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.
95 New variable $lookup_dnssec_authenticated for observability.
97 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
98 Patch submitted by Lars Timman.
100 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
102 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
103 Requires trusted mode and valid format message id, aborts otherwise.
104 Patch contributed by Heiko Schlichting.
110 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
112 PP/02 Make -n do something, by making it not do something.
113 When combined with -bP, the name of an option is not output.
115 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
118 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
119 $sender_host_name and config options to manage this, and basic check
122 PP/05 DSCP support for outbound connections and control modifier for inbound.
124 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
125 (Only plugin which currently uses this is kerberos4, which nobody should
126 be using, but we should make it available and other future plugins might
127 conceivably use it, even though it would break NAT; stuff *should* be
128 using channel bindings instead).
130 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
131 name; added for Sendmail compatibility; requires admin caller.
132 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
133 just ignore it); requires trusted caller.
134 Also parse but ignore: -Ac -Am -X<logfile>
137 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
139 TL/02 Add +smtp_confirmation as a default logging option.
141 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
142 Patch by Magnus Holmgren from 2007-02-20.
144 TL/04 Bugzilla 1281 - Spec typo.
145 Bugzilla 1283 - Spec typo.
146 Bugzilla 1290 - Spec grammar fixes.
148 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
150 TL/06 Add Experimental DMARC support using libopendmarc libraries.
152 TL/07 Fix an out of order global option causing a segfault. Reported to dev
153 mailing list by by Dmitry Isaikin.
155 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
157 JH/02 Support "G" suffix to numbers in ${if comparisons.
159 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
161 NM/01 Bugzilla 1197 - Spec typo
162 Bugzilla 1196 - Spec examples corrections
164 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
166 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
167 gnutls_enable_pkcs11, but renamed to more accurately indicate its
170 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
171 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
173 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
174 "acl {{name}{arg}...}", and optional args on acl condition
177 JH/05 Permit multiple router/transport headers_add/remove lines.
179 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
181 JH/07 Avoid using a waiting database for a single-message-only transport.
182 Performance patch from Paul Fisher. Bugzilla 1262.
184 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
187 JH/09 Add $headers_added variable, with content from use of ACL modifier
188 add_header (but not yet added to the message). Bugzilla 199.
190 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
191 Pulled from Bugzilla 817 by Wolfgang Breyha.
193 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
195 (nb: this is the same fix as in Exim 4.80.1)
197 JH/11 Add A= logging on delivery lines, and a client_set_id option on
200 JH/12 Add optional authenticated_sender logging to A= and a log_selector
203 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
205 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
206 advertise SMTP AUTH mechanism to us, instead of a generic
207 protocol violation error. Also, make Exim more robust to bad
208 data from the Dovecot auth socket.
210 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
212 When a queue runner is handling a message, Exim first routes the
213 recipient addresses, during which it prunes them based on the retry
214 hints database. After that it attempts to deliver the message to
215 any remaining recipients. It then updates the hints database using
218 So if a recipient address works intermittently, it can get repeatedly
219 deferred at routing time. The retry hints record remains fresh so the
220 address never reaches the final cutoff time.
222 This is a fairly common occurrence when a user is bumping up against
223 their storage quota. Exim had some logic in its local delivery code
224 to deal with this. However it did not apply to per-recipient defers
225 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
227 This change adds a proper retry rule check during routing so that the
228 final cutoff time is checked against the message's age. We only do
229 this check if there is an address retry record and there is not a
230 domain retry record; this implies that previous attempts to handle
231 the address had the retry_use_local_parts option turned on. We use
232 this as an approximation for the destination being like a local
233 delivery, as in LMTP.
235 I suspect this new check makes the old local delivery cutoff check
236 redundant, but I have not verified this so I left the code in place.
238 TF/02 Correct gecos expansion when From: is a prefix of the username.
240 Test 0254 submits a message to Exim with the header
244 When I ran the test suite under the user fanf2, Exim expanded
245 the header to contain my full name, whereas it should have added
246 a Resent-Sender: header. It erroneously treats any prefix of the
247 username as equal to the username.
249 This change corrects that bug.
251 GF/01 DCC debug and logging tidyup
252 Error conditions log to paniclog rather than rejectlog.
253 Debug lines prefixed by "DCC: " to remove any ambiguity.
255 TF/03 Avoid unnecessary rebuilds of lookup-related code.
257 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
258 Bug spotted by Jeremy Harris; was flawed since initial commit.
259 Would have resulted in OCSP responses post-SNI triggering an Exim
260 NULL dereference and crash.
262 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
264 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
265 Bug detection, analysis and fix by Samuel Thibault.
266 Bugzilla 1331, Debian bug #698092.
268 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
270 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
271 Server implementation by Todd Lyons, client by JH.
272 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
273 config variable "prdr_enable" controls whether the server
274 advertises the facility. If the client requests PRDR a new
275 acl_data_smtp_prdr ACL is called once for each recipient, after
276 the body content is received and before the acl_smtp_data ACL.
277 The client is controlled by bolth of: a hosts_try_prdr option
278 on the smtp transport, and the server advertisement.
279 Default client logging of deliveries and rejections involving
280 PRDR are flagged with the string "PRDR".
282 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
283 fclose(). Diagnosis by Todd Lyons.
285 PP/17 Update configure.default to handle IPv6 localhost better.
286 Patch by Alain Williams (plus minor tweaks).
289 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
290 This is now consistent with GnuTLS, and is now documented: the
291 previous undocumented portable approach to treating the option as
292 unset was to force an expansion failure. That still works, and
293 an empty string is now equivalent.
295 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
296 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
297 not performing validation itself.
299 PP/20 Added force_command boolean option to pipe transport.
300 Patch from Nick Koston, of cPanel Inc.
302 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
305 TF/04 Added udpsend ACL modifer and hexquote expansion operator
307 PP/21 Fix eximon continuous updating with timestamped log-files.
308 Broken in a format-string cleanup in 4.80, missed when I repaired the
309 other false fix of the same issue.
310 Report and fix from Heiko Schlichting.
313 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
314 Report from Prashanth Katuri.
316 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
317 It's SecureTransport, so affects any MacOS clients which use the
318 system-integrated TLS libraries, including email clients.
320 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
321 using a MIME ACL for non-SMTP local injection.
322 Report and assistance in diagnosis by Warren Baker.
324 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
326 JH/16 Fix comparisons for 64b. Bugzilla 1385.
328 TL/09 Add expansion variable $authenticated_fail_id to keep track of
329 last id that failed so it may be referenced in subsequent ACL's.
331 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
334 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
335 ldap library initialization, allowing self-signed CA's to be
336 used. Also properly sets require_cert option later in code by
337 using NULL (global ldap config) instead of ldap handle (per
338 session). Bug diagnosis and testing by alxgomz.
340 TL/12 Enhanced documentation in the ratelimit.pl script provided in
341 the src/util/ subdirectory.
343 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
344 renamed to Transport Post Delivery Action by Jeremy Harris, as
347 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
348 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
349 redis_servers = needs to be configured which will be used by the redis
350 lookup. Patch from Warren Baker, of The Packet Hub.
352 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
354 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
355 hostname or reverse DNS when processing a host list. Used suggestions
356 from multiple comments on this bug.
358 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
360 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
361 Missed a few lines, added it to make the runtest require no keyboard
364 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
365 contains upper case chars. Make router use caseful_local_part.
367 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
368 support when GnuTLS has been built with p11-kit.
374 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
376 This, or similar/improved, will also be change PP/11 of 4.82.
382 PP/01 Handle short writes when writing local log-files.
383 In practice, only affects FreeBSD (8 onwards).
384 Bugzilla 1053, with thanks to Dmitry Isaikin.
386 NM/01 Bugzilla 949 - Documentation tweak
388 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
391 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
393 PP/02 Implemented gsasl authenticator.
395 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
397 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
398 `pkg-config foo` for cflags/libs.
400 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
401 with rest of GSASL and with heimdal_gssapi.
403 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
404 `pkg-config foo` for cflags/libs for the TLS implementation.
406 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
407 properties get this fed in as external SSF. A number of robustness
408 and debugging improvements to the cyrus_sasl authenticator.
410 PP/08 cyrus_sasl server now expands the server_realm option.
412 PP/09 Bugzilla 1214 - Log authentication information in reject log.
413 Patch by Jeremy Harris.
415 PP/10 Added dbmjz lookup type.
417 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
419 PP/12 MAIL args handles TAB as well as SP, for better interop with
420 non-compliant senders.
421 Analysis and variant patch by Todd Lyons.
423 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
424 Bug report from Lars Müller <lars@samba.org> (via SUSE),
425 Patch from Dirk Mueller <dmueller@suse.com>
427 PP/13 tls_peerdn now print-escaped for spool files.
428 Observed some $tls_peerdn in wild which contained \n, which resulted
429 in spool file corruption.
431 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
432 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
433 or write after TLS renegotiation, which otherwise led to messages
436 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
437 as a tracking header (ie: a signed header comes before the signature).
438 Patch from Wolfgang Breyha.
440 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
441 comma-sep list; embedded commas doubled.
443 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
445 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
447 Report and patch from Dmitry Banschikov.
449 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
450 Removed SSL_clear() after SSL_new() which led to protocol negotiation
451 failures. We appear to now support TLS1.1+ with Exim.
453 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
454 lets Exim select keys and certificates based upon TLS SNI from client.
455 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
456 before an outbound SMTP session. New log_selector, +tls_sni.
458 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
459 NULL dereference. Report and patch from Alun Jones.
461 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
462 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
463 Not seeing resolver debug output on NetBSD, but suspect this is a
464 resolver implementation change.
466 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
467 Left warnings. Added "eximon gdb" invocation mode.
469 PP/21 Defaulting "accept_8bitmime" to true, not false.
471 PP/22 Added -bw for inetd wait mode support.
473 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
474 locate the relevant includes and libraries. Made this the default.
476 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
477 Bugzilla 1246, report and most of solution from Tomasz Kusy.
479 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
480 This may cause build issues on older platforms.
482 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
483 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
484 gnutls_require_mac & gnutls_require_protocols (no longer supported).
485 Added SNI support via GnuTLS too.
486 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
488 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
490 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
491 Applied second patch from Janne, implementing suggestion to default
492 multiple-strings-in-record handling to match SPF spec.
494 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
496 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
497 read-only, out of scope).
498 Patch from Wolfgang Breyha, report from Stuart Northfield.
500 PP/29 Fix three issues highlighted by clang analyser static analysis.
501 Only crash-plausible issue would require the Cambridge-specific
502 iplookup router and a misconfiguration.
503 Report from Marcin Mirosław.
505 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
507 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
508 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
509 As part of this, removing so much warning spew let me fix some minor
510 real issues in debug logging.
512 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
513 assignment on my part. Fixed.
515 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
516 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
517 Janne Snabb (who went above and beyond: thank you).
519 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
520 string otherwise requires a connection and a bunch more work and it's
521 relatively easy to get wrong. Should also expose TLS library linkage
524 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
525 64-bit ${eval} (JH/03).
527 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
528 GNU libc to support some of the 64-bit stuff, should not lead to
529 conflicts. Defined before os.h is pulled in, so if a given platform
530 needs to override this, it can.
532 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
533 protection layer was required, which is not implemented.
534 Bugzilla 1254, patch from Wolfgang Breyha.
536 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
537 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
538 tls_dhparam take prime identifiers. Also unbreak combination of
539 OpenSSL+DH_params+TLSSNI.
541 PP/39 Disable SSLv2 by default in OpenSSL support.
547 PP/01 Solaris build fix for Oracle's LDAP libraries.
548 Bugzilla 1109, patch from Stephen Usher.
550 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
552 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
555 TF/02 Fix a couple more cases where we did not log the error message
556 when unlink() failed. See also change 4.74-TF/03.
558 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
559 lock up or crash if it happened to be inside a call to libc when it
560 got a SIGUSR1 from exiwhat.
562 The SIGUSR1 handler appends the current process status to the process
563 log which is later printed by exiwhat. It used to use the general
564 purpose logging code to do this, but several functions it calls are
565 not safe for signals.
567 The new output code in the SIGUSR1 handler is specific to the process
568 log, and simple enough that it's easy to inspect for signal safety.
569 Removing some special cases also simplifies the general logging code.
570 Removing the spurious timestamps from the process log simplifies
573 TF/04 Improved ratelimit ACL condition.
575 The /noupdate option has been deprecated in favour of /readonly which
576 has clearer semantics. The /leaky, /strict, and /readonly update modes
577 are mutually exclusive. The update mode is no longer included in the
578 database key; it just determines when the database is updated. (This
579 means that when you upgrde Exim will forget old rate measurements.)
581 Exim now checks that the per_* options are used with an update mode that
582 makes sense for the current ACL. For example, when Exim is processing a
583 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
584 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
585 must specify per_mail/readonly. If you omit the update mode it defaults to
586 /leaky where that makes sense (as before) or /readonly where required.
588 The /noupdate option is now undocumented but still supported for
589 backwards compatibility. It is equivalent to /readonly except that in
590 ACLs where /readonly is required you may specify /leaky/noupdate or
591 /strict/noupdate which are treated the same as /readonly.
593 A useful new feature is the /count= option. This is a generalization
594 of the per_byte option, so that you can measure the throughput of other
595 aggregate values. For example, the per_byte option is now equivalent
596 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
598 The per_rcpt option has been generalized using the /count= mechanism
599 (though it's more complicated than the per_byte equivalence). When it is
600 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
601 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
602 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
603 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
604 Note that using per_rcpt with a non-readonly update mode in more than
605 one ACL will cause the recipients to be double-counted. (The per_mail
606 and per_byte options don't have this problem.)
608 The handling of very low rates has changed slightly. If the computed rate
609 is less than the event's count (usually one) then this event is the first
610 after a long gap. In this case the rate is set to the same as this event's
611 count, so that the first message of a spam run is counted properly.
613 The major new feature is a mechanism for counting the rate of unique
614 events. The new per_addr option counts the number of different
615 recipients that someone has sent messages to in the last time period. It
616 behaves like per_rcpt if all the recipient addresses are different, but
617 duplicate recipient addresses do not increase the measured rate. Like
618 the /count= option this is a general mechanism, so the per_addr option
619 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
620 example, measure the rate that a client uses different sender addresses
621 with the options per_mail/unique=$sender_address. There are further
622 details in the main documentation.
624 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
626 TF/06 Removed a few PCRE remnants.
628 TF/07 Automatically extract Exim's version number from tags in the git
629 repository when doing development or release builds.
631 PP/02 Raise smtp_cmd_buffer_size to 16kB.
632 Bugzilla 879. Patch from Paul Fisher.
634 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
635 Heavily based on revision 40f9a89a from Simon Arlott's tree.
638 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
640 PP/05 Variable $av_failed, true if the AV scanner deferred.
641 Bugzilla 1078. Patch from John Horne.
643 PP/06 Stop make process more reliably on build failure.
644 Bugzilla 1087. Patch from Heiko Schlittermann.
646 PP/07 Make maildir_use_size_file an _expandable_ boolean.
647 Bugzilla 1089. Patch from Heiko Schlittermann.
649 PP/08 Handle ${run} returning more data than OS pipe buffer size.
650 Bugzilla 1131. Patch from Holger Weiß.
652 PP/09 Handle IPv6 addresses with SPF.
653 Bugzilla 860. Patch from Wolfgang Breyha.
655 PP/10 GnuTLS: support TLS 1.2 & 1.1.
657 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
660 PP/11 match_* no longer expand right-hand-side by default.
661 New compile-time build option, EXPAND_LISTMATCH_RHS.
662 New expansion conditions, "inlist", "inlisti".
664 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
666 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
668 PP/14 fix log_write() format string regression from TF/03.
669 Bugzilla 1152. Patch from Dmitry Isaikin.
675 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
677 PP/02 Harmonised TLS library version reporting; only show if debugging.
678 Layout now matches that introduced for other libraries in 4.74 PP/03.
680 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
682 PP/04 New "dns_use_edns0" global option.
684 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
687 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
688 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
690 TK/01 Updated PolarSSL code to 0.14.2.
691 Bugzilla 1097. Patch from Andreas Metzler.
693 PP/07 Catch divide-by-zero in ${eval:...}.
696 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
699 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
700 format-string attack -- SECURITY: remote arbitrary code execution.
702 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
703 time unintentionally subject to list matching rules, letting the header
704 cause arbitrary Exim lookups (of items which can occur in lists, *not*
705 arbitrary string expansion). This allowed for information disclosure.
707 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
708 INT_MIN/-1 -- value coerced to INT_MAX.
714 NM/01 Workround for PCRE version dependency in version reporting
717 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
718 This fixes portability to compilers other than gcc, notably
719 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
721 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
722 makefiles for portability to HP-UX and POSIX correctness.
724 PP/01 Permit LOOKUP_foo enabling on the make command-line.
725 Also via indented variable definition in the Makefile.
726 (Debugging by Oliver Heesakkers).
728 PP/02 Restore caching of spamd results with expanded spamd_address.
729 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
731 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
732 Improves build reliability. Fix from: Frank Elsner
734 NM/02 Fix wide character breakage in the rfc2047 coding
735 Fixes bug 1064. Patch from Andrey N. Oktyabrski
737 NM/03 Allow underscore in dnslist lookups
738 Fixes bug 1026. Patch from Graeme Fowler
740 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
741 Code patches from Adam Ciarcinski of NetBSD.
743 NM/04 Fixed exiqgrep to cope with mailq missing size issue
746 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
747 is logged, to avoid truncation. Patch from John Horne.
749 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
750 Patch from Jakob Hirsch.
752 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
753 SQL string expansion failure details.
754 Patch from Andrey Oktyabrski.
756 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
757 Patch from Simon Arlott.
759 PP/09 New lookups functionality failed to compile on old gcc which rejects
760 extern declarations in function scope.
761 Patch from Oliver Fleischmann
763 PP/10 Use sig_atomic_t for flags set from signal handlers.
764 Check getgroups() return and improve debugging.
765 Fixed developed for diagnosis in bug 927 (which turned out to be
768 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
769 Patch from Mark Zealey.
771 PP/12 Bugzilla 1056: Improved spamd server selection.
772 Patch from Mark Zealey.
774 PP/13 Bugzilla 1086: Deal with maildir quota file races.
775 Based on patch from Heiko Schlittermann.
777 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
778 Patch from Uwe Doering, sign-off by Michael Haardt.
780 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
781 variable declaration deep within a block. Bug and patch from
784 PP/15 lookups-Makefile IRIX compatibilty coercion.
786 PP/16 Make DISABLE_DKIM build knob functional.
788 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
789 Patch by Simon Arlott
791 TF/03 Fix valgrind.h portability to C89 compilers that do not support
792 variable argument macros. Our copy now differs from upstream.
798 TF/01 Failure to get a lock on a hints database can have serious
799 consequences so log it to the panic log.
801 TF/02 Log LMTP confirmation messages in the same way as SMTP,
802 controlled using the smtp_confirmation log selector.
804 TF/03 Include the error message when we fail to unlink a spool file.
806 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
807 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
808 for maintaining out-of-tree patches for some time.
810 PP/01 Bugzilla 139: Documentation and portability issues.
811 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
812 Handle per-OS dynamic-module compilation flags.
814 PP/02 Let /dev/null have normal permissions.
815 The 4.73 fixes were a little too stringent and complained about the
816 permissions on /dev/null. Exempt it from some checks.
817 Reported by Andreas M. Kirchwitz.
819 PP/03 Report version information for many libraries, including
820 Exim version information for dynamically loaded libraries. Created
821 version.h, now support a version extension string for distributors
822 who patch heavily. Dynamic module ABI change.
824 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
825 privilege escalation vulnerability whereby the Exim run-time user
826 can cause root to append content of the attacker's choosing to
829 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
832 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
833 If dropping privileges for untrusted macros, we disabled normal logging
834 on the basis that it would fail; for the Exim run-time user, this is not
835 the case, and it resulted in successful deliveries going unlogged.
836 Fixed. Reported by Andreas Metzler.
842 PP/01 Date: & Message-Id: revert to normally being appended to a message,
843 only prepend for the Resent-* case. Fixes regression introduced in
844 Exim 4.70 by NM/22 for Bugzilla 607.
846 PP/02 Include check_rfc2047_length in configure.default because we're seeing
847 increasing numbers of administrators be bitten by this.
849 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
851 PP/03 Bugzilla 994: added openssl_options main configuration option.
853 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
855 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
857 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
859 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
860 without a peer certificate, leading to a segfault because of an
861 assumption that peers always have certificates. Be a little more
862 paranoid. Problem reported by Martin Tscholak.
864 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
865 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
866 NB: ClamAV planning to remove STREAM in "middle of 2010".
867 CL also introduces -bmalware, various -d+acl logging additions and
868 more caution in buffer sizes.
870 PP/09 Implemented reverse_ip expansion operator.
872 PP/10 Bugzilla 937: provide a "debug" ACL control.
874 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
876 PP/12 Bugzilla 973: Implement --version.
878 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
880 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
882 PP/15 Bugzilla 816: support multiple condition rules on Routers.
884 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
885 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
886 ignore trailing whitespace.
888 JJ/02 prevent non-panic DKIM error from being sent to paniclog
890 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
893 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
894 Notification from Dr Andrew Aitchison.
896 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
897 ExtendedDetectionInfo response format.
898 Notification from John Horne.
900 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
903 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
904 XSL and documented dependency on system catalogs, with examples of how
907 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
910 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
911 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
912 configuration file which is writeable by the Exim user or group.
914 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
915 of configuration files to cover files specified with the -C option if
916 they are going to be used with root privileges, not just the default
919 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
920 option (effectively making it always true).
922 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
923 files to be used while preserving root privileges.
925 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
926 that rogue child processes cannot use them.
928 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
929 run-time user, instead of root.
931 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
932 Exim run-time user without dropping privileges.
934 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
935 result string, instead of calling string_vformat() twice with the same
938 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
939 for other users. Others should always drop root privileges if they use
940 -C on the command line, even for a whitelisted configure file.
942 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
944 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
950 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
951 $data_path, and $header_path variables; fixed documentation bugs and
954 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
955 exipick to access non-standard spools, including the "frozen" queue
958 NM/01 Bugzilla 965: Support mysql stored procedures.
959 Patch from Alain Williams
961 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
963 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
964 Patch from Andreas Metzler
966 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
967 Patch from Kirill Miazine
969 NM/05 Bugzilla 671: Added umask to procmail example.
971 JJ/03 installed exipick 20100323.0, fixing doc bug
973 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
974 directory. Notification and patch from Dan Rosenberg.
976 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
978 TK/02 Improve log output when DKIM signing operation fails.
980 MH/01 Treat the transport option dkim_domain as a colon separated
981 list, not as a single string, and sign the message with each element,
982 omitting multiple occurences of the same signer.
984 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
985 Bugzilla 985, 986. Patch by Simon Arlott
987 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
988 Patch by Simon Arlott
990 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
991 MBX locking. Notification from Dan Rosenberg.
997 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
999 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1001 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1003 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1005 NM/04 Bugzilla 915: Flag broken perl installation during build.
1011 TK/01 Added patch by Johannes Berg that expands the main option
1012 "spamd_address" if it starts with a dollar sign.
1014 TK/02 Write list of recipients to X-Envelope-Sender header when building
1015 the mbox-format spool file for content scanning (suggested by Jakob
1018 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1019 (http://www.dcc-servers.net/) support via dccifd. Activated by
1020 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1022 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1023 by Mark Daniel Reidel <mr@df.eu>.
1025 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1026 When building exim an external PCRE library is now needed -
1027 PCRE is a system library on the majority of modern systems.
1028 See entry on PCRE_LIBS in EDITME file.
1030 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1031 conversation. Added nologin parameter to request.
1032 Patch contributed by Kirill Miazine.
1034 TF/01 Do not log submission mode rewrites if they do not change the address.
1036 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1038 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1039 log files in place. Contributed by Roberto Lima.
1041 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1043 TF/03 Bugzilla 615: When checking the local_parts router precondition
1044 after a local_part_suffix or local_part_prefix option, Exim now
1045 does not use the address's named list lookup cache, since this
1046 contains cached lookups for the whole local part.
1048 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1049 Robert Millan. Documentation is in experimental-spec.txt.
1051 TF/04 Bugzilla 668: Fix parallel build (make -j).
1053 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1055 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1056 Patch provided by Jan Srzednicki.
1058 TF/05 Leading white space used to be stripped from $spam_report which
1059 wrecked the formatting. Now it is preserved.
1061 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1062 that they are available at delivery time.
1064 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1066 TF/08 TLS error reporting now respects the incoming_interface and
1067 incoming_port log selectors.
1069 TF/09 Produce a more useful error message if an SMTP transport's hosts
1070 setting expands to an empty string.
1072 NM/06 Bugzilla 744: EXPN did not work under TLS.
1073 Patch provided by Phil Pennock.
1075 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1076 Patch provided by Richard Godbee.
1078 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1079 acl_smtp_notquit, added index entry.
1081 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1082 Patch provided by Eugene Bujak.
1084 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1085 accept(). Patch provided by Maxim Dounin.
1087 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1088 Patch provided by Phil Pennock.
1090 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1092 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1093 Patch provided by Brad "anomie" Jorsch.
1095 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1096 Patch provided by Dean Brooks.
1098 TK/05 Add native DKIM support (does not depend on external libraries).
1100 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1101 Patch provided by Graeme Fowler.
1103 NM/16 Bugzilla 851: Documentation example syntax fix.
1105 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1107 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1110 NM/19 Bugzilla 745: TLS version reporting.
1111 Patch provided by Phil Pennock.
1113 NM/20 Bugzilla 167: bool: condition support.
1114 Patch provided by Phil Pennock.
1116 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1117 clients. Patch provided by Phil Pennock.
1119 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1120 Patch provided by Brad "anomie" Jorsch.
1122 NM/23 Bugzilla 687: Fix misparses in eximstats.
1123 Patch provided by Heiko Schlittermann.
1125 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1126 Patch provided by Heiko Schlittermann.
1128 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1129 plus update to original patch.
1131 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1133 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1134 Patch provided by David Brownlee.
1136 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1138 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1140 NM/30 Bugzilla 888: TLS documentation bugfixes.
1142 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1144 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1145 Unlike the original bugzilla I have changed all shell scripts in src tree.
1147 NM/33 Bugzilla 898: Transport filter timeout fix.
1148 Patch by Todd Rinaldo.
1150 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1151 Patch by Serge Demonchaux.
1153 NM/35 Bugzilla 39: Base64 decode bug fixes.
1154 Patch by Jakob Hirsch.
1156 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1158 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1160 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1162 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1168 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1169 ALT-N's libdkim that I have put here:
1170 http://duncanthrax.net/exim-experimental/
1172 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1173 were called 'true' and it seems that C99 defines that as a reserved
1174 keyword to be used with 'bool' variable types. That means you could
1175 not include C99-style headers which use bools without triggering
1176 build errors in sieve.c.
1178 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1179 as mailq or other aliases. Changed the --help handling significantly
1180 to do whats expected. exim_usage() emits usage/help information.
1182 SC/01 Added the -bylocaldomain option to eximstats.
1184 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1186 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1188 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1194 PH/01 Another patch from the Sieve maintainer.
1196 PH/02 When an IPv6 address is converted to a string for single-key lookup
1197 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1198 dots are used instead of colons so that keys in lsearch files need not
1199 contain colons. This was done some time before quoting was made available
1200 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1201 (notated using the quote facility) so as to distinguish them from IPv4
1202 keys. This meant that lookups for IP addresses in host lists did not work
1203 for iplsearch lookups.
1205 This has been fixed by arranging for IPv6 addresses to be expressed with
1206 colons if the lookup type is iplsearch. This is not incompatible, because
1207 previously such lookups could never work.
1209 The situation is now rather anomolous, since one *can* have colons in
1210 ordinary lsearch keys. However, making the change in all cases is
1211 incompatible and would probably break a number of configurations.
1213 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1216 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1217 conversion specification without a maximum field width, thereby enabling
1218 a rogue spamd server to cause a buffer overflow. While nobody in their
1219 right mind would setup Exim to query an untrusted spamd server, an
1220 attacker that gains access to a server running spamd could potentially
1221 exploit this vulnerability to run arbitrary code as the Exim user.
1223 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1224 $primary_hostname instead of what libspf2 thinks the hosts name is.
1226 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1227 a directory entry by the name of the lookup key. Previously, if a
1228 symlink pointed to a non-existing file or a file in a directory that
1229 Exim lacked permissions to read, a lookup for a key matching that
1230 symlink would fail. Now it is enough that a matching directory entry
1231 exists, symlink or not. (Bugzilla 503.)
1233 PH/03 The body_linecount and body_zerocount variables are now exported in the
1236 PH/04 Added the $dnslist_matched variable.
1238 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1239 This means they are set thereafter only if the connection becomes
1242 PH/06 Added the client_condition to authenticators so that some can be skipped
1243 by clients under certain conditions.
1245 PH/07 The error message for a badly-placed control=no_multiline_responses left
1246 "_responses" off the end of the name.
1248 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1250 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1251 (without spaces) instead of just copying the configuration text.
1253 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1255 PH/11 Added $max_received_linelength.
1257 PH/12 Added +ignore_defer and +include_defer to host lists.
1259 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1260 way in which PCRE > 7.0 is built.
1262 PH/14 Implemented queue_only_load_latch.
1264 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1265 MAIL command. The effect was to mangle the value on 64-bit systems.
1267 PH/16 Another patch from the Sieve maintainer.
1269 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1271 PH/18 If a system quota error occurred while trying to create the file for
1272 a maildir delivery, the message "Mailbox is full" was not appended to the
1273 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1274 only to a quota excession during the actual writing of the file.
1276 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1277 characters?) which causes problems in log lines. The DN values are now
1278 passed through string_printing() before being added to log lines.
1280 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1281 and InterBase are left for another time.)
1283 PH/21 Added message_body_newlines option.
1285 PH/22 Guard against possible overflow in moan_check_errorcopy().
1287 PH/23 POSIX allows open() to be a macro; guard against that.
1289 PH/24 If the recipient of an error message contained an @ in the local part
1290 (suitably quoted, of course), incorrect values were put in $domain and
1291 $local_part during the evaluation of errors_copy.
1297 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1298 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1301 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1302 issue a MAIL command.
1304 PH/02 In an ACL statement such as
1306 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1308 if a client was not listed at all, or was listed with a value other than
1309 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1310 the condition was not true (as it should be), so access was not denied.
1311 The bug was that the ! inversion was incorrectly passed on to the second
1312 item. This has been fixed.
1314 PH/03 Added additional dnslists conditions == and =& which are different from
1315 = and & when the dns lookup returns more than one IP address.
1317 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1318 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1320 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1321 FSYNC, which compiles an option called disable_fsync that allows for
1322 bypassing fsync(). The documentation is heavily laced with warnings.
1324 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1326 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1327 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1328 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1329 including adding "make clean"; (3) Added -fPIC when compiling the test
1330 dynamically loaded module, to get rid of a warning.
1332 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1333 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1334 = 0s. The bug is otherwise harmless.
1336 PH/07 There was a bug in the dovecot authenticator such that the value of
1337 $auth1 could be overwritten, and so not correctly preserved, after a
1338 successful authentication. This usually meant that the value preserved by
1339 the server_setid option was incorrect.
1341 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1343 PH/09 Installed PCRE release 7.0.
1345 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1346 run for batched SMTP input. It is now run at the start of every message
1347 in the batch. While fixing this I discovered that the process information
1348 (output by running exiwhat) was not always getting set for -bs and -bS
1349 input. This is fixed, and it now also says "batched" for BSMTP.
1351 PH/11 Added control=no_pipelining.
1353 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1354 patch, slightly modified), and move the expansion of helo_data till after
1355 the connection is made in the smtp transport (so it can use these
1358 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1360 PH/14 Added log_selector = +pid.
1362 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1364 PH/16 Add ${if forany and ${if forall.
1366 PH/17 Added dsn_from option to vary the From: line in DSNs.
1368 PH/18 Flush SMTP output before performing a callout, unless control =
1369 no_callout_flush is set.
1371 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1372 was true (the default) a successful delivery failed to delete the retry
1373 item, thus causing premature timeout of the address. The bug is now
1376 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1378 PH/21 Long custom messages for fakedefer and fakereject are now split up
1379 into multiline reponses in the same way that messages for "deny" and
1380 other ACL rejections are.
1382 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1383 with slight modification.
1385 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1386 draft, changing the syntax and factoring some duplicate code".
1388 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1389 for deliveries of the second and subsequent messages over the same SMTP
1392 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1393 ${reduce, with only minor "tidies".
1395 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1397 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1398 expansion side effects.
1400 PH/27 When a message times out after an over-quota error from an Exim-imposed
1401 quota, the bounce message says "mailbox is full". This message was not
1402 being given when it was a system quota that was exceeded. It now should
1405 MH/03 Made $recipients available in local_scan(). local_scan() already has
1406 better access to the recipient list through recipients_list[], but
1407 $recipients can be useful in postmaster-provided expansion strings.
1409 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1410 in the case of a MAIL command with additional options following the
1411 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1412 were accidentally chopped off.
1414 PH/29 SMTP synchronization checks are implemented when a command is read -
1415 there is a check that no more input is waiting when there shouldn't be
1416 any. However, for some commands, a delay in an ACL can mean that it is
1417 some time before the response is written. In this time, more input might
1418 arrive, invalidly. So now there are extra checks after an ACL has run for
1419 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1420 pipelining has not been advertised.
1422 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1424 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1425 returned in the SMTP response when smtp_return_error_details was false.
1426 This has been fixed.
1428 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1429 instead of the C I/O that was originally supplied, because problems were
1430 reported on Solaris.
1432 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1433 Exim which did not show up earlier: it was assuming that a call to
1434 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1435 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1436 was a macro that became an assignment, so it seemed to work. This has
1437 changed to a proper function call with a void return, hence the compile
1438 error. Exim's code has been fixed.
1440 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1443 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1445 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1447 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1448 because it turns out to be too big or there is a timeout) there is no
1449 "Completed" line in the log. When some messages of this type were
1450 selected by exigrep, they were listed as "not completed". Others were
1451 picked up by some special patterns. I have improved the selection
1452 criteria to be more general.
1454 PH/38 The host_find_failed option in the manualroute router can now be set
1455 to "ignore", to completely ignore a host whose IP address cannot be
1456 found. If all hosts are ignored, the behaviour is controlled by the new
1457 host_all_ignored option.
1459 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1460 homing or because of multiple MX records with /mx) generated more than
1461 one IP address, and the following item turned out to be the local host,
1462 all the secondary addresses of the first item were incorrectly removed
1463 from the list, along with the local host and any following hosts (which
1464 is what is supposed to happen).
1466 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1467 whoever called Exim into the -H file. In the case of the daemon it was
1468 behaving confusingly. When first started, it used values for whoever
1469 started the daemon, but after a SIGHUP it used the Exim user (because it
1470 calls itself on a restart). I have changed the code so that it now always
1473 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1474 message are rejected with the same error (e.g. no authentication or bad
1475 sender address), and a DATA command is nevertheless sent (as can happen
1476 with PIPELINING or a stupid MUA), the error message that was given to the
1477 RCPT commands is included in the rejection of the DATA command. This is
1478 intended to be helpful for MUAs that show only the final error to their
1481 PH/42 Another patch from the Sieve maintainer.
1483 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1484 Eximstats - Fixed some broken HTML links and added missing column headers
1486 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1487 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1489 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1491 PH/43 Yet another patch from the Sieve maintainer.
1493 PH/44 I found a way to check for a TCP/IP connection going away before sending
1494 the response to the final '.' that terminates a message, but only in the
1495 case where the client has not sent further data following the '.'
1496 (unfortunately, this is allowed). However, in many cases there won't be
1497 any further data because there won't be any more messages to send. A call
1498 to select() can be used: if it shows that the input is "ready", there is
1499 either input waiting, or the socket has been closed. An attempt to read
1500 the next input character can distinguish the two cases. Previously, Exim
1501 would have sent an OK response which the client would never have see.
1502 This could lead to message repetition. This fix should cure that, at
1503 least in a lot of common cases.
1505 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1506 advertised in response to EHLO.
1512 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1513 fixed by 4.65/MH/01 (is this a record?) are fixed:
1515 (i) An empty string was always treated as zero by the numeric comparison
1516 operators. This behaviour has been restored.
1518 (ii) It is documented that the numeric comparison operators always treat
1519 their arguments as decimal numbers. This was broken in that numbers
1520 starting with 0 were being interpreted as octal.
1522 While fixing these problems I realized that there was another issue that
1523 hadn't been noticed. Values of message_size_limit (both the global option
1524 and the transport option) were treated as octal if they started with 0.
1525 The documentation was vague. These values are now always treated as
1526 decimal, and I will make that clear in the documentation.
1532 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1533 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1536 MH/01 Don't check that the operands of numeric comparison operators are
1537 integers when their expansion is in "skipping" mode (fixes bug
1538 introduced by 4.64-PH/07).
1540 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1541 child addresses, Exim now panics and dies. Previously, because the count
1542 is held in a short int, deliveries were likely to be lost. As such a
1543 large number of recipients for a single message is ridiculous
1544 (performance will be very, very poor), I have chosen to impose a limit
1545 rather than extend the field.
1551 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1552 leftover -K file (the existence of which was triggered by #402).
1553 While we were at it, introduced process PID as part of the -K
1554 filename. This should rule out race conditions when creating
1557 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1558 processing considerably. Previous code took too long for large mails,
1559 triggering a timeout which in turn triggers #401.
1561 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1562 in the DK code in transports.c. sendfile() is not really portable,
1563 hence the _LINUX specificness.
1565 TF/01 In the add_headers option to the mail command in an Exim filter,
1566 there was a bug that Exim would claim a syntax error in any
1567 header after the first one which had an odd number of characters
1570 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1571 callout verification, Exim cached a "reject" for the entire domain. This
1572 is correct for most verifications, but it is not correct for a recipient
1573 verification with use_sender or use_postmaster set, because in that case
1574 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1575 case of MAIL FROM:<> rejection from other early rejections (e.g.
1576 rejection of HELO). When verifying a recipient using a non-null MAIL
1577 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1578 Whatever the result of the callout, the value of the domain cache is
1579 left unchanged (for any other kind of callout, getting as far as trying
1580 RCPT means that the domain itself is ok).
1582 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1585 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1586 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1587 session, the daemon ignores ECONNECT errors and logs others; it now
1588 ignores EPIPE as well.
1590 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1591 (quoted-printable decoding).
1593 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1594 later the small subsequent patch to fix an introduced bug.
1596 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1598 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1600 PH/08 An error is now given if message_size_limit is specified negative.
1602 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1603 to be given (somewhat) arbitrary names.
1605 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1608 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1609 miscellaneous code fixes
1611 PH/10 Added the log_reject_target ACL modifier to specify where to log
1614 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1615 hostname. This is wrong, because it relates to the incoming message (and
1616 probably the interface on which it is arriving) and not to the outgoing
1617 callout (which could be using a different interface). This has been
1618 changed to use the value of the helo_data option from the smtp transport
1619 instead - this is what is used when a message is actually being sent. If
1620 there is no remote transport (possible with a router that sets up host
1621 addresses), $smtp_active_hostname is used.
1623 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1624 tweaks were necessary in order to get it to work (see also 21 below):
1625 (a) The code assumed that strncpy() returns a negative number on buffer
1626 overflow, which isn't the case. Replaced with Exim's string_format()
1628 (b) There were several signed/unsigned issues. I just did the minimum
1629 hacking in of casts. There is scope for a larger refactoring.
1630 (c) The code used strcasecmp() which is not a standard C function.
1631 Replaced with Exim's strcmpic() function.
1632 (d) The code set only $1; it now sets $auth1 as well.
1633 (e) A simple test gave the error "authentication client didn't specify
1634 service in request". It would seem that Dovecot has changed its
1635 interface. Fortunately there's a specification; I followed it and
1636 changed what the client sends and it appears to be working now.
1638 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1641 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1642 address A is aliased to B and C, where B exists and C does not. Without
1643 -v the output is "A verified" because verification stops after a
1644 successful redirection if more than one address is generated. However,
1645 with -v the child addresses are also verified. Exim was outputting "A
1646 failed to verify" and then showing the successful verification for C,
1647 with its parentage. It now outputs "B failed to verify", showing B's
1648 parentage before showing the successful verification of C.
1650 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1651 look up a TXT record in a specific list after matching in a combined
1654 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1655 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1656 they consult the DNS. I had assumed they would set it the way they
1657 wanted; and indeed my experiments on Linux seem to show that in some
1658 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1659 To be on the safe side, however, I have now made the interface to
1660 host_find_byname() similar to host_find_bydns(), with an argument
1661 containing the DNS resolver options. The host_find_byname() function now
1662 sets these options at its start, just as host_find_bydns() does. The smtp
1663 transport options dns_qualify_single and dns_search_parents are passed to
1664 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1665 of host_find_byname() use the default settings of RES_DEFNAMES
1666 (qualify_single) but not RES_DNSRCH (search_parents).
1668 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1669 spool_read_header() do less string testing, by means of a preliminary
1670 switch on the second character of optional "-foo" lines. (This is
1671 overdue, caused by the large number of possibilities that now exist.
1672 Originally there were few.) While I was there, I also converted the
1673 str(n)cmp tests so they don't re-test the leading "-" and the first
1674 character, in the hope this might squeeze out yet more improvement.
1676 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1677 flag allowing group syntax was set by the header_syntax check but not
1678 turned off, possible causing trouble later; (2) The flag was not being
1679 set at all for the header_verify test, causing "group"-style headers to
1680 be rejected. I have now set it in this case, and also caused header_
1681 verify to ignore an empty address taken from a group. While doing this, I
1682 came across some other cases where the code for allowing group syntax
1683 while scanning a header line wasn't quite right (mostly, not resetting
1684 the flag correctly in the right place). These bugs could have caused
1685 trouble for malformed header lines. I hope it is now all correct.
1687 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1688 with the "reply" argument non-NULL. The code, however (which originally
1689 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1690 but it didn't always do it. This confused somebody who was copying the
1691 code for some other use. I have removed all the tests.
1693 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1694 feature that was used to support insecure browsers during the U.S. crypto
1695 embargo. It requires special client support, and Exim is probably the
1696 only MTA that supported it -- and would never use it because real RSA is
1697 always available. This code has been removed, because it had the bad
1698 effect of slowing Exim down by computing (never used) parameters for the
1699 RSA_EXPORT functionality.
1701 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1702 authenticator to fail if there's a tab character in the incoming data
1703 (there should never be unless someone is messing about, as it's supposed
1704 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1705 if the connection is using TLS or if the remote IP is the same as the
1706 local IP, and the "valid-client-cert option" if a client certificate has
1709 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1710 authenticators. This can be used for authorization after authentication
1711 succeeds. (In the case of plaintext, it servers for both authentication
1714 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1715 if any retry times were supplied.
1717 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1718 connection, where there is no client IP address to check. In this
1719 situation, the verify now always succeeds.
1721 PH/25 Applied John Jetmore's -Mset patch.
1723 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1725 PH/27 In a string expansion for a processed (not raw) header when multiple
1726 headers of the same name were present, leading whitespace was being
1727 removed from all of them, but trailing whitespace was being removed only
1728 from the last one. Now trailing whitespace is removed from each header
1729 before concatenation. Completely empty headers in a concatenation (as
1730 before) are ignored.
1732 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1733 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1735 PH/29 [Removed. This was a change that I later backed out, and forgot to
1736 correct the ChangeLog entry (that I had efficiently created) before
1737 committing the later change.]
1739 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1740 address errors (4xx response to RCPT) over the same connection as other
1741 messages routed to the same hosts. Such deliveries are always "forced",
1742 so retry times are not inspected. This resulted in far too many retries
1743 for the affected addresses. The effect occurred only when there were more
1744 hosts than the hosts_max_try setting in the smtp transport when it had
1745 the 4xx errors. Those hosts that it had tried were not added to the list
1746 of hosts for which the message was waiting, so if all were tried, there
1747 was no problem. Two fixes have been applied:
1749 (i) If there are any address or message errors in an SMTP delivery, none
1750 of the hosts (tried or untried) are now added to the list of hosts
1751 for which the message is waiting, so the message should not be a
1752 candidate for sending over the same connection that was used for a
1753 successful delivery of some other message. This seems entirely
1754 reasonable: after all the message is NOT "waiting for some host".
1755 This is so "obvious" that I'm not sure why it wasn't done
1756 previously. Hope I haven't missed anything, but it can't do any
1757 harm, as the worst effect is to miss an optimization.
1759 (ii) If, despite (i), such a delivery is accidentally attempted, the
1760 routing retry time is respected, so at least it doesn't keep
1761 hammering the server.
1763 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1764 in ${readsocket because some servers need this prod.
1766 PH/32 Added some extra debug output when updating a wait-xxx database.
1768 PH/33 The hint "could be header name not terminated by colon", which has been
1769 given for certain expansion errors for a long time, was not being given
1770 for the ${if def:h_colon_omitted{... case.
1772 PH/34 The spec says: "With one important exception, whenever a domain list is
1773 being scanned, $domain contains the subject domain." There was at least
1774 one case where this was not true.
1776 PH/35 The error "getsockname() failed: connection reset by peer" was being
1777 written to the panic log as well as the main log, but it isn't really
1778 panic-worthy as it just means the connection died rather early on. I have
1779 removed the panic log writing for the ECONNRESET error when getsockname()
1782 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1783 runs only) independently of the message's sender address. This meant
1784 that, if the 4xx error was in fact related to the sender, a different
1785 message to the same recipient with a different sender could confuse
1786 things. In particualar, this can happen when sending to a greylisting
1787 server, but other circumstances could also provoke similar problems.
1788 I have changed the default so that the retry time for these errors is now
1789 based a combination of the sender and recipient addresses. This change
1790 can be overridden by setting address_retry_include_sender=false in the
1793 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1794 remote server are returned as part of bounce messages. This was not
1795 happening for LMTP over a pipe (the lmtp transport), but now it is the
1796 same for both kinds of LMTP.
1798 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1799 in header lines that were in fact CNAMEs. This is no longer the case.
1801 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1802 and queue runs started by the daemon processed all messages. This has
1803 been fixed so that -R and -S can now usefully be given with -q<time>.
1805 PH/40 Import PCRE release 6.7 (fixes some bugs).
1807 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1809 PH/42 Give an error if -q is specified more than once.
1811 PH/43 Renamed the variables $interface_address and $interface_port as
1812 $received_ip_address and $received_port, to make it clear that these
1813 values apply to message reception, and not to the outgoing interface when
1814 a message is delivered. (The old names remain recognized, of course.)
1816 PH/44 There was no timeout on the connect() call when using a Unix domain
1817 socket in the ${readsocket expansion. There now is.
1819 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1820 be meaningful with "accept".
1822 SC/01 Eximstats V1.43
1823 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1825 SC/02 Eximstats V1.44
1826 Use a glob alias rather than an array ref in the generated
1827 parser. This improves both readability and performance.
1829 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1830 Collect SpamAssassin and rejection statistics.
1831 Don't display local sender or destination tables unless
1832 there is data to show.
1833 Added average volumes into the top table text output.
1835 SC/04 Eximstats V1.46
1836 Collect data on the number of addresses (recipients)
1837 as well as the number of messages.
1839 SC/05 Eximstats V1.47
1840 Added 'Message too big' to the list of mail rejection
1841 reasons (thanks to Marco Gaiarin).
1843 SC/06 Eximstats V1.48
1844 Mainlog lines which have GMT offsets and are too short to
1845 have a flag are now skipped.
1847 SC/07 Eximstats V1.49 (Alain Williams)
1848 Added the -emptyok flag.
1850 SC/08 Eximstats V1.50
1851 Fixes for obtaining the IP address from reject messages.
1853 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1854 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1855 whitesspace changes from 4.64-PH/27
1857 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1860 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1861 are found, allow negative numbers in numeric criteria)
1863 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1865 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1866 to match changes made in 4.64-PH/43
1868 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1870 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1871 "domain" is actually an IP address, and force a failure. This locks out
1872 those revolvers/nameservers that support "A-for-A" lookups, in
1873 contravention of the specifications.
1875 PH/48 When a host name was looked up from an IP address, and the subsequent
1876 forward lookup of the name timed out, the host name was left in
1877 $sender_host_name, contrary to the specification.
1879 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1880 restricted to single-key lookups, Exim was not diagnosing an error if
1881 * or *@ was used with a query-style lookup.
1883 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1885 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1886 long ago, but noone interested enough thought of it. Let's just say that
1887 the "1.1" means that there are some new functions that weren't there at
1888 some point in the past.
1890 PH/51 Error processing for expansion failure of helo_data from an smtp
1891 transport during callout processing was broken.
1893 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1894 tested/used via the -bh/-bhc/-bs options.
1896 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1897 bug, fixed in subsequent PCRE releases).
1899 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1900 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1902 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1908 SC/01 Use a glob alias rather than an array ref in eximstats generated
1909 parser. This improves both readability and performance.
1911 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1912 Don't display local sender or destination tables in eximstats unless
1913 there is data to show.
1914 Added average volumes into the eximstats top table text output.
1916 SC/03 Collect data on the number of addresses (recipients) as well
1917 as the number of messages in eximstats.
1919 TF/01 Correct an error in the documentation for the redirect router. Exim
1920 does (usually) call initgroups() when daemonizing.
1922 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1923 with consistent privilege compared to when running as a daemon.
1925 TF/03 Note in the spec that $authenticated_id is not set for local
1926 submissions from trusted users.
1928 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1929 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1931 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1932 by adding some example configuration directives to the default
1933 configuration file. A little bit of work is required to uncomment the
1934 directives and define how usernames and passwords are checked, but
1935 there is now a framework to start from.
1937 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1938 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1939 without this. I don't know how relevant this is to other LDAP libraries.
1941 PH/02 Add the verb name to the "unknown ACL verb" error.
1943 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1945 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1947 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1948 directory not expanded when it should be if an expanded home directory
1949 was set for the address (which is overridden by the transport).
1951 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1954 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1955 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1956 because it is too late at that time, and has no effect.
1958 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1959 security issue with \' (bugzilla #107). I could not use the
1960 PQescapeStringConn() function, because it needs a PGconn value as one of
1963 PH/08 When testing addresses using -bt, indicate those final addresses that
1964 are duplicates that would not cause an additional delivery. At least one
1965 person was confused, thinking that -bt output corresponded to deliveries.
1966 (Suppressing duplicates isn't a good idea as you lose the information
1967 about possibly different redirections that led to the duplicates.)
1969 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1970 systems where poll() doesn't work, in particular OS X.
1972 PH/10 Added more information to debugging output for retry time not reached.
1974 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1975 operations in malware.c.
1977 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1980 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1981 syslog_duplication set false, log lines that would normally be written
1982 both the the main log and to the reject log were not written to syslog at
1985 PH/14 In the default configuration, change the use of "message" in ACL warn
1986 statements to "add_header".
1988 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1989 not followed by a command (e.g. "seen endif").
1991 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1992 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1995 PH/17 Added extra conditions to the default value of delay_warning_condition
1999 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2000 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2001 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2004 The Auto-Submitted: and various List- headers are standardised, whereas I
2005 don't think Precedence: ever was.
2007 PH/18 Refactored debugging code in route_finduser() to show more information,
2008 in particular, the error code if getpwnam() issues one.
2010 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2011 This is apparently needed in addition to the PH/07 change above to avoid
2012 any possible encoding problems.
2014 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2015 but not after initializing Perl.
2017 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2018 output them only if debugging. By default they are written stderr,
2019 apparently, which is not desirable.
2021 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2024 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2027 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2029 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2030 authenticated or an ident call has been made. Suppress the default
2031 values for $authenticated_id and $authenticated_sender (but permit -oMai
2032 and -oMas) when testing with -bh.
2034 PH/24 Re-jigged the order of the tests in the default configuration so that the
2035 tests for valid domains and recipients precede the DNS black list and CSA
2036 tests, on the grounds that those ones are more expensive.
2038 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2039 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2040 This bug exists in every version of Exim that I still have, right back to
2043 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2044 However, an attempt to turn on case-sensitivity in a regex key by
2045 including (?-i) didn't work because the subject string was already
2046 lowercased, and the effects were non-intuitive. It turns out that a
2047 one-line patch can be used to allow (?-i) to work as expected.
2053 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2054 other effects) broke the use of negated acl sub-conditions.
2056 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2059 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2060 "Deny" causes Exim to reject the incoming connection with a 554 error.
2061 Unfortunately, if there is a major crisis, such as a disk failure,
2062 tcp-wrappers gives "deny", whereas what one would like would be some
2063 kind of temporary error. A kludge has been added to help with this.
2064 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2065 554 error is used if errno is still zero or contains ENOENT (which occurs
2066 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2069 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2071 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2072 errors. Otherwise a message that provokes a temporary error (when other
2073 messages do not) can cause a whole host to time out.
2075 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2076 addresses were routed directly to files or pipes from a redirect router.
2077 File deliveries just didn't batch; pipe deliveries might have suffered
2080 PH/06 A failure to get a lock for a hints database would erroneously always say
2081 "Failed to get write lock", even when it was really a read lock.
2083 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2084 of 0600. This has been changed to use the value of the lockfile_mode
2085 option (which defaults to 0600).
2087 PH/08 Applied small patch from the Sieve maintainer.
2089 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2090 folder from quota calculations, a direct delivery into this folder messed
2091 up the contents of the maildirsize file. This was because the regex was
2092 used only to exclude .Trash (or whatever) when the size of the mailbox
2093 was calculated. There was no check that a delivery was happening into an
2094 excluded directory. This bug has been fixed by ignoring all quota
2095 processing for deliveries into excluded directories.
2097 PH/10 Added the maildirfolder_create_regex option to appendfile.
2103 PH/01 The code for finding all the local interface addresses on a FreeBSD
2104 system running IPv6 was broken. This may well have applied to all BSD
2105 systems, as well as to others that have similar system calls. The broken
2106 code found IPv4 interfaces correctly, but gave incorrect values for the
2107 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2108 that it would not match correctly against @[] and not recognize the IPv6
2111 PH/02 The ipliteral router was not recognizing addresses of the form user@
2112 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2114 PH/03 Added disable_ipv6.
2116 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2117 decoded form, because it is most often used to construct To: headers
2118 lines in autoreplies, and the decoded form may well be syntactically
2119 invalid. However, $reply_address has leading white space removed, and all
2120 newlines turned into spaces so that the autoreply transport does not
2123 PH/05 If group was specified without a user on a router, and no group or user
2124 was specified on a transport, the group from the router was ignored.
2126 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2127 for visible compile-time settings that can be used to change these
2128 numbers, for those that want even more. Backwards compatibility with old
2129 spool files has been maintained. However, going back to a previous Exim
2130 release will lost any variables that are in spool files.
2132 PH/07 Two small changes when running in the test harness: increase delay when
2133 passing a TCP/IP connection to a new process, in case the original
2134 process has to generate a bounce, and remove special handling of
2135 127.0.0.2 (sic), which is no longer necessary.
2137 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2138 be the same on different OS.
2140 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2143 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2144 whether --show-vars was specified or not
2146 JJ/02 exipick: Added support for new ACL variable spool format introduced
2149 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2150 syntactically invalid From: or Reply-to: line, and a filter used this to
2151 generate an autoreply, and therefore failed to obtain an address for the
2152 autoreply, Exim could try to deliver to a non-existent relative file
2153 name, causing unrelated and misleading errors. What now happens is that
2154 it logs this as a hard delivery error, but does not attempt to create a
2157 PH/11 The exinext utility has a -C option for testing purposes, but although
2158 the given file was scanned by exinext itself; it wasn't being passed on
2159 when Exim was called.
2161 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2162 an end-of-file indication when reading a command response.
2164 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2165 compiled. In many other places in Exim, IPv6 addresses are always
2166 recognized, so I have changed this. It also means that IPv4 domain
2167 literals of the form [IPV4:n.n.n.n] are now always recognized.
2169 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2170 used if the router is not running as root, for example, when verifying at
2171 ACL time, or when using -bh. The debugging output from this situation was
2172 non-existent - all you got was a failure to exec. I have made two
2175 (a) Failures to set uid/gid, the current directory, or a process leader
2176 in a subprocess such as that created by queryprogram now generate
2177 suitable debugging ouput when -d is set.
2179 (b) The queryprogram router detects when it is not running as root,
2180 outputs suitable debugging information if -d is set, and then runs
2181 the subprocess without attempting to change uid/gid.
2183 PH/15 Minor change to Makefile for building test_host (undocumented testing
2186 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2187 additional section of a DNS packet that returns MX or SRV records.
2188 Instead, it always explicitly searches for A/AAAA records. This avoids
2189 major problems that occur when a DNS server includes only records of one
2190 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2191 fixed another bug: if SRV records were looked up and the corresponding
2192 address records were *not* found in the additional section, the port
2193 values from the SRV records were lost.
2195 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2196 using the correct key (the original address) when searching the retry
2197 rules in order to find which one to use for generating the retry hint.
2199 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2200 adding the default one. Similarly, if it contains a Reply-To: header, the
2201 errors_reply_to option, if set, is not used.
2203 PH/19 When calculating a retry time, Exim used to measure the "time since
2204 failure" by looking at the "first failed" field in the retry record. Now
2205 it does not use this if it is later than than the arrival time of the
2206 message. Instead it uses the arrival time. This makes for better
2207 behaviour in cases where some deliveries succeed, thus re-setting the
2208 "first failed" field. An example is a quota failure for a huge message
2209 when small messages continue to be delivered. Without this change, the
2210 "time since failure" will always be short, possible causing more frequent
2211 delivery attempts for the huge message than are intended.
2212 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2214 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2215 $1, $2, $3) because the numerical variables can be reset during some
2216 expansion items (e.g. "match"), thereby losing the authentication data.
2218 PH/21 Make -bV show the size of off_t variables so that the test suite can
2219 decide whether to run tests for quotas > 2G.
2221 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2222 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2223 filecount value is greater than 2G or if a quota value is greater than 2G
2224 on a system where the size of off_t is not greater than 4, a panic error
2227 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2228 never match. The debug and -bh output now contains an explicit error
2229 message indicating a malformed IPv4 address or mask.
2231 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2232 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2233 PH/23 above applies.
2235 PH/25 Do not write to syslog when running in the test harness. The only
2236 occasion when this arises is a failure to open the main or panic logs
2237 (for which there is an explicit test).
2239 PH/26 Added the /no_tell option to "control=freeze".
2241 PH/27 If a host name lookup failed very early in a connection, for example, if
2242 the IP address matched host_lookup and the reverse lookup yielded a name
2243 that did not have a forward lookup, an error message of the form "no IP
2244 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2245 could be logged. Now it outputs the IP address instead of "NULL".
2247 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2248 allows the sender and the authenticated sender to be set when
2249 submitting a message from within Exim. Since child_open_exim() is
2250 documented for local_scan(), the new function should be too.
2252 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2253 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2254 results in an empty string is now treated as unset.
2256 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2258 PH/31 Added sender_verify_fail logging option.
2260 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2261 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2262 tidied the source and removed it altogether.
2264 PH/33 When a queue run was abandoned because the load average was too high, a
2265 log line was always written; now it is written only if the queue_run log
2266 selector is set. In addition, the log line for abandonment now contains
2267 information about the queue run such as the pid. This is always present
2268 in "start" and "stop" lines but was omitted from the "abandon" line.
2270 PH/34 Omit spaces between a header name and the colon in the error message that
2271 is given when verify = headers_syntax fails (if there are lots of them,
2272 the message gets confusing).
2274 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2275 names, as there are now some PTR records that contain slashes. This check
2276 is only to protect against broken name servers that fall over on strange
2277 characters, so the fact that it applies to all lookups doesn't matter.
2279 PH/36 Now that the new test suite is complete, we can remove some of the
2280 special code in Exim that was needed for the old test suite. For example,
2281 sorting DNS records because real resolvers return them in an arbitrary
2282 order. The new test suite's fake resolver always returns records in the
2285 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2286 bounces) except when queue_only is set, to avoid logging races between
2287 the different processes.
2289 PH/38 Panic-die if .include specifies a non-absolute path.
2291 PH/39 A tweak to the "H" retry rule from its user.
2293 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2294 a label. They prevented compilation on older perls.
2296 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2297 a warning to be raised on newish perls.
2299 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2300 on queue. Changes to match documented behaviour of showing count of
2301 messages matching specified criteria.
2303 PH/40 Changed the default ident timeout from 30s to 5s.
2305 PH/41 Added support for the use of login_cap features, on those BSD systems
2306 that have them, for controlling the resources used by pipe deliveries.
2308 PH/42 The content-scanning code uses fopen() to create files in which to put
2309 message data. Previously it was not paying any attention to the mode of
2310 the files. Exim runs with umask(0) because the rest of the code creates
2311 files with open(), and sets the required mode explicitly. Thus, these
2312 files were ending up world-writeable. This was not a big issue, because,
2313 being within the spool directory, they were not world-accessible. I have
2314 created a function called modefopen, which takes an additional mode
2315 argument. It sets umask(777), creates the file, chmods it to the required
2316 mode, then resets the umask. All the relevant calls to fopen() in the
2317 content scanning code have been changed to use this function.
2319 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2320 to 24 hours. This avoids potential overflow problems when processing G
2321 and H retry rules. I suspect nobody ever tinkers with this value.
2323 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2325 PH/45 When the plaintext authenticator is running as a client, the server's
2326 challenges are checked to ensure they are valid base64 strings. By
2327 default, the authentication attempt is cancelled if an invalid string is
2328 received. Setting client_ignore_invalid_base64 true ignores these errors.
2329 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2330 they are received. Thus, the responses can be made to depend on the
2331 challenges. If an invalid string is ignored, an empty string is placed in
2334 PH/46 Messages that are created by the autoreply transport now contains a
2335 References: header, in accordance with RFCs 2822 and 3834.
2337 PH/47 Added authenticated_sender_force to the smtp transport.
2339 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2341 PH/49 Installed latest patch from the Sieve maintainer.
2343 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2344 was also set, the appendfile transport was unnecessarily scanning a
2345 directory of message files (e.g. for maildir delivery) to find the count
2346 of files (along with the size), even though it did not need this
2347 information. It now does the scan only if it needs to find either the
2348 size of the count of files.
2350 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2352 PH/52 Two bugs concerned with error handling when the smtp transport is
2355 (i) Exim was not creating retry information for temporary errors given
2356 for individual recipients after the DATA command when the smtp transport
2357 was used in LMTP mode. This meant that they could be retried too
2358 frequently, and not timed out correctly.
2360 (ii) Exim was setting the flag that allows error details to be returned
2361 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2362 recipients that were returned after the DATA command.
2364 PH/53 This is related to PH/52, but is more general: for any failing address,
2365 when detailed error information was permitted to be returned to the
2366 sender, but the error was temporary, then after the final timeout, only
2367 "retry timeout exceeded" was returned. Now it returns the full error as
2368 well as "retry timeout exceeded".
2370 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2371 do this, and (what is worse) MTAs that accept it.
2373 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2374 will now be deprecated.
2376 PH/56 New os.c-cygwin from the Cygwin maintainer.
2378 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2379 formats (previously only available in exim formats via -bpr, -bpru,
2380 and -bpra. Now also available in native and exiqgrep formats)
2382 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2383 with very large, slow to parse queues
2385 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2387 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2389 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2390 responses to authentication challenges, though it was showing the
2391 challenges; (ii) I've removed the CR characters from the debug output for
2394 PH/58 Allow for the insertion of a newline as well as a space when a string
2395 is turned into more than one encoded-word during RFC 2047 encoding. The
2396 Sieve code now uses this.
2398 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2399 data_4xx, lost_connection, tls_required.
2401 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2402 message was being sent as an SMTP response.
2404 PH/61 Add -l and -k options to exicyclog.
2406 PH/62 When verifying, if an address was redirected to one new address, so that
2407 verification continued, and the new address failed or deferred after
2408 having set something in $address_data, the value of $address_data was not
2409 passed back to the ACL. This was different to the case when no
2410 redirection occurred. The value is now passed back in both cases.
2412 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2413 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2414 use login_cap.h, so on its own it isn't the distinguishing feature. The
2415 new name refers directly to the setclassresources() function.
2417 PH/65 Added configuration files for NetBSD3.
2419 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2421 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2422 is preferred over IPv4.
2424 PH/68 The bounce_return_message and bounce_return_body options were not being
2425 honoured for bounces generated during the reception of non-SMTP messages.
2426 In particular, this applied to messages rejected by the ACL. This bug has
2427 been fixed. However, if bounce_return_message is true and bounce_return_
2428 body is false, the headers that are returned for a non-SMTP message
2429 include only those that have been read before the error was detected.
2430 (In the case of an ACL rejection, they have all been read.)
2432 PH/69 The HTML version of the specification is now built in a directory called
2433 spec_html instead of spec.html, because the latter looks like a path with
2434 a MIME-type, and this confuses some software.
2436 PH/70 Catch two compiler warnings in sieve.c.
2438 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2439 function verify_get_ident() calls ip_connect() to connect a socket, but
2440 if the "connect()" function timed out, ip_connect() used to close the
2441 socket. However, verify_get_ident() also closes the socket later, and in
2442 between Exim writes to the log, which may get opened at this point. When
2443 the socket was closed in ip_connect(), the log could get the same file
2444 descriptor number as the socket. This naturally causes chaos. The fix is
2445 not to close the socket in ip_connect(); the socket should be closed by
2446 the function that creates it. There was only one place in the code where
2447 this was missing, in the iplookup router, which I don't think anybody now
2448 uses, but I've fixed it anyway.
2450 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2451 well as to direct DNS lookups. Otherwise the handling of names in host
2452 lists is inconsistent and therefore confusing.
2458 PH/01 Two changes to the default runtime configuration:
2460 (1) Move the checks for relay_from_hosts and authenticated clients from
2461 after to before the (commented out) DNS black list checks.
2463 (2) Add control=submission to the relay_from_hosts and authenticated
2464 clients checks, on the grounds that messages accepted by these
2465 statements are most likely to be submissions.
2467 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2469 (1) Generate an error if the third argument for the ${prvs expansion is
2472 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2475 (3) Reset the variables that are obtained from the first argument of
2476 ${prvscheck and used in the second argument before leaving the code,
2477 because their memory is reclaimed, so using them afterwards may do
2480 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2481 one (it's much easier than Tom thought :-).
2483 (5) Because of (4), we can now allow for the use of $prvscheck_result
2484 inside the third argument.
2486 PH/03 For some reason, the default setting of PATH when running a command from
2487 a pipe transport was just "/usr/bin". I have changed it to
2490 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2491 anything to be listed in the output from -bV.
2493 PH/05 When a filter generated an autoreply, the entire To: header line was
2494 quoted in the delivery log line, like this:
2496 => >A.N.Other <ano@some.domain> <original@ddress> ...
2498 This has been changed so that it extracts the operative address. There
2499 may be more than one such address. If so, they are comma-separated, like
2502 => >ano@some.domain,ona@other.domain <original@ddress> ...
2504 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2505 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2506 not being looked up in the rDNS to get a host name, Exim was showing the
2507 IP address twice in Received: lines, even though the IP addresses were
2508 identical. For example:
2510 Received: from [1.2.3.4] (helo=[1.2.3.4])
2512 However, if the real host name was known, it was omitting the HELO data
2513 if it matched the actual IP address. This has been tidied up so that it
2514 doesn't show the same IP address twice.
2516 PH/07 When both +timestamp and +memory debugging was on, the value given by
2517 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2518 called by the debug printing, thereby overwriting the timestamp buffer.
2519 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2521 PH/08 When the original message was included in an autoreply transport, it
2522 always said "this is a copy of the message, including all the headers",
2523 even if body_only or headers_only was set. It now gives an appropriate
2526 PH/09 Applied a patch from the Sieve maintainer which:
2528 o fixes some comments
2529 o adds the (disabled) notify extension core
2530 o adds some debug output for the result of if/elsif tests
2531 o points to the current vacation draft in the documentation
2532 and documents the missing references header update
2536 o fixes a bug in processing the envelope test (when testing
2537 multiple envelope elements, the last element determinted the
2540 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2541 Electronic Mail") by including:
2543 Auto-submitted: auto-generated
2545 in the messages that it generates (bounce messages and others, such as
2546 warnings). In the case of bounce messages for non-SMTP mesages, there was
2547 also a typo: it was using "Auto_submitted" (underscore instead of
2548 hyphen). Since every message generated by Exim is necessarily in response
2549 to another message, thes have all been changed to:
2551 Auto-Submitted: auto-replied
2553 in accordance with these statements in the RFC:
2555 The auto-replied keyword:
2557 - SHOULD be used on messages sent in direct response to another
2558 message by an automatic process,
2560 - MUST NOT be used on manually-generated messages,
2562 - MAY be used on Delivery Status Notifications (DSNs) and Message
2563 Disposition Notifications (MDNs),
2565 - MUST NOT be used on messages generated by automatic or periodic
2566 processes, except for messages which are automatic responses to
2569 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2570 to the default Received: header definition.
2572 PH/12 Added log selector acl_warn_skipped (default on).
2574 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2575 variables because (a) they are in the wrong storage pool and (b) even if
2576 they were copied, it wouldn't work properly because of the caching.
2578 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2579 checking when decoding. Apparently there are clients that generate
2580 overlong encoded strings. Why am I not surprised?
2582 PH/15 If the first argument of "${if match_address" was not empty, but did not
2583 contain an "@" character, Exim crashed. Now it writes a panic log message
2584 and treats the condition as false.
2586 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2588 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2589 extension "envelope-auth". The code is finished and in agreement with
2590 other implementations, but there is no documentation so far and in fact,
2591 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2592 not changing the active code.
2594 Print executed "if" and "elsif" statements when debugging is used. This
2595 helps a great deal to understand what a filter does.
2597 Document more things not specified clearly in RFC3028. I had all this
2598 sorted out, when out of a sudden new issues came to my mind. Oops."
2600 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2603 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2604 canonical form (as documented). However, after a host name lookup from
2605 the IP address, check_host() was doing a simple string comparison with
2606 addresses acquired from the DNS when checking that the found name did
2607 have the original IP as one of its addresses. Since any found IPv6
2608 addresses are likely to be in abbreviated form, the comparison could
2609 fail. Luckily, there already exists a function for doing the comparison
2610 by converting both addresses to binary, so now that is used instead of
2611 the text comparison.
2613 PH/20 There was another similar case to PH/19, when a complete host name was
2614 given in a host list; looking up its IP address could give an abbreviated
2615 form, whereas the current host's name might or might not be abbreviated.
2616 The same fix has been applied.
2622 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2623 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2626 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2627 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2629 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2631 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2632 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2633 Also, for "Auto-Submitted", treat anything other than "no" as
2634 non-personal, in accordance with RFC 3834. (Previously it treated
2635 anything starting "auto-" as non-personal.)
2637 TF/01 The control=submission/name=... option had a problem with syntax
2638 errors if the name included a slash character. The /name= option
2639 now slurps the rest of the string, so it can include any characters
2640 but it must come last in the list of options (after /sender_retain
2643 PH/05 Some modifications to the interface to the fake nameserver for the new
2651 TK/01 Added the "success_on_redirect" address verification option. See
2652 NewStuff for rationale and an example.
2654 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2656 PH/02 Patch to exigrep to allow it to work on syslog lines.
2658 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2659 fread() to skip over the body file's header line, because in Cygwin the
2660 header line is locked and is inaccessible.
2662 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2663 co-exist for some time) to make it clear that it is the Exim ID that is
2664 referenced, not the Message-ID: header line.
2666 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2667 string_format() function, because snprintf() does not exist on all
2670 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2671 ${stat: expansion item.
2673 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2674 protocol synchronization error", to keep the pedants happy.
2676 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2677 well as for IRIX systems, when gcc is being used. See the host.c source
2680 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2682 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2685 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2686 transport and to the smtp transport in LMTP mode.
2688 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2690 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2691 run a filter in a subprocess. This could lead to confusion in subsequent
2692 lookups in the parent process. There should also be a search_tidyup() at
2693 the end of the subprocess.
2695 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2696 only if the host matched helo_try_verify_hosts, which caused the
2697 verification to occur when the EHLO/HELO command was issued. The ACL just
2698 tested the remembered result. Now, if a previous verification attempt has
2699 not happened, "verify = helo" does it there and then.
2701 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2703 TK/03 Fix log output including CR from clamd.
2705 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2706 exist provoked a memory error which could cause a segfault.
2708 PH/15 Installed PCRE 6.2
2710 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2712 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2713 of the problem. Specifically, suggested +O2 rather than +O1 for the
2716 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2718 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2719 record, the port was not correctly logged when the outgoing_port log
2720 selector was set (it logged the transort's default port).
2722 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2723 fallback_hosts, and "hosts" in the smtp transport.
2725 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2726 host errors such as "Connection refused".
2728 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2729 authentication with radiusclient 0.4.9:
2731 - Error returned from rc_read_config was caught wrongly
2732 - Username/password not passed on to radius server due to wrong length.
2734 The presumption is that some radiusclient API changes for 4.51/PH/17
2735 were not taken care of correctly. The code is still untested by me (my
2736 Linux distribution still has 0.3.2 of radiusclient), but it was
2737 contributed by a Radius user.
2739 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2740 expanding the "port" option of the smtp transport.
2742 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2743 while reading a MIME header. Thanks to Tom Hughes for a patch.
2745 PH/24 Include config.h inside local_scan.h so that configuration settings are
2748 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2749 that in an ACL for RCPT (for example), you can examine exactly what was
2752 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2753 commands, but it was not correctly comparing the address with the actual
2754 client host address. Thus, it would show the EHLO address in Received:
2755 header lines when this was not necessary.
2757 PH/27 Added the % operator to ${eval:}.
2759 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2760 it should be ignoring failures (because with -C, for example, it has lost
2761 privilege). It wasn't ignoring creation failures other than "already
2764 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2767 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2768 because an input line was too long, either on its own, or by virtue of
2769 too many continuations, the temporary file was not being removed, and the
2770 return code was incorrect.
2772 PH/31 Missing "BOOL" in function definition in filtertest.c.
2774 PH/32 Applied Sieve patches from the maintainer.
2776 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2778 PH/33 Added "verify = not_blind".
2780 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2781 Local/Makefile (with some defaults set). These are used in built scripts
2782 such as exicyclog, but they have never been used in the exim_install
2783 script (though there are many overriding facilities there). I have
2784 arranged that the exim_install script now takes note of these two
2787 PH/35 Installed configuration files for Dragonfly.
2789 PH/36 When a locally submitted message by a trusted user did not contain a
2790 From: header, and the sender address was obtained from -f or from an SMTP
2791 MAIL command, and the trusted user did not use -F to supply a sender
2792 name, $originator_name was incorrectly used when constructing a From:
2793 header. Furthermore, $originator_name was used for submission mode
2794 messages from external hosts without From: headers in a similar way,
2795 which is clearly wrong.
2797 PH/37 Added control=suppress_local_fixups.
2799 PH/38 When log_selector = +received_sender was set, and the addition of the
2800 sender made the log line's construction buffer exactly full, or one byte
2801 less than full, an overflow happened when the terminating "\n" was
2804 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2805 when the result of a list match is failure because a DNS lookup failed.
2807 PH/40 RM_COMMAND is now used in the building process.
2809 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2810 the "build-* directories that it finds.
2812 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2813 address in a domain literal was a prefix of an interface address.
2815 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2816 when verifying a sender address, unless rewrite_headers is false.
2818 PH/44 Wrote a long comment about why errors_to addresses are verified as
2819 recipients, not senders.
2821 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2822 the ratelimit ACL was added.
2824 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2826 PH/46 Added extra information about PostgreSQL errors to the error string.
2828 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2829 avoiding the need to install special zones in a real server. This is
2830 backwards compatible; if it can't find the fake resolver, it drops back.
2831 Thus, both old and new test suites can be run.
2833 TF/02 Added util/ratelimit.pl
2835 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2836 clock is set back in time.
2838 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2839 Candler <B.Candler@pobox.com>.
2841 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2842 OK for addresses that are the result of redirections.
2844 PH/48 A number of further additions for the benefit of the new test suite,
2845 including a fake gethostbyname() that interfaces to the fake DNS resolver
2848 TF/06 The fix for widen_domains has also been applied to qualify_single and
2849 search_parents which are the other dnslookup options that can cause
2852 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2855 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2857 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2858 service" if it encountered an MX record with an empty target hostname.
2859 The message is now "an MX or SRV record indicated no SMTP service".
2861 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2862 if the DNS of the sending site is misconfigured. This is quite a
2863 common situation. This change restores the behaviour of treating a
2864 helo verification defer as a failure.
2866 PH/51 If self=fail was set on a router, the bounce message did not include the
2867 actual error message.
2873 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2875 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2876 command itself ended in error, the underlying message about the transport
2877 filter timeout was being overwritten with the pipe command error. Now the
2878 underlying error message should be appended to the second error message.
2880 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2882 PH/02 Reduce the amount of output that "make" produces by default. Full output
2883 can still be requested.
2885 PH/03 The warning log line about a condition test deferring for a "warn" verb
2886 was being output only once per connection, rather than after each
2887 occurrence (because it was using the same function as for successful
2888 "warn" verbs). This seems wrong, so I have changed it.
2890 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2891 it should not have, which might have caused a crash in the right
2892 circumstances, but probably never did.
2894 PH/04 Installed a modified version of Tony Finch's patch to make submission
2895 mode fix the return path as well as the Sender: header line, and to
2896 add a /name= option so that you can make the user's friendly name appear
2899 TF/03 Added the control = fakedefer ACL modifier.
2901 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2902 Mark Lowes for thorough testing.
2904 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2906 TK/03 Merged latest SRS patch from Miles Wilton.
2908 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2909 with the definition in sysexits.h (which is #included earlier).
2910 Fortunately, Exim does not actually use EX_OK. The code used to try to
2911 preserve the sysexits.h value, by assumimg that macro definitions were
2912 scanned for macro replacements. I have been disabused of this notion,
2913 so now the code just undefines EX_OK before #including unistd.h.
2915 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2916 in the smtp transport. When a block could not be written in a single
2917 write() function, the timeout was being re-applied to each part-write.
2918 This seems wrong - if the receiver was accepting one byte at a time it
2919 would take for ever. The timeout is now adjusted when this happens. It
2920 doesn't have to be particularly precise.
2922 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2923 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2925 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2926 without a domain if the check to <postmaster@domain> fails.
2928 SC/01 Eximstats: added -xls and the ability to specify output files
2929 (patch written by Frank Heydlauf).
2931 SC/02 Eximstats: use FileHandles for outputing results.
2933 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2935 SC/04 Eximstats: fixed display of large numbers with -nvr option
2937 SC/05 Eximstats: fixed merging of reports with empty tables.
2939 SC/06 Eximstats: added the -include_original_destination flag
2941 SC/07 Eximstats: removed tabs and trailing whitespace.
2943 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2945 TK/06 MBOX spool code: Add real "From " MBOX separator line
2946 so the .eml file is really in mbox format (even though
2947 most programs do not really care). Patch from Alex Miller.
2949 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2950 The latter is generated from $received_to and is only set if the
2951 message has one envelope recipient. SA can use these headers,
2952 obviously out-of-the-box. Patch from Alex Miller.
2954 PH/08 The ${def test on a variable was returning false if the variable's
2955 value was "0", contrary to what the specification has always said!
2956 The result should be true unless the variable is empty.
2958 PH/09 The syntax error of a character other than { following "${if
2959 def:variable_name" (after optional whitespace) was not being diagnosed.
2960 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2961 accidental colon was present, for example, could give incorrect results.
2963 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2964 result is used (not including appendfile, where other changes are about
2967 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2968 This involved changing a lot of size variables from int to off_t. It
2969 should work with maildirs and everything.
2971 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2972 spamd dying while we are connected to it.
2974 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2977 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2978 function for BDB 4.3.
2980 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2982 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2983 the sources for PCRE have been re-arranged and more files are now
2986 PH/15 The code I had for printing potentially long long variables in PH/11
2987 above was not the best (it lost precision). The length of off_t variables
2988 is now inspected at build time, and an appropriate printing format (%ld
2989 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2990 to be "long long int" or "long int". This is needed for the internal
2991 formatting function string_vformat().
2993 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2994 the configuration file to be ":syslog", then the script "guesses" where
2995 the logs files are, rather than using the compiled in default. In our
2996 case the guess is not the same as the compiled default, so the script
2997 suddenly stopped working when I started to use syslog. The patch checks
2998 to see if log_file_path is "". If so, it attempts to read it from exim
2999 with no configuration file to get the compiled in version, before it
3000 falls back to the previous guessing code."
3002 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3003 implementing BATV in an Exim configuration. See NewStuff for the gory
3006 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3007 Makefile that are specific to HP-UX.
3009 PH/18 If the "use_postmaster" option was set for a recipient callout together
3010 with the "random" option, the postmaster address was used as the MAIL
3011 FROM address for the random test, but not for the subsequent recipient
3012 test. It is now used for both.
3014 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3015 patch removes a few documentation additions to RFC 3028, because the
3016 latest draft now contains them. It adds the new en;ascii-case comparator
3017 and a new error check for 8bit text in MIME parts. Comparator and
3018 require names are now matched exactly. I enabled the subaddress
3019 extension, but it is not well tested yet (read: it works for me)."
3021 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3022 rework some of the code of TK/09 above to avoid the hardwired use of
3023 "%lld" and "long long". Replaced the call to snprintf() with a call to
3026 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3027 records point to non-existent hosts", "retry timeout exceeded", and
3028 "retry time not reached for any host after a long failure period".
3030 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3031 experimental DomainKeys support:
3033 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3034 (2) On an error such as an illegally used "control", the wrong name for
3035 the control was given.
3037 These problems did NOT occur unless DomainKeys support was compiled.
3039 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3041 PH/24 Added ${if match_ip condition.
3043 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3044 databases so that it will be absolutely obvious if a crash occurs in the
3045 DB library. This is a regular occurrence (often caused by mis-matched
3048 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3049 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3050 picked up on a user's system that detects such things. There doesn't seem
3051 to be a gcc warning option for this - only an attribute that has to be
3052 put on the function's prototype. It seems that in Fedora Core 4 they have
3053 set this on a number of new functions. No doubt there will be more in due
3056 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3057 specify a transport. However, if an address that was verified by such a
3058 router was the subject of a callout, Exim crashed because it tried to
3059 read the rcpt_include_affixes from the non-existent transport. Now it
3060 just assumes that the setting of that option is false. This bug was
3061 introduced by 4.51/PH/31.
3063 PH/28 Changed -d+all to exclude +memory, because that information is very
3064 rarely of interest, but it makes the output a lot bigger. People tend to
3065 do -d+all out of habit.
3067 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3068 code in os-type was giving problems when libc.so lives in lib64, like on
3071 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3072 aren't the modern standard, and it seems that some systems' include files
3073 don't always have them. Exim was already checking for some of the newer
3074 ones like T_AAAA, and defining it itself. I've added checks for all the
3075 record types that Exim uses.
3077 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3078 not automatically generating a new one, as it is supposed to. This
3079 prevented TLS from working. If the file did exist, but contained invalid
3080 data, a new version was generated, as expected. It was only the case of a
3081 non-existent file that was broken.
3083 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3084 with a change in libdomainkeys > 0.64.
3086 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3087 from DNS. If the selector record carries the flag, it now has
3088 precedence over the domain-wide flag.
3090 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3092 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3093 the use of a port name that isn't defined in /etc/services) occurred, the
3094 message was deferred as in a normal delivery, and thus remained on the
3095 spool, instead of being failed because of the mua_wrapper setting. This
3096 is now fixed, and I tidied up some of the mua_wrapper messages at the
3099 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3100 the messages in a hash of arrays rather than using individual hashes.
3101 This is a bit cleaner and results in dramatic memory savings, albeit
3102 at a slight CPU cost.
3104 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3105 as requested by Marc Sherman.
3107 SC/10 Eximstats: added histograms for user specified patterns as requested
3110 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3112 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3113 fopen() in the content-scanning modules that did not already have it.
3119 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3120 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3122 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3124 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3126 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3127 to test Sieve filters that use "vacation".
3129 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3130 that changes the way the GnuTLS parameters are stored in the cache file.
3131 The new format can be generated externally. For backward compatibility,
3132 if the data in the cache doesn't make sense, Exim assumes it has read an
3133 old-format file, and it generates new data and writes a new file. This
3134 means that you can't go back to an older release without removing the
3137 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3138 work if there are any delivery delays because "one_time" forces the
3139 parent to be marked "delivered", so its unseen clone is never tried
3140 again. For this reason, Exim now forbids the simultaneous setting of
3143 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3144 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3145 entry if you want to know the details. The fix, however, neglected to
3146 consider the case when local delivery batching is involved. The test for
3147 "previously delivered" was not happening when checking to see if an
3148 address could be batched with a previous (undelivered) one; under
3149 certain circumstances this could lead to multiple deliveries to the same
3152 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3153 in its include files, and this causes problems building Exim.
3155 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3156 header_syntax) but Exim was just ignoring anything given after a slash.
3157 In particular, this caused confusion with an attempt to use "verify =
3158 reverse_host_lookup/defer_ok". An error is now given when options are
3159 supplied for verify items that do not have them. (Maybe reverse_host_
3160 lookup should have a defer_ok option, but that's a different point.)
3162 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3163 defined by RFC 821) to 2048, because there were problems with some AUTH
3164 commands, and RFC 1869 says the size should be increased for extended
3165 SMTP commands that take arguments.
3167 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3170 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3171 "unknown" error; now it says that the functionality isn't in the binary.
3173 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3174 an address' error message when a string expansion fails (syntax or
3175 whatever). Otherwise the password may appear in the log. Following change
3176 PH/42 below, there is no longer a chance of it appearing in a bounce
3179 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3181 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3182 of its addresses was ever tried. (Bugzilla bug #2.)
3184 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3185 the result incorrectly in the debug output. (It correctly added a newline
3186 to what was transported.)
3188 TF/01 Added $received_time.
3190 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3191 commented out examples of how to interface to a virus scanner and to
3192 SpamAssassin. Also added commented examples of av_scanner and
3193 spamd_address settings.
3195 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3196 and controls are allowed in which ACLs. There were a couple of minor
3197 errors. Some of the entries in the conditions table (which is a table of
3198 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3199 negation of where the condition IS allowed.
3201 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3203 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3204 header file does not have a version number, so I've had to invent a new
3205 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3206 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3207 radiusclient), but it was contributed by a Radius user.
3209 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3210 files or directories, for OpenSSL.
3212 PH/19 When an Exim process that is running as root has to create an Exim log
3213 file, it does so in a subprocess that runs as exim:exim so as to get the
3214 ownership right at creation (otherwise, other Exim processes might see
3215 the file with the wrong ownership). There was no test for failure of this
3216 fork() call, which would lead to the process getting stuck as it waited
3217 for a non-existent subprocess. Forks do occasionally fail when resources
3218 run out. I reviewed all the other calls to fork(); they all seem to check
3221 PH/20 When checking for unexpected SMTP input at connect time (before writing
3222 the banner), Exim was not dealing correctly with a non-positive return
3223 from the read() function. If the client had disconnected by this time,
3224 the result was a log entry for a synchronization error with an empty
3225 string after "input=" when read() returned zero. If read() returned -1
3226 (an event I could not check), uninitialized data bytes were printed.
3227 There were reports of junk text (parts of files, etc) appearing after
3230 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3232 PH/22 Added support for macro redefinition, and (re)definition in between
3233 driver and ACL definitions.
3235 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3236 forgetting to use the resulting value; it was using the unexpanded value.
3238 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3239 hadn't been configured. The fix is from Juergen Kreileder, who
3240 understands it better than I do:
3242 "Here's what I see happening with three configured cyrus_sasl
3243 authenticators configured (plain, login, cram-md5):
3245 On startup auth_cyrus_sasl_init() gets called for each of these.
3246 This means three calls to sasl_listmech() without a specified mech_list.
3247 => SASL tests which mechs of all available mechs actually work
3248 => three warnings about OTP not working
3249 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3251 With the patch, sasl_listmech() also gets called three times. But now
3252 SASL's mech_list option is set to the server_mech specified in the the
3253 authenticator. Or in other words, the answer from sasl_listmech()
3254 gets limited to just the mech you're testing for (which is different
3256 => the return list contains just 'plain' or 'login', 'cram-md5' or
3257 nothing depending on the value of ob->server_mech.
3259 I've just tested the patch: Authentication still works fine,
3260 unavailable mechs specified in the exim configuration are still
3261 caught, and the auth.log warnings about OTP are gone."
3263 PH/25 When debugging is enabled, the contents of the command line are added
3264 to the debugging output, even when log_selector=+arguments is not
3267 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3268 answer is "GNU", and only if the return is "GNU/something" is the answer
3271 PH/27 $acl_verify_message is now set immediately after the failure of a
3272 verification in an ACL, and so is available in subsequent modifiers. In
3273 particular, the message can be preserved by coding like this:
3275 warn !verify = sender
3276 set acl_m0 = $acl_verify_message
3278 Previously, $acl_verify_message was set only while expanding "message"
3279 and "log_message" when a very denied access.
3281 PH/28 Modified OS/os.c-Linux with
3283 -#ifndef OS_LOAD_AVERAGE
3284 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3286 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3287 nomenclature these days.)
3289 PH/29 Installed patch from the Sieve maintainer that adds the options
3290 sieve_useraddress and sieve_subaddress to the redirect router.
3292 PH/30 In these circumstances:
3293 . Two addresses routed to the same list of hosts;
3294 . First host does not offer TLS;
3295 . First host accepts first address;
3296 . First host gives temporary error to second address;
3297 . Second host offers TLS and a TLS session is established;
3298 . Second host accepts second address.
3299 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3300 and peerdn, if requested) that were in fact used only for the second
3303 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3304 attention to any local part prefix or suffix that was matched by the
3305 router that accepted the address. It now behaves in the same way as it
3306 does for delivery: the affixes are removed from the local part unless
3307 rcpt_include_affixes is set on the transport.
3309 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3310 timeout during the DATA phase of an incoming message.
3312 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3313 applied a patch sanctioned by the Sieve maintainer.
3315 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3316 the uid or gid is negative. A case of a negative gid caused this to be
3317 noticed. The fix allows for either to be negative.
3319 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3320 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3323 PH/36 The change PH/12 above was broken. Fixed it.
3325 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3326 the grounds that routing the same address twice would always produce the
3327 same answer. This might have been true once, but it is certainly no
3328 longer true now. Routing a child address may depend on the previous
3329 routing that produced that child. Some complicated redirection strategies
3330 went wrong when messages had multiple recipients, and made Exim's
3331 behaviour dependent on the order in which the addresses were given.
3333 I have moved the duplicate checking until after the routing is complete.
3334 Exim scans the addresses that are assigned to local and remote
3335 transports, and removes any duplicates. This means that more work will be
3336 done, as duplicates will always all be routed, but duplicates are
3337 presumably rare, so I don't expect this is of any significance.
3339 For deliveries to pipes, files, and autoreplies, the duplicate checking
3340 still happens during the routing process, since they are not going to be
3343 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3344 It corrects a timeout issue with spamd. This is Ian's comment: "The
3345 background is that sometimes spamd either never reads data from a
3346 connection it has accepted, or it never writes response data. The exiscan
3347 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3348 blindly assumes that writes won't block so it may never time out."
3350 PH/39 Allow G after quota size as well as K and M.
3352 PH/40 The value set for $authenticated_id in an authenticator may not contain
3353 binary zeroes or newlines because the value is written to log lines and
3354 to spool files. There was no check on this. Now the value is run through
3355 the string_printing() function so that such characters are converted to
3356 printable escape sequences.
3358 PH/41 $message_linecount is a new variable that contains the total number of
3359 lines in the message. Compare $body_linecount, which is the count for the
3362 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3363 bounce and delay warning messages, except in certain special cases, which
3366 (a) An SMTP error message from a remote host;
3367 (b) A message specified in a :fail: redirection;
3368 (c) A message specified in a "fail" command in a system filter;
3369 (d) A message specified in a FAIL return from the queryprogram router;
3370 (e) A message specified by the cannot_route_message router option.
3372 In these cases only, Exim does include the error details in bounce and
3373 warning messages. There are also a few cases where bland messages such
3374 as "unrouteable address" or "local delivery error" are given.
3376 PH/43 $value is now also set for the "else" part of a ${run expansion.
3378 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3379 being worked on, but at least Exim now implements the latest version to
3382 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3383 process to complete was treated as a delivery failure, a timeout while
3384 writing the message to the pipe was logged, but erroneously treated as a
3385 successful delivery. Such timeouts include transport filter timeouts. For
3386 consistency with the overall process timeout, these timeouts are now
3387 treated as errors, giving rise to delivery failures by default. However,
3388 there is now a new Boolean option for the pipe transport called
3389 timeout_defer, which, if set TRUE, converts the failures into defers for
3390 both kinds of timeout. A transport filter timeout is now identified in
3393 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3394 systems where "make" and "gmake" are different, calling "gmake" at top
3395 level broke things. I've arranged for the value of $(MAKE) to be passed
3396 from the Makefile to this script so that it can call the same version of
3400 A note about Exim versions 4.44 and 4.50
3401 ----------------------------------------
3403 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3404 changes of various kinds. As a consequence, a big documentation update was
3405 needed. This delayed the release for rather longer than seemed good, especially
3406 in the light of a couple of (minor) security issues. Therefore, the changes
3407 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3408 release. So 4.44 and 4.50 are in effect two different branches that both start
3411 I have left the 4.50 change log unchanged; it contains all the changes since
3412 4.43. The change log for 4.44 is below; many of its items are identical to
3413 those for 4.50. This seems to be the most sensible way to preserve the
3414 historical information.
3420 1. Minor wording change to the doc/README.SIEVE file.
3422 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3423 computation of the current number of files was incorrect.
3425 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3426 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3429 4. Give more explanation in the error message when the command for a transport
3430 filter fails to execute.
3432 5. There are several places where Exim runs a non-Exim command in a
3433 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3434 was being done only for the command run by the queryprogram router. It is
3435 now done for all such subprocesses. The other cases are: ${run, transport
3436 filters, and the commands run by the lmtp and pipe transports.
3438 6. Added CONFIGURE_GROUP build-time option.
3440 7. Some older OS have a limit of 256 on the maximum number of file
3441 descriptors. Exim was using setrlimit() to set 1000 as a large value
3442 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3443 systems. I've change it so that if it can't get 1000, it tries for 256.
3445 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3446 was an oversight, and furthermore, ever since the addition of extra
3447 controls (e.g. 4.43/32), the checks on when to allow different forms of
3448 "control" were broken. There should now be diagnostics for all cases when a
3449 control that does not make sense is encountered.
3451 9. Added the /retain_sender option to "control=submission".
3453 10. $recipients is now available in the predata ACL (oversight).
3455 11. Tidy the search cache before the fork to do a delivery from a message
3456 received from the command line. Otherwise the child will trigger a lookup
3457 failure and thereby defer the delivery if it tries to use (for example) a
3458 cached ldap connection that the parent has called unbind on.
3460 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3461 of $address_data from the recipient verification was clobbered by the
3462 sender verification.
3464 13. The value of address_data from a sender verification is now available in
3465 $sender_address_data in subsequent conditions in the ACL statement.
3467 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3469 15. Added a new option "connect=<time>" to callout options, to set a different
3472 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3473 was its contents. (It was OK if the option was not defined at all.)
3475 17. A "Completed" log line is now written for messages that are removed from
3476 the spool by the -Mrm option.
3478 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3479 information about exactly what failed.
3481 19. Added -dd to debug only the daemon process.
3483 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3484 handles timeouts, both on the server side and network timeouts. Renamed the
3485 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3487 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3488 It is now set to "smtps".
3490 22. $host_address is now set to the target address during the checking of
3491 ignore_target_hosts.
3493 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3494 being passed; this would have caused $sender_host_name to have been used if
3495 matching the list had actually called for a host name (not very likely,
3496 since this list is usually IP addresses). A host name is now passed as
3499 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3500 code that specifies a non-restarting handler (typically sigaction() in
3501 modern systems) in an attempt to fix a rare and obscure crash bug.
3503 25. Narrowed the window for a race in the daemon that could cause it to ignore
3504 SIGCHLD signals. This is not a major problem, because they are used only to
3505 wake it up if nothing else does.
3507 26. A malformed maildirsize file could cause Exim to calculate negative values
3508 for the mailbox size or file count. Odd effects could occur as a result.
3509 The maildirsize information is now recalculated if the size or filecount
3512 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3513 support for a long time. Removed HAVE_SYS_VFS_H.
3515 28. Installed the latest version of exipick from John Jetmore.
3517 29. In an address list, if the pattern was not a regular expression, an empty
3518 subject address (from a bounce message) matched only if the pattern was an
3519 empty string. Non-empty patterns were not even tested. This was the wrong
3520 because it is perfectly reasonable to use an empty address as part of a
3521 database query. An empty address is now tested by patterns that are
3522 lookups. However, all the other forms of pattern expect the subject to
3523 contain a local part and a domain, and therefore, for them, an empty
3524 address still always fails if the pattern is not itself empty.
3526 30. Exim went into a mad DNS loop when attempting to do a callout where the
3527 host was specified on an smtp transport, and looking it up yielded more
3528 than one IP address.
3530 31. Re-factored the code for checking spool and log partition space into a
3531 function that finds that data and another that does the check. The former
3532 is then used to implement four new variables: $spool_space, $log_space,
3533 $spool_inodes, and $log_inodes.
3535 32. The RFC2047 encoding function was originally intended for short strings
3536 such as real names; it was not keeping to the 75-character limit for
3537 encoded words that the RFC imposes. It now respects the limit, and
3538 generates multiple encoded words if necessary. To be on the safe side, I
3539 have increased the buffer size for the ${rfc2047: expansion operator from
3542 33. It is now permitted to omit both strings after an "if" condition; if the
3543 condition is true, the result is "true". As before, when the second string
3544 is omitted, a false condition yields an empty string. This makes it less
3545 cumbersome to write custom ACL and router conditions.
3547 34. Failure to deliver a bounce message always caused it to be frozen, even if
3548 there was an errors_to setting on the router. The errors_to setting is now
3551 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3552 canonical form (fully expanded) before being placed in
3553 $sender_host_address.
3555 36. The table in the code that translates DNS record types into text (T_A to
3556 "A" for instance) was missing entries for NS and CNAME. It is just possible
3557 that this could have caused confusion if both these types were looked up
3558 for the same domain, because the text type is used as part of Exim's
3559 per-process caching. But the chance of anyone hitting this buglet seems
3562 37. The dnsdb lookup has been extended in a number of ways.
3564 (1) There is a new type, "zns", which walks up the domain tree until it
3565 finds some nameserver records. It should be used with care.
3567 (2) There is a new type, "mxh", which is like "mx" except that it returns
3568 just the host names, not the priorities.
3570 (3) It is now possible to give a list of domains (or IP addresses) to be
3571 looked up. The behaviour when one of the lookups defers can be
3572 controlled by a keyword.
3574 (4) It is now possible to specify the separator character for use when
3575 multiple records are returned.
3577 38. The dnslists ACL condition has been extended: it is now possible to supply
3578 a list of IP addresses and/or domains to be looked up in a particular DNS
3581 39. Added log_selector=+queue_time_overall.
3583 40. When running the queue in the test harness, wait just a tad after forking a
3584 delivery process, to get repeatability of debugging output.
3586 41. Include certificate and key file names in error message when GnuTLS fails
3587 to set them up, because the GnuTLS error message doesn't include the name
3588 of the failing file when there is a problem reading it.
3590 42. Allow both -bf and -bF in the same test run.
3592 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3594 44. The "Exiscan patch" is now merged into the mainline Exim source.
3596 45. Sometimes the final signoff response after QUIT could fail to get
3597 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3598 < 0 before doing a fflush(). This bug looks as though it goes back to the
3599 introduction of TLS in release 3.20, but "sometimes" must have been rare
3600 because the tests only now provoked it.
3602 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3603 (this can affect the format of dates).
3605 47. exim_tidydb, when checking for the continued existence of a message for
3606 which it has found a message-specific retry record, was not finding
3607 messages that were in split spool directories. Consequently, it was
3608 deleting retry records that should have stayed in existence.
3610 48. Steve fixed some bugs in eximstats.
3612 49. The SPA authentication driver was not abandoning authentication and moving
3613 on to the next authenticator when an expansion was forced to fail,
3614 contradicting the general specification for all authenticators. Instead it
3615 was generating a temporary error. It now behaves as specified.
3617 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3618 (the order specifies the preference for clients). The order is now AES256,
3619 AES128, 3DES, ARCFOUR128.
3621 51. Small patch to Sieve code - explicitly set From: when generating an
3624 52. Exim crashed if a remote delivery caused a very long error message to be
3625 recorded - for instance if somebody sent an entire SpamAssassin report back
3626 as a large number of 550 error lines. This bug was coincidentally fixed by
3627 increasing the size of one of Exim's internal buffers (big_buffer) that
3628 happened as part of the Exiscan merge. However, to be on the safe side, I
3629 have made the code more robust (and fixed the comments that describe what
3632 53. Now that there can be additional text after "Completed" in log lines (if
3633 the queue_time_overall log selector is set), a one-byte patch to exigrep
3634 was needed to allow it to recognize "Completed" as not the last thing in
3637 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3638 patch that reportedly fixes this has been added. I am not expert enough to
3639 create a test for it. This is what the patch creator wrote:
3641 "I found a little strange behaviour of ldap code when working with
3642 Windows 2003 AD Domain, where users was placed in more than one
3643 Organization Units. When I tried to give exim partial DN, the exit code
3644 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3645 But simultaneously result of request was absolutely normal ldap result,
3646 so I produce this patch..."
3648 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3649 REFERENCE, so I have modified the code to exclude the patch when that macro
3652 55. Some experimental protocols are using DNS PTR records for new purposes. The
3653 keys for these records are domain names, not reversed IP addresses. The
3654 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3655 leaves it alone. Component reversal etc. now happens only for IP addresses.
3658 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3660 57. Double the size of the debug message buffer (to 2048) so that more of very
3661 long debug lines gets shown.
3663 58. The exicyclog utility now does better if the number of log files to keep
3664 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3666 59. Two changes related to the smtp_active_hostname option:
3668 (1) $smtp_active_hostname is now available as a variable.
3669 (2) The default for smtp_banner uses $smtp_active_hostname instead
3670 of $primary_hostname.
3672 60. The host_aton() function is supposed to be passed a string that is known
3673 to be a valid IP address. However, in the case of IPv6 addresses, it was
3674 not checking this. This is a hostage to fortune. Exim now panics and dies
3675 if the condition is not met. A case was found where this could be provoked
3676 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3677 components; fortuitously, this particular loophole had already been fixed
3678 by change 4.50/55 above.
3680 If there are any other similar loopholes, the new check in host_aton()
3681 itself should stop them being exploited. The report I received stated that
3682 data on the command line could provoke the exploit when Exim was running as
3683 exim, but did not say which command line option was involved. All I could
3684 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3685 running as the user.
3688 61. There was a buffer overflow vulnerability in the SPA authentication code
3689 (which came originally from the Samba project). I have added a test to the
3690 spa_base64_to_bits() function which I hope fixes it.
3693 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3694 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3696 63. The daemon start-up calls getloadavg() while still root for those OS that
3697 need the first call to be done as root, but it missed one case: when
3698 deliver_queue_load_max is set with deliver_drop_privilege. This is
3699 necessary for the benefit of the queue runner, because there is no re-exec
3700 when deliver_drop_privilege is set.
3702 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3703 This has been fixed.
3705 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3706 was in use, was not putting the data itself into the right store pool;
3707 consequently, it could be overwritten for a subsequent message in the same
3708 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3711 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3713 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3714 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3715 the function were treating the return as a boolean value, which happened to
3716 work because 0=false and not-0=true, but is not correct code.
3718 68. The host_aton() function was not handling scoped IPv6 addresses (those
3719 with, for example, "%eth0" on the end) correctly.
3721 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3722 negated items (that is, ~something) in unsigned ints. Some compilers
3723 apparently mutter when there is no cast.
3725 70. If an address verification called from an ACL failed, and did not produce a
3726 user-specific message (i.e. there was only a "system" message), nothing was
3727 put in $acl_verify_message. In this situation, it now puts the system
3730 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3731 session; change 4.31/43 added the unwanted input to the log line - except
3732 that it did not do this in the start of session case. It now does.
3734 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3735 This is wrong and can cause the other end to generate a synchronization
3736 error if it is another Exim or anything else that does the synchronization
3737 check. A QUIT command is no longer sent after a timeout.
3739 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3740 during host lookups.
3742 74. The defer_ok option of callout verification was not working if it was used
3743 when verifying addresses in header lines, that is, for this case:
3745 verify = header_sender/callout=defer_ok
3747 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3748 those file descriptors could be used for SMTP connections. If anything
3749 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3750 could be sent to the SMTP client, causing chaos. The daemon now opens
3751 stdin, stdout, and stderr to /dev/null when it puts itself into the
3754 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3755 log by default. The user can override this with suitable Perl magic.
3757 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3758 the log message when discard triggers, was not working for the DATA ACL or
3759 for the non-SMTP ACL.
3761 78. Error message wording change in sieve.c.
3763 79. If smtp_accept_max_per_host was set, the number of connections could be
3764 restricted to fewer than expected, because the daemon was trying to set up
3765 a new connection before checking whether the processes handling previous
3766 connections had finished. The check for completed processes is now done
3767 earlier. On busy systems, this bug wouldn't be noticed because something
3768 else would have woken the daemon, and it would have reaped the completed
3771 80. If a message was submitted locally by a user whose login name contained one
3772 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3773 It caused a spool format error. I have fixed the spool reading code. A
3774 related problem was that the "from" clause in the Received: line became
3775 illegal because of the space(s). It is now covered by ${quote_local_part.
3777 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3780 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3782 83. Patch from Sieve maintainer to fix unterminated string problem in
3783 "vacation" handling.
3785 84. Some minor changes to the Linux configuration files to help with other
3786 OS variants using glibc.
3788 85. One more patch for Sieve to update vacation handling to latest spec.
3791 ----------------------------------------------------
3792 See the note above about the 4.44 and 4.50 releases.
3793 ----------------------------------------------------
3799 1. Change 4.43/35 introduced a bug that caused file counts to be
3800 incorrectly computed when quota_filecount was set in an appendfile
3803 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3804 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3807 3. Give more explanation in the error message when the command for a transport
3808 filter fails to execute.
3810 4. There are several places where Exim runs a non-Exim command in a
3811 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3812 was being done only for the command run by the queryprogram router. It is
3813 now done for all such subprocesses. The other cases are: ${run, transport
3814 filters, and the commands run by the lmtp and pipe transports.
3816 5. Some older OS have a limit of 256 on the maximum number of file
3817 descriptors. Exim was using setrlimit() to set 1000 as a large value
3818 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3819 systems. I've change it so that if it can't get 1000, it tries for 256.
3821 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3822 was an oversight, and furthermore, ever since the addition of extra
3823 controls (e.g. 4.43/32), the checks on when to allow different forms of
3824 "control" were broken. There should now be diagnostics for all cases when a
3825 control that does not make sense is encountered.
3827 7. $recipients is now available in the predata ACL (oversight).
3829 8. Tidy the search cache before the fork to do a delivery from a message
3830 received from the command line. Otherwise the child will trigger a lookup
3831 failure and thereby defer the delivery if it tries to use (for example) a
3832 cached ldap connection that the parent has called unbind on.
3834 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3835 of $address_data from the recipient verification was clobbered by the
3836 sender verification.
3838 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3839 was its contents. (It was OK if the option was not defined at all.)
3841 11. A "Completed" log line is now written for messages that are removed from
3842 the spool by the -Mrm option.
3844 12. $host_address is now set to the target address during the checking of
3845 ignore_target_hosts.
3847 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3848 being passed; this would have caused $sender_host_name to have been used if
3849 matching the list had actually called for a host name (not very likely,
3850 since this list is usually IP addresses). A host name is now passed as
3853 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3854 code that specifies a non-restarting handler (typically sigaction() in
3855 modern systems) in an attempt to fix a rare and obscure crash bug.
3857 15. Narrowed the window for a race in the daemon that could cause it to ignore
3858 SIGCHLD signals. This is not a major problem, because they are used only to
3859 wake it up if nothing else does.
3861 16. A malformed maildirsize file could cause Exim to calculate negative values
3862 for the mailbox size or file count. Odd effects could occur as a result.
3863 The maildirsize information is now recalculated if the size or filecount
3866 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3867 support for a long time. Removed HAVE_SYS_VFS_H.
3869 18. Updated exipick to current release from John Jetmore.
3871 19. Allow an empty sender to be matched against a lookup in an address list.
3872 Previously the only cases considered were a regular expression, or an
3875 20. Exim went into a mad DNS lookup loop when doing a callout where the
3876 host was specified on the transport, if the DNS lookup yielded more than
3879 21. The RFC2047 encoding function was originally intended for short strings
3880 such as real names; it was not keeping to the 75-character limit for
3881 encoded words that the RFC imposes. It now respects the limit, and
3882 generates multiple encoded words if necessary. To be on the safe side, I
3883 have increased the buffer size for the ${rfc2047: expansion operator from
3886 22. Failure to deliver a bounce message always caused it to be frozen, even if
3887 there was an errors_to setting on the router. The errors_to setting is now
3890 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3891 canonical form (fully expanded) before being placed in
3892 $sender_host_address.
3894 24. Updated eximstats to version 1.33
3896 25. Include certificate and key file names in error message when GnuTLS fails
3897 to set them up, because the GnuTLS error message doesn't include the name
3898 of the failing file when there is a problem reading it.
3900 26. Expand error message when OpenSSL has problems setting up cert/key files.
3903 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3904 (this can affect the format of dates).
3906 28. exim_tidydb, when checking for the continued existence of a message for
3907 which it has found a message-specific retry record, was not finding
3908 messages that were in split spool directories. Consequently, it was
3909 deleting retry records that should have stayed in existence.
3911 29. eximstats updated to version 1.35
3912 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3913 1.35 - bugfix such that pie charts by volume are generated correctly
3915 30. The SPA authentication driver was not abandoning authentication and moving
3916 on to the next authenticator when an expansion was forced to fail,
3917 contradicting the general specification for all authenticators. Instead it
3918 was generating a temporary error. It now behaves as specified.
3920 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3921 (the order specifies the preference for clients). The order is now AES256,
3922 AES128, 3DES, ARCFOUR128.
3924 31. Small patch to Sieve code - explicitly set From: when generating an
3927 32. Exim crashed if a remote delivery caused a very long error message to be
3928 recorded - for instance if somebody sent an entire SpamAssassin report back
3929 as a large number of 550 error lines. This bug was coincidentally fixed by
3930 increasing the size of one of Exim's internal buffers (big_buffer) that
3931 happened as part of the Exiscan merge. However, to be on the safe side, I
3932 have made the code more robust (and fixed the comments that describe what
3935 33. Some experimental protocols are using DNS PTR records for new purposes. The
3936 keys for these records are domain names, not reversed IP addresses. The
3937 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3938 leaves it alone. Component reversal etc. now happens only for IP addresses.
3941 34. The host_aton() function is supposed to be passed a string that is known
3942 to be a valid IP address. However, in the case of IPv6 addresses, it was
3943 not checking this. This is a hostage to fortune. Exim now panics and dies
3944 if the condition is not met. A case was found where this could be provoked
3945 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3946 components; fortuitously, this particular loophole had already been fixed
3947 by change 4.50/55 or 4.44/33 above.
3949 If there are any other similar loopholes, the new check in host_aton()
3950 itself should stop them being exploited. The report I received stated that
3951 data on the command line could provoke the exploit when Exim was running as
3952 exim, but did not say which command line option was involved. All I could
3953 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3954 running as the user.
3957 35. There was a buffer overflow vulnerability in the SPA authentication code
3958 (which came originally from the Samba project). I have added a test to the
3959 spa_base64_to_bits() function which I hope fixes it.
3962 36. The daemon start-up calls getloadavg() while still root for those OS that
3963 need the first call to be done as root, but it missed one case: when
3964 deliver_queue_load_max is set with deliver_drop_privilege. This is
3965 necessary for the benefit of the queue runner, because there is no re-exec
3966 when deliver_drop_privilege is set.
3968 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3969 was in use, was not putting the data itself into the right store pool;
3970 consequently, it could be overwritten for a subsequent message in the same
3971 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3974 38. Sometimes the final signoff response after QUIT could fail to get
3975 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3976 < 0 before doing a fflush(). This bug looks as though it goes back to the
3977 introduction of TLS in release 3.20, but "sometimes" must have been rare
3978 because the tests only now provoked it.
3984 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3985 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3986 Now it returns an integer. A number of calls were still expecting a T/F
3987 return. Fortuitously, in all cases, the tests worked in OK situations,
3988 which is the norm. However, things would have gone wrong on any write
3989 failures on the smtp file descriptor. This function is used when sending
3990 messages over SMTP and also when doing verify callouts.
3992 2. When Exim is called to do synchronous delivery of a locally submitted
3993 message (the -odf or -odi options), it no longer closes stderr before doing
3996 3. Implemented the mua_wrapper option.
3998 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4000 5. Implemented the functions header_remove(), header_testname(),
4001 header_add_at_position(), and receive_remove_recipient(), and exported them
4004 6. If an ACL "warn" statement specified the addition of headers, Exim already
4005 inserted X-ACL-Warn: at the start if there was no header name. However, it
4006 was not making this test for the second and subsequent header lines if
4007 there were newlines in the string. This meant that an invalid header could
4008 be inserted if Exim was badly configured.
4010 7. Allow an ACL "warn" statement to add header lines at the start or after all
4011 the Received: headers, as well as at the end.
4013 8. Added the rcpt_4xx retry error code.
4015 9. Added postmaster_mailfrom=xxx to callout verification option.
4017 10. Added mailfrom=xxxx to the callout verification option, for verify=
4020 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4021 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4023 12. Inserted some casts to stop certain compilers warning when using pointer
4024 differences as field lengths or precisions in printf-type calls (mostly
4025 affecting debugging statements).
4027 13. Added optional readline() support for -be (dynamically loaded).
4029 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4030 same clock tick as a message's arrival, so that its received time was the
4031 same as the "first fail" time on the retry record, and that message
4032 remained on the queue past the ultimate address timeout, every queue runner
4033 would try a delivery (because it was past the ultimate address timeout) but
4034 after another failure, the ultimate address timeout, which should have then
4035 bounced the address, did not kick in. This was a "< instead of <=" error;
4036 in most cases the first failure would have been in the next clock tick
4037 after the received time, and all would be well.
4039 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4040 being recognized when the domain list was tested by the match_domain
4041 condition in an expansion string.
4043 16. Added the ${str2b64: operator.
4045 17. Exim was always calling setrlimit() to set a large limit for the number of
4046 processes, without checking whether the existing limit was already
4047 adequate. (It did check for the limit on file descriptors.) Furthermore,
4048 errors from getrlimit() and setrlimit() were being ignored. Now they are
4049 logged to the main and panic logs, but Exim does carry on, to try to do its
4050 job under whatever limits there are.
4052 18. Imported PCRE 5.0.
4054 19. Trivial typo in log message " temporarily refused connection" (the leading
4057 20. If the log selector return_path_on_delivery was set and an address was
4058 redirected to /dev/null, the delivery process crashed because it assumed
4059 that a return path would always be set for a "successful" delivery. In this
4060 case, the whole delivery is bypassed as an optimization, and therefore no
4063 21. Internal re-arrangement: the function for sending a challenge and reading
4064 a response while authentication was assuming a zero-terminated challenge
4065 string. It's now changed to take a pointer and a length, to allow for
4066 binary data in such strings.
4068 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4070 23. Exim was not respecting finduser_retries when seeking the login of the
4071 uid under which it was called; it was always trying 10 times. (The default
4072 setting of finduser_retries is zero.) Also, it was sleeping after the final
4073 failure, which is pointless.
4075 24. Implemented tls_on_connect_ports.
4077 25. Implemented acl_smtp_predata.
4079 26. If the domain in control=submission is set empty, Exim assumes that the
4080 authenticated id is a complete email address when it generates From: or
4081 Sender: header lines.
4083 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4084 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4085 chown and chgrp in /bin and hostname in /usr/bin.
4087 28. Exim was keeping the "process log" file open after each use, just as it
4088 does for the main log. This opens the possibility of it remaining open for
4089 long periods when the USR1 signal hits a daemon. Occasional processlog
4090 errors were reported, that could have been caused by this. Anyway, it seems
4091 much more sensible not to leave this file open at all, so that is what now
4094 29. The long-running daemon process does not normally write to the log once it
4095 has entered its main loop, and it closes the log before doing so. This is
4096 so that log files can straightforwardly be renamed and moved. However,
4097 there are a couple of unusual error situations where the daemon does write
4098 log entries, and I had neglected to close the log afterwards.
4100 30. The text of an SMTP error response that was received during a remote
4101 delivery was being truncated at 512 bytes. This is too short for some of
4102 the long messages that one sometimes sees. I've increased the limit to
4105 31. It is now possible to make retry rules that apply only when a message has a
4106 specific sender, in particular, an empty sender.
4108 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4109 it possible to be selective about when SMTP synchronization is enforced.
4111 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4113 32. Implemented hosts_connection_nolog.
4115 33. Added an ACL for QUIT.
4117 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4120 35. Added mailbox_size and mailbox_filecount to appendfile.
4122 36. Added control = no_multiline_responses to ACLs.
4124 37. There was a bug in the logic of the code that waits for the clock to tick
4125 in the case where the clock went backwards by a substantial amount such
4126 that the microsecond fraction of "now" was more than the microsecond
4127 fraction of "then" (but the whole seconds number was less).
4129 38. Added support for the libradius Radius client library this is found on
4130 FreeBSD (previously only the radiusclient library was supported).
4136 1. When certain lookups returned multiple values in the form name=value, the
4137 quoting of the values was not always being done properly. Specifically:
4138 (a) If the value started with a double quote, but contained no whitespace,
4140 (b) If the value contained whitespace other than a space character (i.e.
4141 tabs or newlines or carriage returns) it was not quoted.
4142 This fix has been applied to the mysql and pgsql lookups by writing a
4143 separate quoting function and calling it from the lookup code. The fix
4144 should probably also be applied to nisplus, ibase and oracle lookups, but
4145 since I cannot test any of those, I have not disturbed their existing code.
4147 2. A hit in the callout cache for a specific address caused a log line with no
4148 reason for rejecting RCPT. Now it says "Previous (cached) callout
4149 verification failure".
4151 3. There was an off-by-one bug in the queryprogram router. An over-long
4152 return line was truncated at 256 instead of 255 characters, thereby
4153 overflowing its buffer with the terminating zero. As well as fixing this, I
4154 have increased the buffer size to 1024 (and made a note to document this).
4156 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4157 when Exim is waiting for an SMTP response from a remote server, Exim
4158 restarts its select() call on the socket, thereby resetting its timeout.
4159 This is not a problem when such interrupts are rare. Somebody set up a cron
4160 job to run exiwhat every 2 minutes, which is less than the normal select()
4161 timeout (5 or 10 minutes). This meant that the select() timeout never
4162 kicked in because it was always reset. I have fixed this by comparing the
4163 time when an interrupt arrives with the time at the start of the first call
4164 to select(). If more time than the timeout has elapsed, the interrupt is
4165 treated as a timeout.
4167 5. Some internal re-factoring in preparation for the addition of Sieve
4168 extensions (by MH). In particular, the "personal" test is moved to a
4169 separate function, and given an option for scanning Cc: and Bcc: (which is
4170 not set for Exim filters).
4172 6. When Exim created an email address using the login of the caller as the
4173 local part (e.g. when creating a From: or Sender: header line), it was not
4174 quoting the local part when it contained special characters such as @.
4176 7. Installed new OpenBSD configuration files.
4178 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4179 try to make them clearer.
4181 9. Callout options, other than the timeout value, were being ignored when
4182 verifying sender addresses in header lines. For example, when using
4184 verify = header_sender/callout=no_cache
4186 the cache was (incorrectly) being used.
4188 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4189 only the Cygwin environment.
4191 11. When return_path_on_delivery was set as a log selector, if different remote
4192 addresses in the same message used different return paths and parallel
4193 remote delivery occurred, the wrong values would sometimes be logged.
4194 (Whenever a remote delivery process finished, the return path value from
4195 the most recently started remote delivery process was logged.)
4197 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4198 lines when AUTH and/or TLS are in use. This is the "received protocol"
4199 field. Exim used to use "asmtp" for authenticated SMTP, without any
4200 indication (in the protocol name) for TLS use. Now it follows the RFC and
4201 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4202 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4203 names appear in log lines as well as in Received: header lines.
4205 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4206 extensions, and comparison tests, and to fix some bugs.
4208 14. Changes to the "personal" filter test:
4210 (1) The test was buggy in that it was just doing the equivalent of
4211 "contains" tests on header lines. For example, if a user's address was
4212 anne@some.where, the "personal" test would incorrectly be true for
4214 To: susanne@some.where
4216 This test is now done by extracting each address from the header in turn,
4217 and checking the entire address. Other tests that are part of "personal"
4218 are now done using regular expressions (for example, to check local parts
4219 of addresses in From: header lines).
4221 (2) The list of non-personal local parts in From: addresses has been
4222 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4223 taken from the Sieve specification recommendations.
4225 (3) If the message contains any header line starting with "List-" it is
4226 treated as non-personal.
4228 (4) The test for "circular" in the Subject: header line has been removed
4229 because it now seems ill-conceived.
4231 15. Minor typos in src/EDITME comments corrected.
4233 16. Installed latest exipick from John Jetmore.
4235 17. If headers_add on a router specified a text string that was too long for
4236 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4237 of string_sprintf() is now avoided.
4239 18. $message_body_size was not set (it was always zero) when running the DATA
4240 ACL and the local_scan() function.
4242 19. For the "mail" command in an Exim filter, no default was being set for
4243 the once_repeat time, causing a random time value to be used if "once" was
4244 specified. (If the value happened to be <= 0, no repeat happened.) The
4245 default is now 0s, meaning "never repeat". The "vacation" command was OK
4246 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4247 (I found it when inspecting the code).
4249 20. There is now an overall timeout for performing a callout verification. It
4250 defaults to 4 times the callout timeout, which applies to individual SMTP
4251 commands during the callout. The overall timeout applies when there is more
4252 than one host that can be tried. The timeout is checked before trying the
4253 next host. This prevents very long delays if there are a large number of
4254 hosts and all are timing out (e.g. when the network connections are timing
4255 out). The value of the overall timeout can be changed by specifying an
4256 additional sub-option for "callout", called "maxwait". For example:
4258 verify = sender/callout=5s,maxwait=20s
4260 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4261 to the end before writing, but this should make it even safer).
4263 22. Exim was forgetting that it had advertised PIPELINING for the second and
4264 subsequent messages on an SMTP connection. It was also not resetting its
4265 memory on STARTTLS and an internal HELO.
4267 23. When Exim logs an SMTP synchronization error within a session, it now
4268 records whether PIPELINING has been advertised or not.
4270 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4271 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4272 rather than long int.
4274 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4276 26. Added the never_mail option to autoreply.
4282 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4283 crash if the getsockname() call failed; this can happen if a connection is
4284 closed very soon after it is established. The problem was simply in the
4285 order in which certain operations were done, causing Exim to try to write
4286 to the SMTP stream before it had set up the file descriptor. The bug has
4287 been fixed by making things happen in the correct order.
4293 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4294 before the connection was closed, thus losing the rejection response.
4296 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4297 some early Solaris releases, but causes trouble in current releases where
4298 socklen_t is defined.
4300 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4303 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4306 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4307 that releasing the top of it at the end releases what was used for sub-
4308 expansions (unless the block got too big). However, discard this block if
4309 the first thing is a variable or header, so that we can use its block when
4310 it is dynamic (useful for very large $message_headers, for example).
4312 6. Lookups now cache *every* query, not just the most recent. A new, separate
4313 store pool is used for this. It can be recovered when all lookup caches are
4314 flushed. Lookups now release memory at the end of their result strings.
4315 This has involved some general refactoring of the lookup sources.
4317 7. Some code has been added to the store_xxx() functions to reduce the amount
4318 of flapping under certain conditions.
4320 8. log_incoming_interface used to affect only the <= reception log lines. Now
4321 it causes the local interface and port to be added to several more SMTP log
4322 lines, for example "SMTP connection from", and rejection lines.
4324 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4326 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4328 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4329 could be overwritten at the end of the current message (or the start of a
4330 new message if it was set in a HELO ACL). The value is now preserved for
4331 the duration of the SMTP connection.
4333 12. If a transport had a headers_rewrite setting, and a matching header line
4334 contained an unqualified address, that address was qualified, even if it
4335 did not match any rewriting rules. The underlying bug was that the values
4336 of the flags that permit the existence of unqualified sender and recipient
4337 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4338 non-local messages, and by -bnq for local messages) were not being
4339 preserved with the message after it was received.
4341 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4342 "next input=" as part of the text comprising the host identity instead of
4343 the correct text. The code was using the same buffer for two different
4344 strings. However, depending on which order the printing function evaluated
4345 its arguments, the bug did not always show up. Under Linux, for example, my
4346 test suite worked just fine.
4348 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4349 doesn't work with some older versions of Perl. It has been changed to "my",
4350 which in any case is probably the better facility to use.
4352 15. A really picky compiler found some instances of statements for creating
4353 error messages that either had too many or two few arguments for the format
4356 16. The size of the buffer for calls to the DNS resolver has been increased
4357 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4358 for addresses that have a lot of PTR records. This alleviates a problem; it
4359 does not fully solve it.
4361 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4362 buffer now truncates the list and logs the incident, which is the same
4363 action as happens when Exim is looking up a host name and its aliases.
4364 Previously in this situation something unpredictable would happen;
4365 sometimes it was "internal error: store_reset failed".
4367 18. If a server dropped the connection unexpectedly when an Exim client was
4368 using GnuTLS and trying to read a response, the client delivery process
4369 crashed while trying to generate an error log message.
4371 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4372 string, for example:
4374 warn message = H1: something\nH2: something
4376 the text was added as a single header line from Exim's point of view
4377 though it ended up OK in the delivered message. However, searching for the
4378 second and subsequent header lines using $h_h2: did not work. This has been
4379 fixed. Similarly, if a system filter added multiple headers in this way,
4380 the routers could not see them.
4382 20. Expanded the error message when iplsearch is called with an invalid key to
4383 suggest using net-iplsearch in a host list.
4385 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4386 ACLs are no longer actually imposed (and a message to that effect is
4389 22. If a "gecos" field in a passwd entry contained escaped characters, in
4390 particular, if it contained a \" sequence, Exim got it wrong when building
4391 a From: or a Sender: header from that name. A second bug also caused
4392 incorrect handling when an unquoted " was present following a character
4393 that needed quoting.
4395 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4396 was not being matched caselessly.
4398 24. Arranged for all hyphens in the exim.8 source to be escaped with
4401 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4402 properly. Recipient callout cache records were still being keyed to include
4403 the sender, even when use_sender was set false. This led to far more
4404 callouts that were necessary. The sender is no longer included in the key
4405 when use_sender is false.
4407 26. Added "control = submission" modifier to ACLs.
4409 27. Added the ${base62d: operator to decode base 62 numbers.
4411 28. dnsdb lookups can now access SRV records.
4413 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4414 the configuration file.
4416 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4417 (-v) mode. This makes the output for a verbose queue run more intelligible.
4419 31. Added a use_postmaster feature to recipient callouts.
4421 32. Added the $body_zerocount variable, containing the number of binary zero
4422 bytes in the message body.
4424 33. The time of last modification of the "new" subdirectory is now used as the
4425 "mailbox time last read" when there is a quota error for a maildir
4428 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4430 35. Added +ignore_unknown as a special item in host lists.
4432 36. Code for decoding IPv6 addresses in host lists is now included, even if
4433 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4434 address was recognized as an IP address, but was then not correctly decoded
4435 into binary, causing unexpected and incorrect effects when compared with
4442 1. Very minor rewording of debugging text in manualroute to say "list of
4443 hosts" instead of "hostlist".
4445 2. If verify=header_syntax was set, and a header line with an unqualified
4446 address (no domain) and a large number of spaces between the end of the
4447 name and the colon was received, the reception process suffered a buffer
4448 overflow, and (when I tested it) crashed. This was caused by some obsolete
4449 code that should have been removed. The fix is to remove it!
4451 3. When running in the test harness, delay a bit after writing a bounce
4452 message to get a bit more predictability in the log output.
4454 4. Added a call to search_tidyup() just before forking a reception process. In
4455 theory, someone could use a lookup in the expansion of smtp_accept_max_
4456 per_host which, without the tidyup, could leave open a database connection.
4458 5. Added the variables $recipient_data and $sender_data which get set from a
4459 lookup success in an ACL "recipients" or "senders" condition, or a router
4460 "senders" option, similar to $domain_data and $local_part_data.
4462 6. Moved the writing of debug_print from before to after the "senders" test
4465 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4466 problems for message scanning, either using a data ACL, or using
4467 local_scan() because the Received: header was not generated till after they
4468 were called (in order to set the time as the time of reception completion).
4469 I have revised the way this works. The header is now generated after the
4470 body is received, but before the ACL or local_scan() are called. After they
4471 are run, the timestamp in the header is updated.
4477 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4478 before starting a queue runner without re-exec. This happened only when
4479 deliver_drop_privilege was set or when the Exim user was set to root. The
4480 effect of the bug was that timeouts during subsequent deliveries caused
4481 crashes instead of being properly handled. The handler is now left at its
4482 default (and expected) setting.
4484 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4485 message, again when deliver_drop_privilege is set or Exim is run as root.
4486 The bug described in (1) was not present in this case, but the tidying up
4487 of the other signals was missing. I have made the two cases consistent.
4489 3. The ignore_target_hosts setting on a manualroute router was being ignored
4490 for hosts that were looked up using the /MX notation.
4492 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4495 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4496 operated on the sender address. After changing the $sender_address to <>
4497 for the sender address verify, Exim was re-instated it as the original
4498 (before rewriting) address, but remembering that it had rewritten it, so it
4499 wasn't rewriting it again. This bug also had the effect of breaking the
4500 sender address verification caching when the sender address was rewritten.
4502 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4503 This has been changed so that if the ip literal address matches
4504 ignore_target_hosts, the router declines.
4506 7. Added expansion conditions match_domain, match_address, and match_local_
4507 part (NOT match_host).
4509 8. The placeholder for the Received: header didn't have a length field set.
4511 9. Added code to Exim itself and to exim_lock to test for a specific race
4512 condition that could lead to file corruption when using MBX delivery. The
4513 issue is with the lockfile that is created in /tmp. If this file is removed
4514 after a process has opened it but before that process has acquired a lock,
4515 there is the potential for a second process to recreate the file and also
4516 acquire a lock. This could lead to two Exim processes writing to the file
4517 at the same time. The added code performs the same test as UW imapd; it
4518 checks after acquiring the lock that its file descriptor still refers to
4519 the same named file.
4521 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4522 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4525 11. Added the smtp_active_hostname option. If used, this will typically be made
4526 to depend on the incoming interface address. Because $interface_address is
4527 not set up until the daemon has forked a reception process, error responses
4528 that can happen earlier (such as "too many connections") no longer contain
4531 12. If an expansion in a condition on a "warn" statement fails because a lookup
4532 defers, the "warn" statement is abandoned, and the next ACL statement is
4533 processed. Previously this caused the whole ACL to be aborted.
4535 13. Added the iplsearch lookup type.
4537 14. Added ident_timeout as a log selector.
4539 15. Added tls_certificate_verified as a log selector.
4541 16. Added a global option tls_require_ciphers (compare the smtp transport
4542 option of the same name). This controls incoming TLS connections.
4544 17. I finally figured out how to make tls_require_ciphers do a similar thing
4545 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4546 before starting the TLS session.
4548 18. Tabs are now shown as \t in -bP output.
4550 19. If the log selector return_path_on_delivery was set, Exim crashed when
4551 bouncing a message because it had too many Received: header lines.
4553 20. If two routers both had headers_remove settings, and the first one included
4554 a superfluous trailing colon, the final name in the first list and the
4555 first name in the second list were incorrectly joined into one item (with a
4556 colon in the middle).
4562 1. Added -C and -D options to the exinext utility, mainly to make it easier
4563 to include in the automated testing, but these could be helpful when
4564 multiple configurations are in use.
4566 2. The exinext utility was not formatting the output nicely when there was
4567 an alternate port involved in the retry record key, nor when there was a
4568 message id as well (for retries that were specific to a specific message
4569 and a specific host). It was also confused by IPv6 addresses, because of
4570 the additional colons they contain. I have fixed the IPv4 problem, and
4571 patched it up to do a reasonable job for IPv6.
4573 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4574 delivery, the log line now contains "pipelined" if PIPELINING was used.
4576 4. An SMTP transport process used to panic and die if the bind() call to set
4577 an explicit outgoing interface failed. This has been changed; it is now
4578 treated in the same way as a connect() failure.
4580 5. A reference to $sender_host_name in the part of a conditional expansion
4581 that was being skipped was still causing a DNS lookup. This no longer
4584 6. The def: expansion condition was not recognizing references to header lines
4585 that used bh_ and bheader_.
4587 7. Added the _cache feature to named lists.
4589 8. The code for checking quota_filecount in the appendfile transport was
4590 allowing one more file than it should have been.
4592 9. For compatibility with Sendmail, the command line option
4600 and sets the incoming protocol and host name (for trusted callers). The
4601 host name and its colon can be omitted when only the protocol is to be set.
4602 Note the Exim already has two private options, -pd and -ps, that refer to
4603 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4604 "s", but I don't think that's a major issue.
4606 10. A number of refactoring changes to the code, none of which should affect
4609 (a) The number of logging options was getting close to filling up the
4610 32-bit word that was used as a bit map. I have split them into two classes:
4611 those that are passed in the argument to log_write(), and those that are
4612 only ever tested independently outside of that function. These are now in
4613 separate 32-bit words, so there is plenty of room for expansion again.
4614 There is no change in the user interface or the logging behaviour.
4616 (b) When building, for example, log lines, the code previously used a
4617 macro that called string_cat() twice, in order to add two strings. This is
4618 not really sufficiently general. Furthermore, there was one instance where
4619 it was actually wrong because one of the argument was used twice, and in
4620 one call a function was used. (As it happened, calling the function twice
4621 did not affect the overall behaviour.) The macro has been replaced by a
4622 function that can join an arbitrary number of extra strings onto a growing
4625 (c) The code for expansion conditions now uses a table and a binary chop
4626 instead of a serial search (which was left over from when there were very
4627 few conditions). Also, it now recognizes conditions like "pam" even when
4628 the relevant support is not compiled in: a suitably worded error message is
4629 given if an attempt is made to use such a condition.
4631 11. Added ${time_interval:xxxxx}.
4633 12. A bug was causing one of the ddress fields not to be passed back correctly
4634 from remote delivery subprocesses. The field in question was not being
4635 subsequently used, so this caused to problems in practice.
4637 13. Added new log selectors queue_time and deliver_time.
4639 14. Might have fixed a bug in maildirsizefile handling that threw up
4640 "unexpected character" debug warnings, and recalculated the data
4641 unnecessarily. In any case, I expanded the warning message to give more
4644 15. Added the message "Restricted characters in address" to the statements in
4645 the default ACL that block characters like @ and % in local parts.
4647 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4648 Three changes have been made:
4650 (a) There was a serious bug; a negative response to MAIL caused the whole
4651 recipient domain to be cached as invalid, thereby blocking all messages
4652 to all local parts at the same domain, from all senders. This bug has
4653 been fixed. The domain is no longer cached after a negative response to
4654 MAIL if the sender used is not empty.
4656 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4659 (c) A new callout option, "use_sender" has been added for people who want
4660 the modified behaviour.
4666 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4669 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4670 indeed breaks things for older releases.
4672 3. Added additional logging to the case where there is a problem reading data
4673 from a filter that is running in a subprocess using a pipe, in order to
4674 try to track down a specific problem.
4676 4. Testing facility fudge: when running in the test harness and attempting
4677 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4678 getting "No route to host". Convert this to a timeout.
4680 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4683 6. Some OS don't have socklen_t but use size_t instead. This affects the
4684 fifth argument of getsockopt() amongst other things. This is now
4685 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4686 can be set for individual OS. I have set it for SunOS5, OSF1, and
4687 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4688 some earlier ones do not.
4690 7. Change 4.30/15 was not doing the test caselessly.
4692 8. The standard form for an IPv6 address literal was being rejected by address
4693 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4694 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4695 this, as well as the form without the "IPv6" on the front (but only when
4696 address literals are enabled, of course).
4698 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4700 10. Exim crashed if a message with an empty sender address specified by -f
4701 encountered a router with an errors_to setting. This could be provoked only
4702 by a command such as
4706 where an empty string was supplied; "<>" did not hit this bug.
4708 11. Installed PCRE release 4.5.
4710 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4711 remained set. It is now erased.
4713 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4714 times from message ids (which are base 36 rather than the normal 62).
4716 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4717 were being counted as actual protocol errors, and logged if the log
4718 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4719 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4720 and DATA following a set of rejected RCPTs do not count as protocol errors.
4721 In other words, Exim assumes they were pipelined, though this may not
4722 actually be the case. Of course, in all cases the client gets an
4723 appropriate error code.
4725 15. If a lookup fails in an ACL condition, a message about the failure may
4726 be available; it is used if testing the ACL cannot continue, because most
4727 such messages specify what the cause of the deferral is. However, some
4728 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4729 that caused an old message to be retained and used if a later statement
4730 caused a defer, replacing the real cause of the deferral.
4732 16. If an IP address had so many PTR records that the DNS lookup buffer
4733 was not large enough to hold them, Exim could crash while trying to process
4734 the truncated data. It now detects and logs this case.
4736 17. Further to 4.21/58, another change has been made: if (and only if) the
4737 first line of a message (the first header line) ends with CRLF, a bare LF
4738 in a subsequent header line has a space inserted after it, so as not to
4739 terminate the header.
4741 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4742 unnecessarily, used atoi() instead of strtol(), and didn't check the
4743 termination when getting file sizes from file names by regex.
4745 19. Completely re-implemented the support for maildirsize files, in the light
4746 of a number of problems with the previous contributed implementation
4747 (4.30/29). In particular:
4749 . If the quota is zero, the maildirsize file is maintained, but no quota is
4752 . If the maildir directory does not exist, it is created before any attempt
4753 to write a maildirsize file.
4755 . The quota value in the file is just a cache; if the quota is changed in
4756 the transport, the new value overrides.
4758 . A regular expression is available for excluding directories from the
4761 20. The autoreply transport checks the characters in options that define the
4762 message's headers; it allows continued headers, but it was checking with
4763 isspace() after an embedded newline instead of explicitly looking for a
4766 21. If all the "regular" hosts to which an address was routed had passed their
4767 expiry times, and had not reached their retry times, the address was
4768 bounced, even if fallback hosts were defined. Now Exim should go on to try
4771 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4772 equivalent code in the SMTP transport. Some hosts send humungous responses
4773 to HELO/EHLO, more than 1024 it seems.
4775 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4776 gives compiler warnings in some environments. I've now done it "properly",
4779 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4780 (because of problems with the built-in one) was declared to return uschar *
4781 instead of char *, causing compiler failure.
4783 25. Fixed a file descriptor leak when processing alias/forward files.
4785 26. Fixed a minor format string issue in dbfn.c.
4787 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4789 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4790 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4793 29. When a sender address is verified, it is cached, to save repeating the test
4794 when there is more than one recipient in a message. However, when the
4795 verification involves a callout, it is possible for different callout
4796 options to be set for different recipients. It is too complicated to keep
4797 track of this in the cache, so now Exim always runs a verification when a
4798 callout is required, relying on the callout cache for the optimization.
4799 The overhead is duplication of the address routing, but this should not be
4802 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4803 to be verified with callout=postmaster, and the main callout worked but the
4804 postmaster check failed, the verification correctly failed. However, if a
4805 subsequent RCPT command asked for sender verification *without* the
4806 postmaster check, incorrect caching caused this verification also to fail,
4809 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4810 it was not caching the DNS options (qualify_single, search_parents) that
4811 were used when the lookup failed. A subsequent lookup with different
4812 options therefore always gave the same answer, though there were cases
4813 where it should not have. (Example: a "domains = !$mx_any" option on a
4814 dnslookup router: the "domains" option is always processed without any
4815 widening, but the router might have qualify_single set.) Now Exim uses the
4816 cached value only when the same options are set.
4818 32. Added John Jetmore's "exipick" utility to the distribution.
4820 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4821 than a timeout (e.g. a certificate is required, and is not provided), an
4822 Exim server now closes the connection immediately. Previously it waited for
4823 the client to close - but if the client is SSL, it seems that they each
4824 wait for each other, leading to a delay before one of them times out.
4826 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4827 maintained 0.8.x compatibility because I don't think many are using it, and
4828 it is clearly obsolete.
4830 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4833 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4834 client certificate was expired. A simple patch fixes this, though I don't
4835 understand the full logic of why the verify callback is called multiple
4838 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4839 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4840 which causes problems with some clients (such as the Certicom SSL Plus
4841 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4842 disables the coutermeasure allowing Eudora to connect."
4844 38. Exim was not checking that a write() to a log file succeeded. This could
4845 lead to Bad Things if a log got too big, in particular if it hit a file
4846 size limit. Exim now panics and dies if it cannot write to a log file, just
4847 as it does if it cannot open a log file.
4849 39. Modified OS/Makefile-Linux so that it now contains
4851 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4853 The two -D definitions ensure that Exim is compiled with large file
4854 support, which makes it possible to handle log files that are bigger than
4857 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4858 instance) a domain was checked against a named list that involved a lookup,
4859 causing $domain_data to be set, then another domain was checked against the
4860 same list, then the first domain was re-checked, the value of $domain_data
4861 after the final check could be wrong. In particular, if the second check
4862 failed, it could be set empty. This bug probably also applied to
4865 41. The strip_trailing_dot option was not being applied to the address given
4866 with the -f command-line option.
4868 42. The code for reading a message's header from the spool was incrementing
4869 $received_count, but never initializing it. This meant that the value was
4870 incorrect (doubled) while delivering a message in the same process in which
4871 it was received. In the most common configuration of Exim, this never
4872 happens - a fresh exec is done - but it can happen when
4873 deliver_drop_privilege is set.
4875 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4876 it now includes up to 150 characters of the unexpected data in the log
4879 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4880 and building data strings. The size of both of these buffers was 10 000
4881 bytes - far larger than anybody would *ever* want, thought I. Needless to
4882 say, somebody hit the limit. I have increased the maximum line length to
4883 20 000 and the maximum data length of concatenated lines to 100 000. I have
4884 also fixed two bugs, because there was no checking on these buffers. Tsk,
4885 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4886 buffer is too small.
4888 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4889 lsearch lookups. Now it does.
4891 46. When parsing a route_list item in a manualroute router, a fixed-length
4892 buffer was used for the list of hosts. I made this 1024 bytes long,
4893 thinking that nobody would ever have a list of hosts that long. Wrong.
4894 Somebody had a whole pile of complicated expansion conditions, and the
4895 string was silently truncated, leading to an expansion error. It turns out
4896 that it is easier to change to an unlimited length (owing to other changes
4897 that have happened since this code was originally written) than to build
4898 structure for giving a limitation error. The length of the item that
4899 expands into the list of hosts is now unlimited.
4901 47. The lsearch lookup could not handle data where the length of text line was
4902 more than 4095 characters. Such lines were truncated, leading to shortened
4903 data being returned. It should now handle lines of any length.
4905 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4906 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4909 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4910 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4911 no longer happens in comment lines. A list of replacements is now placed
4912 at the head of all of the source files, except those whose only change is
4913 to replace PERL_COMMAND in the very first #! line.
4915 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4916 messages on the queue, with a bottom-up merge sort, using code contributed
4917 by Michael Haardt. This should make operations like -bp somewhat faster on
4918 large queues. It won't affect queue runners, except when queue_run_in_order
4921 51. Installed eximstats 1.31 in the distribution.
4923 52. Added support for SRV lookups to the dnslookup router.
4925 53. If an ACL referred to $message_body or $message_body_end, the value was not
4926 reset for any messages that followed in the same SMTP session.
4928 54. The store-handling optimization for building very long strings was not
4929 differentiating between the different store pools. I don't think this
4930 actually made any difference in practice, but I've tidied it.
4932 55. While running the routers to verify a sender address, $sender_address
4933 was still set to the sender address. This is wrong, because when routing to
4934 send a bounce to the sender, it would be empty. Therefore, I have changed
4935 it so that, while verifying a sender address, $sender_address is set to <>.
4936 (There is no change to what happens when verifying a recipient address.)
4938 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4939 target A or AAAA records (if not already returned) without resetting the
4940 qualify_single or search_parents options of the DNS resolver. These are
4941 inappropriate in this case because the targets of MX and SRV records must
4942 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4943 target that, when qualified, matched something in the local domain. These
4944 two options are now turned off when doing these lookups.
4946 57. It seems that at least some releases of Reiserfs (which does not have the
4947 concept of a fixed number of inodes) returns zero and not -1 for the
4948 number of available inodes. This interacted badly with check_spool_inodes,
4949 which assumed that -1 was the "no such thing" setting. What I have done is
4950 to check that the total number of inodes is greater than zero before doing
4951 the test of how many are available.
4953 58. When a "warn" ACL statement has a log_message modifier, the message is
4954 remembered, and not repeated. This is to avoid a lot of repetition when a
4955 message has many recipients that cause the same warning to be written.
4956 Howewer, Exim was preserving the list of already written lines for an
4957 entire SMTP session, which doesn't seem right. The memory is now reset if a
4958 new message is started.
4960 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4961 debugging output unless log_rewrite was also set.
4963 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4964 of (int)(handle) when we know that handle contains (void *)(-1).
4966 61. The Exim daemon panic-logs an error return when it closes the incoming
4967 connection. However "connection reset by peer" seems to be common, and
4968 isn't really an error worthy of noting specially, so that particular error
4971 62. When Exim is trying to find all the local interfaces, it used to panic and
4972 die if the ioctl to get the interface flags failed. However, it seems that
4973 on at least one OS (Solaris 9) it is possible to have an interface that is
4974 included in the list of interfaces, but for which you get a failure error
4975 for this call. This happens when the interface is not "plumbed" into a
4976 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4977 failure of the "get flags" call assumes that the interface is down.
4979 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4980 makes life easier for people who are doing arithmetic on fields extracted
4981 from dates, where you often get leading zeros that should not be
4982 interpreted as octal.
4984 64. Added qualify_domain to the redirect router, to override the global
4987 65. If a pathologically long header line contained very many addresses (the
4988 report of this problem mentioned 10 000) and each of them was rewritten,
4989 Exim could use up a very large amount of memory. (It kept on making new
4990 copies of the header line as it rewrote, and never released the old ones.)
4991 At the expense of a bit more processing, the header rewriting function has
4992 been changed so that it no longer eats memory in this way.
4994 66. The generation of the Received: header has been moved from the time that a
4995 message starts to be received, to the time that it finishes. The timestamp
4996 in the Received: header should now be very close to that of the <= log
4997 line. There are two side-effects of this change:
4999 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5000 logged header lines no longer include the local Received: line, because
5001 it has not yet been created. The same applies to a copy of the message
5002 that is returned to a non-SMTP sender when a message is rejected.
5004 (b) When a filter file is tested using -bf, no additional Received: header
5005 is added to the test message. After some thought, I decided that this
5008 This change does not affect the value of $received_for. It is still set
5009 after address rewriting, but before local_scan() is called.
5011 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5013 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5014 gave an unhelpful panic error message, and a defer error. I have managed to
5015 change this behaviour so that it now rejects any supplied certificate,
5016 which seems right, as the list of acceptable certificates is empty.
5018 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5019 gave an unhelpful defer error. I have not managed to make this reject any
5020 supplied certificates, but the error message it gives is "no certificate
5021 supplied", which is not helpful.
5023 70. exigrep's output now also includes lines that are not associated with any
5024 message, but which match the given pattern. Implemented by a patch from
5025 Martin Sluka, which also tidied up the Perl a bit.
5027 71. Recipient callout verification, like sender verification, was using <> in
5028 the MAIL FROM command. This isn't really the right thing, since the actual
5029 sender may affect whether the remote host accepts the recipient or not. I
5030 have changed it to use the actual sender in the callout; this means that
5031 the cache record is now keyed on a recipient/sender pair, not just the
5032 recipient address. There doesn't seem to be a real danger of callout loops,
5033 since a callout by the remote host to check the sender would use <>.
5034 [SEE ABOVE: changed after hitting problems.]
5036 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5037 temporary errors. However, in the case of such a code being given after
5038 the end of a data transmission (i.e. after ".") Exim was failing to write
5039 a retry record for the message. (Yes, there was some broken host that was
5040 actually sending 8xx at this point.)
5042 73. An unknown lookup type in a host list could cause Exim to panic-die when
5043 the list was checked. (An example that provoked this was putting <; in the
5044 middle of a list instead of at the start.) If this happened during a DATA
5045 ACL check, a -D file could be left lying around. This kind of configuration
5046 error no longer causes Exim to die; instead it causes a defer errror. The
5047 incident is still logged to the main and panic logs.
5049 74. Buglet left over from Exim 3 conversion. The message "too many messages
5050 in one connection" was written to the rejectlog but not the mainlog, except
5051 when address rewriting (yes!) was being logged.
5053 75. Added write_rejectlog option.
5055 76. When a system filter was run not as root (that is, when system_filter_user
5056 was set), the values of the $n variables were not being returned to the
5057 main process; thus, they were not subsequently available in the $sn
5060 77. Added +return_path_on_delivery log selector.
5062 78. A connection timeout was being treated differently from recipients deferred
5063 when testing hosts_max_try with a message that was older than the host's
5064 retry timeout. (The host should not be counted, thus allowing all hosts to
5065 be tried at least once before bouncing.) This may have been the cause of an
5066 occasionally reported bug whereby a message would remain on the queue
5067 longer than the retry timeout, but would be bounced if a delivery was
5068 forced. I say "may" because I never totally pinned down the problem;
5069 setting up timeout/retry tests is difficult. See also the next item.
5071 79. The ultimate address timeout was not being applied to errors that involved
5072 a combination of host plus message (for example, a timeout on a MAIL
5073 command). When an address resolved to a number of possible hosts, and they
5074 were not all tried for each delivery (e.g. because of hosts_max_try), a
5075 message could remain on the queue longer than the retry timeout.
5077 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5080 81. Fixed an obscure SMTP outgoing bug which required at least the following
5081 conditions: (a) there was another message waiting for the same server;
5082 (b) the server returned 5xx to all RCPT commands in the first message so
5083 that the message was not completed; (c) the server dropped the connection
5084 or gave a negative response to the RSET that Exim sends to abort the
5085 transaction. The observed case was a dropped connection after DATA that had
5086 been sent in pipelining mode. That is, the server had advertised PIPELINING
5087 but was not implementing it correctly. The effect of the bug was incorrect
5088 behaviour, such as trying another host, and this could lead to a crash.
5094 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5095 and daemon.c were passed as pointers to ints; they should have been
5096 pointers to socklen_t variables (which are typically unsigned ints).
5098 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5101 3. Fixed a really odd bug that affected only the testing scheme; patching a
5102 certain fixed string in the binary changed the value of another string that
5103 happened to be identical to the end of the original first string.
5105 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5106 name", it returns that address as the IP address. On some operating
5107 systems (e.g. Solaris), it also passes back the IP address string as the
5108 "host name". However, on others (e.g. Linux), it passes back an empty
5109 string. Exim wasn't checking for this, and was changing the host name to an
5110 empty string, assuming it had been canonicized.
5112 5. Although rare, it is permitted to have more than one PTR record for a given
5113 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5114 all the names associated with an address, because they do in Solaris.
5115 However, it seems that they do not in Linux for data that comes from the
5116 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5117 I found this out when I moved to a new Linux workstation and tried to run
5118 the Exim test suite.
5120 To get round this problem I have changed the code so that it now does its
5121 own call to the DNS to look up PTR records when searching for a host name.
5122 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5123 addresses that are only in /etc/hosts are still found.
5125 This behaviour is, however, controlled by an option called host_lookup_
5126 order, which defaults to "bydns:byaddr". If people want to use the other
5127 order, or indeed, just use one or the other means of lookup, they can
5128 specify it in this variable.
5130 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5131 some operating systems, this comes back from gethostbyaddr() as an empty
5132 string, and this is what Exim used to test for. However, it seems that in
5133 other systems, "." is yielded. Exim now tests for this case too.
5135 7. The values of check_spool_space and check_log_space are now held internally
5136 as a number of kilobytes instead of an absolute number of bytes. If a
5137 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5138 kilobyte. This means that much larger values can be stored.
5140 8. Exim monitor: an attempt to get the action menu when not actually pointing
5141 at a message produces an empty menu entitled "No message selected". This
5142 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5143 no entries in it ("Shell widget menu has zero width and/or height"). So I
5144 have added a single, blank menu entry in this case.
5146 9. Added ${quote_local_part.
5148 10. MIME decoding is now applied to the contents of Subject: header lines when
5151 11. Now that a reference to $sender_host_address automatically causes a reverse
5152 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5153 host lookup before query-style lookups in lists that might use this
5154 variable. This has therefore been abolished, and the "net-" prefix is no
5155 longer necessary for query-style lookups.
5157 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5158 have been a typo for LFLAGS, so it has been changed.
5160 13. The install script calls Exim with "-C /dev/null" in order to find the
5161 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5162 to be output. Howeve, since Exim outputs its version number before the
5163 error, it didn't break the script. It just looked ugly. I fixed this by
5164 always allowing "-C /dev/null" if the caller is root.
5166 14. Ignore overlarge ACL variable number when reading spool file - insurance
5167 against a later release with more variables having written the file.
5169 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5170 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5171 this, as well as the form without the "IPv6" on the front.
5173 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5174 OS/Makefile-Darwin file.
5176 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5177 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5179 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5180 variable. However, some OS use ints for inodes. I've added cast to long int
5181 to get rid of the compiler warning.
5183 19. I had forgotten to lock out "/../" in configuration file names when
5184 ALT_CONFIG_PREFIX was set.
5186 20. Routers used for verification do not need to specify transports. However,
5187 if such a router generated a host list, and callout was configured, Exim
5188 crashed, because it could not find a port number from the (non-existent)
5189 transport. It now assumes port 25 in this circumstance.
5191 21. Added the -t option to exigrep.
5193 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5194 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5195 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5196 them is set, LOOKUP_LSEARCH is forced.
5198 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5200 24. Added sender and host information to the "rejected by local_scan()" log
5201 line; previously there was no indication of these.
5203 25. Added .include_if_exists.
5205 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5206 Linux. It turns out that not all file systems support this. Apparently some
5207 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5208 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5209 not supported on the file descriptor, is now ignored when Exim is trying to
5210 sync a directory. This applies only to Linux.
5212 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5214 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5215 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5218 29. Added support for maildirsize files from supplied patch (modified a bit).
5220 30. The use of :fail: followed by an empty string could lead Exim to respond to
5221 sender verification failures with (e.g.):
5223 550 Verification failed for <xxx>
5224 550 Sender verify failed
5226 where the first response line was missing the '-' that indicates it is not
5227 the final line of the response.
5229 31. The loop for finding the name of the user that called Exim had a hardwired
5230 limit of 10; it now uses the value of finduser_retries, which is used for
5231 all other user lookups.
5233 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5236 33. Exim was neglecting to zero errno before one call of strtol() when
5237 expanding a string and expecting an integer value. On some systems this
5238 resulted in spurious "integer overflow" errors. Also, it was casting the
5239 result into an int without checking.
5241 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5242 did not work. The code looks as if it has *never* worked, though it appears
5243 to have been documented since at least releast 1.62. I have made it work.
5245 35. The "timeout_DNS" error in retry rules, also documented since at least
5246 1.62, also never worked. As it isn't clear exactly what this means, and
5247 clearly it isn't a major issue, I have abolished the feature by treating it
5248 as "timeout", and writing a warning to the main and panic logs.
5250 36. The display of retry rules for -brt wasn't always showing the error code
5253 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5254 timeout_connect_A, timeout_connect_MX.
5256 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5257 to the empty sender.
5259 39. The daemon was not analysing the content of -oX till after it had closed
5260 stderr and disconnected from the controlling terminal. This meant that any
5261 syntax errors were only noted on the panic log, and the return code from
5262 the command was 0. By re-arranging the code a little, I've made the
5263 decoding happen first, so such errors now appear on stderr, and the return
5264 code is 1. However, the actual setting up of the sockets still happens in
5265 the disconnected process, so errors there are still only recorded on the
5268 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5269 connections (as happens on some IP stacks) was logged at start up time as
5270 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5271 it from "IPv6 and IPv4", which means that two separate sockets are being
5274 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5275 says whether AF_INET or AF_INET6 was passed as an argument.
5277 42. Exiwhat output was messed up when time zones were included in log
5280 43. Exiwhat now gives more information about the daemon's listening ports,
5281 and whether -tls-on-connect was used.
5283 44. The "port" option of the smtp transport is now expanded.
5285 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5286 silently ignored. Now an error message is written to the main and panic
5289 46. There's a new ACL modifier called "logwrite" which writes to a log file
5290 as soon as it is encountered.
5292 47. Added $local_user_uid and $local_user_gid at routing time.
5294 48. Exim crashed when trying to verify a sender address that was being
5297 49. Exim was recognizing only a space character after ".include". It now also
5298 recognizes a tab character.
5300 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5301 extracting the relevant information from the specification. The man page no
5302 longer contains scrambled data for the -d option, and I've added a section
5303 at the front about calling Exim under different names.
5305 51. Added "extra_headers" argument to the "mail" command in filter files.
5307 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5310 53. Installed eximstats 1.29.
5312 54. Added transport_filter_timeout as a generic transport option.
5314 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5315 Cc: header lines. This was required by RFC 822, but it not required by RFC
5318 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5319 incoming messages that did not have them. Now it does so only if the
5320 message originates locally, that is, if there is no associated remote host
5321 address. When Resent- header lines are present, this applies to the Resent-
5322 lines rather than the non-Resent- lines.
5324 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5325 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5327 58. Messages for configuration errors now include the name of the main
5328 configuration file - useful now that there may be more than one file in a
5329 list (.included file names were always shown).
5331 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5332 for those rare installations that do not start the daemon as root or run it
5333 setuid root. I've cut out the call to initgroups() if the daemon is not
5336 60. The Exim user and group can now be bound into the binary as text strings
5337 that are looked up at the start of Exim's processing.
5339 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5341 62. Added $mailstore_basename variable.
5343 63. Installed patch to sieve.c from Michael Haardt.
5345 64. When Exim failed to open the panic log after failing to open the main log,
5346 the original message it was trying to log was written to stderr and debug
5347 output, but if they were not available (the usual case in production), it
5348 was lost. Now it is written to syslog before the two lines that record the
5349 failures to open the logs.
5351 65. Users' Exim filters run in subprocesses under the user's uid. It is
5352 possible for a "deliver" command or an alias in a "personal" command to
5353 provoke an address rewrite. If logging of address rewriting is configured,
5354 this fails because the process is not running as root or exim. There may be
5355 a better way of dealing with this, but for the moment (because 4.30 needs
5356 to be released), I have disabled address rewrite logging when running a
5357 filter in a non-root, non-exim process.
5363 1. The buildconfig auxiliary program wasn't quoting the value set for
5364 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5365 not defined. This bug was masked in 4.22 by the effect that was fixed in
5368 2. Some messages that were rejected after a message id was allocated were
5369 shown as "incomplete" by exigrep. It no longer does this for messages that
5370 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5372 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5373 have allow_domain_literals set, the ID did not get logged in the <= line.
5374 Domain literals are now always recognized in Message-ID: header lines.
5376 4. The first argument for a ${extract expansion item is the key name or field
5377 number. Leading and trailing spaces in this item were not being ignored,
5378 causing some misleading effects.
5380 5. When deliver_drop_privilege was set, single queue runner processes started
5381 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5382 same command in the process it spins off) were not dropping privilege.
5384 6. When the daemon running as "exim" started a queue runner, it always
5385 re-executed Exim in the spun-off process. This is a waste of effort when
5386 deliver_drop_privilege is set. The new process now just calls the
5387 queue-runner function directly.
5393 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5396 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5397 ignored. Though the use of -oP was forcing the writing of a pid file, it
5398 was always written to the default place.
5400 3. If the message "no IP address found for host xxxx" is generated during
5401 incoming verification, it is now followed by identification of the incoming
5402 connection (so you can more easily find what provoked it).
5404 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5406 5. Added some features to "harden" Exim a bit more against certain attacks:
5408 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5409 be put in Local/Makefile. This is like the never_users runtime option,
5410 but it cannot be overridden. The default setting is "root".
5412 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5413 prefix string with which any file named in a -C command line option
5416 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5417 is retained for -C and -D only if the caller of Exim is root. Without
5418 it, the exim user may also use -C and -D and retain privilege.
5420 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5421 command line option is disabled.
5423 6. Macro names set by the -D option must start with an upper case letter, just
5424 like macro names defined in the configuration file.
5426 7. Added "dereference=" facility to LDAP.
5428 8. Two instances of the typo "uknown" in the source files are fixed.
5430 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5431 the Configure-Makefile script screwed up while processing it.
5433 10. Incorporated PCRE 4.4.
5435 11. The SMTP synchronization check was not operating right at the start of an
5436 SMTP session. For example, it could not catch a HELO sent before the client
5437 waited for the greeting. There is now a check for outstanding input at the
5438 point when the greeting is written. Because of the duplex, asynchronous
5439 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5440 way, but not yet received, when the check is performed.
5442 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5443 on TCP/IP sockets, because this apparently causes some broken clients to
5446 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5447 unchanged) from the Cygwin maintainer.
5449 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5450 was supported instead of testing for mbx. Effectively a typo.
5452 15. The spa authenticator server code was not checking that the input it
5453 received was valid base64.
5455 16. The debug output line for the "set" modifier in ACLs was not showing the
5456 name of the variable that was being set.
5458 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5460 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5461 lookup on its own. Something else was needed to trigger the lookup. For
5462 example, a match in host_lookup or the need for a host name in a host list.
5463 Now, if $sender_host_name is referenced and the host name has not yet been
5464 looked up, a lookup is performed. If the lookup fails, the variable remains
5465 empty, and $host_lookup_failed is set to "1".
5467 19. Added "eqi" as a case-independent comparison operator.
5469 20. The saslauthd authentication condition could segfault if neither service
5470 nor realm was specified.
5472 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5473 error message that was logged was misleading, and incoming SMTP
5474 connections were dropped. The message is now more accurate, and temporary
5475 errors are given to SMTP connections.
5477 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5478 (except RSET) with a 421 error, until QUIT is received. However, it was
5479 failing to send a response to QUIT.
5481 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5482 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5483 ACL is now run after the helo_try_verify_hosts code.
5485 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5486 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5487 case-independent, so other case variants are also recognized). Apparently
5488 some systems use these upper case variants.
5490 25. If more than two messages were waiting for the same host, and a transport
5491 filter was specified for the transport, Exim sent two messages over the
5492 same TCP/IP connection, and then failed with "socket operation on non-
5493 socket" when it tried to send the third.
5495 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5497 27. The extern definition of crypt16() in expand.c was not being excluded when
5498 the OS had its own crypt16() function.
5500 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5501 as a preferred synonym for return_size_limit, both as an option and as an
5504 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5506 30. Changed the default configuration ACL to relax the local part checking rule
5507 for addresses that are not in any local domains. For these addresses,
5508 slashes and pipe symbols are allowed within local parts, but the sequence
5509 /../ is explicitly forbidden.
5511 31. SPA server authentication was not clearing the challenge buffer before
5514 32. log_message in a "warn" ACL statement was writing to the reject log as
5515 well as to the main log, which contradicts the documentation and doesn't
5516 seem right (because no rejection is happening). So I have stopped it.
5518 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5519 However, I am unable to do any testing of this.
5521 34. Fixed an infelicity in the appendfile transport. When checking directories
5522 for a mailbox, to see if any needed to be created, it was accidentally
5523 using path names with one or more superfluous leading slashes; tracing
5524 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5526 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5527 contents are added to the log line that is written for every discarded
5528 recipient. (Previously a log_message setting was ignored.)
5530 36. The ${quote: operator now quotes the string if it is empty.
5532 37. The install script runs exim in order to find its version number. If for
5533 some reason other than non-existence or emptiness, which it checks, it
5534 could not run './exim', it was installing it with an empty version number,
5535 i.e. as "exim-". This error state is now caught, and the installation is
5538 38. An argument was missing from the function that creates an error message
5539 when Exim fails to connect to the socket for saslauthd authentication.
5540 This could cause Exim to crash, or give a corrupted message.
5542 39. Added isip, isip4, and isip6 to ${if conditions.
5544 40. The ACL variables $acl_xx are now saved with the message, and can be
5545 accessed later in routers, transports, and filters.
5547 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5548 strings in the file are not string-expanded.
5550 42. If a MAIL command specified a SIZE value that was too large to fit into an
5551 int variable, the check against message_size_limit failed. Such values are
5552 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5553 day this will have to be increased, but I don't think I want to be around
5554 when emails are that large.
5561 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5562 iconv() is not standard in FreeBSD.
5564 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5565 IPv6 enabled. The observed symptom was a segmentation fault on return from
5566 the function os_common_find_running_interfaces() in src/os.c.
5568 3. In the check_special_case() function in daemon.c I had used "errno" as an
5569 argument name, which causes warnings on some systems. This was basically a
5570 typo, since it was named "eno" in the comments!
5572 4. The code that waits for the clock to tick (at a resolution of some fraction
5573 of a second) so as to ensure message-id uniqueness was always waiting for
5574 at least one whole tick, when it could have waited for less. [This is
5575 almost certainly not relevant at current processor speeds, where it is
5576 unlikely to ever wait at all. But we try to future-proof.]
5578 5. The function that sleeps for a time interval that includes fractions of a
5579 second contained a race. It did not block SIGALRM between setting the
5580 timer, and suspending (a couple of lines later). If the interval was short
5581 and the sigsuspend() was delayed until after it had expired, the suspension
5582 never ended. On busy systems this could lead to processes getting stuck for
5585 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5586 process, before it forks any delivery processes. The open lookup caching
5587 mechanism meant that the open file or database connection was passed into
5588 the delivery process. The problem was that delivery processes always tidy
5589 up cached lookup data. This could cause a problem for the next delivery
5590 process started by the queue runner, because the external queue runner
5591 process does not know about the closure. So the next delivery process
5592 still has data in the lookup cache. In the case of a file lookup, there was
5593 no problem because closing a file descriptor in a subprocess doesn't affect
5594 the parent. However, if the lookup was caching a connection to a database,
5595 the connection was closed, and the second delivery process was likely to
5596 see errors such as "PGSQL: query failed: server closed the connection
5597 unexpectedly". The problem has been fixed by closing all cached lookups
5598 in a queue runner before running a delivery process.
5600 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5601 seem to have the "const" qualifier which it has on other OS. I've
5604 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5605 *of the same type* (client or server) with the same public name that an
5606 error should be diagnosed.
5608 9. When Exim looked up a host name for an IP address, but failed to find the
5609 original IP address when looking up the host name (a safety check), it
5610 output the message "<ip address> does not match any IP for NULL", which was
5611 confusing, to say the least. The bug was that the host name should have
5612 appeared instead of "NULL".
5614 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5615 user that is built into the binary, and the -C or -D options is used, root
5616 privilege is dropped before the configuration file is read. In addition,
5617 logging is switched to stderr instead of the normal log files. If the
5618 configuration then re-defines the Exim user, the unprivileged environment
5619 is probably not what is expected, so Exim logs a panic warning message (but
5622 However, if deliver_drop_privilege is set, the unprivileged state may well
5623 be exactly what is intended, so the warning has been cut out in that case,
5624 and Exim is allowed to try to write to its normal log files.
5630 1. smtp_return_error_details was not giving details for temporary sender
5631 or receiver verification errors.
5633 2. Diagnose a configuration error if two authenticators have the same public
5636 3. Exim used not to create the message log file for a message until the first
5637 delivery attempt. This could be confusing when incoming messages were held
5638 for policy or load reasons. The message log file is now created at the time
5639 the message is received, and an initial "Received" line is written to it.
5641 4. The automatically generated man page for command line options had a minor
5642 bug that caused no ill effects; however, a more serious problem was that
5643 the procedure for building the man page automatically didn't always
5644 operate. Consequently, release 4.20 contains an out-of-date version. This
5645 shouldn't happen again.
5647 5. When building Exim with embedded Perl support, the script that builds the
5648 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5649 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5651 6. The freeze_tell option was not being used for messages that were frozen on
5652 arrival, either by an ACL or by local_scan().
5654 7. Added the smtp_incomplete_transaction log selector.
5656 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5657 was accepting AUTH without a new EHLO.
5659 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5660 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5663 10. Logging of TCP/IP connections (when configured) now happens in the main
5664 daemon process instead of the child process, so that the TCP/IP connection
5665 count is more accurate (but it can never be perfect).
5667 11. The use of "drop" in a nested ACL was not being handled correctly in the
5668 outer ACL. Now, if condition failure induced by the nested "drop" causes
5669 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5670 or "require"), the connection is dropped.
5672 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5673 that yield "discard" can now be used with an "accept" or a "discard" verb,
5674 but an error is generated for any others (because I can't see a useful way
5675 to define what should happen).
5677 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5678 are now processed in the same way as lines in the Exim configuration file.
5679 In particular, continuation lines are supported.
5681 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5683 15. Added -ti meaning -t -i.
5685 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5686 domains, and warn by logging if others are found.
5688 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5689 returned from GIFCONF when Exim is trying to find the list of interfaces on
5690 a host. The code in os.c has been modified to copy each ifreq to an aligned
5691 structure in all cases.
5693 Also, in some cases, the returned ifreq's were being copied to a 'struct
5694 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5695 means the last couple of bytes of an IPv6 address could be chopped if the
5696 ifreq contained only a normal sockaddr (14 bytes storage).
5698 18. Named domain lists were not supported in the hosts_treat_as_local option.
5699 An entry such as +xxxx was not recognized, and was treated as a literal
5702 19. Ensure that header lines added by a DATA ACL are included in the reject log
5703 if the ACL subsequently rejects the message.
5705 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5706 MD5 (which is deprecated).
5708 21. When testing a filter file using -bf, Exim was writing a message when it
5709 took the sender from a "From " line in the message, but it was not doing so
5710 when it took $return_path from a Return-Path: header line. It now does.
5712 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5713 with a valid header line field name (a series of printing characters
5714 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5716 23. Changed "disc" in the source to "disk" to conform to the documentation and
5717 the book and for uniformity.
5719 24. Ignore Sendmail's -Ooption=value command line item.
5721 25. When execve() failed while trying to run a command in a pipe transport,
5722 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5723 could be confused with a return value of 69 from the command itself. This
5724 has been changed to 127, the value the shell returns if it is asked to run
5725 a non-existent command. The wording for the related log line suggests a
5726 non-existent command as the problem.
5728 26. If received_header_text expands to an empty string, do not add a Received:
5729 header line to the message. (Well, it adds a token one on the spool, but
5730 marks it "old" so that it doesn't get used or transmitted.)
5732 27. Installed eximstats 1.28 (addition of -nt option).
5734 28. There was no check for failure on the call to getsockname() in the daemon
5735 code. This can fail if there is a shortage of resources on the system, with
5736 ENOMEM, for example. A temporary error is now given on failure.
5738 29. Contrary to the C standard, it seems that in some environments, the
5739 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5740 program. Exim now does this explicitly; it affects the formatting of
5741 timestamps using strftime().
5743 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5744 complaints. I've now initialized all the variables, to avoid this.
5746 32. Header lines added by a system filter were not being "seen" during
5747 transport-time rewrites.
5749 33. The info_callback() function passed to OpenSSL is set up with type void
5750 (*)(SSL *, int, int), as described somewhere. However, when calling the
5751 function (actually a macro) that sets it up, the type void(*)() is
5752 expected. I've put in a cast to prevent warnings from picky compilers.
5754 34. If a DNS black list lookup found a CNAME record, but there were no A
5755 records associated with the domain it pointed at, Exim crashed.
5757 35. If a DNS black list lookup returned more than one A record, Exim ignored
5758 all but the first. It now scans all returned addresses if a particular IP
5759 value is being sought. In this situation, the contents of the
5760 $dnslist_value variable are a list of all the addresses, separated by a
5763 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5764 to accept a host name and all its aliases if the forward lookup for any of
5765 them yielded the IP address of the incoming connection. Now it accepts only
5766 those names whose forward lookup yields the correct IP address. Any other
5767 names are discarded. This closes a loophole whereby a rogue DNS
5768 administrator could create reverse DNS records to break through a
5769 wildcarded host restriction in an ACL.
5771 37. If a user filter or a system filter that ran in a subprocess used any of
5772 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5773 the wrong values were passed to the pipe command ($thisaddress had the
5774 value of $0, $0 had the value of $1, etc). This bug was introduced by
5775 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5777 38. Improved the line breaking for long SMTP error messages from ACLs.
5778 Previously, if there was no break point between 40 and 75 characters, Exim
5779 left the rest of the message alone. Two changes have been made: (a) I've
5780 reduced the minimum length to 35 characters; (b) if it can't find a break
5781 point between 35 and 75 characters, it looks ahead and uses the first one
5782 that it finds. This may give the occasional overlong line, but at least the
5783 remaining text gets split now.
5785 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5786 file descriptors might be low, and that setting 1000 would always raise it.
5787 It turns out that in some environments, the limit is already over 1000 and
5788 that lowering it causes trouble. So now Exim takes care not to decrease it.
5790 40. When delivering a message, the value of $return_path is set to $sender_
5791 address at the start of routing (routers may change the value). By an
5792 oversight, this default was not being set up when an address was tested by
5793 -bt or -bv, which affected the outcome if any router or filter referred to
5796 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5797 log only when "message" or "log_message" are set. However, if one of the
5798 conditions was an address verification, or a call to a nested ACL, the
5799 messages generated by the underlying test were being passed through. This
5800 no longer happens. The underlying message is available in $acl_verify_
5801 message for both "message" and "log_message" expansions, so it can be
5802 passed through if needed.
5804 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5805 new expansion $bh_ to give the encoded byte string without charset
5806 translation. Translation happens only if iconv() is available; HAVE_ICONV
5807 indicates this at build time. HEADERS_CHARSET gives the charset to
5808 translate to; headers_charset can change it in the configuration, and
5809 "headers charset" can change it in an individual filter file.
5811 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5812 that creates RFC 2047 encoded "words" labels them as that charset instead
5813 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5814 expansion operator; (ii) when Exim creates a From: line for a local
5815 message; (iii) when a header line is rewritten to include a "phrase" part.
5817 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5818 buggy, causing it to skip the first lines of messages whose message ID
5819 ended in 'D'. This would not have bitten before Exim release 4.14, because
5820 message IDs were unlikely to end in 'D' before then. The effect was to have
5821 incorrect size information for certain domains.
5823 45. #include "config.h" was missing at the start of the crypt16.c module. This
5824 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5827 46. If there was a timeout during a "random" callout check, Exim treated it as
5828 a failure of the random address, and carried on sending RSET and the real
5829 address. If the delay was just some slowness somewhere, the response to the
5830 original RCPT would be taken as a response to RSET and so on, causing
5831 mayhem of various kinds.
5833 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5834 when I implemented it. It didn't allow for the fact that some option values
5835 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5836 the right test for positive values.
5838 48. Domain names in DNS records are case-independent. Exim always looks them up
5839 in lower case. Some resolvers return domain names in exactly the case they
5840 appear in the zone file, that is, they may contain uppercase letters. Not
5841 all resolvers do this - some return always lower case. Exim was treating a
5842 change of case by a resolver as a change of domain, similar to a widening
5843 of a domain abbreviation. This triggered its re-routing code and so it was
5844 trying to route what was effectively the same domain again. This normally
5845 caused routing to fail (because the router wouldn't handle the domain
5846 twice). Now Exim checks for this case specially, and just changes the
5847 casing of the domain that it ultimately uses when it transmits the message
5850 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5853 50. If a filter generated a file delivery with a non-absolute name (possible if
5854 no home directory exists for the router), the forbid_file option was not
5857 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5858 the existing equality matching.
5860 52. Exim was using ints instead of ino_t variables in some places where it was
5861 dealing with inode numbers.
5863 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5864 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5865 called TMPDIR, and if it finds it is different, it changes its value.
5867 54. The smtp_printf() function is now made available to local_scan() so
5868 additional output lines can be written before returning. There is also an
5869 smtp_fflush() function to enable the detection of a dropped connection.
5870 The variables smtp_input and smtp_batched_input are exported to
5873 55. Changed the default runtime configuration: the message "Unknown user"
5874 has been removed from the ACL, and instead placed on the localuser router,
5875 using the cannot_route_message feature. This means that any verification
5876 failures that generate their own messages won't get overridden. Similarly,
5877 the "Unrouteable address" message that was in the ACL for unverifiable
5878 relay addresses has also been removed.
5880 56. Added hosts_avoid_esmtp to the smtp transport.
5882 57. The exicyclog script was not checking for the esoteric option
5883 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5884 will work only if exicyclog is run under the appropriate euid.
5886 58. Following a discussion on the list, the rules by which Exim recognises line
5887 endings on incoming messages have been changed. The -dropcr and drop_cr
5888 options are now no-ops, retained only for backwards compatibility. The
5889 following line terminators are recognized: LF CRLF CR. However, special
5890 processing applies to CR:
5892 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5893 nor a local message in the state where . is a terminator.
5895 (ii) If a bare CR is encountered in a header line, an extra space is added
5896 after the line terminator so as not to end the header. The reasoning
5897 behind this is that bare CRs in header lines are most likely either
5898 to be mistakes, or people trying to play silly games.
5900 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5901 was being incorrectly given as 18 bytes larger than it should have been.
5902 This is a VOB (very old bug).
5904 60. This may never have affected anything current, but just in case it has:
5905 When the local host is found other than at the start of a list of hosts,
5906 the local host, those with the same MX, and any that follow, are discarded.
5907 When the list in question was part of a longer list of hosts, the following
5908 hosts (not currently being processed) were also being discarded. This no
5909 longer happens. I'm not sure if this situation could ever has previously
5912 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5915 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5916 header. This is something that is recommended in a new Internet Draft, and
5917 is something that is documented as being done by Sendmail. There are two
5918 possible values. For messages generated by the autoreply transport, Exim
5921 Auto-Submitted: auto-replied
5923 whereas for all other generated messages (e.g. bounces) it adds
5925 Auto-Submitted: auto-generated
5927 63. The "personal" condition in filters now includes a test for the
5928 Auto-Submitted: header. If it contains the string "auto-" the message it
5929 not considered personal.
5931 64. Added rcpt_include_affixes as a generic transport option.
5933 65. Added queue_only_override (default true).
5935 66. Added the syslog_duplication option.
5937 67. If what should have been the first header line of a message consisted of
5938 a space followed by a colon, Exim was mis-interpreting it as a header line.
5939 It isn't of course - it is syntactically invalid and should therefore be
5940 treated as the start of the message body. The misbehaviour could have
5941 caused a number of strange effects, including loss of data in subsequent
5942 header lines, and spool format errors.
5944 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5945 client host had authenticated. This control can now be exercised by an ACL
5946 for more flexibility.
5948 69. By default, callouts do not happen when testing with -bh. There is now a
5949 variant, -bhc, which does actually run the callout code, including
5950 consulting and updating the callout cache.
5952 70. Added support for saslauthd authentication, courtesy of Alexander
5955 71. If statvfs() failed on the spool or log directories while checking their
5956 size for availability, Exim confusingly gave the error "space shortage".
5957 Furthermore, in debugging mode it crashed with a floating point exception.
5958 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5959 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5960 serious problem, Exim now writes to the main and panic logs when this
5961 happens, with details of the failure. It then refuses to accept the
5962 incoming message, giving the message "spool directory problem" or "log
5963 directory problem" with a 421 code for SMTP messages.
5965 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5966 0, 1, and 2 exist, because some OS complain for execs without them (see
5967 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5968 descriptors. However, the code omitted to check that the open succeeded,
5969 causing mysterious errors if for some reason the permissions on /dev/null
5970 got screwed. Now Exim writes a message to the main and panic logs, and
5971 bombs out if it can't open /dev/null.
5973 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5974 interact so that it is all more flexible. It is supposed to remain
5975 backwards compatible. Also added extra_local_interfaces.
5977 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5978 to bomb out with an assertion failure - to the client this appears as a
5979 connection drop. This problem occurs in the part of the code that was taken
5980 from the Samba project. Fortunately, the assertion is in a very simple
5981 function, so I have fixed this by reproducing the function inline in the
5982 one place where it is called, and arranging for authentication to fail
5983 instead of killing the process with assert().
5985 75. The SPA client code was not working when the server requested OEM rather
5986 than Unicode encoding.
5988 76. Added code to make require_files with a specific uid setting more usable in
5989 the case where statting the file as root fails - usually a non-root-mounted
5990 NFS file system. When this happens and the failure is EACCES, Exim now
5991 forks a subprocess and does the per-uid checking as the relevant uid.
5993 77. Added process_log_path.
5995 78. If log_file_path was not explicitly set, a setting of check_log_space or
5996 check_log_inodes was ignored.
5998 79. If a space check for the spool or log partitions fails, the incident is now
5999 logged. Of course, in the latter case the data may get lost...
6001 80. Added the %p formatting code to string_format() so that it can be used to
6002 print addresses in debug_print(). Adjusted all the address printing in the
6003 debugging in store.c to use %p rather than %d.
6005 81. There was a concern that a line of code in smtp_in.c could overflow a
6006 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6007 initially expressed, the concern was not well-founded, because trailing
6008 spaces are removed early. However, if the trailing spaces were followed by
6009 a NULL, they did not get removed, so the overflow was possible. Two fixes
6012 (a) I re-wrote the offending code in a cleaner fashion.
6013 (b) If an incoming SMTP command contains a NULL character, it is rejected
6016 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6017 runs initgroups(), so that if the Exim user is in any additional groups,
6018 they will be used during message reception.
6024 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob