the time as the time of reception completion).
4462 I have revised the way this works. The header is now generated after the
4463 body is received, but before the ACL or local_scan() are called. After they
4464 are run, the timestamp in the header is updated.
4470 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4471 before starting a queue runner without re-exec. This happened only when
4472 deliver_drop_privilege was set or when the Exim user was set to root. The
4473 effect of the bug was that timeouts during subsequent deliveries caused
4474 crashes instead of being properly handled. The handler is now left at its
4475 default (and expected) setting.
4477 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4478 message, again when deliver_drop_privilege is set or Exim is run as root.
4479 The bug described in (1) was not present in this case, but the tidying up
4480 of the other signals was missing. I have made the two cases consistent.
4482 3. The ignore_target_hosts setting on a manualroute router was being ignored
4483 for hosts that were looked up using the /MX notation.
4485 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4488 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4489 operated on the sender address. After changing the $sender_address to <>
4490 for the sender address verify, Exim was re-instated it as the original
4491 (before rewriting) address, but remembering that it had rewritten it, so it
4492 wasn't rewriting it again. This bug also had the effect of breaking the
4493 sender address verification caching when the sender address was rewritten.
4495 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4496 This has been changed so that if the ip literal address matches
4497 ignore_target_hosts, the router declines.
4499 7. Added expansion conditions match_domain, match_address, and match_local_
4500 part (NOT match_host).
4502 8. The placeholder for the Received: header didn't have a length field set.
4504 9. Added code to Exim itself and to exim_lock to test for a specific race
4505 condition that could lead to file corruption when using MBX delivery. The
4506 issue is with the lockfile that is created in /tmp. If this file is removed
4507 after a process has opened it but before that process has acquired a lock,
4508 there is the potential for a second process to recreate the file and also
4509 acquire a lock. This could lead to two Exim processes writing to the file
4510 at the same time. The added code performs the same test as UW imapd; it
4511 checks after acquiring the lock that its file descriptor still refers to
4512 the same named file.
4514 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4515 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4518 11. Added the smtp_active_hostname option. If used, this will typically be made
4519 to depend on the incoming interface address. Because $interface_address is
4520 not set up until the daemon has forked a reception process, error responses
4521 that can happen earlier (such as "too many connections") no longer contain
4524 12. If an expansion in a condition on a "warn" statement fails because a lookup
4525 defers, the "warn" statement is abandoned, and the next ACL statement is
4526 processed. Previously this caused the whole ACL to be aborted.
4528 13. Added the iplsearch lookup type.
4530 14. Added ident_timeout as a log selector.
4532 15. Added tls_certificate_verified as a log selector.
4534 16. Added a global option tls_require_ciphers (compare the smtp transport
4535 option of the same name). This controls incoming TLS connections.
4537 17. I finally figured out how to make tls_require_ciphers do a similar thing
4538 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4539 before starting the TLS session.
4541 18. Tabs are now shown as \t in -bP output.
4543 19. If the log selector return_path_on_delivery was set, Exim crashed when
4544 bouncing a message because it had too many Received: header lines.
4546 20. If two routers both had headers_remove settings, and the first one included
4547 a superfluous trailing colon, the final name in the first list and the
4548 first name in the second list were incorrectly joined into one item (with a
4549 colon in the middle).
4555 1. Added -C and -D options to the exinext utility, mainly to make it easier
4556 to include in the automated testing, but these could be helpful when
4557 multiple configurations are in use.
4559 2. The exinext utility was not formatting the output nicely when there was
4560 an alternate port involved in the retry record key, nor when there was a
4561 message id as well (for retries that were specific to a specific message
4562 and a specific host). It was also confused by IPv6 addresses, because of
4563 the additional colons they contain. I have fixed the IPv4 problem, and
4564 patched it up to do a reasonable job for IPv6.
4566 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4567 delivery, the log line now contains "pipelined" if PIPELINING was used.
4569 4. An SMTP transport process used to panic and die if the bind() call to set
4570 an explicit outgoing interface failed. This has been changed; it is now
4571 treated in the same way as a connect() failure.
4573 5. A reference to $sender_host_name in the part of a conditional expansion
4574 that was being skipped was still causing a DNS lookup. This no longer
4577 6. The def: expansion condition was not recognizing references to header lines
4578 that used bh_ and bheader_.
4580 7. Added the _cache feature to named lists.
4582 8. The code for checking quota_filecount in the appendfile transport was
4583 allowing one more file than it should have been.
4585 9. For compatibility with Sendmail, the command line option
4593 and sets the incoming protocol and host name (for trusted callers). The
4594 host name and its colon can be omitted when only the protocol is to be set.
4595 Note the Exim already has two private options, -pd and -ps, that refer to
4596 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4597 "s", but I don't think that's a major issue.
4599 10. A number of refactoring changes to the code, none of which should affect
4602 (a) The number of logging options was getting close to filling up the
4603 32-bit word that was used as a bit map. I have split them into two classes:
4604 those that are passed in the argument to log_write(), and those that are
4605 only ever tested independently outside of that function. These are now in
4606 separate 32-bit words, so there is plenty of room for expansion again.
4607 There is no change in the user interface or the logging behaviour.
4609 (b) When building, for example, log lines, the code previously used a
4610 macro that called string_cat() twice, in order to add two strings. This is
4611 not really sufficiently general. Furthermore, there was one instance where
4612 it was actually wrong because one of the argument was used twice, and in
4613 one call a function was used. (As it happened, calling the function twice
4614 did not affect the overall behaviour.) The macro has been replaced by a
4615 function that can join an arbitrary number of extra strings onto a growing
4618 (c) The code for expansion conditions now uses a table and a binary chop
4619 instead of a serial search (which was left over from when there were very
4620 few conditions). Also, it now recognizes conditions like "pam" even when
4621 the relevant support is not compiled in: a suitably worded error message is
4622 given if an attempt is made to use such a condition.
4624 11. Added ${time_interval:xxxxx}.
4626 12. A bug was causing one of the ddress fields not to be passed back correctly
4627 from remote delivery subprocesses. The field in question was not being
4628 subsequently used, so this caused to problems in practice.
4630 13. Added new log selectors queue_time and deliver_time.
4632 14. Might have fixed a bug in maildirsizefile handling that threw up
4633 "unexpected character" debug warnings, and recalculated the data
4634 unnecessarily. In any case, I expanded the warning message to give more
4637 15. Added the message "Restricted characters in address" to the statements in
4638 the default ACL that block characters like @ and % in local parts.
4640 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4641 Three changes have been made:
4643 (a) There was a serious bug; a negative response to MAIL caused the whole
4644 recipient domain to be cached as invalid, thereby blocking all messages
4645 to all local parts at the same domain, from all senders. This bug has
4646 been fixed. The domain is no longer cached after a negative response to
4647 MAIL if the sender used is not empty.
4649 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4652 (c) A new callout option, "use_sender" has been added for people who want
4653 the modified behaviour.
4659 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4662 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4663 indeed breaks things for older releases.
4665 3. Added additional logging to the case where there is a problem reading data
4666 from a filter that is running in a subprocess using a pipe, in order to
4667 try to track down a specific problem.
4669 4. Testing facility fudge: when running in the test harness and attempting
4670 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4671 getting "No route to host". Convert this to a timeout.
4673 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4676 6. Some OS don't have socklen_t but use size_t instead. This affects the
4677 fifth argument of getsockopt() amongst other things. This is now
4678 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4679 can be set for individual OS. I have set it for SunOS5, OSF1, and
4680 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4681 some earlier ones do not.
4683 7. Change 4.30/15 was not doing the test caselessly.
4685 8. The standard form for an IPv6 address literal was being rejected by address
4686 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4687 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4688 this, as well as the form without the "IPv6" on the front (but only when
4689 address literals are enabled, of course).
4691 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4693 10. Exim crashed if a message with an empty sender address specified by -f
4694 encountered a router with an errors_to setting. This could be provoked only
4695 by a command such as
4699 where an empty string was supplied; "<>" did not hit this bug.
4701 11. Installed PCRE release 4.5.
4703 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4704 remained set. It is now erased.
4706 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4707 times from message ids (which are base 36 rather than the normal 62).
4709 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4710 were being counted as actual protocol errors, and logged if the log
4711 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4712 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4713 and DATA following a set of rejected RCPTs do not count as protocol errors.
4714 In other words, Exim assumes they were pipelined, though this may not
4715 actually be the case. Of course, in all cases the client gets an
4716 appropriate error code.
4718 15. If a lookup fails in an ACL condition, a message about the failure may
4719 be available; it is used if testing the ACL cannot continue, because most
4720 such messages specify what the cause of the deferral is. However, some
4721 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4722 that caused an old message to be retained and used if a later statement
4723 caused a defer, replacing the real cause of the deferral.
4725 16. If an IP address had so many PTR records that the DNS lookup buffer
4726 was not large enough to hold them, Exim could crash while trying to process
4727 the truncated data. It now detects and logs this case.
4729 17. Further to 4.21/58, another change has been made: if (and only if) the
4730 first line of a message (the first header line) ends with CRLF, a bare LF
4731 in a subsequent header line has a space inserted after it, so as not to
4732 terminate the header.
4734 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4735 unnecessarily, used atoi() instead of strtol(), and didn't check the
4736 termination when getting file sizes from file names by regex.
4738 19. Completely re-implemented the support for maildirsize files, in the light
4739 of a number of problems with the previous contributed implementation
4740 (4.30/29). In particular:
4742 . If the quota is zero, the maildirsize file is maintained, but no quota is
4745 . If the maildir directory does not exist, it is created before any attempt
4746 to write a maildirsize file.
4748 . The quota value in the file is just a cache; if the quota is changed in
4749 the transport, the new value overrides.
4751 . A regular expression is available for excluding directories from the
4754 20. The autoreply transport checks the characters in options that define the
4755 message's headers; it allows continued headers, but it was checking with
4756 isspace() after an embedded newline instead of explicitly looking for a
4759 21. If all the "regular" hosts to which an address was routed had passed their
4760 expiry times, and had not reached their retry times, the address was
4761 bounced, even if fallback hosts were defined. Now Exim should go on to try
4764 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4765 equivalent code in the SMTP transport. Some hosts send humungous responses
4766 to HELO/EHLO, more than 1024 it seems.
4768 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4769 gives compiler warnings in some environments. I've now done it "properly",
4772 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4773 (because of problems with the built-in one) was declared to return uschar *
4774 instead of char *, causing compiler failure.
4776 25. Fixed a file descriptor leak when processing alias/forward files.
4778 26. Fixed a minor format string issue in dbfn.c.
4780 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4782 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4783 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4786 29. When a sender address is verified, it is cached, to save repeating the test
4787 when there is more than one recipient in a message. However, when the
4788 verification involves a callout, it is possible for different callout
4789 options to be set for different recipients. It is too complicated to keep
4790 track of this in the cache, so now Exim always runs a verification when a
4791 callout is required, relying on the callout cache for the optimization.
4792 The overhead is duplication of the address routing, but this should not be
4795 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4796 to be verified with callout=postmaster, and the main callout worked but the
4797 postmaster check failed, the verification correctly failed. However, if a
4798 subsequent RCPT command asked for sender verification *without* the
4799 postmaster check, incorrect caching caused this verification also to fail,
4802 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4803 it was not caching the DNS options (qualify_single, search_parents) that
4804 were used when the lookup failed. A subsequent lookup with different
4805 options therefore always gave the same answer, though there were cases
4806 where it should not have. (Example: a "domains = !$mx_any" option on a
4807 dnslookup router: the "domains" option is always processed without any
4808 widening, but the router might have qualify_single set.) Now Exim uses the
4809 cached value only when the same options are set.
4811 32. Added John Jetmore's "exipick" utility to the distribution.
4813 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4814 than a timeout (e.g. a certificate is required, and is not provided), an
4815 Exim server now closes the connection immediately. Previously it waited for
4816 the client to close - but if the client is SSL, it seems that they each
4817 wait for each other, leading to a delay before one of them times out.
4819 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4820 maintained 0.8.x compatibility because I don't think many are using it, and
4821 it is clearly obsolete.
4823 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4826 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4827 client certificate was expired. A simple patch fixes this, though I don't
4828 understand the full logic of why the verify callback is called multiple
4831 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4832 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4833 which causes problems with some clients (such as the Certicom SSL Plus
4834 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4835 disables the coutermeasure allowing Eudora to connect."
4837 38. Exim was not checking that a write() to a log file succeeded. This could
4838 lead to Bad Things if a log got too big, in particular if it hit a file
4839 size limit. Exim now panics and dies if it cannot write to a log file, just
4840 as it does if it cannot open a log file.
4842 39. Modified OS/Makefile-Linux so that it now contains
4844 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4846 The two -D definitions ensure that Exim is compiled with large file
4847 support, which makes it possible to handle log files that are bigger than
4850 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4851 instance) a domain was checked against a named list that involved a lookup,
4852 causing $domain_data to be set, then another domain was checked against the
4853 same list, then the first domain was re-checked, the value of $domain_data
4854 after the final check could be wrong. In particular, if the second check
4855 failed, it could be set empty. This bug probably also applied to
4858 41. The strip_trailing_dot option was not being applied to the address given
4859 with the -f command-line option.
4861 42. The code for reading a message's header from the spool was incrementing
4862 $received_count, but never initializing it. This meant that the value was
4863 incorrect (doubled) while delivering a message in the same process in which
4864 it was received. In the most common configuration of Exim, this never
4865 happens - a fresh exec is done - but it can happen when
4866 deliver_drop_privilege is set.
4868 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4869 it now includes up to 150 characters of the unexpected data in the log
4872 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4873 and building data strings. The size of both of these buffers was 10 000
4874 bytes - far larger than anybody would *ever* want, thought I. Needless to
4875 say, somebody hit the limit. I have increased the maximum line length to
4876 20 000 and the maximum data length of concatenated lines to 100 000. I have
4877 also fixed two bugs, because there was no checking on these buffers. Tsk,
4878 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4879 buffer is too small.
4881 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4882 lsearch lookups. Now it does.
4884 46. When parsing a route_list item in a manualroute router, a fixed-length
4885 buffer was used for the list of hosts. I made this 1024 bytes long,
4886 thinking that nobody would ever have a list of hosts that long. Wrong.
4887 Somebody had a whole pile of complicated expansion conditions, and the
4888 string was silently truncated, leading to an expansion error. It turns out
4889 that it is easier to change to an unlimited length (owing to other changes
4890 that have happened since this code was originally written) than to build
4891 structure for giving a limitation error. The length of the item that
4892 expands into the list of hosts is now unlimited.
4894 47. The lsearch lookup could not handle data where the length of text line was
4895 more than 4095 characters. Such lines were truncated, leading to shortened
4896 data being returned. It should now handle lines of any length.
4898 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4899 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4902 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4903 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4904 no longer happens in comment lines. A list of replacements is now placed
4905 at the head of all of the source files, except those whose only change is
4906 to replace PERL_COMMAND in the very first #! line.
4908 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4909 messages on the queue, with a bottom-up merge sort, using code contributed
4910 by Michael Haardt. This should make operations like -bp somewhat faster on
4911 large queues. It won't affect queue runners, except when queue_run_in_order
4914 51. Installed eximstats 1.31 in the distribution.
4916 52. Added support for SRV lookups to the dnslookup router.
4918 53. If an ACL referred to $message_body or $message_body_end, the value was not
4919 reset for any messages that followed in the same SMTP session.
4921 54. The store-handling optimization for building very long strings was not
4922 differentiating between the different store pools. I don't think this
4923 actually made any difference in practice, but I've tidied it.
4925 55. While running the routers to verify a sender address, $sender_address
4926 was still set to the sender address. This is wrong, because when routing to
4927 send a bounce to the sender, it would be empty. Therefore, I have changed
4928 it so that, while verifying a sender address, $sender_address is set to <>.
4929 (There is no change to what happens when verifying a recipient address.)
4931 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4932 target A or AAAA records (if not already returned) without resetting the
4933 qualify_single or search_parents options of the DNS resolver. These are
4934 inappropriate in this case because the targets of MX and SRV records must
4935 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4936 target that, when qualified, matched something in the local domain. These
4937 two options are now turned off when doing these lookups.
4939 57. It seems that at least some releases of Reiserfs (which does not have the
4940 concept of a fixed number of inodes) returns zero and not -1 for the
4941 number of available inodes. This interacted badly with check_spool_inodes,
4942 which assumed that -1 was the "no such thing" setting. What I have done is
4943 to check that the total number of inodes is greater than zero before doing
4944 the test of how many are available.
4946 58. When a "warn" ACL statement has a log_message modifier, the message is
4947 remembered, and not repeated. This is to avoid a lot of repetition when a
4948 message has many recipients that cause the same warning to be written.
4949 Howewer, Exim was preserving the list of already written lines for an
4950 entire SMTP session, which doesn't seem right. The memory is now reset if a
4951 new message is started.
4953 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4954 debugging output unless log_rewrite was also set.
4956 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4957 of (int)(handle) when we know that handle contains (void *)(-1).
4959 61. The Exim daemon panic-logs an error return when it closes the incoming
4960 connection. However "connection reset by peer" seems to be common, and
4961 isn't really an error worthy of noting specially, so that particular error
4964 62. When Exim is trying to find all the local interfaces, it used to panic and
4965 die if the ioctl to get the interface flags failed. However, it seems that
4966 on at least one OS (Solaris 9) it is possible to have an interface that is
4967 included in the list of interfaces, but for which you get a failure error
4968 for this call. This happens when the interface is not "plumbed" into a
4969 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4970 failure of the "get flags" call assumes that the interface is down.
4972 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4973 makes life easier for people who are doing arithmetic on fields extracted
4974 from dates, where you often get leading zeros that should not be
4975 interpreted as octal.
4977 64. Added qualify_domain to the redirect router, to override the global
4980 65. If a pathologically long header line contained very many addresses (the
4981 report of this problem mentioned 10 000) and each of them was rewritten,
4982 Exim could use up a very large amount of memory. (It kept on making new
4983 copies of the header line as it rewrote, and never released the old ones.)
4984 At the expense of a bit more processing, the header rewriting function has
4985 been changed so that it no longer eats memory in this way.
4987 66. The generation of the Received: header has been moved from the time that a
4988 message starts to be received, to the time that it finishes. The timestamp
4989 in the Received: header should now be very close to that of the <= log
4990 line. There are two side-effects of this change:
4992 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4993 logged header lines no longer include the local Received: line, because
4994 it has not yet been created. The same applies to a copy of the message
4995 that is returned to a non-SMTP sender when a message is rejected.
4997 (b) When a filter file is tested using -bf, no additional Received: header
4998 is added to the test message. After some thought, I decided that this
5001 This change does not affect the value of $received_for. It is still set
5002 after address rewriting, but before local_scan() is called.
5004 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5006 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5007 gave an unhelpful panic error message, and a defer error. I have managed to
5008 change this behaviour so that it now rejects any supplied certificate,
5009 which seems right, as the list of acceptable certificates is empty.
5011 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5012 gave an unhelpful defer error. I have not managed to make this reject any
5013 supplied certificates, but the error message it gives is "no certificate
5014 supplied", which is not helpful.
5016 70. exigrep's output now also includes lines that are not associated with any
5017 message, but which match the given pattern. Implemented by a patch from
5018 Martin Sluka, which also tidied up the Perl a bit.
5020 71. Recipient callout verification, like sender verification, was using <> in
5021 the MAIL FROM command. This isn't really the right thing, since the actual
5022 sender may affect whether the remote host accepts the recipient or not. I
5023 have changed it to use the actual sender in the callout; this means that
5024 the cache record is now keyed on a recipient/sender pair, not just the
5025 recipient address. There doesn't seem to be a real danger of callout loops,
5026 since a callout by the remote host to check the sender would use <>.
5027 [SEE ABOVE: changed after hitting problems.]
5029 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5030 temporary errors. However, in the case of such a code being given after
5031 the end of a data transmission (i.e. after ".") Exim was failing to write
5032 a retry record for the message. (Yes, there was some broken host that was
5033 actually sending 8xx at this point.)
5035 73. An unknown lookup type in a host list could cause Exim to panic-die when
5036 the list was checked. (An example that provoked this was putting <; in the
5037 middle of a list instead of at the start.) If this happened during a DATA
5038 ACL check, a -D file could be left lying around. This kind of configuration
5039 error no longer causes Exim to die; instead it causes a defer errror. The
5040 incident is still logged to the main and panic logs.
5042 74. Buglet left over from Exim 3 conversion. The message "too many messages
5043 in one connection" was written to the rejectlog but not the mainlog, except
5044 when address rewriting (yes!) was being logged.
5046 75. Added write_rejectlog option.
5048 76. When a system filter was run not as root (that is, when system_filter_user
5049 was set), the values of the $n variables were not being returned to the
5050 main process; thus, they were not subsequently available in the $sn
5053 77. Added +return_path_on_delivery log selector.
5055 78. A connection timeout was being treated differently from recipients deferred
5056 when testing hosts_max_try with a message that was older than the host's
5057 retry timeout. (The host should not be counted, thus allowing all hosts to
5058 be tried at least once before bouncing.) This may have been the cause of an
5059 occasionally reported bug whereby a message would remain on the queue
5060 longer than the retry timeout, but would be bounced if a delivery was
5061 forced. I say "may" because I never totally pinned down the problem;
5062 setting up timeout/retry tests is difficult. See also the next item.
5064 79. The ultimate address timeout was not being applied to errors that involved
5065 a combination of host plus message (for example, a timeout on a MAIL
5066 command). When an address resolved to a number of possible hosts, and they
5067 were not all tried for each delivery (e.g. because of hosts_max_try), a
5068 message could remain on the queue longer than the retry timeout.
5070 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5073 81. Fixed an obscure SMTP outgoing bug which required at least the following
5074 conditions: (a) there was another message waiting for the same server;
5075 (b) the server returned 5xx to all RCPT commands in the first message so
5076 that the message was not completed; (c) the server dropped the connection
5077 or gave a negative response to the RSET that Exim sends to abort the
5078 transaction. The observed case was a dropped connection after DATA that had
5079 been sent in pipelining mode. That is, the server had advertised PIPELINING
5080 but was not implementing it correctly. The effect of the bug was incorrect
5081 behaviour, such as trying another host, and this could lead to a crash.
5087 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5088 and daemon.c were passed as pointers to ints; they should have been
5089 pointers to socklen_t variables (which are typically unsigned ints).
5091 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5094 3. Fixed a really odd bug that affected only the testing scheme; patching a
5095 certain fixed string in the binary changed the value of another string that
5096 happened to be identical to the end of the original first string.
5098 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5099 name", it returns that address as the IP address. On some operating
5100 systems (e.g. Solaris), it also passes back the IP address string as the
5101 "host name". However, on others (e.g. Linux), it passes back an empty
5102 string. Exim wasn't checking for this, and was changing the host name to an
5103 empty string, assuming it had been canonicized.
5105 5. Although rare, it is permitted to have more than one PTR record for a given
5106 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5107 all the names associated with an address, because they do in Solaris.
5108 However, it seems that they do not in Linux for data that comes from the
5109 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5110 I found this out when I moved to a new Linux workstation and tried to run
5111 the Exim test suite.
5113 To get round this problem I have changed the code so that it now does its
5114 own call to the DNS to look up PTR records when searching for a host name.
5115 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5116 addresses that are only in /etc/hosts are still found.
5118 This behaviour is, however, controlled by an option called host_lookup_
5119 order, which defaults to "bydns:byaddr". If people want to use the other
5120 order, or indeed, just use one or the other means of lookup, they can
5121 specify it in this variable.
5123 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5124 some operating systems, this comes back from gethostbyaddr() as an empty
5125 string, and this is what Exim used to test for. However, it seems that in
5126 other systems, "." is yielded. Exim now tests for this case too.
5128 7. The values of check_spool_space and check_log_space are now held internally
5129 as a number of kilobytes instead of an absolute number of bytes. If a
5130 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5131 kilobyte. This means that much larger values can be stored.
5133 8. Exim monitor: an attempt to get the action menu when not actually pointing
5134 at a message produces an empty menu entitled "No message selected". This
5135 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5136 no entries in it ("Shell widget menu has zero width and/or height"). So I
5137 have added a single, blank menu entry in this case.
5139 9. Added ${quote_local_part.
5141 10. MIME decoding is now applied to the contents of Subject: header lines when
5144 11. Now that a reference to $sender_host_address automatically causes a reverse
5145 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5146 host lookup before query-style lookups in lists that might use this
5147 variable. This has therefore been abolished, and the "net-" prefix is no
5148 longer necessary for query-style lookups.
5150 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5151 have been a typo for LFLAGS, so it has been changed.
5153 13. The install script calls Exim with "-C /dev/null" in order to find the
5154 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5155 to be output. Howeve, since Exim outputs its version number before the
5156 error, it didn't break the script. It just looked ugly. I fixed this by
5157 always allowing "-C /dev/null" if the caller is root.
5159 14. Ignore overlarge ACL variable number when reading spool file - insurance
5160 against a later release with more variables having written the file.
5162 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5163 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5164 this, as well as the form without the "IPv6" on the front.
5166 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5167 OS/Makefile-Darwin file.
5169 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5170 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5172 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5173 variable. However, some OS use ints for inodes. I've added cast to long int
5174 to get rid of the compiler warning.
5176 19. I had forgotten to lock out "/../" in configuration file names when
5177 ALT_CONFIG_PREFIX was set.
5179 20. Routers used for verification do not need to specify transports. However,
5180 if such a router generated a host list, and callout was configured, Exim
5181 crashed, because it could not find a port number from the (non-existent)
5182 transport. It now assumes port 25 in this circumstance.
5184 21. Added the -t option to exigrep.
5186 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5187 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5188 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5189 them is set, LOOKUP_LSEARCH is forced.
5191 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5193 24. Added sender and host information to the "rejected by local_scan()" log
5194 line; previously there was no indication of these.
5196 25. Added .include_if_exists.
5198 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5199 Linux. It turns out that not all file systems support this. Apparently some
5200 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5201 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5202 not supported on the file descriptor, is now ignored when Exim is trying to
5203 sync a directory. This applies only to Linux.
5205 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5207 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5208 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5211 29. Added support for maildirsize files from supplied patch (modified a bit).
5213 30. The use of :fail: followed by an empty string could lead Exim to respond to
5214 sender verification failures with (e.g.):
5216 550 Verification failed for <xxx>
5217 550 Sender verify failed
5219 where the first response line was missing the '-' that indicates it is not
5220 the final line of the response.
5222 31. The loop for finding the name of the user that called Exim had a hardwired
5223 limit of 10; it now uses the value of finduser_retries, which is used for
5224 all other user lookups.
5226 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5229 33. Exim was neglecting to zero errno before one call of strtol() when
5230 expanding a string and expecting an integer value. On some systems this
5231 resulted in spurious "integer overflow" errors. Also, it was casting the
5232 result into an int without checking.
5234 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5235 did not work. The code looks as if it has *never* worked, though it appears
5236 to have been documented since at least releast 1.62. I have made it work.
5238 35. The "timeout_DNS" error in retry rules, also documented since at least
5239 1.62, also never worked. As it isn't clear exactly what this means, and
5240 clearly it isn't a major issue, I have abolished the feature by treating it
5241 as "timeout", and writing a warning to the main and panic logs.
5243 36. The display of retry rules for -brt wasn't always showing the error code
5246 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5247 timeout_connect_A, timeout_connect_MX.
5249 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5250 to the empty sender.
5252 39. The daemon was not analysing the content of -oX till after it had closed
5253 stderr and disconnected from the controlling terminal. This meant that any
5254 syntax errors were only noted on the panic log, and the return code from
5255 the command was 0. By re-arranging the code a little, I've made the
5256 decoding happen first, so such errors now appear on stderr, and the return
5257 code is 1. However, the actual setting up of the sockets still happens in
5258 the disconnected process, so errors there are still only recorded on the
5261 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5262 connections (as happens on some IP stacks) was logged at start up time as
5263 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5264 it from "IPv6 and IPv4", which means that two separate sockets are being
5267 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5268 says whether AF_INET or AF_INET6 was passed as an argument.
5270 42. Exiwhat output was messed up when time zones were included in log
5273 43. Exiwhat now gives more information about the daemon's listening ports,
5274 and whether -tls-on-connect was used.
5276 44. The "port" option of the smtp transport is now expanded.
5278 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5279 silently ignored. Now an error message is written to the main and panic
5282 46. There's a new ACL modifier called "logwrite" which writes to a log file
5283 as soon as it is encountered.
5285 47. Added $local_user_uid and $local_user_gid at routing time.
5287 48. Exim crashed when trying to verify a sender address that was being
5290 49. Exim was recognizing only a space character after ".include". It now also
5291 recognizes a tab character.
5293 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5294 extracting the relevant information from the specification. The man page no
5295 longer contains scrambled data for the -d option, and I've added a section
5296 at the front about calling Exim under different names.
5298 51. Added "extra_headers" argument to the "mail" command in filter files.
5300 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5303 53. Installed eximstats 1.29.
5305 54. Added transport_filter_timeout as a generic transport option.
5307 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5308 Cc: header lines. This was required by RFC 822, but it not required by RFC
5311 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5312 incoming messages that did not have them. Now it does so only if the
5313 message originates locally, that is, if there is no associated remote host
5314 address. When Resent- header lines are present, this applies to the Resent-
5315 lines rather than the non-Resent- lines.
5317 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5318 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5320 58. Messages for configuration errors now include the name of the main
5321 configuration file - useful now that there may be more than one file in a
5322 list (.included file names were always shown).
5324 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5325 for those rare installations that do not start the daemon as root or run it
5326 setuid root. I've cut out the call to initgroups() if the daemon is not
5329 60. The Exim user and group can now be bound into the binary as text strings
5330 that are looked up at the start of Exim's processing.
5332 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5334 62. Added $mailstore_basename variable.
5336 63. Installed patch to sieve.c from Michael Haardt.
5338 64. When Exim failed to open the panic log after failing to open the main log,
5339 the original message it was trying to log was written to stderr and debug
5340 output, but if they were not available (the usual case in production), it
5341 was lost. Now it is written to syslog before the two lines that record the
5342 failures to open the logs.
5344 65. Users' Exim filters run in subprocesses under the user's uid. It is
5345 possible for a "deliver" command or an alias in a "personal" command to
5346 provoke an address rewrite. If logging of address rewriting is configured,
5347 this fails because the process is not running as root or exim. There may be
5348 a better way of dealing with this, but for the moment (because 4.30 needs
5349 to be released), I have disabled address rewrite logging when running a
5350 filter in a non-root, non-exim process.
5356 1. The buildconfig auxiliary program wasn't quoting the value set for
5357 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5358 not defined. This bug was masked in 4.22 by the effect that was fixed in
5361 2. Some messages that were rejected after a message id was allocated were
5362 shown as "incomplete" by exigrep. It no longer does this for messages that
5363 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5365 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5366 have allow_domain_literals set, the ID did not get logged in the <= line.
5367 Domain literals are now always recognized in Message-ID: header lines.
5369 4. The first argument for a ${extract expansion item is the key name or field
5370 number. Leading and trailing spaces in this item were not being ignored,
5371 causing some misleading effects.
5373 5. When deliver_drop_privilege was set, single queue runner processes started
5374 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5375 same command in the process it spins off) were not dropping privilege.
5377 6. When the daemon running as "exim" started a queue runner, it always
5378 re-executed Exim in the spun-off process. This is a waste of effort when
5379 deliver_drop_privilege is set. The new process now just calls the
5380 queue-runner function directly.
5386 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5389 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5390 ignored. Though the use of -oP was forcing the writing of a pid file, it
5391 was always written to the default place.
5393 3. If the message "no IP address found for host xxxx" is generated during
5394 incoming verification, it is now followed by identification of the incoming
5395 connection (so you can more easily find what provoked it).
5397 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5399 5. Added some features to "harden" Exim a bit more against certain attacks:
5401 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5402 be put in Local/Makefile. This is like the never_users runtime option,
5403 but it cannot be overridden. The default setting is "root".
5405 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5406 prefix string with which any file named in a -C command line option
5409 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5410 is retained for -C and -D only if the caller of Exim is root. Without
5411 it, the exim user may also use -C and -D and retain privilege.
5413 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5414 command line option is disabled.
5416 6. Macro names set by the -D option must start with an upper case letter, just
5417 like macro names defined in the configuration file.
5419 7. Added "dereference=" facility to LDAP.
5421 8. Two instances of the typo "uknown" in the source files are fixed.
5423 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5424 the Configure-Makefile script screwed up while processing it.
5426 10. Incorporated PCRE 4.4.
5428 11. The SMTP synchronization check was not operating right at the start of an
5429 SMTP session. For example, it could not catch a HELO sent before the client
5430 waited for the greeting. There is now a check for outstanding input at the
5431 point when the greeting is written. Because of the duplex, asynchronous
5432 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5433 way, but not yet received, when the check is performed.
5435 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5436 on TCP/IP sockets, because this apparently causes some broken clients to
5439 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5440 unchanged) from the Cygwin maintainer.
5442 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5443 was supported instead of testing for mbx. Effectively a typo.
5445 15. The spa authenticator server code was not checking that the input it
5446 received was valid base64.
5448 16. The debug output line for the "set" modifier in ACLs was not showing the
5449 name of the variable that was being set.
5451 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5453 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5454 lookup on its own. Something else was needed to trigger the lookup. For
5455 example, a match in host_lookup or the need for a host name in a host list.
5456 Now, if $sender_host_name is referenced and the host name has not yet been
5457 looked up, a lookup is performed. If the lookup fails, the variable remains
5458 empty, and $host_lookup_failed is set to "1".
5460 19. Added "eqi" as a case-independent comparison operator.
5462 20. The saslauthd authentication condition could segfault if neither service
5463 nor realm was specified.
5465 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5466 error message that was logged was misleading, and incoming SMTP
5467 connections were dropped. The message is now more accurate, and temporary
5468 errors are given to SMTP connections.
5470 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5471 (except RSET) with a 421 error, until QUIT is received. However, it was
5472 failing to send a response to QUIT.
5474 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5475 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5476 ACL is now run after the helo_try_verify_hosts code.
5478 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5479 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5480 case-independent, so other case variants are also recognized). Apparently
5481 some systems use these upper case variants.
5483 25. If more than two messages were waiting for the same host, and a transport
5484 filter was specified for the transport, Exim sent two messages over the
5485 same TCP/IP connection, and then failed with "socket operation on non-
5486 socket" when it tried to send the third.
5488 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5490 27. The extern definition of crypt16() in expand.c was not being excluded when
5491 the OS had its own crypt16() function.
5493 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5494 as a preferred synonym for return_size_limit, both as an option and as an
5497 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5499 30. Changed the default configuration ACL to relax the local part checking rule
5500 for addresses that are not in any local domains. For these addresses,
5501 slashes and pipe symbols are allowed within local parts, but the sequence
5502 /../ is explicitly forbidden.
5504 31. SPA server authentication was not clearing the challenge buffer before
5507 32. log_message in a "warn" ACL statement was writing to the reject log as
5508 well as to the main log, which contradicts the documentation and doesn't
5509 seem right (because no rejection is happening). So I have stopped it.
5511 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5512 However, I am unable to do any testing of this.
5514 34. Fixed an infelicity in the appendfile transport. When checking directories
5515 for a mailbox, to see if any needed to be created, it was accidentally
5516 using path names with one or more superfluous leading slashes; tracing
5517 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5519 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5520 contents are added to the log line that is written for every discarded
5521 recipient. (Previously a log_message setting was ignored.)
5523 36. The ${quote: operator now quotes the string if it is empty.
5525 37. The install script runs exim in order to find its version number. If for
5526 some reason other than non-existence or emptiness, which it checks, it
5527 could not run './exim', it was installing it with an empty version number,
5528 i.e. as "exim-". This error state is now caught, and the installation is
5531 38. An argument was missing from the function that creates an error message
5532 when Exim fails to connect to the socket for saslauthd authentication.
5533 This could cause Exim to crash, or give a corrupted message.
5535 39. Added isip, isip4, and isip6 to ${if conditions.
5537 40. The ACL variables $acl_xx are now saved with the message, and can be
5538 accessed later in routers, transports, and filters.
5540 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5541 strings in the file are not string-expanded.
5543 42. If a MAIL command specified a SIZE value that was too large to fit into an
5544 int variable, the check against message_size_limit failed. Such values are
5545 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5546 day this will have to be increased, but I don't think I want to be around
5547 when emails are that large.
5554 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5555 iconv() is not standard in FreeBSD.
5557 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5558 IPv6 enabled. The observed symptom was a segmentation fault on return from
5559 the function os_common_find_running_interfaces() in src/os.c.
5561 3. In the check_special_case() function in daemon.c I had used "errno" as an
5562 argument name, which causes warnings on some systems. This was basically a
5563 typo, since it was named "eno" in the comments!
5565 4. The code that waits for the clock to tick (at a resolution of some fraction
5566 of a second) so as to ensure message-id uniqueness was always waiting for
5567 at least one whole tick, when it could have waited for less. [This is
5568 almost certainly not relevant at current processor speeds, where it is
5569 unlikely to ever wait at all. But we try to future-proof.]
5571 5. The function that sleeps for a time interval that includes fractions of a
5572 second contained a race. It did not block SIGALRM between setting the
5573 timer, and suspending (a couple of lines later). If the interval was short
5574 and the sigsuspend() was delayed until after it had expired, the suspension
5575 never ended. On busy systems this could lead to processes getting stuck for
5578 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5579 process, before it forks any delivery processes. The open lookup caching
5580 mechanism meant that the open file or database connection was passed into
5581 the delivery process. The problem was that delivery processes always tidy
5582 up cached lookup data. This could cause a problem for the next delivery
5583 process started by the queue runner, because the external queue runner
5584 process does not know about the closure. So the next delivery process
5585 still has data in the lookup cache. In the case of a file lookup, there was
5586 no problem because closing a file descriptor in a subprocess doesn't affect
5587 the parent. However, if the lookup was caching a connection to a database,
5588 the connection was closed, and the second delivery process was likely to
5589 see errors such as "PGSQL: query failed: server closed the connection
5590 unexpectedly". The problem has been fixed by closing all cached lookups
5591 in a queue runner before running a delivery process.
5593 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5594 seem to have the "const" qualifier which it has on other OS. I've
5597 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5598 *of the same type* (client or server) with the same public name that an
5599 error should be diagnosed.
5601 9. When Exim looked up a host name for an IP address, but failed to find the
5602 original IP address when looking up the host name (a safety check), it
5603 output the message "<ip address> does not match any IP for NULL", which was
5604 confusing, to say the least. The bug was that the host name should have
5605 appeared instead of "NULL".
5607 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5608 user that is built into the binary, and the -C or -D options is used, root
5609 privilege is dropped before the configuration file is read. In addition,
5610 logging is switched to stderr instead of the normal log files. If the
5611 configuration then re-defines the Exim user, the unprivileged environment
5612 is probably not what is expected, so Exim logs a panic warning message (but
5615 However, if deliver_drop_privilege is set, the unprivileged state may well
5616 be exactly what is intended, so the warning has been cut out in that case,
5617 and Exim is allowed to try to write to its normal log files.
5623 1. smtp_return_error_details was not giving details for temporary sender
5624 or receiver verification errors.
5626 2. Diagnose a configuration error if two authenticators have the same public
5629 3. Exim used not to create the message log file for a message until the first
5630 delivery attempt. This could be confusing when incoming messages were held
5631 for policy or load reasons. The message log file is now created at the time
5632 the message is received, and an initial "Received" line is written to it.
5634 4. The automatically generated man page for command line options had a minor
5635 bug that caused no ill effects; however, a more serious problem was that
5636 the procedure for building the man page automatically didn't always
5637 operate. Consequently, release 4.20 contains an out-of-date version. This
5638 shouldn't happen again.
5640 5. When building Exim with embedded Perl support, the script that builds the
5641 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5642 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5644 6. The freeze_tell option was not being used for messages that were frozen on
5645 arrival, either by an ACL or by local_scan().
5647 7. Added the smtp_incomplete_transaction log selector.
5649 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5650 was accepting AUTH without a new EHLO.
5652 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5653 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5656 10. Logging of TCP/IP connections (when configured) now happens in the main
5657 daemon process instead of the child process, so that the TCP/IP connection
5658 count is more accurate (but it can never be perfect).
5660 11. The use of "drop" in a nested ACL was not being handled correctly in the
5661 outer ACL. Now, if condition failure induced by the nested "drop" causes
5662 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5663 or "require"), the connection is dropped.
5665 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5666 that yield "discard" can now be used with an "accept" or a "discard" verb,
5667 but an error is generated for any others (because I can't see a useful way
5668 to define what should happen).
5670 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5671 are now processed in the same way as lines in the Exim configuration file.
5672 In particular, continuation lines are supported.
5674 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5676 15. Added -ti meaning -t -i.
5678 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5679 domains, and warn by logging if others are found.
5681 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5682 returned from GIFCONF when Exim is trying to find the list of interfaces on
5683 a host. The code in os.c has been modified to copy each ifreq to an aligned
5684 structure in all cases.
5686 Also, in some cases, the returned ifreq's were being copied to a 'struct
5687 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5688 means the last couple of bytes of an IPv6 address could be chopped if the
5689 ifreq contained only a normal sockaddr (14 bytes storage).
5691 18. Named domain lists were not supported in the hosts_treat_as_local option.
5692 An entry such as +xxxx was not recognized, and was treated as a literal
5695 19. Ensure that header lines added by a DATA ACL are included in the reject log
5696 if the ACL subsequently rejects the message.
5698 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5699 MD5 (which is deprecated).
5701 21. When testing a filter file using -bf, Exim was writing a message when it
5702 took the sender from a "From " line in the message, but it was not doing so
5703 when it took $return_path from a Return-Path: header line. It now does.
5705 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5706 with a valid header line field name (a series of printing characters
5707 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5709 23. Changed "disc" in the source to "disk" to conform to the documentation and
5710 the book and for uniformity.
5712 24. Ignore Sendmail's -Ooption=value command line item.
5714 25. When execve() failed while trying to run a command in a pipe transport,
5715 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5716 could be confused with a return value of 69 from the command itself. This
5717 has been changed to 127, the value the shell returns if it is asked to run
5718 a non-existent command. The wording for the related log line suggests a
5719 non-existent command as the problem.
5721 26. If received_header_text expands to an empty string, do not add a Received:
5722 header line to the message. (Well, it adds a token one on the spool, but
5723 marks it "old" so that it doesn't get used or transmitted.)
5725 27. Installed eximstats 1.28 (addition of -nt option).
5727 28. There was no check for failure on the call to getsockname() in the daemon
5728 code. This can fail if there is a shortage of resources on the system, with
5729 ENOMEM, for example. A temporary error is now given on failure.
5731 29. Contrary to the C standard, it seems that in some environments, the
5732 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5733 program. Exim now does this explicitly; it affects the formatting of
5734 timestamps using strftime().
5736 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5737 complaints. I've now initialized all the variables, to avoid this.
5739 32. Header lines added by a system filter were not being "seen" during
5740 transport-time rewrites.
5742 33. The info_callback() function passed to OpenSSL is set up with type void
5743 (*)(SSL *, int, int), as described somewhere. However, when calling the
5744 function (actually a macro) that sets it up, the type void(*)() is
5745 expected. I've put in a cast to prevent warnings from picky compilers.
5747 34. If a DNS black list lookup found a CNAME record, but there were no A
5748 records associated with the domain it pointed at, Exim crashed.
5750 35. If a DNS black list lookup returned more than one A record, Exim ignored
5751 all but the first. It now scans all returned addresses if a particular IP
5752 value is being sought. In this situation, the contents of the
5753 $dnslist_value variable are a list of all the addresses, separated by a
5756 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5757 to accept a host name and all its aliases if the forward lookup for any of
5758 them yielded the IP address of the incoming connection. Now it accepts only
5759 those names whose forward lookup yields the correct IP address. Any other
5760 names are discarded. This closes a loophole whereby a rogue DNS
5761 administrator could create reverse DNS records to break through a
5762 wildcarded host restriction in an ACL.
5764 37. If a user filter or a system filter that ran in a subprocess used any of
5765 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5766 the wrong values were passed to the pipe command ($thisaddress had the
5767 value of $0, $0 had the value of $1, etc). This bug was introduced by
5768 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5770 38. Improved the line breaking for long SMTP error messages from ACLs.
5771 Previously, if there was no break point between 40 and 75 characters, Exim
5772 left the rest of the message alone. Two changes have been made: (a) I've
5773 reduced the minimum length to 35 characters; (b) if it can't find a break
5774 point between 35 and 75 characters, it looks ahead and uses the first one
5775 that it finds. This may give the occasional overlong line, but at least the
5776 remaining text gets split now.
5778 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5779 file descriptors might be low, and that setting 1000 would always raise it.
5780 It turns out that in some environments, the limit is already over 1000 and
5781 that lowering it causes trouble. So now Exim takes care not to decrease it.
5783 40. When delivering a message, the value of $return_path is set to $sender_
5784 address at the start of routing (routers may change the value). By an
5785 oversight, this default was not being set up when an address was tested by
5786 -bt or -bv, which affected the outcome if any router or filter referred to
5789 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5790 log only when "message" or "log_message" are set. However, if one of the
5791 conditions was an address verification, or a call to a nested ACL, the
5792 messages generated by the underlying test were being passed through. This
5793 no longer happens. The underlying message is available in $acl_verify_
5794 message for both "message" and "log_message" expansions, so it can be
5795 passed through if needed.
5797 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5798 new expansion $bh_ to give the encoded byte string without charset
5799 translation. Translation happens only if iconv() is available; HAVE_ICONV
5800 indicates this at build time. HEADERS_CHARSET gives the charset to
5801 translate to; headers_charset can change it in the configuration, and
5802 "headers charset" can change it in an individual filter file.
5804 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5805 that creates RFC 2047 encoded "words" labels them as that charset instead
5806 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5807 expansion operator; (ii) when Exim creates a From: line for a local
5808 message; (iii) when a header line is rewritten to include a "phrase" part.
5810 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5811 buggy, causing it to skip the first lines of messages whose message ID
5812 ended in 'D'. This would not have bitten before Exim release 4.14, because
5813 message IDs were unlikely to end in 'D' before then. The effect was to have
5814 incorrect size information for certain domains.
5816 45. #include "config.h" was missing at the start of the crypt16.c module. This
5817 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5820 46. If there was a timeout during a "random" callout check, Exim treated it as
5821 a failure of the random address, and carried on sending RSET and the real
5822 address. If the delay was just some slowness somewhere, the response to the
5823 original RCPT would be taken as a response to RSET and so on, causing
5824 mayhem of various kinds.
5826 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5827 when I implemented it. It didn't allow for the fact that some option values
5828 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5829 the right test for positive values.
5831 48. Domain names in DNS records are case-independent. Exim always looks them up
5832 in lower case. Some resolvers return domain names in exactly the case they
5833 appear in the zone file, that is, they may contain uppercase letters. Not
5834 all resolvers do this - some return always lower case. Exim was treating a
5835 change of case by a resolver as a change of domain, similar to a widening
5836 of a domain abbreviation. This triggered its re-routing code and so it was
5837 trying to route what was effectively the same domain again. This normally
5838 caused routing to fail (because the router wouldn't handle the domain
5839 twice). Now Exim checks for this case specially, and just changes the
5840 casing of the domain that it ultimately uses when it transmits the message
5843 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5846 50. If a filter generated a file delivery with a non-absolute name (possible if
5847 no home directory exists for the router), the forbid_file option was not
5850 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5851 the existing equality matching.
5853 52. Exim was using ints instead of ino_t variables in some places where it was
5854 dealing with inode numbers.
5856 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5857 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5858 called TMPDIR, and if it finds it is different, it changes its value.
5860 54. The smtp_printf() function is now made available to local_scan() so
5861 additional output lines can be written before returning. There is also an
5862 smtp_fflush() function to enable the detection of a dropped connection.
5863 The variables smtp_input and smtp_batched_input are exported to
5866 55. Changed the default runtime configuration: the message "Unknown user"
5867 has been removed from the ACL, and instead placed on the localuser router,
5868 using the cannot_route_message feature. This means that any verification
5869 failures that generate their own messages won't get overridden. Similarly,
5870 the "Unrouteable address" message that was in the ACL for unverifiable
5871 relay addresses has also been removed.
5873 56. Added hosts_avoid_esmtp to the smtp transport.
5875 57. The exicyclog script was not checking for the esoteric option
5876 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5877 will work only if exicyclog is run under the appropriate euid.
5879 58. Following a discussion on the list, the rules by which Exim recognises line
5880 endings on incoming messages have been changed. The -dropcr and drop_cr
5881 options are now no-ops, retained only for backwards compatibility. The
5882 following line terminators are recognized: LF CRLF CR. However, special
5883 processing applies to CR:
5885 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5886 nor a local message in the state where . is a terminator.
5888 (ii) If a bare CR is encountered in a header line, an extra space is added
5889 after the line terminator so as not to end the header. The reasoning
5890 behind this is that bare CRs in header lines are most likely either
5891 to be mistakes, or people trying to play silly games.
5893 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5894 was being incorrectly given as 18 bytes larger than it should have been.
5895 This is a VOB (very old bug).
5897 60. This may never have affected anything current, but just in case it has:
5898 When the local host is found other than at the start of a list of hosts,
5899 the local host, those with the same MX, and any that follow, are discarded.
5900 When the list in question was part of a longer list of hosts, the following
5901 hosts (not currently being processed) were also being discarded. This no
5902 longer happens. I'm not sure if this situation could ever has previously
5905 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5908 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5909 header. This is something that is recommended in a new Internet Draft, and
5910 is something that is documented as being done by Sendmail. There are two
5911 possible values. For messages generated by the autoreply transport, Exim
5914 Auto-Submitted: auto-replied
5916 whereas for all other generated messages (e.g. bounces) it adds
5918 Auto-Submitted: auto-generated
5920 63. The "personal" condition in filters now includes a test for the
5921 Auto-Submitted: header. If it contains the string "auto-" the message it
5922 not considered personal.
5924 64. Added rcpt_include_affixes as a generic transport option.
5926 65. Added queue_only_override (default true).
5928 66. Added the syslog_duplication option.
5930 67. If what should have been the first header line of a message consisted of
5931 a space followed by a colon, Exim was mis-interpreting it as a header line.
5932 It isn't of course - it is syntactically invalid and should therefore be
5933 treated as the start of the message body. The misbehaviour could have
5934 caused a number of strange effects, including loss of data in subsequent
5935 header lines, and spool format errors.
5937 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5938 client host had authenticated. This control can now be exercised by an ACL
5939 for more flexibility.
5941 69. By default, callouts do not happen when testing with -bh. There is now a
5942 variant, -bhc, which does actually run the callout code, including
5943 consulting and updating the callout cache.
5945 70. Added support for saslauthd authentication, courtesy of Alexander
5948 71. If statvfs() failed on the spool or log directories while checking their
5949 size for availability, Exim confusingly gave the error "space shortage".
5950 Furthermore, in debugging mode it crashed with a floating point exception.
5951 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5952 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5953 serious problem, Exim now writes to the main and panic logs when this
5954 happens, with details of the failure. It then refuses to accept the
5955 incoming message, giving the message "spool directory problem" or "log
5956 directory problem" with a 421 code for SMTP messages.
5958 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5959 0, 1, and 2 exist, because some OS complain for execs without them (see
5960 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5961 descriptors. However, the code omitted to check that the open succeeded,
5962 causing mysterious errors if for some reason the permissions on /dev/null
5963 got screwed. Now Exim writes a message to the main and panic logs, and
5964 bombs out if it can't open /dev/null.
5966 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5967 interact so that it is all more flexible. It is supposed to remain
5968 backwards compatible. Also added extra_local_interfaces.
5970 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5971 to bomb out with an assertion failure - to the client this appears as a
5972 connection drop. This problem occurs in the part of the code that was taken
5973 from the Samba project. Fortunately, the assertion is in a very simple
5974 function, so I have fixed this by reproducing the function inline in the
5975 one place where it is called, and arranging for authentication to fail
5976 instead of killing the process with assert().
5978 75. The SPA client code was not working when the server requested OEM rather
5979 than Unicode encoding.
5981 76. Added code to make require_files with a specific uid setting more usable in
5982 the case where statting the file as root fails - usually a non-root-mounted
5983 NFS file system. When this happens and the failure is EACCES, Exim now
5984 forks a subprocess and does the per-uid checking as the relevant uid.
5986 77. Added process_log_path.
5988 78. If log_file_path was not explicitly set, a setting of check_log_space or
5989 check_log_inodes was ignored.
5991 79. If a space check for the spool or log partitions fails, the incident is now
5992 logged. Of course, in the latter case the data may get lost...
5994 80. Added the %p formatting code to string_format() so that it can be used to
5995 print addresses in debug_print(). Adjusted all the address printing in the
5996 debugging in store.c to use %p rather than %d.
5998 81. There was a concern that a line of code in smtp_in.c could overflow a
5999 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6000 initially expressed, the concern was not well-founded, because trailing
6001 spaces are removed early. However, if the trailing spaces were followed by
6002 a NULL, they did not get removed, so the overflow was possible. Two fixes
6005 (a) I re-wrote the offending code in a cleaner fashion.
6006 (b) If an incoming SMTP command contains a NULL character, it is rejected
6009 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6010 runs initgroups(), so that if the Exim user is in any additional groups,
6011 they will be used during message reception.
6017 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob