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.
104 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
106 PP/02 Make -n do something, by making it not do something.
107 When combined with -bP, the name of an option is not output.
109 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
112 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
113 $sender_host_name and config options to manage this, and basic check
116 PP/05 DSCP support for outbound connections and control modifier for inbound.
118 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
119 (Only plugin which currently uses this is kerberos4, which nobody should
120 be using, but we should make it available and other future plugins might
121 conceivably use it, even though it would break NAT; stuff *should* be
122 using channel bindings instead).
124 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
125 name; added for Sendmail compatibility; requires admin caller.
126 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
127 just ignore it); requires trusted caller.
128 Also parse but ignore: -Ac -Am -X<logfile>
131 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
133 TL/02 Add +smtp_confirmation as a default logging option.
135 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
136 Patch by Magnus Holmgren from 2007-02-20.
138 TL/04 Bugzilla 1281 - Spec typo.
139 Bugzilla 1283 - Spec typo.
140 Bugzilla 1290 - Spec grammar fixes.
142 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
144 TL/06 Add Experimental DMARC support using libopendmarc libraries.
146 TL/07 Fix an out of order global option causing a segfault. Reported to dev
147 mailing list by by Dmitry Isaikin.
149 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
151 JH/02 Support "G" suffix to numbers in ${if comparisons.
153 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
155 NM/01 Bugzilla 1197 - Spec typo
156 Bugzilla 1196 - Spec examples corrections
158 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
160 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
161 gnutls_enable_pkcs11, but renamed to more accurately indicate its
164 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
165 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
167 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
168 "acl {{name}{arg}...}", and optional args on acl condition
171 JH/05 Permit multiple router/transport headers_add/remove lines.
173 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
175 JH/07 Avoid using a waiting database for a single-message-only transport.
176 Performance patch from Paul Fisher. Bugzilla 1262.
178 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
181 JH/09 Add $headers_added variable, with content from use of ACL modifier
182 add_header (but not yet added to the message). Bugzilla 199.
184 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
185 Pulled from Bugzilla 817 by Wolfgang Breyha.
187 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
189 (nb: this is the same fix as in Exim 4.80.1)
191 JH/11 Add A= logging on delivery lines, and a client_set_id option on
194 JH/12 Add optional authenticated_sender logging to A= and a log_selector
197 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
199 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
200 advertise SMTP AUTH mechanism to us, instead of a generic
201 protocol violation error. Also, make Exim more robust to bad
202 data from the Dovecot auth socket.
204 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
206 When a queue runner is handling a message, Exim first routes the
207 recipient addresses, during which it prunes them based on the retry
208 hints database. After that it attempts to deliver the message to
209 any remaining recipients. It then updates the hints database using
212 So if a recipient address works intermittently, it can get repeatedly
213 deferred at routing time. The retry hints record remains fresh so the
214 address never reaches the final cutoff time.
216 This is a fairly common occurrence when a user is bumping up against
217 their storage quota. Exim had some logic in its local delivery code
218 to deal with this. However it did not apply to per-recipient defers
219 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
221 This change adds a proper retry rule check during routing so that the
222 final cutoff time is checked against the message's age. We only do
223 this check if there is an address retry record and there is not a
224 domain retry record; this implies that previous attempts to handle
225 the address had the retry_use_local_parts option turned on. We use
226 this as an approximation for the destination being like a local
227 delivery, as in LMTP.
229 I suspect this new check makes the old local delivery cutoff check
230 redundant, but I have not verified this so I left the code in place.
232 TF/02 Correct gecos expansion when From: is a prefix of the username.
234 Test 0254 submits a message to Exim with the header
238 When I ran the test suite under the user fanf2, Exim expanded
239 the header to contain my full name, whereas it should have added
240 a Resent-Sender: header. It erroneously treats any prefix of the
241 username as equal to the username.
243 This change corrects that bug.
245 GF/01 DCC debug and logging tidyup
246 Error conditions log to paniclog rather than rejectlog.
247 Debug lines prefixed by "DCC: " to remove any ambiguity.
249 TF/03 Avoid unnecessary rebuilds of lookup-related code.
251 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
252 Bug spotted by Jeremy Harris; was flawed since initial commit.
253 Would have resulted in OCSP responses post-SNI triggering an Exim
254 NULL dereference and crash.
256 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
258 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
259 Bug detection, analysis and fix by Samuel Thibault.
260 Bugzilla 1331, Debian bug #698092.
262 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
264 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
265 Server implementation by Todd Lyons, client by JH.
266 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
267 config variable "prdr_enable" controls whether the server
268 advertises the facility. If the client requests PRDR a new
269 acl_data_smtp_prdr ACL is called once for each recipient, after
270 the body content is received and before the acl_smtp_data ACL.
271 The client is controlled by bolth of: a hosts_try_prdr option
272 on the smtp transport, and the server advertisement.
273 Default client logging of deliveries and rejections involving
274 PRDR are flagged with the string "PRDR".
276 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
277 fclose(). Diagnosis by Todd Lyons.
279 PP/17 Update configure.default to handle IPv6 localhost better.
280 Patch by Alain Williams (plus minor tweaks).
283 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
284 This is now consistent with GnuTLS, and is now documented: the
285 previous undocumented portable approach to treating the option as
286 unset was to force an expansion failure. That still works, and
287 an empty string is now equivalent.
289 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
290 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
291 not performing validation itself.
293 PP/20 Added force_command boolean option to pipe transport.
294 Patch from Nick Koston, of cPanel Inc.
296 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
299 TF/04 Added udpsend ACL modifer and hexquote expansion operator
301 PP/21 Fix eximon continuous updating with timestamped log-files.
302 Broken in a format-string cleanup in 4.80, missed when I repaired the
303 other false fix of the same issue.
304 Report and fix from Heiko Schlichting.
307 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
308 Report from Prashanth Katuri.
310 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
311 It's SecureTransport, so affects any MacOS clients which use the
312 system-integrated TLS libraries, including email clients.
314 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
315 using a MIME ACL for non-SMTP local injection.
316 Report and assistance in diagnosis by Warren Baker.
318 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
320 JH/16 Fix comparisons for 64b. Bugzilla 1385.
322 TL/09 Add expansion variable $authenticated_fail_id to keep track of
323 last id that failed so it may be referenced in subsequent ACL's.
325 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
328 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
329 ldap library initialization, allowing self-signed CA's to be
330 used. Also properly sets require_cert option later in code by
331 using NULL (global ldap config) instead of ldap handle (per
332 session). Bug diagnosis and testing by alxgomz.
334 TL/12 Enhanced documentation in the ratelimit.pl script provided in
335 the src/util/ subdirectory.
337 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
338 renamed to Transport Post Delivery Action by Jeremy Harris, as
341 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
342 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
343 redis_servers = needs to be configured which will be used by the redis
344 lookup. Patch from Warren Baker, of The Packet Hub.
346 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
348 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
349 hostname or reverse DNS when processing a host list. Used suggestions
350 from multiple comments on this bug.
352 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
354 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
355 Missed a few lines, added it to make the runtest require no keyboard
358 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
359 contains upper case chars. Make router use caseful_local_part.
361 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
362 support when GnuTLS has been built with p11-kit.
368 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
370 This, or similar/improved, will also be change PP/11 of 4.82.
376 PP/01 Handle short writes when writing local log-files.
377 In practice, only affects FreeBSD (8 onwards).
378 Bugzilla 1053, with thanks to Dmitry Isaikin.
380 NM/01 Bugzilla 949 - Documentation tweak
382 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
385 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
387 PP/02 Implemented gsasl authenticator.
389 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
391 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
392 `pkg-config foo` for cflags/libs.
394 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
395 with rest of GSASL and with heimdal_gssapi.
397 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
398 `pkg-config foo` for cflags/libs for the TLS implementation.
400 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
401 properties get this fed in as external SSF. A number of robustness
402 and debugging improvements to the cyrus_sasl authenticator.
404 PP/08 cyrus_sasl server now expands the server_realm option.
406 PP/09 Bugzilla 1214 - Log authentication information in reject log.
407 Patch by Jeremy Harris.
409 PP/10 Added dbmjz lookup type.
411 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
413 PP/12 MAIL args handles TAB as well as SP, for better interop with
414 non-compliant senders.
415 Analysis and variant patch by Todd Lyons.
417 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
418 Bug report from Lars Müller <lars@samba.org> (via SUSE),
419 Patch from Dirk Mueller <dmueller@suse.com>
421 PP/13 tls_peerdn now print-escaped for spool files.
422 Observed some $tls_peerdn in wild which contained \n, which resulted
423 in spool file corruption.
425 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
426 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
427 or write after TLS renegotiation, which otherwise led to messages
430 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
431 as a tracking header (ie: a signed header comes before the signature).
432 Patch from Wolfgang Breyha.
434 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
435 comma-sep list; embedded commas doubled.
437 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
439 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
441 Report and patch from Dmitry Banschikov.
443 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
444 Removed SSL_clear() after SSL_new() which led to protocol negotiation
445 failures. We appear to now support TLS1.1+ with Exim.
447 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
448 lets Exim select keys and certificates based upon TLS SNI from client.
449 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
450 before an outbound SMTP session. New log_selector, +tls_sni.
452 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
453 NULL dereference. Report and patch from Alun Jones.
455 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
456 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
457 Not seeing resolver debug output on NetBSD, but suspect this is a
458 resolver implementation change.
460 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
461 Left warnings. Added "eximon gdb" invocation mode.
463 PP/21 Defaulting "accept_8bitmime" to true, not false.
465 PP/22 Added -bw for inetd wait mode support.
467 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
468 locate the relevant includes and libraries. Made this the default.
470 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
471 Bugzilla 1246, report and most of solution from Tomasz Kusy.
473 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
474 This may cause build issues on older platforms.
476 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
477 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
478 gnutls_require_mac & gnutls_require_protocols (no longer supported).
479 Added SNI support via GnuTLS too.
480 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
482 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
484 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
485 Applied second patch from Janne, implementing suggestion to default
486 multiple-strings-in-record handling to match SPF spec.
488 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
490 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
491 read-only, out of scope).
492 Patch from Wolfgang Breyha, report from Stuart Northfield.
494 PP/29 Fix three issues highlighted by clang analyser static analysis.
495 Only crash-plausible issue would require the Cambridge-specific
496 iplookup router and a misconfiguration.
497 Report from Marcin Mirosław.
499 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
501 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
502 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
503 As part of this, removing so much warning spew let me fix some minor
504 real issues in debug logging.
506 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
507 assignment on my part. Fixed.
509 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
510 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
511 Janne Snabb (who went above and beyond: thank you).
513 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
514 string otherwise requires a connection and a bunch more work and it's
515 relatively easy to get wrong. Should also expose TLS library linkage
518 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
519 64-bit ${eval} (JH/03).
521 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
522 GNU libc to support some of the 64-bit stuff, should not lead to
523 conflicts. Defined before os.h is pulled in, so if a given platform
524 needs to override this, it can.
526 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
527 protection layer was required, which is not implemented.
528 Bugzilla 1254, patch from Wolfgang Breyha.
530 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
531 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
532 tls_dhparam take prime identifiers. Also unbreak combination of
533 OpenSSL+DH_params+TLSSNI.
535 PP/39 Disable SSLv2 by default in OpenSSL support.
541 PP/01 Solaris build fix for Oracle's LDAP libraries.
542 Bugzilla 1109, patch from Stephen Usher.
544 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
546 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
549 TF/02 Fix a couple more cases where we did not log the error message
550 when unlink() failed. See also change 4.74-TF/03.
552 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
553 lock up or crash if it happened to be inside a call to libc when it
554 got a SIGUSR1 from exiwhat.
556 The SIGUSR1 handler appends the current process status to the process
557 log which is later printed by exiwhat. It used to use the general
558 purpose logging code to do this, but several functions it calls are
559 not safe for signals.
561 The new output code in the SIGUSR1 handler is specific to the process
562 log, and simple enough that it's easy to inspect for signal safety.
563 Removing some special cases also simplifies the general logging code.
564 Removing the spurious timestamps from the process log simplifies
567 TF/04 Improved ratelimit ACL condition.
569 The /noupdate option has been deprecated in favour of /readonly which
570 has clearer semantics. The /leaky, /strict, and /readonly update modes
571 are mutually exclusive. The update mode is no longer included in the
572 database key; it just determines when the database is updated. (This
573 means that when you upgrde Exim will forget old rate measurements.)
575 Exim now checks that the per_* options are used with an update mode that
576 makes sense for the current ACL. For example, when Exim is processing a
577 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
578 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
579 must specify per_mail/readonly. If you omit the update mode it defaults to
580 /leaky where that makes sense (as before) or /readonly where required.
582 The /noupdate option is now undocumented but still supported for
583 backwards compatibility. It is equivalent to /readonly except that in
584 ACLs where /readonly is required you may specify /leaky/noupdate or
585 /strict/noupdate which are treated the same as /readonly.
587 A useful new feature is the /count= option. This is a generalization
588 of the per_byte option, so that you can measure the throughput of other
589 aggregate values. For example, the per_byte option is now equivalent
590 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
592 The per_rcpt option has been generalized using the /count= mechanism
593 (though it's more complicated than the per_byte equivalence). When it is
594 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
595 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
596 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
597 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
598 Note that using per_rcpt with a non-readonly update mode in more than
599 one ACL will cause the recipients to be double-counted. (The per_mail
600 and per_byte options don't have this problem.)
602 The handling of very low rates has changed slightly. If the computed rate
603 is less than the event's count (usually one) then this event is the first
604 after a long gap. In this case the rate is set to the same as this event's
605 count, so that the first message of a spam run is counted properly.
607 The major new feature is a mechanism for counting the rate of unique
608 events. The new per_addr option counts the number of different
609 recipients that someone has sent messages to in the last time period. It
610 behaves like per_rcpt if all the recipient addresses are different, but
611 duplicate recipient addresses do not increase the measured rate. Like
612 the /count= option this is a general mechanism, so the per_addr option
613 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
614 example, measure the rate that a client uses different sender addresses
615 with the options per_mail/unique=$sender_address. There are further
616 details in the main documentation.
618 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
620 TF/06 Removed a few PCRE remnants.
622 TF/07 Automatically extract Exim's version number from tags in the git
623 repository when doing development or release builds.
625 PP/02 Raise smtp_cmd_buffer_size to 16kB.
626 Bugzilla 879. Patch from Paul Fisher.
628 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
629 Heavily based on revision 40f9a89a from Simon Arlott's tree.
632 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
634 PP/05 Variable $av_failed, true if the AV scanner deferred.
635 Bugzilla 1078. Patch from John Horne.
637 PP/06 Stop make process more reliably on build failure.
638 Bugzilla 1087. Patch from Heiko Schlittermann.
640 PP/07 Make maildir_use_size_file an _expandable_ boolean.
641 Bugzilla 1089. Patch from Heiko Schlittermann.
643 PP/08 Handle ${run} returning more data than OS pipe buffer size.
644 Bugzilla 1131. Patch from Holger Weiß.
646 PP/09 Handle IPv6 addresses with SPF.
647 Bugzilla 860. Patch from Wolfgang Breyha.
649 PP/10 GnuTLS: support TLS 1.2 & 1.1.
651 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
654 PP/11 match_* no longer expand right-hand-side by default.
655 New compile-time build option, EXPAND_LISTMATCH_RHS.
656 New expansion conditions, "inlist", "inlisti".
658 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
660 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
662 PP/14 fix log_write() format string regression from TF/03.
663 Bugzilla 1152. Patch from Dmitry Isaikin.
669 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
671 PP/02 Harmonised TLS library version reporting; only show if debugging.
672 Layout now matches that introduced for other libraries in 4.74 PP/03.
674 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
676 PP/04 New "dns_use_edns0" global option.
678 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
681 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
682 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
684 TK/01 Updated PolarSSL code to 0.14.2.
685 Bugzilla 1097. Patch from Andreas Metzler.
687 PP/07 Catch divide-by-zero in ${eval:...}.
690 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
693 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
694 format-string attack -- SECURITY: remote arbitrary code execution.
696 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
697 time unintentionally subject to list matching rules, letting the header
698 cause arbitrary Exim lookups (of items which can occur in lists, *not*
699 arbitrary string expansion). This allowed for information disclosure.
701 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
702 INT_MIN/-1 -- value coerced to INT_MAX.
708 NM/01 Workround for PCRE version dependency in version reporting
711 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
712 This fixes portability to compilers other than gcc, notably
713 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
715 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
716 makefiles for portability to HP-UX and POSIX correctness.
718 PP/01 Permit LOOKUP_foo enabling on the make command-line.
719 Also via indented variable definition in the Makefile.
720 (Debugging by Oliver Heesakkers).
722 PP/02 Restore caching of spamd results with expanded spamd_address.
723 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
725 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
726 Improves build reliability. Fix from: Frank Elsner
728 NM/02 Fix wide character breakage in the rfc2047 coding
729 Fixes bug 1064. Patch from Andrey N. Oktyabrski
731 NM/03 Allow underscore in dnslist lookups
732 Fixes bug 1026. Patch from Graeme Fowler
734 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
735 Code patches from Adam Ciarcinski of NetBSD.
737 NM/04 Fixed exiqgrep to cope with mailq missing size issue
740 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
741 is logged, to avoid truncation. Patch from John Horne.
743 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
744 Patch from Jakob Hirsch.
746 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
747 SQL string expansion failure details.
748 Patch from Andrey Oktyabrski.
750 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
751 Patch from Simon Arlott.
753 PP/09 New lookups functionality failed to compile on old gcc which rejects
754 extern declarations in function scope.
755 Patch from Oliver Fleischmann
757 PP/10 Use sig_atomic_t for flags set from signal handlers.
758 Check getgroups() return and improve debugging.
759 Fixed developed for diagnosis in bug 927 (which turned out to be
762 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
763 Patch from Mark Zealey.
765 PP/12 Bugzilla 1056: Improved spamd server selection.
766 Patch from Mark Zealey.
768 PP/13 Bugzilla 1086: Deal with maildir quota file races.
769 Based on patch from Heiko Schlittermann.
771 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
772 Patch from Uwe Doering, sign-off by Michael Haardt.
774 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
775 variable declaration deep within a block. Bug and patch from
778 PP/15 lookups-Makefile IRIX compatibilty coercion.
780 PP/16 Make DISABLE_DKIM build knob functional.
782 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
783 Patch by Simon Arlott
785 TF/03 Fix valgrind.h portability to C89 compilers that do not support
786 variable argument macros. Our copy now differs from upstream.
792 TF/01 Failure to get a lock on a hints database can have serious
793 consequences so log it to the panic log.
795 TF/02 Log LMTP confirmation messages in the same way as SMTP,
796 controlled using the smtp_confirmation log selector.
798 TF/03 Include the error message when we fail to unlink a spool file.
800 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
801 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
802 for maintaining out-of-tree patches for some time.
804 PP/01 Bugzilla 139: Documentation and portability issues.
805 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
806 Handle per-OS dynamic-module compilation flags.
808 PP/02 Let /dev/null have normal permissions.
809 The 4.73 fixes were a little too stringent and complained about the
810 permissions on /dev/null. Exempt it from some checks.
811 Reported by Andreas M. Kirchwitz.
813 PP/03 Report version information for many libraries, including
814 Exim version information for dynamically loaded libraries. Created
815 version.h, now support a version extension string for distributors
816 who patch heavily. Dynamic module ABI change.
818 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
819 privilege escalation vulnerability whereby the Exim run-time user
820 can cause root to append content of the attacker's choosing to
823 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
826 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
827 If dropping privileges for untrusted macros, we disabled normal logging
828 on the basis that it would fail; for the Exim run-time user, this is not
829 the case, and it resulted in successful deliveries going unlogged.
830 Fixed. Reported by Andreas Metzler.
836 PP/01 Date: & Message-Id: revert to normally being appended to a message,
837 only prepend for the Resent-* case. Fixes regression introduced in
838 Exim 4.70 by NM/22 for Bugzilla 607.
840 PP/02 Include check_rfc2047_length in configure.default because we're seeing
841 increasing numbers of administrators be bitten by this.
843 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
845 PP/03 Bugzilla 994: added openssl_options main configuration option.
847 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
849 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
851 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
853 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
854 without a peer certificate, leading to a segfault because of an
855 assumption that peers always have certificates. Be a little more
856 paranoid. Problem reported by Martin Tscholak.
858 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
859 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
860 NB: ClamAV planning to remove STREAM in "middle of 2010".
861 CL also introduces -bmalware, various -d+acl logging additions and
862 more caution in buffer sizes.
864 PP/09 Implemented reverse_ip expansion operator.
866 PP/10 Bugzilla 937: provide a "debug" ACL control.
868 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
870 PP/12 Bugzilla 973: Implement --version.
872 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
874 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
876 PP/15 Bugzilla 816: support multiple condition rules on Routers.
878 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
879 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
880 ignore trailing whitespace.
882 JJ/02 prevent non-panic DKIM error from being sent to paniclog
884 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
887 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
888 Notification from Dr Andrew Aitchison.
890 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
891 ExtendedDetectionInfo response format.
892 Notification from John Horne.
894 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
897 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
898 XSL and documented dependency on system catalogs, with examples of how
901 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
904 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
905 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
906 configuration file which is writeable by the Exim user or group.
908 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
909 of configuration files to cover files specified with the -C option if
910 they are going to be used with root privileges, not just the default
913 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
914 option (effectively making it always true).
916 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
917 files to be used while preserving root privileges.
919 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
920 that rogue child processes cannot use them.
922 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
923 run-time user, instead of root.
925 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
926 Exim run-time user without dropping privileges.
928 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
929 result string, instead of calling string_vformat() twice with the same
932 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
933 for other users. Others should always drop root privileges if they use
934 -C on the command line, even for a whitelisted configure file.
936 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
938 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
944 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
945 $data_path, and $header_path variables; fixed documentation bugs and
948 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
949 exipick to access non-standard spools, including the "frozen" queue
952 NM/01 Bugzilla 965: Support mysql stored procedures.
953 Patch from Alain Williams
955 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
957 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
958 Patch from Andreas Metzler
960 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
961 Patch from Kirill Miazine
963 NM/05 Bugzilla 671: Added umask to procmail example.
965 JJ/03 installed exipick 20100323.0, fixing doc bug
967 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
968 directory. Notification and patch from Dan Rosenberg.
970 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
972 TK/02 Improve log output when DKIM signing operation fails.
974 MH/01 Treat the transport option dkim_domain as a colon separated
975 list, not as a single string, and sign the message with each element,
976 omitting multiple occurences of the same signer.
978 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
979 Bugzilla 985, 986. Patch by Simon Arlott
981 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
982 Patch by Simon Arlott
984 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
985 MBX locking. Notification from Dan Rosenberg.
991 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
993 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
995 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
997 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
999 NM/04 Bugzilla 915: Flag broken perl installation during build.
1005 TK/01 Added patch by Johannes Berg that expands the main option
1006 "spamd_address" if it starts with a dollar sign.
1008 TK/02 Write list of recipients to X-Envelope-Sender header when building
1009 the mbox-format spool file for content scanning (suggested by Jakob
1012 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1013 (http://www.dcc-servers.net/) support via dccifd. Activated by
1014 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1016 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1017 by Mark Daniel Reidel <mr@df.eu>.
1019 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1020 When building exim an external PCRE library is now needed -
1021 PCRE is a system library on the majority of modern systems.
1022 See entry on PCRE_LIBS in EDITME file.
1024 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1025 conversation. Added nologin parameter to request.
1026 Patch contributed by Kirill Miazine.
1028 TF/01 Do not log submission mode rewrites if they do not change the address.
1030 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1032 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1033 log files in place. Contributed by Roberto Lima.
1035 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1037 TF/03 Bugzilla 615: When checking the local_parts router precondition
1038 after a local_part_suffix or local_part_prefix option, Exim now
1039 does not use the address's named list lookup cache, since this
1040 contains cached lookups for the whole local part.
1042 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1043 Robert Millan. Documentation is in experimental-spec.txt.
1045 TF/04 Bugzilla 668: Fix parallel build (make -j).
1047 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1049 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1050 Patch provided by Jan Srzednicki.
1052 TF/05 Leading white space used to be stripped from $spam_report which
1053 wrecked the formatting. Now it is preserved.
1055 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1056 that they are available at delivery time.
1058 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1060 TF/08 TLS error reporting now respects the incoming_interface and
1061 incoming_port log selectors.
1063 TF/09 Produce a more useful error message if an SMTP transport's hosts
1064 setting expands to an empty string.
1066 NM/06 Bugzilla 744: EXPN did not work under TLS.
1067 Patch provided by Phil Pennock.
1069 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1070 Patch provided by Richard Godbee.
1072 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1073 acl_smtp_notquit, added index entry.
1075 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1076 Patch provided by Eugene Bujak.
1078 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1079 accept(). Patch provided by Maxim Dounin.
1081 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1082 Patch provided by Phil Pennock.
1084 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1086 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1087 Patch provided by Brad "anomie" Jorsch.
1089 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1090 Patch provided by Dean Brooks.
1092 TK/05 Add native DKIM support (does not depend on external libraries).
1094 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1095 Patch provided by Graeme Fowler.
1097 NM/16 Bugzilla 851: Documentation example syntax fix.
1099 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1101 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1104 NM/19 Bugzilla 745: TLS version reporting.
1105 Patch provided by Phil Pennock.
1107 NM/20 Bugzilla 167: bool: condition support.
1108 Patch provided by Phil Pennock.
1110 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1111 clients. Patch provided by Phil Pennock.
1113 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1114 Patch provided by Brad "anomie" Jorsch.
1116 NM/23 Bugzilla 687: Fix misparses in eximstats.
1117 Patch provided by Heiko Schlittermann.
1119 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1120 Patch provided by Heiko Schlittermann.
1122 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1123 plus update to original patch.
1125 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1127 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1128 Patch provided by David Brownlee.
1130 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1132 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1134 NM/30 Bugzilla 888: TLS documentation bugfixes.
1136 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1138 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1139 Unlike the original bugzilla I have changed all shell scripts in src tree.
1141 NM/33 Bugzilla 898: Transport filter timeout fix.
1142 Patch by Todd Rinaldo.
1144 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1145 Patch by Serge Demonchaux.
1147 NM/35 Bugzilla 39: Base64 decode bug fixes.
1148 Patch by Jakob Hirsch.
1150 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1152 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1154 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1156 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1162 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1163 ALT-N's libdkim that I have put here:
1164 http://duncanthrax.net/exim-experimental/
1166 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1167 were called 'true' and it seems that C99 defines that as a reserved
1168 keyword to be used with 'bool' variable types. That means you could
1169 not include C99-style headers which use bools without triggering
1170 build errors in sieve.c.
1172 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1173 as mailq or other aliases. Changed the --help handling significantly
1174 to do whats expected. exim_usage() emits usage/help information.
1176 SC/01 Added the -bylocaldomain option to eximstats.
1178 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1180 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1182 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1188 PH/01 Another patch from the Sieve maintainer.
1190 PH/02 When an IPv6 address is converted to a string for single-key lookup
1191 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1192 dots are used instead of colons so that keys in lsearch files need not
1193 contain colons. This was done some time before quoting was made available
1194 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1195 (notated using the quote facility) so as to distinguish them from IPv4
1196 keys. This meant that lookups for IP addresses in host lists did not work
1197 for iplsearch lookups.
1199 This has been fixed by arranging for IPv6 addresses to be expressed with
1200 colons if the lookup type is iplsearch. This is not incompatible, because
1201 previously such lookups could never work.
1203 The situation is now rather anomolous, since one *can* have colons in
1204 ordinary lsearch keys. However, making the change in all cases is
1205 incompatible and would probably break a number of configurations.
1207 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1210 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1211 conversion specification without a maximum field width, thereby enabling
1212 a rogue spamd server to cause a buffer overflow. While nobody in their
1213 right mind would setup Exim to query an untrusted spamd server, an
1214 attacker that gains access to a server running spamd could potentially
1215 exploit this vulnerability to run arbitrary code as the Exim user.
1217 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1218 $primary_hostname instead of what libspf2 thinks the hosts name is.
1220 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1221 a directory entry by the name of the lookup key. Previously, if a
1222 symlink pointed to a non-existing file or a file in a directory that
1223 Exim lacked permissions to read, a lookup for a key matching that
1224 symlink would fail. Now it is enough that a matching directory entry
1225 exists, symlink or not. (Bugzilla 503.)
1227 PH/03 The body_linecount and body_zerocount variables are now exported in the
1230 PH/04 Added the $dnslist_matched variable.
1232 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1233 This means they are set thereafter only if the connection becomes
1236 PH/06 Added the client_condition to authenticators so that some can be skipped
1237 by clients under certain conditions.
1239 PH/07 The error message for a badly-placed control=no_multiline_responses left
1240 "_responses" off the end of the name.
1242 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1244 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1245 (without spaces) instead of just copying the configuration text.
1247 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1249 PH/11 Added $max_received_linelength.
1251 PH/12 Added +ignore_defer and +include_defer to host lists.
1253 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1254 way in which PCRE > 7.0 is built.
1256 PH/14 Implemented queue_only_load_latch.
1258 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1259 MAIL command. The effect was to mangle the value on 64-bit systems.
1261 PH/16 Another patch from the Sieve maintainer.
1263 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1265 PH/18 If a system quota error occurred while trying to create the file for
1266 a maildir delivery, the message "Mailbox is full" was not appended to the
1267 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1268 only to a quota excession during the actual writing of the file.
1270 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1271 characters?) which causes problems in log lines. The DN values are now
1272 passed through string_printing() before being added to log lines.
1274 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1275 and InterBase are left for another time.)
1277 PH/21 Added message_body_newlines option.
1279 PH/22 Guard against possible overflow in moan_check_errorcopy().
1281 PH/23 POSIX allows open() to be a macro; guard against that.
1283 PH/24 If the recipient of an error message contained an @ in the local part
1284 (suitably quoted, of course), incorrect values were put in $domain and
1285 $local_part during the evaluation of errors_copy.
1291 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1292 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1295 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1296 issue a MAIL command.
1298 PH/02 In an ACL statement such as
1300 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1302 if a client was not listed at all, or was listed with a value other than
1303 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1304 the condition was not true (as it should be), so access was not denied.
1305 The bug was that the ! inversion was incorrectly passed on to the second
1306 item. This has been fixed.
1308 PH/03 Added additional dnslists conditions == and =& which are different from
1309 = and & when the dns lookup returns more than one IP address.
1311 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1312 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1314 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1315 FSYNC, which compiles an option called disable_fsync that allows for
1316 bypassing fsync(). The documentation is heavily laced with warnings.
1318 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1320 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1321 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1322 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1323 including adding "make clean"; (3) Added -fPIC when compiling the test
1324 dynamically loaded module, to get rid of a warning.
1326 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1327 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1328 = 0s. The bug is otherwise harmless.
1330 PH/07 There was a bug in the dovecot authenticator such that the value of
1331 $auth1 could be overwritten, and so not correctly preserved, after a
1332 successful authentication. This usually meant that the value preserved by
1333 the server_setid option was incorrect.
1335 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1337 PH/09 Installed PCRE release 7.0.
1339 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1340 run for batched SMTP input. It is now run at the start of every message
1341 in the batch. While fixing this I discovered that the process information
1342 (output by running exiwhat) was not always getting set for -bs and -bS
1343 input. This is fixed, and it now also says "batched" for BSMTP.
1345 PH/11 Added control=no_pipelining.
1347 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1348 patch, slightly modified), and move the expansion of helo_data till after
1349 the connection is made in the smtp transport (so it can use these
1352 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1354 PH/14 Added log_selector = +pid.
1356 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1358 PH/16 Add ${if forany and ${if forall.
1360 PH/17 Added dsn_from option to vary the From: line in DSNs.
1362 PH/18 Flush SMTP output before performing a callout, unless control =
1363 no_callout_flush is set.
1365 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1366 was true (the default) a successful delivery failed to delete the retry
1367 item, thus causing premature timeout of the address. The bug is now
1370 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1372 PH/21 Long custom messages for fakedefer and fakereject are now split up
1373 into multiline reponses in the same way that messages for "deny" and
1374 other ACL rejections are.
1376 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1377 with slight modification.
1379 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1380 draft, changing the syntax and factoring some duplicate code".
1382 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1383 for deliveries of the second and subsequent messages over the same SMTP
1386 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1387 ${reduce, with only minor "tidies".
1389 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1391 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1392 expansion side effects.
1394 PH/27 When a message times out after an over-quota error from an Exim-imposed
1395 quota, the bounce message says "mailbox is full". This message was not
1396 being given when it was a system quota that was exceeded. It now should
1399 MH/03 Made $recipients available in local_scan(). local_scan() already has
1400 better access to the recipient list through recipients_list[], but
1401 $recipients can be useful in postmaster-provided expansion strings.
1403 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1404 in the case of a MAIL command with additional options following the
1405 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1406 were accidentally chopped off.
1408 PH/29 SMTP synchronization checks are implemented when a command is read -
1409 there is a check that no more input is waiting when there shouldn't be
1410 any. However, for some commands, a delay in an ACL can mean that it is
1411 some time before the response is written. In this time, more input might
1412 arrive, invalidly. So now there are extra checks after an ACL has run for
1413 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1414 pipelining has not been advertised.
1416 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1418 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1419 returned in the SMTP response when smtp_return_error_details was false.
1420 This has been fixed.
1422 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1423 instead of the C I/O that was originally supplied, because problems were
1424 reported on Solaris.
1426 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1427 Exim which did not show up earlier: it was assuming that a call to
1428 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1429 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1430 was a macro that became an assignment, so it seemed to work. This has
1431 changed to a proper function call with a void return, hence the compile
1432 error. Exim's code has been fixed.
1434 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1437 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1439 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1441 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1442 because it turns out to be too big or there is a timeout) there is no
1443 "Completed" line in the log. When some messages of this type were
1444 selected by exigrep, they were listed as "not completed". Others were
1445 picked up by some special patterns. I have improved the selection
1446 criteria to be more general.
1448 PH/38 The host_find_failed option in the manualroute router can now be set
1449 to "ignore", to completely ignore a host whose IP address cannot be
1450 found. If all hosts are ignored, the behaviour is controlled by the new
1451 host_all_ignored option.
1453 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1454 homing or because of multiple MX records with /mx) generated more than
1455 one IP address, and the following item turned out to be the local host,
1456 all the secondary addresses of the first item were incorrectly removed
1457 from the list, along with the local host and any following hosts (which
1458 is what is supposed to happen).
1460 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1461 whoever called Exim into the -H file. In the case of the daemon it was
1462 behaving confusingly. When first started, it used values for whoever
1463 started the daemon, but after a SIGHUP it used the Exim user (because it
1464 calls itself on a restart). I have changed the code so that it now always
1467 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1468 message are rejected with the same error (e.g. no authentication or bad
1469 sender address), and a DATA command is nevertheless sent (as can happen
1470 with PIPELINING or a stupid MUA), the error message that was given to the
1471 RCPT commands is included in the rejection of the DATA command. This is
1472 intended to be helpful for MUAs that show only the final error to their
1475 PH/42 Another patch from the Sieve maintainer.
1477 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1478 Eximstats - Fixed some broken HTML links and added missing column headers
1480 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1481 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1483 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1485 PH/43 Yet another patch from the Sieve maintainer.
1487 PH/44 I found a way to check for a TCP/IP connection going away before sending
1488 the response to the final '.' that terminates a message, but only in the
1489 case where the client has not sent further data following the '.'
1490 (unfortunately, this is allowed). However, in many cases there won't be
1491 any further data because there won't be any more messages to send. A call
1492 to select() can be used: if it shows that the input is "ready", there is
1493 either input waiting, or the socket has been closed. An attempt to read
1494 the next input character can distinguish the two cases. Previously, Exim
1495 would have sent an OK response which the client would never have see.
1496 This could lead to message repetition. This fix should cure that, at
1497 least in a lot of common cases.
1499 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1500 advertised in response to EHLO.
1506 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1507 fixed by 4.65/MH/01 (is this a record?) are fixed:
1509 (i) An empty string was always treated as zero by the numeric comparison
1510 operators. This behaviour has been restored.
1512 (ii) It is documented that the numeric comparison operators always treat
1513 their arguments as decimal numbers. This was broken in that numbers
1514 starting with 0 were being interpreted as octal.
1516 While fixing these problems I realized that there was another issue that
1517 hadn't been noticed. Values of message_size_limit (both the global option
1518 and the transport option) were treated as octal if they started with 0.
1519 The documentation was vague. These values are now always treated as
1520 decimal, and I will make that clear in the documentation.
1526 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1527 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1530 MH/01 Don't check that the operands of numeric comparison operators are
1531 integers when their expansion is in "skipping" mode (fixes bug
1532 introduced by 4.64-PH/07).
1534 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1535 child addresses, Exim now panics and dies. Previously, because the count
1536 is held in a short int, deliveries were likely to be lost. As such a
1537 large number of recipients for a single message is ridiculous
1538 (performance will be very, very poor), I have chosen to impose a limit
1539 rather than extend the field.
1545 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1546 leftover -K file (the existence of which was triggered by #402).
1547 While we were at it, introduced process PID as part of the -K
1548 filename. This should rule out race conditions when creating
1551 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1552 processing considerably. Previous code took too long for large mails,
1553 triggering a timeout which in turn triggers #401.
1555 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1556 in the DK code in transports.c. sendfile() is not really portable,
1557 hence the _LINUX specificness.
1559 TF/01 In the add_headers option to the mail command in an Exim filter,
1560 there was a bug that Exim would claim a syntax error in any
1561 header after the first one which had an odd number of characters
1564 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1565 callout verification, Exim cached a "reject" for the entire domain. This
1566 is correct for most verifications, but it is not correct for a recipient
1567 verification with use_sender or use_postmaster set, because in that case
1568 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1569 case of MAIL FROM:<> rejection from other early rejections (e.g.
1570 rejection of HELO). When verifying a recipient using a non-null MAIL
1571 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1572 Whatever the result of the callout, the value of the domain cache is
1573 left unchanged (for any other kind of callout, getting as far as trying
1574 RCPT means that the domain itself is ok).
1576 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1579 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1580 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1581 session, the daemon ignores ECONNECT errors and logs others; it now
1582 ignores EPIPE as well.
1584 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1585 (quoted-printable decoding).
1587 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1588 later the small subsequent patch to fix an introduced bug.
1590 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1592 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1594 PH/08 An error is now given if message_size_limit is specified negative.
1596 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1597 to be given (somewhat) arbitrary names.
1599 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1602 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1603 miscellaneous code fixes
1605 PH/10 Added the log_reject_target ACL modifier to specify where to log
1608 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1609 hostname. This is wrong, because it relates to the incoming message (and
1610 probably the interface on which it is arriving) and not to the outgoing
1611 callout (which could be using a different interface). This has been
1612 changed to use the value of the helo_data option from the smtp transport
1613 instead - this is what is used when a message is actually being sent. If
1614 there is no remote transport (possible with a router that sets up host
1615 addresses), $smtp_active_hostname is used.
1617 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1618 tweaks were necessary in order to get it to work (see also 21 below):
1619 (a) The code assumed that strncpy() returns a negative number on buffer
1620 overflow, which isn't the case. Replaced with Exim's string_format()
1622 (b) There were several signed/unsigned issues. I just did the minimum
1623 hacking in of casts. There is scope for a larger refactoring.
1624 (c) The code used strcasecmp() which is not a standard C function.
1625 Replaced with Exim's strcmpic() function.
1626 (d) The code set only $1; it now sets $auth1 as well.
1627 (e) A simple test gave the error "authentication client didn't specify
1628 service in request". It would seem that Dovecot has changed its
1629 interface. Fortunately there's a specification; I followed it and
1630 changed what the client sends and it appears to be working now.
1632 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1635 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1636 address A is aliased to B and C, where B exists and C does not. Without
1637 -v the output is "A verified" because verification stops after a
1638 successful redirection if more than one address is generated. However,
1639 with -v the child addresses are also verified. Exim was outputting "A
1640 failed to verify" and then showing the successful verification for C,
1641 with its parentage. It now outputs "B failed to verify", showing B's
1642 parentage before showing the successful verification of C.
1644 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1645 look up a TXT record in a specific list after matching in a combined
1648 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1649 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1650 they consult the DNS. I had assumed they would set it the way they
1651 wanted; and indeed my experiments on Linux seem to show that in some
1652 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1653 To be on the safe side, however, I have now made the interface to
1654 host_find_byname() similar to host_find_bydns(), with an argument
1655 containing the DNS resolver options. The host_find_byname() function now
1656 sets these options at its start, just as host_find_bydns() does. The smtp
1657 transport options dns_qualify_single and dns_search_parents are passed to
1658 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1659 of host_find_byname() use the default settings of RES_DEFNAMES
1660 (qualify_single) but not RES_DNSRCH (search_parents).
1662 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1663 spool_read_header() do less string testing, by means of a preliminary
1664 switch on the second character of optional "-foo" lines. (This is
1665 overdue, caused by the large number of possibilities that now exist.
1666 Originally there were few.) While I was there, I also converted the
1667 str(n)cmp tests so they don't re-test the leading "-" and the first
1668 character, in the hope this might squeeze out yet more improvement.
1670 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1671 flag allowing group syntax was set by the header_syntax check but not
1672 turned off, possible causing trouble later; (2) The flag was not being
1673 set at all for the header_verify test, causing "group"-style headers to
1674 be rejected. I have now set it in this case, and also caused header_
1675 verify to ignore an empty address taken from a group. While doing this, I
1676 came across some other cases where the code for allowing group syntax
1677 while scanning a header line wasn't quite right (mostly, not resetting
1678 the flag correctly in the right place). These bugs could have caused
1679 trouble for malformed header lines. I hope it is now all correct.
1681 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1682 with the "reply" argument non-NULL. The code, however (which originally
1683 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1684 but it didn't always do it. This confused somebody who was copying the
1685 code for some other use. I have removed all the tests.
1687 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1688 feature that was used to support insecure browsers during the U.S. crypto
1689 embargo. It requires special client support, and Exim is probably the
1690 only MTA that supported it -- and would never use it because real RSA is
1691 always available. This code has been removed, because it had the bad
1692 effect of slowing Exim down by computing (never used) parameters for the
1693 RSA_EXPORT functionality.
1695 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1696 authenticator to fail if there's a tab character in the incoming data
1697 (there should never be unless someone is messing about, as it's supposed
1698 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1699 if the connection is using TLS or if the remote IP is the same as the
1700 local IP, and the "valid-client-cert option" if a client certificate has
1703 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1704 authenticators. This can be used for authorization after authentication
1705 succeeds. (In the case of plaintext, it servers for both authentication
1708 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1709 if any retry times were supplied.
1711 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1712 connection, where there is no client IP address to check. In this
1713 situation, the verify now always succeeds.
1715 PH/25 Applied John Jetmore's -Mset patch.
1717 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1719 PH/27 In a string expansion for a processed (not raw) header when multiple
1720 headers of the same name were present, leading whitespace was being
1721 removed from all of them, but trailing whitespace was being removed only
1722 from the last one. Now trailing whitespace is removed from each header
1723 before concatenation. Completely empty headers in a concatenation (as
1724 before) are ignored.
1726 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1727 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1729 PH/29 [Removed. This was a change that I later backed out, and forgot to
1730 correct the ChangeLog entry (that I had efficiently created) before
1731 committing the later change.]
1733 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1734 address errors (4xx response to RCPT) over the same connection as other
1735 messages routed to the same hosts. Such deliveries are always "forced",
1736 so retry times are not inspected. This resulted in far too many retries
1737 for the affected addresses. The effect occurred only when there were more
1738 hosts than the hosts_max_try setting in the smtp transport when it had
1739 the 4xx errors. Those hosts that it had tried were not added to the list
1740 of hosts for which the message was waiting, so if all were tried, there
1741 was no problem. Two fixes have been applied:
1743 (i) If there are any address or message errors in an SMTP delivery, none
1744 of the hosts (tried or untried) are now added to the list of hosts
1745 for which the message is waiting, so the message should not be a
1746 candidate for sending over the same connection that was used for a
1747 successful delivery of some other message. This seems entirely
1748 reasonable: after all the message is NOT "waiting for some host".
1749 This is so "obvious" that I'm not sure why it wasn't done
1750 previously. Hope I haven't missed anything, but it can't do any
1751 harm, as the worst effect is to miss an optimization.
1753 (ii) If, despite (i), such a delivery is accidentally attempted, the
1754 routing retry time is respected, so at least it doesn't keep
1755 hammering the server.
1757 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1758 in ${readsocket because some servers need this prod.
1760 PH/32 Added some extra debug output when updating a wait-xxx database.
1762 PH/33 The hint "could be header name not terminated by colon", which has been
1763 given for certain expansion errors for a long time, was not being given
1764 for the ${if def:h_colon_omitted{... case.
1766 PH/34 The spec says: "With one important exception, whenever a domain list is
1767 being scanned, $domain contains the subject domain." There was at least
1768 one case where this was not true.
1770 PH/35 The error "getsockname() failed: connection reset by peer" was being
1771 written to the panic log as well as the main log, but it isn't really
1772 panic-worthy as it just means the connection died rather early on. I have
1773 removed the panic log writing for the ECONNRESET error when getsockname()
1776 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1777 runs only) independently of the message's sender address. This meant
1778 that, if the 4xx error was in fact related to the sender, a different
1779 message to the same recipient with a different sender could confuse
1780 things. In particualar, this can happen when sending to a greylisting
1781 server, but other circumstances could also provoke similar problems.
1782 I have changed the default so that the retry time for these errors is now
1783 based a combination of the sender and recipient addresses. This change
1784 can be overridden by setting address_retry_include_sender=false in the
1787 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1788 remote server are returned as part of bounce messages. This was not
1789 happening for LMTP over a pipe (the lmtp transport), but now it is the
1790 same for both kinds of LMTP.
1792 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1793 in header lines that were in fact CNAMEs. This is no longer the case.
1795 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1796 and queue runs started by the daemon processed all messages. This has
1797 been fixed so that -R and -S can now usefully be given with -q<time>.
1799 PH/40 Import PCRE release 6.7 (fixes some bugs).
1801 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1803 PH/42 Give an error if -q is specified more than once.
1805 PH/43 Renamed the variables $interface_address and $interface_port as
1806 $received_ip_address and $received_port, to make it clear that these
1807 values apply to message reception, and not to the outgoing interface when
1808 a message is delivered. (The old names remain recognized, of course.)
1810 PH/44 There was no timeout on the connect() call when using a Unix domain
1811 socket in the ${readsocket expansion. There now is.
1813 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1814 be meaningful with "accept".
1816 SC/01 Eximstats V1.43
1817 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1819 SC/02 Eximstats V1.44
1820 Use a glob alias rather than an array ref in the generated
1821 parser. This improves both readability and performance.
1823 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1824 Collect SpamAssassin and rejection statistics.
1825 Don't display local sender or destination tables unless
1826 there is data to show.
1827 Added average volumes into the top table text output.
1829 SC/04 Eximstats V1.46
1830 Collect data on the number of addresses (recipients)
1831 as well as the number of messages.
1833 SC/05 Eximstats V1.47
1834 Added 'Message too big' to the list of mail rejection
1835 reasons (thanks to Marco Gaiarin).
1837 SC/06 Eximstats V1.48
1838 Mainlog lines which have GMT offsets and are too short to
1839 have a flag are now skipped.
1841 SC/07 Eximstats V1.49 (Alain Williams)
1842 Added the -emptyok flag.
1844 SC/08 Eximstats V1.50
1845 Fixes for obtaining the IP address from reject messages.
1847 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1848 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1849 whitesspace changes from 4.64-PH/27
1851 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1854 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1855 are found, allow negative numbers in numeric criteria)
1857 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1859 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1860 to match changes made in 4.64-PH/43
1862 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1864 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1865 "domain" is actually an IP address, and force a failure. This locks out
1866 those revolvers/nameservers that support "A-for-A" lookups, in
1867 contravention of the specifications.
1869 PH/48 When a host name was looked up from an IP address, and the subsequent
1870 forward lookup of the name timed out, the host name was left in
1871 $sender_host_name, contrary to the specification.
1873 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1874 restricted to single-key lookups, Exim was not diagnosing an error if
1875 * or *@ was used with a query-style lookup.
1877 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1879 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1880 long ago, but noone interested enough thought of it. Let's just say that
1881 the "1.1" means that there are some new functions that weren't there at
1882 some point in the past.
1884 PH/51 Error processing for expansion failure of helo_data from an smtp
1885 transport during callout processing was broken.
1887 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1888 tested/used via the -bh/-bhc/-bs options.
1890 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1891 bug, fixed in subsequent PCRE releases).
1893 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1894 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1896 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1902 SC/01 Use a glob alias rather than an array ref in eximstats generated
1903 parser. This improves both readability and performance.
1905 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1906 Don't display local sender or destination tables in eximstats unless
1907 there is data to show.
1908 Added average volumes into the eximstats top table text output.
1910 SC/03 Collect data on the number of addresses (recipients) as well
1911 as the number of messages in eximstats.
1913 TF/01 Correct an error in the documentation for the redirect router. Exim
1914 does (usually) call initgroups() when daemonizing.
1916 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1917 with consistent privilege compared to when running as a daemon.
1919 TF/03 Note in the spec that $authenticated_id is not set for local
1920 submissions from trusted users.
1922 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1923 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1925 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1926 by adding some example configuration directives to the default
1927 configuration file. A little bit of work is required to uncomment the
1928 directives and define how usernames and passwords are checked, but
1929 there is now a framework to start from.
1931 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1932 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1933 without this. I don't know how relevant this is to other LDAP libraries.
1935 PH/02 Add the verb name to the "unknown ACL verb" error.
1937 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1939 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1941 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1942 directory not expanded when it should be if an expanded home directory
1943 was set for the address (which is overridden by the transport).
1945 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1948 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1949 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1950 because it is too late at that time, and has no effect.
1952 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1953 security issue with \' (bugzilla #107). I could not use the
1954 PQescapeStringConn() function, because it needs a PGconn value as one of
1957 PH/08 When testing addresses using -bt, indicate those final addresses that
1958 are duplicates that would not cause an additional delivery. At least one
1959 person was confused, thinking that -bt output corresponded to deliveries.
1960 (Suppressing duplicates isn't a good idea as you lose the information
1961 about possibly different redirections that led to the duplicates.)
1963 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1964 systems where poll() doesn't work, in particular OS X.
1966 PH/10 Added more information to debugging output for retry time not reached.
1968 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1969 operations in malware.c.
1971 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1974 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1975 syslog_duplication set false, log lines that would normally be written
1976 both the the main log and to the reject log were not written to syslog at
1979 PH/14 In the default configuration, change the use of "message" in ACL warn
1980 statements to "add_header".
1982 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1983 not followed by a command (e.g. "seen endif").
1985 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1986 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1989 PH/17 Added extra conditions to the default value of delay_warning_condition
1993 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1994 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1995 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1998 The Auto-Submitted: and various List- headers are standardised, whereas I
1999 don't think Precedence: ever was.
2001 PH/18 Refactored debugging code in route_finduser() to show more information,
2002 in particular, the error code if getpwnam() issues one.
2004 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2005 This is apparently needed in addition to the PH/07 change above to avoid
2006 any possible encoding problems.
2008 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2009 but not after initializing Perl.
2011 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2012 output them only if debugging. By default they are written stderr,
2013 apparently, which is not desirable.
2015 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2018 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2021 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2023 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2024 authenticated or an ident call has been made. Suppress the default
2025 values for $authenticated_id and $authenticated_sender (but permit -oMai
2026 and -oMas) when testing with -bh.
2028 PH/24 Re-jigged the order of the tests in the default configuration so that the
2029 tests for valid domains and recipients precede the DNS black list and CSA
2030 tests, on the grounds that those ones are more expensive.
2032 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2033 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2034 This bug exists in every version of Exim that I still have, right back to
2037 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2038 However, an attempt to turn on case-sensitivity in a regex key by
2039 including (?-i) didn't work because the subject string was already
2040 lowercased, and the effects were non-intuitive. It turns out that a
2041 one-line patch can be used to allow (?-i) to work as expected.
2047 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2048 other effects) broke the use of negated acl sub-conditions.
2050 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2053 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2054 "Deny" causes Exim to reject the incoming connection with a 554 error.
2055 Unfortunately, if there is a major crisis, such as a disk failure,
2056 tcp-wrappers gives "deny", whereas what one would like would be some
2057 kind of temporary error. A kludge has been added to help with this.
2058 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2059 554 error is used if errno is still zero or contains ENOENT (which occurs
2060 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2063 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2065 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2066 errors. Otherwise a message that provokes a temporary error (when other
2067 messages do not) can cause a whole host to time out.
2069 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2070 addresses were routed directly to files or pipes from a redirect router.
2071 File deliveries just didn't batch; pipe deliveries might have suffered
2074 PH/06 A failure to get a lock for a hints database would erroneously always say
2075 "Failed to get write lock", even when it was really a read lock.
2077 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2078 of 0600. This has been changed to use the value of the lockfile_mode
2079 option (which defaults to 0600).
2081 PH/08 Applied small patch from the Sieve maintainer.
2083 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2084 folder from quota calculations, a direct delivery into this folder messed
2085 up the contents of the maildirsize file. This was because the regex was
2086 used only to exclude .Trash (or whatever) when the size of the mailbox
2087 was calculated. There was no check that a delivery was happening into an
2088 excluded directory. This bug has been fixed by ignoring all quota
2089 processing for deliveries into excluded directories.
2091 PH/10 Added the maildirfolder_create_regex option to appendfile.
2097 PH/01 The code for finding all the local interface addresses on a FreeBSD
2098 system running IPv6 was broken. This may well have applied to all BSD
2099 systems, as well as to others that have similar system calls. The broken
2100 code found IPv4 interfaces correctly, but gave incorrect values for the
2101 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2102 that it would not match correctly against @[] and not recognize the IPv6
2105 PH/02 The ipliteral router was not recognizing addresses of the form user@
2106 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2108 PH/03 Added disable_ipv6.
2110 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2111 decoded form, because it is most often used to construct To: headers
2112 lines in autoreplies, and the decoded form may well be syntactically
2113 invalid. However, $reply_address has leading white space removed, and all
2114 newlines turned into spaces so that the autoreply transport does not
2117 PH/05 If group was specified without a user on a router, and no group or user
2118 was specified on a transport, the group from the router was ignored.
2120 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2121 for visible compile-time settings that can be used to change these
2122 numbers, for those that want even more. Backwards compatibility with old
2123 spool files has been maintained. However, going back to a previous Exim
2124 release will lost any variables that are in spool files.
2126 PH/07 Two small changes when running in the test harness: increase delay when
2127 passing a TCP/IP connection to a new process, in case the original
2128 process has to generate a bounce, and remove special handling of
2129 127.0.0.2 (sic), which is no longer necessary.
2131 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2132 be the same on different OS.
2134 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2137 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2138 whether --show-vars was specified or not
2140 JJ/02 exipick: Added support for new ACL variable spool format introduced
2143 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2144 syntactically invalid From: or Reply-to: line, and a filter used this to
2145 generate an autoreply, and therefore failed to obtain an address for the
2146 autoreply, Exim could try to deliver to a non-existent relative file
2147 name, causing unrelated and misleading errors. What now happens is that
2148 it logs this as a hard delivery error, but does not attempt to create a
2151 PH/11 The exinext utility has a -C option for testing purposes, but although
2152 the given file was scanned by exinext itself; it wasn't being passed on
2153 when Exim was called.
2155 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2156 an end-of-file indication when reading a command response.
2158 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2159 compiled. In many other places in Exim, IPv6 addresses are always
2160 recognized, so I have changed this. It also means that IPv4 domain
2161 literals of the form [IPV4:n.n.n.n] are now always recognized.
2163 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2164 used if the router is not running as root, for example, when verifying at
2165 ACL time, or when using -bh. The debugging output from this situation was
2166 non-existent - all you got was a failure to exec. I have made two
2169 (a) Failures to set uid/gid, the current directory, or a process leader
2170 in a subprocess such as that created by queryprogram now generate
2171 suitable debugging ouput when -d is set.
2173 (b) The queryprogram router detects when it is not running as root,
2174 outputs suitable debugging information if -d is set, and then runs
2175 the subprocess without attempting to change uid/gid.
2177 PH/15 Minor change to Makefile for building test_host (undocumented testing
2180 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2181 additional section of a DNS packet that returns MX or SRV records.
2182 Instead, it always explicitly searches for A/AAAA records. This avoids
2183 major problems that occur when a DNS server includes only records of one
2184 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2185 fixed another bug: if SRV records were looked up and the corresponding
2186 address records were *not* found in the additional section, the port
2187 values from the SRV records were lost.
2189 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2190 using the correct key (the original address) when searching the retry
2191 rules in order to find which one to use for generating the retry hint.
2193 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2194 adding the default one. Similarly, if it contains a Reply-To: header, the
2195 errors_reply_to option, if set, is not used.
2197 PH/19 When calculating a retry time, Exim used to measure the "time since
2198 failure" by looking at the "first failed" field in the retry record. Now
2199 it does not use this if it is later than than the arrival time of the
2200 message. Instead it uses the arrival time. This makes for better
2201 behaviour in cases where some deliveries succeed, thus re-setting the
2202 "first failed" field. An example is a quota failure for a huge message
2203 when small messages continue to be delivered. Without this change, the
2204 "time since failure" will always be short, possible causing more frequent
2205 delivery attempts for the huge message than are intended.
2206 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2208 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2209 $1, $2, $3) because the numerical variables can be reset during some
2210 expansion items (e.g. "match"), thereby losing the authentication data.
2212 PH/21 Make -bV show the size of off_t variables so that the test suite can
2213 decide whether to run tests for quotas > 2G.
2215 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2216 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2217 filecount value is greater than 2G or if a quota value is greater than 2G
2218 on a system where the size of off_t is not greater than 4, a panic error
2221 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2222 never match. The debug and -bh output now contains an explicit error
2223 message indicating a malformed IPv4 address or mask.
2225 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2226 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2227 PH/23 above applies.
2229 PH/25 Do not write to syslog when running in the test harness. The only
2230 occasion when this arises is a failure to open the main or panic logs
2231 (for which there is an explicit test).
2233 PH/26 Added the /no_tell option to "control=freeze".
2235 PH/27 If a host name lookup failed very early in a connection, for example, if
2236 the IP address matched host_lookup and the reverse lookup yielded a name
2237 that did not have a forward lookup, an error message of the form "no IP
2238 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2239 could be logged. Now it outputs the IP address instead of "NULL".
2241 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2242 allows the sender and the authenticated sender to be set when
2243 submitting a message from within Exim. Since child_open_exim() is
2244 documented for local_scan(), the new function should be too.
2246 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2247 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2248 results in an empty string is now treated as unset.
2250 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2252 PH/31 Added sender_verify_fail logging option.
2254 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2255 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2256 tidied the source and removed it altogether.
2258 PH/33 When a queue run was abandoned because the load average was too high, a
2259 log line was always written; now it is written only if the queue_run log
2260 selector is set. In addition, the log line for abandonment now contains
2261 information about the queue run such as the pid. This is always present
2262 in "start" and "stop" lines but was omitted from the "abandon" line.
2264 PH/34 Omit spaces between a header name and the colon in the error message that
2265 is given when verify = headers_syntax fails (if there are lots of them,
2266 the message gets confusing).
2268 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2269 names, as there are now some PTR records that contain slashes. This check
2270 is only to protect against broken name servers that fall over on strange
2271 characters, so the fact that it applies to all lookups doesn't matter.
2273 PH/36 Now that the new test suite is complete, we can remove some of the
2274 special code in Exim that was needed for the old test suite. For example,
2275 sorting DNS records because real resolvers return them in an arbitrary
2276 order. The new test suite's fake resolver always returns records in the
2279 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2280 bounces) except when queue_only is set, to avoid logging races between
2281 the different processes.
2283 PH/38 Panic-die if .include specifies a non-absolute path.
2285 PH/39 A tweak to the "H" retry rule from its user.
2287 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2288 a label. They prevented compilation on older perls.
2290 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2291 a warning to be raised on newish perls.
2293 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2294 on queue. Changes to match documented behaviour of showing count of
2295 messages matching specified criteria.
2297 PH/40 Changed the default ident timeout from 30s to 5s.
2299 PH/41 Added support for the use of login_cap features, on those BSD systems
2300 that have them, for controlling the resources used by pipe deliveries.
2302 PH/42 The content-scanning code uses fopen() to create files in which to put
2303 message data. Previously it was not paying any attention to the mode of
2304 the files. Exim runs with umask(0) because the rest of the code creates
2305 files with open(), and sets the required mode explicitly. Thus, these
2306 files were ending up world-writeable. This was not a big issue, because,
2307 being within the spool directory, they were not world-accessible. I have
2308 created a function called modefopen, which takes an additional mode
2309 argument. It sets umask(777), creates the file, chmods it to the required
2310 mode, then resets the umask. All the relevant calls to fopen() in the
2311 content scanning code have been changed to use this function.
2313 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2314 to 24 hours. This avoids potential overflow problems when processing G
2315 and H retry rules. I suspect nobody ever tinkers with this value.
2317 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2319 PH/45 When the plaintext authenticator is running as a client, the server's
2320 challenges are checked to ensure they are valid base64 strings. By
2321 default, the authentication attempt is cancelled if an invalid string is
2322 received. Setting client_ignore_invalid_base64 true ignores these errors.
2323 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2324 they are received. Thus, the responses can be made to depend on the
2325 challenges. If an invalid string is ignored, an empty string is placed in
2328 PH/46 Messages that are created by the autoreply transport now contains a
2329 References: header, in accordance with RFCs 2822 and 3834.
2331 PH/47 Added authenticated_sender_force to the smtp transport.
2333 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2335 PH/49 Installed latest patch from the Sieve maintainer.
2337 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2338 was also set, the appendfile transport was unnecessarily scanning a
2339 directory of message files (e.g. for maildir delivery) to find the count
2340 of files (along with the size), even though it did not need this
2341 information. It now does the scan only if it needs to find either the
2342 size of the count of files.
2344 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2346 PH/52 Two bugs concerned with error handling when the smtp transport is
2349 (i) Exim was not creating retry information for temporary errors given
2350 for individual recipients after the DATA command when the smtp transport
2351 was used in LMTP mode. This meant that they could be retried too
2352 frequently, and not timed out correctly.
2354 (ii) Exim was setting the flag that allows error details to be returned
2355 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2356 recipients that were returned after the DATA command.
2358 PH/53 This is related to PH/52, but is more general: for any failing address,
2359 when detailed error information was permitted to be returned to the
2360 sender, but the error was temporary, then after the final timeout, only
2361 "retry timeout exceeded" was returned. Now it returns the full error as
2362 well as "retry timeout exceeded".
2364 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2365 do this, and (what is worse) MTAs that accept it.
2367 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2368 will now be deprecated.
2370 PH/56 New os.c-cygwin from the Cygwin maintainer.
2372 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2373 formats (previously only available in exim formats via -bpr, -bpru,
2374 and -bpra. Now also available in native and exiqgrep formats)
2376 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2377 with very large, slow to parse queues
2379 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2381 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2383 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2384 responses to authentication challenges, though it was showing the
2385 challenges; (ii) I've removed the CR characters from the debug output for
2388 PH/58 Allow for the insertion of a newline as well as a space when a string
2389 is turned into more than one encoded-word during RFC 2047 encoding. The
2390 Sieve code now uses this.
2392 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2393 data_4xx, lost_connection, tls_required.
2395 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2396 message was being sent as an SMTP response.
2398 PH/61 Add -l and -k options to exicyclog.
2400 PH/62 When verifying, if an address was redirected to one new address, so that
2401 verification continued, and the new address failed or deferred after
2402 having set something in $address_data, the value of $address_data was not
2403 passed back to the ACL. This was different to the case when no
2404 redirection occurred. The value is now passed back in both cases.
2406 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2407 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2408 use login_cap.h, so on its own it isn't the distinguishing feature. The
2409 new name refers directly to the setclassresources() function.
2411 PH/65 Added configuration files for NetBSD3.
2413 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2415 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2416 is preferred over IPv4.
2418 PH/68 The bounce_return_message and bounce_return_body options were not being
2419 honoured for bounces generated during the reception of non-SMTP messages.
2420 In particular, this applied to messages rejected by the ACL. This bug has
2421 been fixed. However, if bounce_return_message is true and bounce_return_
2422 body is false, the headers that are returned for a non-SMTP message
2423 include only those that have been read before the error was detected.
2424 (In the case of an ACL rejection, they have all been read.)
2426 PH/69 The HTML version of the specification is now built in a directory called
2427 spec_html instead of spec.html, because the latter looks like a path with
2428 a MIME-type, and this confuses some software.
2430 PH/70 Catch two compiler warnings in sieve.c.
2432 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2433 function verify_get_ident() calls ip_connect() to connect a socket, but
2434 if the "connect()" function timed out, ip_connect() used to close the
2435 socket. However, verify_get_ident() also closes the socket later, and in
2436 between Exim writes to the log, which may get opened at this point. When
2437 the socket was closed in ip_connect(), the log could get the same file
2438 descriptor number as the socket. This naturally causes chaos. The fix is
2439 not to close the socket in ip_connect(); the socket should be closed by
2440 the function that creates it. There was only one place in the code where
2441 this was missing, in the iplookup router, which I don't think anybody now
2442 uses, but I've fixed it anyway.
2444 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2445 well as to direct DNS lookups. Otherwise the handling of names in host
2446 lists is inconsistent and therefore confusing.
2452 PH/01 Two changes to the default runtime configuration:
2454 (1) Move the checks for relay_from_hosts and authenticated clients from
2455 after to before the (commented out) DNS black list checks.
2457 (2) Add control=submission to the relay_from_hosts and authenticated
2458 clients checks, on the grounds that messages accepted by these
2459 statements are most likely to be submissions.
2461 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2463 (1) Generate an error if the third argument for the ${prvs expansion is
2466 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2469 (3) Reset the variables that are obtained from the first argument of
2470 ${prvscheck and used in the second argument before leaving the code,
2471 because their memory is reclaimed, so using them afterwards may do
2474 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2475 one (it's much easier than Tom thought :-).
2477 (5) Because of (4), we can now allow for the use of $prvscheck_result
2478 inside the third argument.
2480 PH/03 For some reason, the default setting of PATH when running a command from
2481 a pipe transport was just "/usr/bin". I have changed it to
2484 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2485 anything to be listed in the output from -bV.
2487 PH/05 When a filter generated an autoreply, the entire To: header line was
2488 quoted in the delivery log line, like this:
2490 => >A.N.Other <ano@some.domain> <original@ddress> ...
2492 This has been changed so that it extracts the operative address. There
2493 may be more than one such address. If so, they are comma-separated, like
2496 => >ano@some.domain,ona@other.domain <original@ddress> ...
2498 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2499 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2500 not being looked up in the rDNS to get a host name, Exim was showing the
2501 IP address twice in Received: lines, even though the IP addresses were
2502 identical. For example:
2504 Received: from [1.2.3.4] (helo=[1.2.3.4])
2506 However, if the real host name was known, it was omitting the HELO data
2507 if it matched the actual IP address. This has been tidied up so that it
2508 doesn't show the same IP address twice.
2510 PH/07 When both +timestamp and +memory debugging was on, the value given by
2511 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2512 called by the debug printing, thereby overwriting the timestamp buffer.
2513 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2515 PH/08 When the original message was included in an autoreply transport, it
2516 always said "this is a copy of the message, including all the headers",
2517 even if body_only or headers_only was set. It now gives an appropriate
2520 PH/09 Applied a patch from the Sieve maintainer which:
2522 o fixes some comments
2523 o adds the (disabled) notify extension core
2524 o adds some debug output for the result of if/elsif tests
2525 o points to the current vacation draft in the documentation
2526 and documents the missing references header update
2530 o fixes a bug in processing the envelope test (when testing
2531 multiple envelope elements, the last element determinted the
2534 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2535 Electronic Mail") by including:
2537 Auto-submitted: auto-generated
2539 in the messages that it generates (bounce messages and others, such as
2540 warnings). In the case of bounce messages for non-SMTP mesages, there was
2541 also a typo: it was using "Auto_submitted" (underscore instead of
2542 hyphen). Since every message generated by Exim is necessarily in response
2543 to another message, thes have all been changed to:
2545 Auto-Submitted: auto-replied
2547 in accordance with these statements in the RFC:
2549 The auto-replied keyword:
2551 - SHOULD be used on messages sent in direct response to another
2552 message by an automatic process,
2554 - MUST NOT be used on manually-generated messages,
2556 - MAY be used on Delivery Status Notifications (DSNs) and Message
2557 Disposition Notifications (MDNs),
2559 - MUST NOT be used on messages generated by automatic or periodic
2560 processes, except for messages which are automatic responses to
2563 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2564 to the default Received: header definition.
2566 PH/12 Added log selector acl_warn_skipped (default on).
2568 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2569 variables because (a) they are in the wrong storage pool and (b) even if
2570 they were copied, it wouldn't work properly because of the caching.
2572 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2573 checking when decoding. Apparently there are clients that generate
2574 overlong encoded strings. Why am I not surprised?
2576 PH/15 If the first argument of "${if match_address" was not empty, but did not
2577 contain an "@" character, Exim crashed. Now it writes a panic log message
2578 and treats the condition as false.
2580 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2582 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2583 extension "envelope-auth". The code is finished and in agreement with
2584 other implementations, but there is no documentation so far and in fact,
2585 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2586 not changing the active code.
2588 Print executed "if" and "elsif" statements when debugging is used. This
2589 helps a great deal to understand what a filter does.
2591 Document more things not specified clearly in RFC3028. I had all this
2592 sorted out, when out of a sudden new issues came to my mind. Oops."
2594 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2597 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2598 canonical form (as documented). However, after a host name lookup from
2599 the IP address, check_host() was doing a simple string comparison with
2600 addresses acquired from the DNS when checking that the found name did
2601 have the original IP as one of its addresses. Since any found IPv6
2602 addresses are likely to be in abbreviated form, the comparison could
2603 fail. Luckily, there already exists a function for doing the comparison
2604 by converting both addresses to binary, so now that is used instead of
2605 the text comparison.
2607 PH/20 There was another similar case to PH/19, when a complete host name was
2608 given in a host list; looking up its IP address could give an abbreviated
2609 form, whereas the current host's name might or might not be abbreviated.
2610 The same fix has been applied.
2616 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2617 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2620 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2621 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2623 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2625 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2626 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2627 Also, for "Auto-Submitted", treat anything other than "no" as
2628 non-personal, in accordance with RFC 3834. (Previously it treated
2629 anything starting "auto-" as non-personal.)
2631 TF/01 The control=submission/name=... option had a problem with syntax
2632 errors if the name included a slash character. The /name= option
2633 now slurps the rest of the string, so it can include any characters
2634 but it must come last in the list of options (after /sender_retain
2637 PH/05 Some modifications to the interface to the fake nameserver for the new
2645 TK/01 Added the "success_on_redirect" address verification option. See
2646 NewStuff for rationale and an example.
2648 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2650 PH/02 Patch to exigrep to allow it to work on syslog lines.
2652 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2653 fread() to skip over the body file's header line, because in Cygwin the
2654 header line is locked and is inaccessible.
2656 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2657 co-exist for some time) to make it clear that it is the Exim ID that is
2658 referenced, not the Message-ID: header line.
2660 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2661 string_format() function, because snprintf() does not exist on all
2664 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2665 ${stat: expansion item.
2667 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2668 protocol synchronization error", to keep the pedants happy.
2670 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2671 well as for IRIX systems, when gcc is being used. See the host.c source
2674 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2676 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2679 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2680 transport and to the smtp transport in LMTP mode.
2682 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2684 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2685 run a filter in a subprocess. This could lead to confusion in subsequent
2686 lookups in the parent process. There should also be a search_tidyup() at
2687 the end of the subprocess.
2689 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2690 only if the host matched helo_try_verify_hosts, which caused the
2691 verification to occur when the EHLO/HELO command was issued. The ACL just
2692 tested the remembered result. Now, if a previous verification attempt has
2693 not happened, "verify = helo" does it there and then.
2695 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2697 TK/03 Fix log output including CR from clamd.
2699 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2700 exist provoked a memory error which could cause a segfault.
2702 PH/15 Installed PCRE 6.2
2704 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2706 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2707 of the problem. Specifically, suggested +O2 rather than +O1 for the
2710 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2712 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2713 record, the port was not correctly logged when the outgoing_port log
2714 selector was set (it logged the transort's default port).
2716 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2717 fallback_hosts, and "hosts" in the smtp transport.
2719 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2720 host errors such as "Connection refused".
2722 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2723 authentication with radiusclient 0.4.9:
2725 - Error returned from rc_read_config was caught wrongly
2726 - Username/password not passed on to radius server due to wrong length.
2728 The presumption is that some radiusclient API changes for 4.51/PH/17
2729 were not taken care of correctly. The code is still untested by me (my
2730 Linux distribution still has 0.3.2 of radiusclient), but it was
2731 contributed by a Radius user.
2733 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2734 expanding the "port" option of the smtp transport.
2736 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2737 while reading a MIME header. Thanks to Tom Hughes for a patch.
2739 PH/24 Include config.h inside local_scan.h so that configuration settings are
2742 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2743 that in an ACL for RCPT (for example), you can examine exactly what was
2746 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2747 commands, but it was not correctly comparing the address with the actual
2748 client host address. Thus, it would show the EHLO address in Received:
2749 header lines when this was not necessary.
2751 PH/27 Added the % operator to ${eval:}.
2753 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2754 it should be ignoring failures (because with -C, for example, it has lost
2755 privilege). It wasn't ignoring creation failures other than "already
2758 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2761 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2762 because an input line was too long, either on its own, or by virtue of
2763 too many continuations, the temporary file was not being removed, and the
2764 return code was incorrect.
2766 PH/31 Missing "BOOL" in function definition in filtertest.c.
2768 PH/32 Applied Sieve patches from the maintainer.
2770 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2772 PH/33 Added "verify = not_blind".
2774 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2775 Local/Makefile (with some defaults set). These are used in built scripts
2776 such as exicyclog, but they have never been used in the exim_install
2777 script (though there are many overriding facilities there). I have
2778 arranged that the exim_install script now takes note of these two
2781 PH/35 Installed configuration files for Dragonfly.
2783 PH/36 When a locally submitted message by a trusted user did not contain a
2784 From: header, and the sender address was obtained from -f or from an SMTP
2785 MAIL command, and the trusted user did not use -F to supply a sender
2786 name, $originator_name was incorrectly used when constructing a From:
2787 header. Furthermore, $originator_name was used for submission mode
2788 messages from external hosts without From: headers in a similar way,
2789 which is clearly wrong.
2791 PH/37 Added control=suppress_local_fixups.
2793 PH/38 When log_selector = +received_sender was set, and the addition of the
2794 sender made the log line's construction buffer exactly full, or one byte
2795 less than full, an overflow happened when the terminating "\n" was
2798 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2799 when the result of a list match is failure because a DNS lookup failed.
2801 PH/40 RM_COMMAND is now used in the building process.
2803 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2804 the "build-* directories that it finds.
2806 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2807 address in a domain literal was a prefix of an interface address.
2809 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2810 when verifying a sender address, unless rewrite_headers is false.
2812 PH/44 Wrote a long comment about why errors_to addresses are verified as
2813 recipients, not senders.
2815 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2816 the ratelimit ACL was added.
2818 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2820 PH/46 Added extra information about PostgreSQL errors to the error string.
2822 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2823 avoiding the need to install special zones in a real server. This is
2824 backwards compatible; if it can't find the fake resolver, it drops back.
2825 Thus, both old and new test suites can be run.
2827 TF/02 Added util/ratelimit.pl
2829 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2830 clock is set back in time.
2832 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2833 Candler <B.Candler@pobox.com>.
2835 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2836 OK for addresses that are the result of redirections.
2838 PH/48 A number of further additions for the benefit of the new test suite,
2839 including a fake gethostbyname() that interfaces to the fake DNS resolver
2842 TF/06 The fix for widen_domains has also been applied to qualify_single and
2843 search_parents which are the other dnslookup options that can cause
2846 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2849 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2851 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2852 service" if it encountered an MX record with an empty target hostname.
2853 The message is now "an MX or SRV record indicated no SMTP service".
2855 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2856 if the DNS of the sending site is misconfigured. This is quite a
2857 common situation. This change restores the behaviour of treating a
2858 helo verification defer as a failure.
2860 PH/51 If self=fail was set on a router, the bounce message did not include the
2861 actual error message.
2867 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2869 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2870 command itself ended in error, the underlying message about the transport
2871 filter timeout was being overwritten with the pipe command error. Now the
2872 underlying error message should be appended to the second error message.
2874 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2876 PH/02 Reduce the amount of output that "make" produces by default. Full output
2877 can still be requested.
2879 PH/03 The warning log line about a condition test deferring for a "warn" verb
2880 was being output only once per connection, rather than after each
2881 occurrence (because it was using the same function as for successful
2882 "warn" verbs). This seems wrong, so I have changed it.
2884 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2885 it should not have, which might have caused a crash in the right
2886 circumstances, but probably never did.
2888 PH/04 Installed a modified version of Tony Finch's patch to make submission
2889 mode fix the return path as well as the Sender: header line, and to
2890 add a /name= option so that you can make the user's friendly name appear
2893 TF/03 Added the control = fakedefer ACL modifier.
2895 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2896 Mark Lowes for thorough testing.
2898 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2900 TK/03 Merged latest SRS patch from Miles Wilton.
2902 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2903 with the definition in sysexits.h (which is #included earlier).
2904 Fortunately, Exim does not actually use EX_OK. The code used to try to
2905 preserve the sysexits.h value, by assumimg that macro definitions were
2906 scanned for macro replacements. I have been disabused of this notion,
2907 so now the code just undefines EX_OK before #including unistd.h.
2909 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2910 in the smtp transport. When a block could not be written in a single
2911 write() function, the timeout was being re-applied to each part-write.
2912 This seems wrong - if the receiver was accepting one byte at a time it
2913 would take for ever. The timeout is now adjusted when this happens. It
2914 doesn't have to be particularly precise.
2916 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2917 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2919 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2920 without a domain if the check to <postmaster@domain> fails.
2922 SC/01 Eximstats: added -xls and the ability to specify output files
2923 (patch written by Frank Heydlauf).
2925 SC/02 Eximstats: use FileHandles for outputing results.
2927 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2929 SC/04 Eximstats: fixed display of large numbers with -nvr option
2931 SC/05 Eximstats: fixed merging of reports with empty tables.
2933 SC/06 Eximstats: added the -include_original_destination flag
2935 SC/07 Eximstats: removed tabs and trailing whitespace.
2937 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2939 TK/06 MBOX spool code: Add real "From " MBOX separator line
2940 so the .eml file is really in mbox format (even though
2941 most programs do not really care). Patch from Alex Miller.
2943 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2944 The latter is generated from $received_to and is only set if the
2945 message has one envelope recipient. SA can use these headers,
2946 obviously out-of-the-box. Patch from Alex Miller.
2948 PH/08 The ${def test on a variable was returning false if the variable's
2949 value was "0", contrary to what the specification has always said!
2950 The result should be true unless the variable is empty.
2952 PH/09 The syntax error of a character other than { following "${if
2953 def:variable_name" (after optional whitespace) was not being diagnosed.
2954 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2955 accidental colon was present, for example, could give incorrect results.
2957 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2958 result is used (not including appendfile, where other changes are about
2961 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2962 This involved changing a lot of size variables from int to off_t. It
2963 should work with maildirs and everything.
2965 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2966 spamd dying while we are connected to it.
2968 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2971 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2972 function for BDB 4.3.
2974 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2976 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2977 the sources for PCRE have been re-arranged and more files are now
2980 PH/15 The code I had for printing potentially long long variables in PH/11
2981 above was not the best (it lost precision). The length of off_t variables
2982 is now inspected at build time, and an appropriate printing format (%ld
2983 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2984 to be "long long int" or "long int". This is needed for the internal
2985 formatting function string_vformat().
2987 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2988 the configuration file to be ":syslog", then the script "guesses" where
2989 the logs files are, rather than using the compiled in default. In our
2990 case the guess is not the same as the compiled default, so the script
2991 suddenly stopped working when I started to use syslog. The patch checks
2992 to see if log_file_path is "". If so, it attempts to read it from exim
2993 with no configuration file to get the compiled in version, before it
2994 falls back to the previous guessing code."
2996 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2997 implementing BATV in an Exim configuration. See NewStuff for the gory
3000 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3001 Makefile that are specific to HP-UX.
3003 PH/18 If the "use_postmaster" option was set for a recipient callout together
3004 with the "random" option, the postmaster address was used as the MAIL
3005 FROM address for the random test, but not for the subsequent recipient
3006 test. It is now used for both.
3008 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3009 patch removes a few documentation additions to RFC 3028, because the
3010 latest draft now contains them. It adds the new en;ascii-case comparator
3011 and a new error check for 8bit text in MIME parts. Comparator and
3012 require names are now matched exactly. I enabled the subaddress
3013 extension, but it is not well tested yet (read: it works for me)."
3015 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3016 rework some of the code of TK/09 above to avoid the hardwired use of
3017 "%lld" and "long long". Replaced the call to snprintf() with a call to
3020 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3021 records point to non-existent hosts", "retry timeout exceeded", and
3022 "retry time not reached for any host after a long failure period".
3024 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3025 experimental DomainKeys support:
3027 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3028 (2) On an error such as an illegally used "control", the wrong name for
3029 the control was given.
3031 These problems did NOT occur unless DomainKeys support was compiled.
3033 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3035 PH/24 Added ${if match_ip condition.
3037 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3038 databases so that it will be absolutely obvious if a crash occurs in the
3039 DB library. This is a regular occurrence (often caused by mis-matched
3042 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3043 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3044 picked up on a user's system that detects such things. There doesn't seem
3045 to be a gcc warning option for this - only an attribute that has to be
3046 put on the function's prototype. It seems that in Fedora Core 4 they have
3047 set this on a number of new functions. No doubt there will be more in due
3050 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3051 specify a transport. However, if an address that was verified by such a
3052 router was the subject of a callout, Exim crashed because it tried to
3053 read the rcpt_include_affixes from the non-existent transport. Now it
3054 just assumes that the setting of that option is false. This bug was
3055 introduced by 4.51/PH/31.
3057 PH/28 Changed -d+all to exclude +memory, because that information is very
3058 rarely of interest, but it makes the output a lot bigger. People tend to
3059 do -d+all out of habit.
3061 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3062 code in os-type was giving problems when libc.so lives in lib64, like on
3065 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3066 aren't the modern standard, and it seems that some systems' include files
3067 don't always have them. Exim was already checking for some of the newer
3068 ones like T_AAAA, and defining it itself. I've added checks for all the
3069 record types that Exim uses.
3071 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3072 not automatically generating a new one, as it is supposed to. This
3073 prevented TLS from working. If the file did exist, but contained invalid
3074 data, a new version was generated, as expected. It was only the case of a
3075 non-existent file that was broken.
3077 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3078 with a change in libdomainkeys > 0.64.
3080 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3081 from DNS. If the selector record carries the flag, it now has
3082 precedence over the domain-wide flag.
3084 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3086 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3087 the use of a port name that isn't defined in /etc/services) occurred, the
3088 message was deferred as in a normal delivery, and thus remained on the
3089 spool, instead of being failed because of the mua_wrapper setting. This
3090 is now fixed, and I tidied up some of the mua_wrapper messages at the
3093 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3094 the messages in a hash of arrays rather than using individual hashes.
3095 This is a bit cleaner and results in dramatic memory savings, albeit
3096 at a slight CPU cost.
3098 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3099 as requested by Marc Sherman.
3101 SC/10 Eximstats: added histograms for user specified patterns as requested
3104 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3106 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3107 fopen() in the content-scanning modules that did not already have it.
3113 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3114 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3116 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3118 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3120 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3121 to test Sieve filters that use "vacation".
3123 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3124 that changes the way the GnuTLS parameters are stored in the cache file.
3125 The new format can be generated externally. For backward compatibility,
3126 if the data in the cache doesn't make sense, Exim assumes it has read an
3127 old-format file, and it generates new data and writes a new file. This
3128 means that you can't go back to an older release without removing the
3131 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3132 work if there are any delivery delays because "one_time" forces the
3133 parent to be marked "delivered", so its unseen clone is never tried
3134 again. For this reason, Exim now forbids the simultaneous setting of
3137 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3138 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3139 entry if you want to know the details. The fix, however, neglected to
3140 consider the case when local delivery batching is involved. The test for
3141 "previously delivered" was not happening when checking to see if an
3142 address could be batched with a previous (undelivered) one; under
3143 certain circumstances this could lead to multiple deliveries to the same
3146 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3147 in its include files, and this causes problems building Exim.
3149 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3150 header_syntax) but Exim was just ignoring anything given after a slash.
3151 In particular, this caused confusion with an attempt to use "verify =
3152 reverse_host_lookup/defer_ok". An error is now given when options are
3153 supplied for verify items that do not have them. (Maybe reverse_host_
3154 lookup should have a defer_ok option, but that's a different point.)
3156 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3157 defined by RFC 821) to 2048, because there were problems with some AUTH
3158 commands, and RFC 1869 says the size should be increased for extended
3159 SMTP commands that take arguments.
3161 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3164 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3165 "unknown" error; now it says that the functionality isn't in the binary.
3167 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3168 an address' error message when a string expansion fails (syntax or
3169 whatever). Otherwise the password may appear in the log. Following change
3170 PH/42 below, there is no longer a chance of it appearing in a bounce
3173 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3175 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3176 of its addresses was ever tried. (Bugzilla bug #2.)
3178 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3179 the result incorrectly in the debug output. (It correctly added a newline
3180 to what was transported.)
3182 TF/01 Added $received_time.
3184 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3185 commented out examples of how to interface to a virus scanner and to
3186 SpamAssassin. Also added commented examples of av_scanner and
3187 spamd_address settings.
3189 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3190 and controls are allowed in which ACLs. There were a couple of minor
3191 errors. Some of the entries in the conditions table (which is a table of
3192 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3193 negation of where the condition IS allowed.
3195 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3197 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3198 header file does not have a version number, so I've had to invent a new
3199 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3200 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3201 radiusclient), but it was contributed by a Radius user.
3203 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3204 files or directories, for OpenSSL.
3206 PH/19 When an Exim process that is running as root has to create an Exim log
3207 file, it does so in a subprocess that runs as exim:exim so as to get the
3208 ownership right at creation (otherwise, other Exim processes might see
3209 the file with the wrong ownership). There was no test for failure of this
3210 fork() call, which would lead to the process getting stuck as it waited
3211 for a non-existent subprocess. Forks do occasionally fail when resources
3212 run out. I reviewed all the other calls to fork(); they all seem to check
3215 PH/20 When checking for unexpected SMTP input at connect time (before writing
3216 the banner), Exim was not dealing correctly with a non-positive return
3217 from the read() function. If the client had disconnected by this time,
3218 the result was a log entry for a synchronization error with an empty
3219 string after "input=" when read() returned zero. If read() returned -1
3220 (an event I could not check), uninitialized data bytes were printed.
3221 There were reports of junk text (parts of files, etc) appearing after
3224 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3226 PH/22 Added support for macro redefinition, and (re)definition in between
3227 driver and ACL definitions.
3229 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3230 forgetting to use the resulting value; it was using the unexpanded value.
3232 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3233 hadn't been configured. The fix is from Juergen Kreileder, who
3234 understands it better than I do:
3236 "Here's what I see happening with three configured cyrus_sasl
3237 authenticators configured (plain, login, cram-md5):
3239 On startup auth_cyrus_sasl_init() gets called for each of these.
3240 This means three calls to sasl_listmech() without a specified mech_list.
3241 => SASL tests which mechs of all available mechs actually work
3242 => three warnings about OTP not working
3243 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3245 With the patch, sasl_listmech() also gets called three times. But now
3246 SASL's mech_list option is set to the server_mech specified in the the
3247 authenticator. Or in other words, the answer from sasl_listmech()
3248 gets limited to just the mech you're testing for (which is different
3250 => the return list contains just 'plain' or 'login', 'cram-md5' or
3251 nothing depending on the value of ob->server_mech.
3253 I've just tested the patch: Authentication still works fine,
3254 unavailable mechs specified in the exim configuration are still
3255 caught, and the auth.log warnings about OTP are gone."
3257 PH/25 When debugging is enabled, the contents of the command line are added
3258 to the debugging output, even when log_selector=+arguments is not
3261 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3262 answer is "GNU", and only if the return is "GNU/something" is the answer
3265 PH/27 $acl_verify_message is now set immediately after the failure of a
3266 verification in an ACL, and so is available in subsequent modifiers. In
3267 particular, the message can be preserved by coding like this:
3269 warn !verify = sender
3270 set acl_m0 = $acl_verify_message
3272 Previously, $acl_verify_message was set only while expanding "message"
3273 and "log_message" when a very denied access.
3275 PH/28 Modified OS/os.c-Linux with
3277 -#ifndef OS_LOAD_AVERAGE
3278 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3280 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3281 nomenclature these days.)
3283 PH/29 Installed patch from the Sieve maintainer that adds the options
3284 sieve_useraddress and sieve_subaddress to the redirect router.
3286 PH/30 In these circumstances:
3287 . Two addresses routed to the same list of hosts;
3288 . First host does not offer TLS;
3289 . First host accepts first address;
3290 . First host gives temporary error to second address;
3291 . Second host offers TLS and a TLS session is established;
3292 . Second host accepts second address.
3293 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3294 and peerdn, if requested) that were in fact used only for the second
3297 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3298 attention to any local part prefix or suffix that was matched by the
3299 router that accepted the address. It now behaves in the same way as it
3300 does for delivery: the affixes are removed from the local part unless
3301 rcpt_include_affixes is set on the transport.
3303 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3304 timeout during the DATA phase of an incoming message.
3306 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3307 applied a patch sanctioned by the Sieve maintainer.
3309 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3310 the uid or gid is negative. A case of a negative gid caused this to be
3311 noticed. The fix allows for either to be negative.
3313 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3314 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3317 PH/36 The change PH/12 above was broken. Fixed it.
3319 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3320 the grounds that routing the same address twice would always produce the
3321 same answer. This might have been true once, but it is certainly no
3322 longer true now. Routing a child address may depend on the previous
3323 routing that produced that child. Some complicated redirection strategies
3324 went wrong when messages had multiple recipients, and made Exim's
3325 behaviour dependent on the order in which the addresses were given.
3327 I have moved the duplicate checking until after the routing is complete.
3328 Exim scans the addresses that are assigned to local and remote
3329 transports, and removes any duplicates. This means that more work will be
3330 done, as duplicates will always all be routed, but duplicates are
3331 presumably rare, so I don't expect this is of any significance.
3333 For deliveries to pipes, files, and autoreplies, the duplicate checking
3334 still happens during the routing process, since they are not going to be
3337 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3338 It corrects a timeout issue with spamd. This is Ian's comment: "The
3339 background is that sometimes spamd either never reads data from a
3340 connection it has accepted, or it never writes response data. The exiscan
3341 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3342 blindly assumes that writes won't block so it may never time out."
3344 PH/39 Allow G after quota size as well as K and M.
3346 PH/40 The value set for $authenticated_id in an authenticator may not contain
3347 binary zeroes or newlines because the value is written to log lines and
3348 to spool files. There was no check on this. Now the value is run through
3349 the string_printing() function so that such characters are converted to
3350 printable escape sequences.
3352 PH/41 $message_linecount is a new variable that contains the total number of
3353 lines in the message. Compare $body_linecount, which is the count for the
3356 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3357 bounce and delay warning messages, except in certain special cases, which
3360 (a) An SMTP error message from a remote host;
3361 (b) A message specified in a :fail: redirection;
3362 (c) A message specified in a "fail" command in a system filter;
3363 (d) A message specified in a FAIL return from the queryprogram router;
3364 (e) A message specified by the cannot_route_message router option.
3366 In these cases only, Exim does include the error details in bounce and
3367 warning messages. There are also a few cases where bland messages such
3368 as "unrouteable address" or "local delivery error" are given.
3370 PH/43 $value is now also set for the "else" part of a ${run expansion.
3372 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3373 being worked on, but at least Exim now implements the latest version to
3376 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3377 process to complete was treated as a delivery failure, a timeout while
3378 writing the message to the pipe was logged, but erroneously treated as a
3379 successful delivery. Such timeouts include transport filter timeouts. For
3380 consistency with the overall process timeout, these timeouts are now
3381 treated as errors, giving rise to delivery failures by default. However,
3382 there is now a new Boolean option for the pipe transport called
3383 timeout_defer, which, if set TRUE, converts the failures into defers for
3384 both kinds of timeout. A transport filter timeout is now identified in
3387 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3388 systems where "make" and "gmake" are different, calling "gmake" at top
3389 level broke things. I've arranged for the value of $(MAKE) to be passed
3390 from the Makefile to this script so that it can call the same version of
3394 A note about Exim versions 4.44 and 4.50
3395 ----------------------------------------
3397 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3398 changes of various kinds. As a consequence, a big documentation update was
3399 needed. This delayed the release for rather longer than seemed good, especially
3400 in the light of a couple of (minor) security issues. Therefore, the changes
3401 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3402 release. So 4.44 and 4.50 are in effect two different branches that both start
3405 I have left the 4.50 change log unchanged; it contains all the changes since
3406 4.43. The change log for 4.44 is below; many of its items are identical to
3407 those for 4.50. This seems to be the most sensible way to preserve the
3408 historical information.
3414 1. Minor wording change to the doc/README.SIEVE file.
3416 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3417 computation of the current number of files was incorrect.
3419 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3420 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3423 4. Give more explanation in the error message when the command for a transport
3424 filter fails to execute.
3426 5. There are several places where Exim runs a non-Exim command in a
3427 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3428 was being done only for the command run by the queryprogram router. It is
3429 now done for all such subprocesses. The other cases are: ${run, transport
3430 filters, and the commands run by the lmtp and pipe transports.
3432 6. Added CONFIGURE_GROUP build-time option.
3434 7. Some older OS have a limit of 256 on the maximum number of file
3435 descriptors. Exim was using setrlimit() to set 1000 as a large value
3436 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3437 systems. I've change it so that if it can't get 1000, it tries for 256.
3439 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3440 was an oversight, and furthermore, ever since the addition of extra
3441 controls (e.g. 4.43/32), the checks on when to allow different forms of
3442 "control" were broken. There should now be diagnostics for all cases when a
3443 control that does not make sense is encountered.
3445 9. Added the /retain_sender option to "control=submission".
3447 10. $recipients is now available in the predata ACL (oversight).
3449 11. Tidy the search cache before the fork to do a delivery from a message
3450 received from the command line. Otherwise the child will trigger a lookup
3451 failure and thereby defer the delivery if it tries to use (for example) a
3452 cached ldap connection that the parent has called unbind on.
3454 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3455 of $address_data from the recipient verification was clobbered by the
3456 sender verification.
3458 13. The value of address_data from a sender verification is now available in
3459 $sender_address_data in subsequent conditions in the ACL statement.
3461 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3463 15. Added a new option "connect=<time>" to callout options, to set a different
3466 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3467 was its contents. (It was OK if the option was not defined at all.)
3469 17. A "Completed" log line is now written for messages that are removed from
3470 the spool by the -Mrm option.
3472 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3473 information about exactly what failed.
3475 19. Added -dd to debug only the daemon process.
3477 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3478 handles timeouts, both on the server side and network timeouts. Renamed the
3479 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3481 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3482 It is now set to "smtps".
3484 22. $host_address is now set to the target address during the checking of
3485 ignore_target_hosts.
3487 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3488 being passed; this would have caused $sender_host_name to have been used if
3489 matching the list had actually called for a host name (not very likely,
3490 since this list is usually IP addresses). A host name is now passed as
3493 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3494 code that specifies a non-restarting handler (typically sigaction() in
3495 modern systems) in an attempt to fix a rare and obscure crash bug.
3497 25. Narrowed the window for a race in the daemon that could cause it to ignore
3498 SIGCHLD signals. This is not a major problem, because they are used only to
3499 wake it up if nothing else does.
3501 26. A malformed maildirsize file could cause Exim to calculate negative values
3502 for the mailbox size or file count. Odd effects could occur as a result.
3503 The maildirsize information is now recalculated if the size or filecount
3506 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3507 support for a long time. Removed HAVE_SYS_VFS_H.
3509 28. Installed the latest version of exipick from John Jetmore.
3511 29. In an address list, if the pattern was not a regular expression, an empty
3512 subject address (from a bounce message) matched only if the pattern was an
3513 empty string. Non-empty patterns were not even tested. This was the wrong
3514 because it is perfectly reasonable to use an empty address as part of a
3515 database query. An empty address is now tested by patterns that are
3516 lookups. However, all the other forms of pattern expect the subject to
3517 contain a local part and a domain, and therefore, for them, an empty
3518 address still always fails if the pattern is not itself empty.
3520 30. Exim went into a mad DNS loop when attempting to do a callout where the
3521 host was specified on an smtp transport, and looking it up yielded more
3522 than one IP address.
3524 31. Re-factored the code for checking spool and log partition space into a
3525 function that finds that data and another that does the check. The former
3526 is then used to implement four new variables: $spool_space, $log_space,
3527 $spool_inodes, and $log_inodes.
3529 32. The RFC2047 encoding function was originally intended for short strings
3530 such as real names; it was not keeping to the 75-character limit for
3531 encoded words that the RFC imposes. It now respects the limit, and
3532 generates multiple encoded words if necessary. To be on the safe side, I
3533 have increased the buffer size for the ${rfc2047: expansion operator from
3536 33. It is now permitted to omit both strings after an "if" condition; if the
3537 condition is true, the result is "true". As before, when the second string
3538 is omitted, a false condition yields an empty string. This makes it less
3539 cumbersome to write custom ACL and router conditions.
3541 34. Failure to deliver a bounce message always caused it to be frozen, even if
3542 there was an errors_to setting on the router. The errors_to setting is now
3545 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3546 canonical form (fully expanded) before being placed in
3547 $sender_host_address.
3549 36. The table in the code that translates DNS record types into text (T_A to
3550 "A" for instance) was missing entries for NS and CNAME. It is just possible
3551 that this could have caused confusion if both these types were looked up
3552 for the same domain, because the text type is used as part of Exim's
3553 per-process caching. But the chance of anyone hitting this buglet seems
3556 37. The dnsdb lookup has been extended in a number of ways.
3558 (1) There is a new type, "zns", which walks up the domain tree until it
3559 finds some nameserver records. It should be used with care.
3561 (2) There is a new type, "mxh", which is like "mx" except that it returns
3562 just the host names, not the priorities.
3564 (3) It is now possible to give a list of domains (or IP addresses) to be
3565 looked up. The behaviour when one of the lookups defers can be
3566 controlled by a keyword.
3568 (4) It is now possible to specify the separator character for use when
3569 multiple records are returned.
3571 38. The dnslists ACL condition has been extended: it is now possible to supply
3572 a list of IP addresses and/or domains to be looked up in a particular DNS
3575 39. Added log_selector=+queue_time_overall.
3577 40. When running the queue in the test harness, wait just a tad after forking a
3578 delivery process, to get repeatability of debugging output.
3580 41. Include certificate and key file names in error message when GnuTLS fails
3581 to set them up, because the GnuTLS error message doesn't include the name
3582 of the failing file when there is a problem reading it.
3584 42. Allow both -bf and -bF in the same test run.
3586 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3588 44. The "Exiscan patch" is now merged into the mainline Exim source.
3590 45. Sometimes the final signoff response after QUIT could fail to get
3591 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3592 < 0 before doing a fflush(). This bug looks as though it goes back to the
3593 introduction of TLS in release 3.20, but "sometimes" must have been rare
3594 because the tests only now provoked it.
3596 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3597 (this can affect the format of dates).
3599 47. exim_tidydb, when checking for the continued existence of a message for
3600 which it has found a message-specific retry record, was not finding
3601 messages that were in split spool directories. Consequently, it was
3602 deleting retry records that should have stayed in existence.
3604 48. Steve fixed some bugs in eximstats.
3606 49. The SPA authentication driver was not abandoning authentication and moving
3607 on to the next authenticator when an expansion was forced to fail,
3608 contradicting the general specification for all authenticators. Instead it
3609 was generating a temporary error. It now behaves as specified.
3611 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3612 (the order specifies the preference for clients). The order is now AES256,
3613 AES128, 3DES, ARCFOUR128.
3615 51. Small patch to Sieve code - explicitly set From: when generating an
3618 52. Exim crashed if a remote delivery caused a very long error message to be
3619 recorded - for instance if somebody sent an entire SpamAssassin report back
3620 as a large number of 550 error lines. This bug was coincidentally fixed by
3621 increasing the size of one of Exim's internal buffers (big_buffer) that
3622 happened as part of the Exiscan merge. However, to be on the safe side, I
3623 have made the code more robust (and fixed the comments that describe what
3626 53. Now that there can be additional text after "Completed" in log lines (if
3627 the queue_time_overall log selector is set), a one-byte patch to exigrep
3628 was needed to allow it to recognize "Completed" as not the last thing in
3631 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3632 patch that reportedly fixes this has been added. I am not expert enough to
3633 create a test for it. This is what the patch creator wrote:
3635 "I found a little strange behaviour of ldap code when working with
3636 Windows 2003 AD Domain, where users was placed in more than one
3637 Organization Units. When I tried to give exim partial DN, the exit code
3638 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3639 But simultaneously result of request was absolutely normal ldap result,
3640 so I produce this patch..."
3642 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3643 REFERENCE, so I have modified the code to exclude the patch when that macro
3646 55. Some experimental protocols are using DNS PTR records for new purposes. The
3647 keys for these records are domain names, not reversed IP addresses. The
3648 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3649 leaves it alone. Component reversal etc. now happens only for IP addresses.
3652 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3654 57. Double the size of the debug message buffer (to 2048) so that more of very
3655 long debug lines gets shown.
3657 58. The exicyclog utility now does better if the number of log files to keep
3658 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3660 59. Two changes related to the smtp_active_hostname option:
3662 (1) $smtp_active_hostname is now available as a variable.
3663 (2) The default for smtp_banner uses $smtp_active_hostname instead
3664 of $primary_hostname.
3666 60. The host_aton() function is supposed to be passed a string that is known
3667 to be a valid IP address. However, in the case of IPv6 addresses, it was
3668 not checking this. This is a hostage to fortune. Exim now panics and dies
3669 if the condition is not met. A case was found where this could be provoked
3670 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3671 components; fortuitously, this particular loophole had already been fixed
3672 by change 4.50/55 above.
3674 If there are any other similar loopholes, the new check in host_aton()
3675 itself should stop them being exploited. The report I received stated that
3676 data on the command line could provoke the exploit when Exim was running as
3677 exim, but did not say which command line option was involved. All I could
3678 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3679 running as the user.
3682 61. There was a buffer overflow vulnerability in the SPA authentication code
3683 (which came originally from the Samba project). I have added a test to the
3684 spa_base64_to_bits() function which I hope fixes it.
3687 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3688 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3690 63. The daemon start-up calls getloadavg() while still root for those OS that
3691 need the first call to be done as root, but it missed one case: when
3692 deliver_queue_load_max is set with deliver_drop_privilege. This is
3693 necessary for the benefit of the queue runner, because there is no re-exec
3694 when deliver_drop_privilege is set.
3696 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3697 This has been fixed.
3699 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3700 was in use, was not putting the data itself into the right store pool;
3701 consequently, it could be overwritten for a subsequent message in the same
3702 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3705 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3707 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3708 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3709 the function were treating the return as a boolean value, which happened to
3710 work because 0=false and not-0=true, but is not correct code.
3712 68. The host_aton() function was not handling scoped IPv6 addresses (those
3713 with, for example, "%eth0" on the end) correctly.
3715 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3716 negated items (that is, ~something) in unsigned ints. Some compilers
3717 apparently mutter when there is no cast.
3719 70. If an address verification called from an ACL failed, and did not produce a
3720 user-specific message (i.e. there was only a "system" message), nothing was
3721 put in $acl_verify_message. In this situation, it now puts the system
3724 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3725 session; change 4.31/43 added the unwanted input to the log line - except
3726 that it did not do this in the start of session case. It now does.
3728 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3729 This is wrong and can cause the other end to generate a synchronization
3730 error if it is another Exim or anything else that does the synchronization
3731 check. A QUIT command is no longer sent after a timeout.
3733 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3734 during host lookups.
3736 74. The defer_ok option of callout verification was not working if it was used
3737 when verifying addresses in header lines, that is, for this case:
3739 verify = header_sender/callout=defer_ok
3741 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3742 those file descriptors could be used for SMTP connections. If anything
3743 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3744 could be sent to the SMTP client, causing chaos. The daemon now opens
3745 stdin, stdout, and stderr to /dev/null when it puts itself into the
3748 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3749 log by default. The user can override this with suitable Perl magic.
3751 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3752 the log message when discard triggers, was not working for the DATA ACL or
3753 for the non-SMTP ACL.
3755 78. Error message wording change in sieve.c.
3757 79. If smtp_accept_max_per_host was set, the number of connections could be
3758 restricted to fewer than expected, because the daemon was trying to set up
3759 a new connection before checking whether the processes handling previous
3760 connections had finished. The check for completed processes is now done
3761 earlier. On busy systems, this bug wouldn't be noticed because something
3762 else would have woken the daemon, and it would have reaped the completed
3765 80. If a message was submitted locally by a user whose login name contained one
3766 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3767 It caused a spool format error. I have fixed the spool reading code. A
3768 related problem was that the "from" clause in the Received: line became
3769 illegal because of the space(s). It is now covered by ${quote_local_part.
3771 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3774 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3776 83. Patch from Sieve maintainer to fix unterminated string problem in
3777 "vacation" handling.
3779 84. Some minor changes to the Linux configuration files to help with other
3780 OS variants using glibc.
3782 85. One more patch for Sieve to update vacation handling to latest spec.
3785 ----------------------------------------------------
3786 See the note above about the 4.44 and 4.50 releases.
3787 ----------------------------------------------------
3793 1. Change 4.43/35 introduced a bug that caused file counts to be
3794 incorrectly computed when quota_filecount was set in an appendfile
3797 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3798 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3801 3. Give more explanation in the error message when the command for a transport
3802 filter fails to execute.
3804 4. There are several places where Exim runs a non-Exim command in a
3805 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3806 was being done only for the command run by the queryprogram router. It is
3807 now done for all such subprocesses. The other cases are: ${run, transport
3808 filters, and the commands run by the lmtp and pipe transports.
3810 5. Some older OS have a limit of 256 on the maximum number of file
3811 descriptors. Exim was using setrlimit() to set 1000 as a large value
3812 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3813 systems. I've change it so that if it can't get 1000, it tries for 256.
3815 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3816 was an oversight, and furthermore, ever since the addition of extra
3817 controls (e.g. 4.43/32), the checks on when to allow different forms of
3818 "control" were broken. There should now be diagnostics for all cases when a
3819 control that does not make sense is encountered.
3821 7. $recipients is now available in the predata ACL (oversight).
3823 8. Tidy the search cache before the fork to do a delivery from a message
3824 received from the command line. Otherwise the child will trigger a lookup
3825 failure and thereby defer the delivery if it tries to use (for example) a
3826 cached ldap connection that the parent has called unbind on.
3828 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3829 of $address_data from the recipient verification was clobbered by the
3830 sender verification.
3832 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3833 was its contents. (It was OK if the option was not defined at all.)
3835 11. A "Completed" log line is now written for messages that are removed from
3836 the spool by the -Mrm option.
3838 12. $host_address is now set to the target address during the checking of
3839 ignore_target_hosts.
3841 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3842 being passed; this would have caused $sender_host_name to have been used if
3843 matching the list had actually called for a host name (not very likely,
3844 since this list is usually IP addresses). A host name is now passed as
3847 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3848 code that specifies a non-restarting handler (typically sigaction() in
3849 modern systems) in an attempt to fix a rare and obscure crash bug.
3851 15. Narrowed the window for a race in the daemon that could cause it to ignore
3852 SIGCHLD signals. This is not a major problem, because they are used only to
3853 wake it up if nothing else does.
3855 16. A malformed maildirsize file could cause Exim to calculate negative values
3856 for the mailbox size or file count. Odd effects could occur as a result.
3857 The maildirsize information is now recalculated if the size or filecount
3860 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3861 support for a long time. Removed HAVE_SYS_VFS_H.
3863 18. Updated exipick to current release from John Jetmore.
3865 19. Allow an empty sender to be matched against a lookup in an address list.
3866 Previously the only cases considered were a regular expression, or an
3869 20. Exim went into a mad DNS lookup loop when doing a callout where the
3870 host was specified on the transport, if the DNS lookup yielded more than
3873 21. The RFC2047 encoding function was originally intended for short strings
3874 such as real names; it was not keeping to the 75-character limit for
3875 encoded words that the RFC imposes. It now respects the limit, and
3876 generates multiple encoded words if necessary. To be on the safe side, I
3877 have increased the buffer size for the ${rfc2047: expansion operator from
3880 22. Failure to deliver a bounce message always caused it to be frozen, even if
3881 there was an errors_to setting on the router. The errors_to setting is now
3884 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3885 canonical form (fully expanded) before being placed in
3886 $sender_host_address.
3888 24. Updated eximstats to version 1.33
3890 25. Include certificate and key file names in error message when GnuTLS fails
3891 to set them up, because the GnuTLS error message doesn't include the name
3892 of the failing file when there is a problem reading it.
3894 26. Expand error message when OpenSSL has problems setting up cert/key files.
3897 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3898 (this can affect the format of dates).
3900 28. exim_tidydb, when checking for the continued existence of a message for
3901 which it has found a message-specific retry record, was not finding
3902 messages that were in split spool directories. Consequently, it was
3903 deleting retry records that should have stayed in existence.
3905 29. eximstats updated to version 1.35
3906 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3907 1.35 - bugfix such that pie charts by volume are generated correctly
3909 30. The SPA authentication driver was not abandoning authentication and moving
3910 on to the next authenticator when an expansion was forced to fail,
3911 contradicting the general specification for all authenticators. Instead it
3912 was generating a temporary error. It now behaves as specified.
3914 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3915 (the order specifies the preference for clients). The order is now AES256,
3916 AES128, 3DES, ARCFOUR128.
3918 31. Small patch to Sieve code - explicitly set From: when generating an
3921 32. Exim crashed if a remote delivery caused a very long error message to be
3922 recorded - for instance if somebody sent an entire SpamAssassin report back
3923 as a large number of 550 error lines. This bug was coincidentally fixed by
3924 increasing the size of one of Exim's internal buffers (big_buffer) that
3925 happened as part of the Exiscan merge. However, to be on the safe side, I
3926 have made the code more robust (and fixed the comments that describe what
3929 33. Some experimental protocols are using DNS PTR records for new purposes. The
3930 keys for these records are domain names, not reversed IP addresses. The
3931 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3932 leaves it alone. Component reversal etc. now happens only for IP addresses.
3935 34. The host_aton() function is supposed to be passed a string that is known
3936 to be a valid IP address. However, in the case of IPv6 addresses, it was
3937 not checking this. This is a hostage to fortune. Exim now panics and dies
3938 if the condition is not met. A case was found where this could be provoked
3939 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3940 components; fortuitously, this particular loophole had already been fixed
3941 by change 4.50/55 or 4.44/33 above.
3943 If there are any other similar loopholes, the new check in host_aton()
3944 itself should stop them being exploited. The report I received stated that
3945 data on the command line could provoke the exploit when Exim was running as
3946 exim, but did not say which command line option was involved. All I could
3947 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3948 running as the user.
3951 35. There was a buffer overflow vulnerability in the SPA authentication code
3952 (which came originally from the Samba project). I have added a test to the
3953 spa_base64_to_bits() function which I hope fixes it.
3956 36. The daemon start-up calls getloadavg() while still root for those OS that
3957 need the first call to be done as root, but it missed one case: when
3958 deliver_queue_load_max is set with deliver_drop_privilege. This is
3959 necessary for the benefit of the queue runner, because there is no re-exec
3960 when deliver_drop_privilege is set.
3962 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3963 was in use, was not putting the data itself into the right store pool;
3964 consequently, it could be overwritten for a subsequent message in the same
3965 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3968 38. Sometimes the final signoff response after QUIT could fail to get
3969 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3970 < 0 before doing a fflush(). This bug looks as though it goes back to the
3971 introduction of TLS in release 3.20, but "sometimes" must have been rare
3972 because the tests only now provoked it.
3978 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3979 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3980 Now it returns an integer. A number of calls were still expecting a T/F
3981 return. Fortuitously, in all cases, the tests worked in OK situations,
3982 which is the norm. However, things would have gone wrong on any write
3983 failures on the smtp file descriptor. This function is used when sending
3984 messages over SMTP and also when doing verify callouts.
3986 2. When Exim is called to do synchronous delivery of a locally submitted
3987 message (the -odf or -odi options), it no longer closes stderr before doing
3990 3. Implemented the mua_wrapper option.
3992 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3994 5. Implemented the functions header_remove(), header_testname(),
3995 header_add_at_position(), and receive_remove_recipient(), and exported them
3998 6. If an ACL "warn" statement specified the addition of headers, Exim already
3999 inserted X-ACL-Warn: at the start if there was no header name. However, it
4000 was not making this test for the second and subsequent header lines if
4001 there were newlines in the string. This meant that an invalid header could
4002 be inserted if Exim was badly configured.
4004 7. Allow an ACL "warn" statement to add header lines at the start or after all
4005 the Received: headers, as well as at the end.
4007 8. Added the rcpt_4xx retry error code.
4009 9. Added postmaster_mailfrom=xxx to callout verification option.
4011 10. Added mailfrom=xxxx to the callout verification option, for verify=
4014 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4015 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4017 12. Inserted some casts to stop certain compilers warning when using pointer
4018 differences as field lengths or precisions in printf-type calls (mostly
4019 affecting debugging statements).
4021 13. Added optional readline() support for -be (dynamically loaded).
4023 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4024 same clock tick as a message's arrival, so that its received time was the
4025 same as the "first fail" time on the retry record, and that message
4026 remained on the queue past the ultimate address timeout, every queue runner
4027 would try a delivery (because it was past the ultimate address timeout) but
4028 after another failure, the ultimate address timeout, which should have then
4029 bounced the address, did not kick in. This was a "< instead of <=" error;
4030 in most cases the first failure would have been in the next clock tick
4031 after the received time, and all would be well.
4033 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4034 being recognized when the domain list was tested by the match_domain
4035 condition in an expansion string.
4037 16. Added the ${str2b64: operator.
4039 17. Exim was always calling setrlimit() to set a large limit for the number of
4040 processes, without checking whether the existing limit was already
4041 adequate. (It did check for the limit on file descriptors.) Furthermore,
4042 errors from getrlimit() and setrlimit() were being ignored. Now they are
4043 logged to the main and panic logs, but Exim does carry on, to try to do its
4044 job under whatever limits there are.
4046 18. Imported PCRE 5.0.
4048 19. Trivial typo in log message " temporarily refused connection" (the leading
4051 20. If the log selector return_path_on_delivery was set and an address was
4052 redirected to /dev/null, the delivery process crashed because it assumed
4053 that a return path would always be set for a "successful" delivery. In this
4054 case, the whole delivery is bypassed as an optimization, and therefore no
4057 21. Internal re-arrangement: the function for sending a challenge and reading
4058 a response while authentication was assuming a zero-terminated challenge
4059 string. It's now changed to take a pointer and a length, to allow for
4060 binary data in such strings.
4062 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4064 23. Exim was not respecting finduser_retries when seeking the login of the
4065 uid under which it was called; it was always trying 10 times. (The default
4066 setting of finduser_retries is zero.) Also, it was sleeping after the final
4067 failure, which is pointless.
4069 24. Implemented tls_on_connect_ports.
4071 25. Implemented acl_smtp_predata.
4073 26. If the domain in control=submission is set empty, Exim assumes that the
4074 authenticated id is a complete email address when it generates From: or
4075 Sender: header lines.
4077 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4078 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4079 chown and chgrp in /bin and hostname in /usr/bin.
4081 28. Exim was keeping the "process log" file open after each use, just as it
4082 does for the main log. This opens the possibility of it remaining open for
4083 long periods when the USR1 signal hits a daemon. Occasional processlog
4084 errors were reported, that could have been caused by this. Anyway, it seems
4085 much more sensible not to leave this file open at all, so that is what now
4088 29. The long-running daemon process does not normally write to the log once it
4089 has entered its main loop, and it closes the log before doing so. This is
4090 so that log files can straightforwardly be renamed and moved. However,
4091 there are a couple of unusual error situations where the daemon does write
4092 log entries, and I had neglected to close the log afterwards.
4094 30. The text of an SMTP error response that was received during a remote
4095 delivery was being truncated at 512 bytes. This is too short for some of
4096 the long messages that one sometimes sees. I've increased the limit to
4099 31. It is now possible to make retry rules that apply only when a message has a
4100 specific sender, in particular, an empty sender.
4102 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4103 it possible to be selective about when SMTP synchronization is enforced.
4105 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4107 32. Implemented hosts_connection_nolog.
4109 33. Added an ACL for QUIT.
4111 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4114 35. Added mailbox_size and mailbox_filecount to appendfile.
4116 36. Added control = no_multiline_responses to ACLs.
4118 37. There was a bug in the logic of the code that waits for the clock to tick
4119 in the case where the clock went backwards by a substantial amount such
4120 that the microsecond fraction of "now" was more than the microsecond
4121 fraction of "then" (but the whole seconds number was less).
4123 38. Added support for the libradius Radius client library this is found on
4124 FreeBSD (previously only the radiusclient library was supported).
4130 1. When certain lookups returned multiple values in the form name=value, the
4131 quoting of the values was not always being done properly. Specifically:
4132 (a) If the value started with a double quote, but contained no whitespace,
4134 (b) If the value contained whitespace other than a space character (i.e.
4135 tabs or newlines or carriage returns) it was not quoted.
4136 This fix has been applied to the mysql and pgsql lookups by writing a
4137 separate quoting function and calling it from the lookup code. The fix
4138 should probably also be applied to nisplus, ibase and oracle lookups, but
4139 since I cannot test any of those, I have not disturbed their existing code.
4141 2. A hit in the callout cache for a specific address caused a log line with no
4142 reason for rejecting RCPT. Now it says "Previous (cached) callout
4143 verification failure".
4145 3. There was an off-by-one bug in the queryprogram router. An over-long
4146 return line was truncated at 256 instead of 255 characters, thereby
4147 overflowing its buffer with the terminating zero. As well as fixing this, I
4148 have increased the buffer size to 1024 (and made a note to document this).
4150 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4151 when Exim is waiting for an SMTP response from a remote server, Exim
4152 restarts its select() call on the socket, thereby resetting its timeout.
4153 This is not a problem when such interrupts are rare. Somebody set up a cron
4154 job to run exiwhat every 2 minutes, which is less than the normal select()
4155 timeout (5 or 10 minutes). This meant that the select() timeout never
4156 kicked in because it was always reset. I have fixed this by comparing the
4157 time when an interrupt arrives with the time at the start of the first call
4158 to select(). If more time than the timeout has elapsed, the interrupt is
4159 treated as a timeout.
4161 5. Some internal re-factoring in preparation for the addition of Sieve
4162 extensions (by MH). In particular, the "personal" test is moved to a
4163 separate function, and given an option for scanning Cc: and Bcc: (which is
4164 not set for Exim filters).
4166 6. When Exim created an email address using the login of the caller as the
4167 local part (e.g. when creating a From: or Sender: header line), it was not
4168 quoting the local part when it contained special characters such as @.
4170 7. Installed new OpenBSD configuration files.
4172 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4173 try to make them clearer.
4175 9. Callout options, other than the timeout value, were being ignored when
4176 verifying sender addresses in header lines. For example, when using
4178 verify = header_sender/callout=no_cache
4180 the cache was (incorrectly) being used.
4182 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4183 only the Cygwin environment.
4185 11. When return_path_on_delivery was set as a log selector, if different remote
4186 addresses in the same message used different return paths and parallel
4187 remote delivery occurred, the wrong values would sometimes be logged.
4188 (Whenever a remote delivery process finished, the return path value from
4189 the most recently started remote delivery process was logged.)
4191 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4192 lines when AUTH and/or TLS are in use. This is the "received protocol"
4193 field. Exim used to use "asmtp" for authenticated SMTP, without any
4194 indication (in the protocol name) for TLS use. Now it follows the RFC and
4195 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4196 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4197 names appear in log lines as well as in Received: header lines.
4199 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4200 extensions, and comparison tests, and to fix some bugs.
4202 14. Changes to the "personal" filter test:
4204 (1) The test was buggy in that it was just doing the equivalent of
4205 "contains" tests on header lines. For example, if a user's address was
4206 anne@some.where, the "personal" test would incorrectly be true for
4208 To: susanne@some.where
4210 This test is now done by extracting each address from the header in turn,
4211 and checking the entire address. Other tests that are part of "personal"
4212 are now done using regular expressions (for example, to check local parts
4213 of addresses in From: header lines).
4215 (2) The list of non-personal local parts in From: addresses has been
4216 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4217 taken from the Sieve specification recommendations.
4219 (3) If the message contains any header line starting with "List-" it is
4220 treated as non-personal.
4222 (4) The test for "circular" in the Subject: header line has been removed
4223 because it now seems ill-conceived.
4225 15. Minor typos in src/EDITME comments corrected.
4227 16. Installed latest exipick from John Jetmore.
4229 17. If headers_add on a router specified a text string that was too long for
4230 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4231 of string_sprintf() is now avoided.
4233 18. $message_body_size was not set (it was always zero) when running the DATA
4234 ACL and the local_scan() function.
4236 19. For the "mail" command in an Exim filter, no default was being set for
4237 the once_repeat time, causing a random time value to be used if "once" was
4238 specified. (If the value happened to be <= 0, no repeat happened.) The
4239 default is now 0s, meaning "never repeat". The "vacation" command was OK
4240 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4241 (I found it when inspecting the code).
4243 20. There is now an overall timeout for performing a callout verification. It
4244 defaults to 4 times the callout timeout, which applies to individual SMTP
4245 commands during the callout. The overall timeout applies when there is more
4246 than one host that can be tried. The timeout is checked before trying the
4247 next host. This prevents very long delays if there are a large number of
4248 hosts and all are timing out (e.g. when the network connections are timing
4249 out). The value of the overall timeout can be changed by specifying an
4250 additional sub-option for "callout", called "maxwait". For example:
4252 verify = sender/callout=5s,maxwait=20s
4254 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4255 to the end before writing, but this should make it even safer).
4257 22. Exim was forgetting that it had advertised PIPELINING for the second and
4258 subsequent messages on an SMTP connection. It was also not resetting its
4259 memory on STARTTLS and an internal HELO.
4261 23. When Exim logs an SMTP synchronization error within a session, it now
4262 records whether PIPELINING has been advertised or not.
4264 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4265 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4266 rather than long int.
4268 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4270 26. Added the never_mail option to autoreply.
4276 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4277 crash if the getsockname() call failed; this can happen if a connection is
4278 closed very soon after it is established. The problem was simply in the
4279 order in which certain operations were done, causing Exim to try to write
4280 to the SMTP stream before it had set up the file descriptor. The bug has
4281 been fixed by making things happen in the correct order.
4287 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4288 before the connection was closed, thus losing the rejection response.
4290 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4291 some early Solaris releases, but causes trouble in current releases where
4292 socklen_t is defined.
4294 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4297 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4300 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4301 that releasing the top of it at the end releases what was used for sub-
4302 expansions (unless the block got too big). However, discard this block if
4303 the first thing is a variable or header, so that we can use its block when
4304 it is dynamic (useful for very large $message_headers, for example).
4306 6. Lookups now cache *every* query, not just the most recent. A new, separate
4307 store pool is used for this. It can be recovered when all lookup caches are
4308 flushed. Lookups now release memory at the end of their result strings.
4309 This has involved some general refactoring of the lookup sources.
4311 7. Some code has been added to the store_xxx() functions to reduce the amount
4312 of flapping under certain conditions.
4314 8. log_incoming_interface used to affect only the <= reception log lines. Now
4315 it causes the local interface and port to be added to several more SMTP log
4316 lines, for example "SMTP connection from", and rejection lines.
4318 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4320 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4322 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4323 could be overwritten at the end of the current message (or the start of a
4324 new message if it was set in a HELO ACL). The value is now preserved for
4325 the duration of the SMTP connection.
4327 12. If a transport had a headers_rewrite setting, and a matching header line
4328 contained an unqualified address, that address was qualified, even if it
4329 did not match any rewriting rules. The underlying bug was that the values
4330 of the flags that permit the existence of unqualified sender and recipient
4331 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4332 non-local messages, and by -bnq for local messages) were not being
4333 preserved with the message after it was received.
4335 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4336 "next input=" as part of the text comprising the host identity instead of
4337 the correct text. The code was using the same buffer for two different
4338 strings. However, depending on which order the printing function evaluated
4339 its arguments, the bug did not always show up. Under Linux, for example, my
4340 test suite worked just fine.
4342 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4343 doesn't work with some older versions of Perl. It has been changed to "my",
4344 which in any case is probably the better facility to use.
4346 15. A really picky compiler found some instances of statements for creating
4347 error messages that either had too many or two few arguments for the format
4350 16. The size of the buffer for calls to the DNS resolver has been increased
4351 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4352 for addresses that have a lot of PTR records. This alleviates a problem; it
4353 does not fully solve it.
4355 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4356 buffer now truncates the list and logs the incident, which is the same
4357 action as happens when Exim is looking up a host name and its aliases.
4358 Previously in this situation something unpredictable would happen;
4359 sometimes it was "internal error: store_reset failed".
4361 18. If a server dropped the connection unexpectedly when an Exim client was
4362 using GnuTLS and trying to read a response, the client delivery process
4363 crashed while trying to generate an error log message.
4365 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4366 string, for example:
4368 warn message = H1: something\nH2: something
4370 the text was added as a single header line from Exim's point of view
4371 though it ended up OK in the delivered message. However, searching for the
4372 second and subsequent header lines using $h_h2: did not work. This has been
4373 fixed. Similarly, if a system filter added multiple headers in this way,
4374 the routers could not see them.
4376 20. Expanded the error message when iplsearch is called with an invalid key to
4377 suggest using net-iplsearch in a host list.
4379 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4380 ACLs are no longer actually imposed (and a message to that effect is
4383 22. If a "gecos" field in a passwd entry contained escaped characters, in
4384 particular, if it contained a \" sequence, Exim got it wrong when building
4385 a From: or a Sender: header from that name. A second bug also caused
4386 incorrect handling when an unquoted " was present following a character
4387 that needed quoting.
4389 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4390 was not being matched caselessly.
4392 24. Arranged for all hyphens in the exim.8 source to be escaped with
4395 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4396 properly. Recipient callout cache records were still being keyed to include
4397 the sender, even when use_sender was set false. This led to far more
4398 callouts that were necessary. The sender is no longer included in the key
4399 when use_sender is false.
4401 26. Added "control = submission" modifier to ACLs.
4403 27. Added the ${base62d: operator to decode base 62 numbers.
4405 28. dnsdb lookups can now access SRV records.
4407 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4408 the configuration file.
4410 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4411 (-v) mode. This makes the output for a verbose queue run more intelligible.
4413 31. Added a use_postmaster feature to recipient callouts.
4415 32. Added the $body_zerocount variable, containing the number of binary zero
4416 bytes in the message body.
4418 33. The time of last modification of the "new" subdirectory is now used as the
4419 "mailbox time last read" when there is a quota error for a maildir
4422 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4424 35. Added +ignore_unknown as a special item in host lists.
4426 36. Code for decoding IPv6 addresses in host lists is now included, even if
4427 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4428 address was recognized as an IP address, but was then not correctly decoded
4429 into binary, causing unexpected and incorrect effects when compared with
4436 1. Very minor rewording of debugging text in manualroute to say "list of
4437 hosts" instead of "hostlist".
4439 2. If verify=header_syntax was set, and a header line with an unqualified
4440 address (no domain) and a large number of spaces between the end of the
4441 name and the colon was received, the reception process suffered a buffer
4442 overflow, and (when I tested it) crashed. This was caused by some obsolete
4443 code that should have been removed. The fix is to remove it!
4445 3. When running in the test harness, delay a bit after writing a bounce
4446 message to get a bit more predictability in the log output.
4448 4. Added a call to search_tidyup() just before forking a reception process. In
4449 theory, someone could use a lookup in the expansion of smtp_accept_max_
4450 per_host which, without the tidyup, could leave open a database connection.
4452 5. Added the variables $recipient_data and $sender_data which get set from a
4453 lookup success in an ACL "recipients" or "senders" condition, or a router
4454 "senders" option, similar to $domain_data and $local_part_data.
4456 6. Moved the writing of debug_print from before to after the "senders" test
4459 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4460 problems for message scanning, either using a data ACL, or using
4461 local_scan() because the Received: header was not generated till after they
4462 were called (in order to set the time as the time of reception completion).
4463 I have revised the way this works. The header is now generated after the
4464 body is received, but before the ACL or local_scan() are called. After they
4465 are run, the timestamp in the header is updated.
4471 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4472 before starting a queue runner without re-exec. This happened only when
4473 deliver_drop_privilege was set or when the Exim user was set to root. The
4474 effect of the bug was that timeouts during subsequent deliveries caused
4475 crashes instead of being properly handled. The handler is now left at its
4476 default (and expected) setting.
4478 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4479 message, again when deliver_drop_privilege is set or Exim is run as root.
4480 The bug described in (1) was not present in this case, but the tidying up
4481 of the other signals was missing. I have made the two cases consistent.
4483 3. The ignore_target_hosts setting on a manualroute router was being ignored
4484 for hosts that were looked up using the /MX notation.
4486 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4489 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4490 operated on the sender address. After changing the $sender_address to <>
4491 for the sender address verify, Exim was re-instated it as the original
4492 (before rewriting) address, but remembering that it had rewritten it, so it
4493 wasn't rewriting it again. This bug also had the effect of breaking the
4494 sender address verification caching when the sender address was rewritten.
4496 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4497 This has been changed so that if the ip literal address matches
4498 ignore_target_hosts, the router declines.
4500 7. Added expansion conditions match_domain, match_address, and match_local_
4501 part (NOT match_host).
4503 8. The placeholder for the Received: header didn't have a length field set.
4505 9. Added code to Exim itself and to exim_lock to test for a specific race
4506 condition that could lead to file corruption when using MBX delivery. The
4507 issue is with the lockfile that is created in /tmp. If this file is removed
4508 after a process has opened it but before that process has acquired a lock,
4509 there is the potential for a second process to recreate the file and also
4510 acquire a lock. This could lead to two Exim processes writing to the file
4511 at the same time. The added code performs the same test as UW imapd; it
4512 checks after acquiring the lock that its file descriptor still refers to
4513 the same named file.
4515 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4516 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4519 11. Added the smtp_active_hostname option. If used, this will typically be made
4520 to depend on the incoming interface address. Because $interface_address is
4521 not set up until the daemon has forked a reception process, error responses
4522 that can happen earlier (such as "too many connections") no longer contain
4525 12. If an expansion in a condition on a "warn" statement fails because a lookup
4526 defers, the "warn" statement is abandoned, and the next ACL statement is
4527 processed. Previously this caused the whole ACL to be aborted.
4529 13. Added the iplsearch lookup type.
4531 14. Added ident_timeout as a log selector.
4533 15. Added tls_certificate_verified as a log selector.
4535 16. Added a global option tls_require_ciphers (compare the smtp transport
4536 option of the same name). This controls incoming TLS connections.
4538 17. I finally figured out how to make tls_require_ciphers do a similar thing
4539 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4540 before starting the TLS session.
4542 18. Tabs are now shown as \t in -bP output.
4544 19. If the log selector return_path_on_delivery was set, Exim crashed when
4545 bouncing a message because it had too many Received: header lines.
4547 20. If two routers both had headers_remove settings, and the first one included
4548 a superfluous trailing colon, the final name in the first list and the
4549 first name in the second list were incorrectly joined into one item (with a
4550 colon in the middle).
4556 1. Added -C and -D options to the exinext utility, mainly to make it easier
4557 to include in the automated testing, but these could be helpful when
4558 multiple configurations are in use.
4560 2. The exinext utility was not formatting the output nicely when there was
4561 an alternate port involved in the retry record key, nor when there was a
4562 message id as well (for retries that were specific to a specific message
4563 and a specific host). It was also confused by IPv6 addresses, because of
4564 the additional colons they contain. I have fixed the IPv4 problem, and
4565 patched it up to do a reasonable job for IPv6.
4567 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4568 delivery, the log line now contains "pipelined" if PIPELINING was used.
4570 4. An SMTP transport process used to panic and die if the bind() call to set
4571 an explicit outgoing interface failed. This has been changed; it is now
4572 treated in the same way as a connect() failure.
4574 5. A reference to $sender_host_name in the part of a conditional expansion
4575 that was being skipped was still causing a DNS lookup. This no longer
4578 6. The def: expansion condition was not recognizing references to header lines
4579 that used bh_ and bheader_.
4581 7. Added the _cache feature to named lists.
4583 8. The code for checking quota_filecount in the appendfile transport was
4584 allowing one more file than it should have been.
4586 9. For compatibility with Sendmail, the command line option
4594 and sets the incoming protocol and host name (for trusted callers). The
4595 host name and its colon can be omitted when only the protocol is to be set.
4596 Note the Exim already has two private options, -pd and -ps, that refer to
4597 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4598 "s", but I don't think that's a major issue.
4600 10. A number of refactoring changes to the code, none of which should affect
4603 (a) The number of logging options was getting close to filling up the
4604 32-bit word that was used as a bit map. I have split them into two classes:
4605 those that are passed in the argument to log_write(), and those that are
4606 only ever tested independently outside of that function. These are now in
4607 separate 32-bit words, so there is plenty of room for expansion again.
4608 There is no change in the user interface or the logging behaviour.
4610 (b) When building, for example, log lines, the code previously used a
4611 macro that called string_cat() twice, in order to add two strings. This is
4612 not really sufficiently general. Furthermore, there was one instance where
4613 it was actually wrong because one of the argument was used twice, and in
4614 one call a function was used. (As it happened, calling the function twice
4615 did not affect the overall behaviour.) The macro has been replaced by a
4616 function that can join an arbitrary number of extra strings onto a growing
4619 (c) The code for expansion conditions now uses a table and a binary chop
4620 instead of a serial search (which was left over from when there were very
4621 few conditions). Also, it now recognizes conditions like "pam" even when
4622 the relevant support is not compiled in: a suitably worded error message is
4623 given if an attempt is made to use such a condition.
4625 11. Added ${time_interval:xxxxx}.
4627 12. A bug was causing one of the ddress fields not to be passed back correctly
4628 from remote delivery subprocesses. The field in question was not being
4629 subsequently used, so this caused to problems in practice.
4631 13. Added new log selectors queue_time and deliver_time.
4633 14. Might have fixed a bug in maildirsizefile handling that threw up
4634 "unexpected character" debug warnings, and recalculated the data
4635 unnecessarily. In any case, I expanded the warning message to give more
4638 15. Added the message "Restricted characters in address" to the statements in
4639 the default ACL that block characters like @ and % in local parts.
4641 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4642 Three changes have been made:
4644 (a) There was a serious bug; a negative response to MAIL caused the whole
4645 recipient domain to be cached as invalid, thereby blocking all messages
4646 to all local parts at the same domain, from all senders. This bug has
4647 been fixed. The domain is no longer cached after a negative response to
4648 MAIL if the sender used is not empty.
4650 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4653 (c) A new callout option, "use_sender" has been added for people who want
4654 the modified behaviour.
4660 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4663 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4664 indeed breaks things for older releases.
4666 3. Added additional logging to the case where there is a problem reading data
4667 from a filter that is running in a subprocess using a pipe, in order to
4668 try to track down a specific problem.
4670 4. Testing facility fudge: when running in the test harness and attempting
4671 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4672 getting "No route to host". Convert this to a timeout.
4674 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4677 6. Some OS don't have socklen_t but use size_t instead. This affects the
4678 fifth argument of getsockopt() amongst other things. This is now
4679 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4680 can be set for individual OS. I have set it for SunOS5, OSF1, and
4681 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4682 some earlier ones do not.
4684 7. Change 4.30/15 was not doing the test caselessly.
4686 8. The standard form for an IPv6 address literal was being rejected by address
4687 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4688 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4689 this, as well as the form without the "IPv6" on the front (but only when
4690 address literals are enabled, of course).
4692 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4694 10. Exim crashed if a message with an empty sender address specified by -f
4695 encountered a router with an errors_to setting. This could be provoked only
4696 by a command such as
4700 where an empty string was supplied; "<>" did not hit this bug.
4702 11. Installed PCRE release 4.5.
4704 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4705 remained set. It is now erased.
4707 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4708 times from message ids (which are base 36 rather than the normal 62).
4710 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4711 were being counted as actual protocol errors, and logged if the log
4712 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4713 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4714 and DATA following a set of rejected RCPTs do not count as protocol errors.
4715 In other words, Exim assumes they were pipelined, though this may not
4716 actually be the case. Of course, in all cases the client gets an
4717 appropriate error code.
4719 15. If a lookup fails in an ACL condition, a message about the failure may
4720 be available; it is used if testing the ACL cannot continue, because most
4721 such messages specify what the cause of the deferral is. However, some
4722 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4723 that caused an old message to be retained and used if a later statement
4724 caused a defer, replacing the real cause of the deferral.
4726 16. If an IP address had so many PTR records that the DNS lookup buffer
4727 was not large enough to hold them, Exim could crash while trying to process
4728 the truncated data. It now detects and logs this case.
4730 17. Further to 4.21/58, another change has been made: if (and only if) the
4731 first line of a message (the first header line) ends with CRLF, a bare LF
4732 in a subsequent header line has a space inserted after it, so as not to
4733 terminate the header.
4735 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4736 unnecessarily, used atoi() instead of strtol(), and didn't check the
4737 termination when getting file sizes from file names by regex.
4739 19. Completely re-implemented the support for maildirsize files, in the light
4740 of a number of problems with the previous contributed implementation
4741 (4.30/29). In particular:
4743 . If the quota is zero, the maildirsize file is maintained, but no quota is
4746 . If the maildir directory does not exist, it is created before any attempt
4747 to write a maildirsize file.
4749 . The quota value in the file is just a cache; if the quota is changed in
4750 the transport, the new value overrides.
4752 . A regular expression is available for excluding directories from the
4755 20. The autoreply transport checks the characters in options that define the
4756 message's headers; it allows continued headers, but it was checking with
4757 isspace() after an embedded newline instead of explicitly looking for a
4760 21. If all the "regular" hosts to which an address was routed had passed their
4761 expiry times, and had not reached their retry times, the address was
4762 bounced, even if fallback hosts were defined. Now Exim should go on to try
4765 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4766 equivalent code in the SMTP transport. Some hosts send humungous responses
4767 to HELO/EHLO, more than 1024 it seems.
4769 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4770 gives compiler warnings in some environments. I've now done it "properly",
4773 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4774 (because of problems with the built-in one) was declared to return uschar *
4775 instead of char *, causing compiler failure.
4777 25. Fixed a file descriptor leak when processing alias/forward files.
4779 26. Fixed a minor format string issue in dbfn.c.
4781 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4783 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4784 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4787 29. When a sender address is verified, it is cached, to save repeating the test
4788 when there is more than one recipient in a message. However, when the
4789 verification involves a callout, it is possible for different callout
4790 options to be set for different recipients. It is too complicated to keep
4791 track of this in the cache, so now Exim always runs a verification when a
4792 callout is required, relying on the callout cache for the optimization.
4793 The overhead is duplication of the address routing, but this should not be
4796 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4797 to be verified with callout=postmaster, and the main callout worked but the
4798 postmaster check failed, the verification correctly failed. However, if a
4799 subsequent RCPT command asked for sender verification *without* the
4800 postmaster check, incorrect caching caused this verification also to fail,
4803 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4804 it was not caching the DNS options (qualify_single, search_parents) that
4805 were used when the lookup failed. A subsequent lookup with different
4806 options therefore always gave the same answer, though there were cases
4807 where it should not have. (Example: a "domains = !$mx_any" option on a
4808 dnslookup router: the "domains" option is always processed without any
4809 widening, but the router might have qualify_single set.) Now Exim uses the
4810 cached value only when the same options are set.
4812 32. Added John Jetmore's "exipick" utility to the distribution.
4814 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4815 than a timeout (e.g. a certificate is required, and is not provided), an
4816 Exim server now closes the connection immediately. Previously it waited for
4817 the client to close - but if the client is SSL, it seems that they each
4818 wait for each other, leading to a delay before one of them times out.
4820 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4821 maintained 0.8.x compatibility because I don't think many are using it, and
4822 it is clearly obsolete.
4824 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4827 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4828 client certificate was expired. A simple patch fixes this, though I don't
4829 understand the full logic of why the verify callback is called multiple
4832 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4833 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4834 which causes problems with some clients (such as the Certicom SSL Plus
4835 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4836 disables the coutermeasure allowing Eudora to connect."
4838 38. Exim was not checking that a write() to a log file succeeded. This could
4839 lead to Bad Things if a log got too big, in particular if it hit a file
4840 size limit. Exim now panics and dies if it cannot write to a log file, just
4841 as it does if it cannot open a log file.
4843 39. Modified OS/Makefile-Linux so that it now contains
4845 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4847 The two -D definitions ensure that Exim is compiled with large file
4848 support, which makes it possible to handle log files that are bigger than
4851 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4852 instance) a domain was checked against a named list that involved a lookup,
4853 causing $domain_data to be set, then another domain was checked against the
4854 same list, then the first domain was re-checked, the value of $domain_data
4855 after the final check could be wrong. In particular, if the second check
4856 failed, it could be set empty. This bug probably also applied to
4859 41. The strip_trailing_dot option was not being applied to the address given
4860 with the -f command-line option.
4862 42. The code for reading a message's header from the spool was incrementing
4863 $received_count, but never initializing it. This meant that the value was
4864 incorrect (doubled) while delivering a message in the same process in which
4865 it was received. In the most common configuration of Exim, this never
4866 happens - a fresh exec is done - but it can happen when
4867 deliver_drop_privilege is set.
4869 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4870 it now includes up to 150 characters of the unexpected data in the log
4873 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4874 and building data strings. The size of both of these buffers was 10 000
4875 bytes - far larger than anybody would *ever* want, thought I. Needless to
4876 say, somebody hit the limit. I have increased the maximum line length to
4877 20 000 and the maximum data length of concatenated lines to 100 000. I have
4878 also fixed two bugs, because there was no checking on these buffers. Tsk,
4879 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4880 buffer is too small.
4882 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4883 lsearch lookups. Now it does.
4885 46. When parsing a route_list item in a manualroute router, a fixed-length
4886 buffer was used for the list of hosts. I made this 1024 bytes long,
4887 thinking that nobody would ever have a list of hosts that long. Wrong.
4888 Somebody had a whole pile of complicated expansion conditions, and the
4889 string was silently truncated, leading to an expansion error. It turns out
4890 that it is easier to change to an unlimited length (owing to other changes
4891 that have happened since this code was originally written) than to build
4892 structure for giving a limitation error. The length of the item that
4893 expands into the list of hosts is now unlimited.
4895 47. The lsearch lookup could not handle data where the length of text line was
4896 more than 4095 characters. Such lines were truncated, leading to shortened
4897 data being returned. It should now handle lines of any length.
4899 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4900 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4903 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4904 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4905 no longer happens in comment lines. A list of replacements is now placed
4906 at the head of all of the source files, except those whose only change is
4907 to replace PERL_COMMAND in the very first #! line.
4909 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4910 messages on the queue, with a bottom-up merge sort, using code contributed
4911 by Michael Haardt. This should make operations like -bp somewhat faster on
4912 large queues. It won't affect queue runners, except when queue_run_in_order
4915 51. Installed eximstats 1.31 in the distribution.
4917 52. Added support for SRV lookups to the dnslookup router.
4919 53. If an ACL referred to $message_body or $message_body_end, the value was not
4920 reset for any messages that followed in the same SMTP session.
4922 54. The store-handling optimization for building very long strings was not
4923 differentiating between the different store pools. I don't think this
4924 actually made any difference in practice, but I've tidied it.
4926 55. While running the routers to verify a sender address, $sender_address
4927 was still set to the sender address. This is wrong, because when routing to
4928 send a bounce to the sender, it would be empty. Therefore, I have changed
4929 it so that, while verifying a sender address, $sender_address is set to <>.
4930 (There is no change to what happens when verifying a recipient address.)
4932 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4933 target A or AAAA records (if not already returned) without resetting the
4934 qualify_single or search_parents options of the DNS resolver. These are
4935 inappropriate in this case because the targets of MX and SRV records must
4936 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4937 target that, when qualified, matched something in the local domain. These
4938 two options are now turned off when doing these lookups.
4940 57. It seems that at least some releases of Reiserfs (which does not have the
4941 concept of a fixed number of inodes) returns zero and not -1 for the
4942 number of available inodes. This interacted badly with check_spool_inodes,
4943 which assumed that -1 was the "no such thing" setting. What I have done is
4944 to check that the total number of inodes is greater than zero before doing
4945 the test of how many are available.
4947 58. When a "warn" ACL statement has a log_message modifier, the message is
4948 remembered, and not repeated. This is to avoid a lot of repetition when a
4949 message has many recipients that cause the same warning to be written.
4950 Howewer, Exim was preserving the list of already written lines for an
4951 entire SMTP session, which doesn't seem right. The memory is now reset if a
4952 new message is started.
4954 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4955 debugging output unless log_rewrite was also set.
4957 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4958 of (int)(handle) when we know that handle contains (void *)(-1).
4960 61. The Exim daemon panic-logs an error return when it closes the incoming
4961 connection. However "connection reset by peer" seems to be common, and
4962 isn't really an error worthy of noting specially, so that particular error
4965 62. When Exim is trying to find all the local interfaces, it used to panic and
4966 die if the ioctl to get the interface flags failed. However, it seems that
4967 on at least one OS (Solaris 9) it is possible to have an interface that is
4968 included in the list of interfaces, but for which you get a failure error
4969 for this call. This happens when the interface is not "plumbed" into a
4970 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4971 failure of the "get flags" call assumes that the interface is down.
4973 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4974 makes life easier for people who are doing arithmetic on fields extracted
4975 from dates, where you often get leading zeros that should not be
4976 interpreted as octal.
4978 64. Added qualify_domain to the redirect router, to override the global
4981 65. If a pathologically long header line contained very many addresses (the
4982 report of this problem mentioned 10 000) and each of them was rewritten,
4983 Exim could use up a very large amount of memory. (It kept on making new
4984 copies of the header line as it rewrote, and never released the old ones.)
4985 At the expense of a bit more processing, the header rewriting function has
4986 been changed so that it no longer eats memory in this way.
4988 66. The generation of the Received: header has been moved from the time that a
4989 message starts to be received, to the time that it finishes. The timestamp
4990 in the Received: header should now be very close to that of the <= log
4991 line. There are two side-effects of this change:
4993 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4994 logged header lines no longer include the local Received: line, because
4995 it has not yet been created. The same applies to a copy of the message
4996 that is returned to a non-SMTP sender when a message is rejected.
4998 (b) When a filter file is tested using -bf, no additional Received: header
4999 is added to the test message. After some thought, I decided that this
5002 This change does not affect the value of $received_for. It is still set
5003 after address rewriting, but before local_scan() is called.
5005 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5007 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5008 gave an unhelpful panic error message, and a defer error. I have managed to
5009 change this behaviour so that it now rejects any supplied certificate,
5010 which seems right, as the list of acceptable certificates is empty.
5012 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5013 gave an unhelpful defer error. I have not managed to make this reject any
5014 supplied certificates, but the error message it gives is "no certificate
5015 supplied", which is not helpful.
5017 70. exigrep's output now also includes lines that are not associated with any
5018 message, but which match the given pattern. Implemented by a patch from
5019 Martin Sluka, which also tidied up the Perl a bit.
5021 71. Recipient callout verification, like sender verification, was using <> in
5022 the MAIL FROM command. This isn't really the right thing, since the actual
5023 sender may affect whether the remote host accepts the recipient or not. I
5024 have changed it to use the actual sender in the callout; this means that
5025 the cache record is now keyed on a recipient/sender pair, not just the
5026 recipient address. There doesn't seem to be a real danger of callout loops,
5027 since a callout by the remote host to check the sender would use <>.
5028 [SEE ABOVE: changed after hitting problems.]
5030 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5031 temporary errors. However, in the case of such a code being given after
5032 the end of a data transmission (i.e. after ".") Exim was failing to write
5033 a retry record for the message. (Yes, there was some broken host that was
5034 actually sending 8xx at this point.)
5036 73. An unknown lookup type in a host list could cause Exim to panic-die when
5037 the list was checked. (An example that provoked this was putting <; in the
5038 middle of a list instead of at the start.) If this happened during a DATA
5039 ACL check, a -D file could be left lying around. This kind of configuration
5040 error no longer causes Exim to die; instead it causes a defer errror. The
5041 incident is still logged to the main and panic logs.
5043 74. Buglet left over from Exim 3 conversion. The message "too many messages
5044 in one connection" was written to the rejectlog but not the mainlog, except
5045 when address rewriting (yes!) was being logged.
5047 75. Added write_rejectlog option.
5049 76. When a system filter was run not as root (that is, when system_filter_user
5050 was set), the values of the $n variables were not being returned to the
5051 main process; thus, they were not subsequently available in the $sn
5054 77. Added +return_path_on_delivery log selector.
5056 78. A connection timeout was being treated differently from recipients deferred
5057 when testing hosts_max_try with a message that was older than the host's
5058 retry timeout. (The host should not be counted, thus allowing all hosts to
5059 be tried at least once before bouncing.) This may have been the cause of an
5060 occasionally reported bug whereby a message would remain on the queue
5061 longer than the retry timeout, but would be bounced if a delivery was
5062 forced. I say "may" because I never totally pinned down the problem;
5063 setting up timeout/retry tests is difficult. See also the next item.
5065 79. The ultimate address timeout was not being applied to errors that involved
5066 a combination of host plus message (for example, a timeout on a MAIL
5067 command). When an address resolved to a number of possible hosts, and they
5068 were not all tried for each delivery (e.g. because of hosts_max_try), a
5069 message could remain on the queue longer than the retry timeout.
5071 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5074 81. Fixed an obscure SMTP outgoing bug which required at least the following
5075 conditions: (a) there was another message waiting for the same server;
5076 (b) the server returned 5xx to all RCPT commands in the first message so
5077 that the message was not completed; (c) the server dropped the connection
5078 or gave a negative response to the RSET that Exim sends to abort the
5079 transaction. The observed case was a dropped connection after DATA that had
5080 been sent in pipelining mode. That is, the server had advertised PIPELINING
5081 but was not implementing it correctly. The effect of the bug was incorrect
5082 behaviour, such as trying another host, and this could lead to a crash.
5088 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5089 and daemon.c were passed as pointers to ints; they should have been
5090 pointers to socklen_t variables (which are typically unsigned ints).
5092 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5095 3. Fixed a really odd bug that affected only the testing scheme; patching a
5096 certain fixed string in the binary changed the value of another string that
5097 happened to be identical to the end of the original first string.
5099 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5100 name", it returns that address as the IP address. On some operating
5101 systems (e.g. Solaris), it also passes back the IP address string as the
5102 "host name". However, on others (e.g. Linux), it passes back an empty
5103 string. Exim wasn't checking for this, and was changing the host name to an
5104 empty string, assuming it had been canonicized.
5106 5. Although rare, it is permitted to have more than one PTR record for a given
5107 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5108 all the names associated with an address, because they do in Solaris.
5109 However, it seems that they do not in Linux for data that comes from the
5110 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5111 I found this out when I moved to a new Linux workstation and tried to run
5112 the Exim test suite.
5114 To get round this problem I have changed the code so that it now does its
5115 own call to the DNS to look up PTR records when searching for a host name.
5116 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5117 addresses that are only in /etc/hosts are still found.
5119 This behaviour is, however, controlled by an option called host_lookup_
5120 order, which defaults to "bydns:byaddr". If people want to use the other
5121 order, or indeed, just use one or the other means of lookup, they can
5122 specify it in this variable.
5124 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5125 some operating systems, this comes back from gethostbyaddr() as an empty
5126 string, and this is what Exim used to test for. However, it seems that in
5127 other systems, "." is yielded. Exim now tests for this case too.
5129 7. The values of check_spool_space and check_log_space are now held internally
5130 as a number of kilobytes instead of an absolute number of bytes. If a
5131 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5132 kilobyte. This means that much larger values can be stored.
5134 8. Exim monitor: an attempt to get the action menu when not actually pointing
5135 at a message produces an empty menu entitled "No message selected". This
5136 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5137 no entries in it ("Shell widget menu has zero width and/or height"). So I
5138 have added a single, blank menu entry in this case.
5140 9. Added ${quote_local_part.
5142 10. MIME decoding is now applied to the contents of Subject: header lines when
5145 11. Now that a reference to $sender_host_address automatically causes a reverse
5146 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5147 host lookup before query-style lookups in lists that might use this
5148 variable. This has therefore been abolished, and the "net-" prefix is no
5149 longer necessary for query-style lookups.
5151 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5152 have been a typo for LFLAGS, so it has been changed.
5154 13. The install script calls Exim with "-C /dev/null" in order to find the
5155 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5156 to be output. Howeve, since Exim outputs its version number before the
5157 error, it didn't break the script. It just looked ugly. I fixed this by
5158 always allowing "-C /dev/null" if the caller is root.
5160 14. Ignore overlarge ACL variable number when reading spool file - insurance
5161 against a later release with more variables having written the file.
5163 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5164 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5165 this, as well as the form without the "IPv6" on the front.
5167 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5168 OS/Makefile-Darwin file.
5170 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5171 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5173 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5174 variable. However, some OS use ints for inodes. I've added cast to long int
5175 to get rid of the compiler warning.
5177 19. I had forgotten to lock out "/../" in configuration file names when
5178 ALT_CONFIG_PREFIX was set.
5180 20. Routers used for verification do not need to specify transports. However,
5181 if such a router generated a host list, and callout was configured, Exim
5182 crashed, because it could not find a port number from the (non-existent)
5183 transport. It now assumes port 25 in this circumstance.
5185 21. Added the -t option to exigrep.
5187 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5188 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5189 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5190 them is set, LOOKUP_LSEARCH is forced.
5192 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5194 24. Added sender and host information to the "rejected by local_scan()" log
5195 line; previously there was no indication of these.
5197 25. Added .include_if_exists.
5199 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5200 Linux. It turns out that not all file systems support this. Apparently some
5201 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5202 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5203 not supported on the file descriptor, is now ignored when Exim is trying to
5204 sync a directory. This applies only to Linux.
5206 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5208 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5209 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5212 29. Added support for maildirsize files from supplied patch (modified a bit).
5214 30. The use of :fail: followed by an empty string could lead Exim to respond to
5215 sender verification failures with (e.g.):
5217 550 Verification failed for <xxx>
5218 550 Sender verify failed
5220 where the first response line was missing the '-' that indicates it is not
5221 the final line of the response.
5223 31. The loop for finding the name of the user that called Exim had a hardwired
5224 limit of 10; it now uses the value of finduser_retries, which is used for
5225 all other user lookups.
5227 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5230 33. Exim was neglecting to zero errno before one call of strtol() when
5231 expanding a string and expecting an integer value. On some systems this
5232 resulted in spurious "integer overflow" errors. Also, it was casting the
5233 result into an int without checking.
5235 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5236 did not work. The code looks as if it has *never* worked, though it appears
5237 to have been documented since at least releast 1.62. I have made it work.
5239 35. The "timeout_DNS" error in retry rules, also documented since at least
5240 1.62, also never worked. As it isn't clear exactly what this means, and
5241 clearly it isn't a major issue, I have abolished the feature by treating it
5242 as "timeout", and writing a warning to the main and panic logs.
5244 36. The display of retry rules for -brt wasn't always showing the error code
5247 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5248 timeout_connect_A, timeout_connect_MX.
5250 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5251 to the empty sender.
5253 39. The daemon was not analysing the content of -oX till after it had closed
5254 stderr and disconnected from the controlling terminal. This meant that any
5255 syntax errors were only noted on the panic log, and the return code from
5256 the command was 0. By re-arranging the code a little, I've made the
5257 decoding happen first, so such errors now appear on stderr, and the return
5258 code is 1. However, the actual setting up of the sockets still happens in
5259 the disconnected process, so errors there are still only recorded on the
5262 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5263 connections (as happens on some IP stacks) was logged at start up time as
5264 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5265 it from "IPv6 and IPv4", which means that two separate sockets are being
5268 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5269 says whether AF_INET or AF_INET6 was passed as an argument.
5271 42. Exiwhat output was messed up when time zones were included in log
5274 43. Exiwhat now gives more information about the daemon's listening ports,
5275 and whether -tls-on-connect was used.
5277 44. The "port" option of the smtp transport is now expanded.
5279 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5280 silently ignored. Now an error message is written to the main and panic
5283 46. There's a new ACL modifier called "logwrite" which writes to a log file
5284 as soon as it is encountered.
5286 47. Added $local_user_uid and $local_user_gid at routing time.
5288 48. Exim crashed when trying to verify a sender address that was being
5291 49. Exim was recognizing only a space character after ".include". It now also
5292 recognizes a tab character.
5294 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5295 extracting the relevant information from the specification. The man page no
5296 longer contains scrambled data for the -d option, and I've added a section
5297 at the front about calling Exim under different names.
5299 51. Added "extra_headers" argument to the "mail" command in filter files.
5301 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5304 53. Installed eximstats 1.29.
5306 54. Added transport_filter_timeout as a generic transport option.
5308 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5309 Cc: header lines. This was required by RFC 822, but it not required by RFC
5312 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5313 incoming messages that did not have them. Now it does so only if the
5314 message originates locally, that is, if there is no associated remote host
5315 address. When Resent- header lines are present, this applies to the Resent-
5316 lines rather than the non-Resent- lines.
5318 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5319 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5321 58. Messages for configuration errors now include the name of the main
5322 configuration file - useful now that there may be more than one file in a
5323 list (.included file names were always shown).
5325 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5326 for those rare installations that do not start the daemon as root or run it
5327 setuid root. I've cut out the call to initgroups() if the daemon is not
5330 60. The Exim user and group can now be bound into the binary as text strings
5331 that are looked up at the start of Exim's processing.
5333 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5335 62. Added $mailstore_basename variable.
5337 63. Installed patch to sieve.c from Michael Haardt.
5339 64. When Exim failed to open the panic log after failing to open the main log,
5340 the original message it was trying to log was written to stderr and debug
5341 output, but if they were not available (the usual case in production), it
5342 was lost. Now it is written to syslog before the two lines that record the
5343 failures to open the logs.
5345 65. Users' Exim filters run in subprocesses under the user's uid. It is
5346 possible for a "deliver" command or an alias in a "personal" command to
5347 provoke an address rewrite. If logging of address rewriting is configured,
5348 this fails because the process is not running as root or exim. There may be
5349 a better way of dealing with this, but for the moment (because 4.30 needs
5350 to be released), I have disabled address rewrite logging when running a
5351 filter in a non-root, non-exim process.
5357 1. The buildconfig auxiliary program wasn't quoting the value set for
5358 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5359 not defined. This bug was masked in 4.22 by the effect that was fixed in
5362 2. Some messages that were rejected after a message id was allocated were
5363 shown as "incomplete" by exigrep. It no longer does this for messages that
5364 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5366 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5367 have allow_domain_literals set, the ID did not get logged in the <= line.
5368 Domain literals are now always recognized in Message-ID: header lines.
5370 4. The first argument for a ${extract expansion item is the key name or field
5371 number. Leading and trailing spaces in this item were not being ignored,
5372 causing some misleading effects.
5374 5. When deliver_drop_privilege was set, single queue runner processes started
5375 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5376 same command in the process it spins off) were not dropping privilege.
5378 6. When the daemon running as "exim" started a queue runner, it always
5379 re-executed Exim in the spun-off process. This is a waste of effort when
5380 deliver_drop_privilege is set. The new process now just calls the
5381 queue-runner function directly.
5387 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5390 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5391 ignored. Though the use of -oP was forcing the writing of a pid file, it
5392 was always written to the default place.
5394 3. If the message "no IP address found for host xxxx" is generated during
5395 incoming verification, it is now followed by identification of the incoming
5396 connection (so you can more easily find what provoked it).
5398 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5400 5. Added some features to "harden" Exim a bit more against certain attacks:
5402 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5403 be put in Local/Makefile. This is like the never_users runtime option,
5404 but it cannot be overridden. The default setting is "root".
5406 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5407 prefix string with which any file named in a -C command line option
5410 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5411 is retained for -C and -D only if the caller of Exim is root. Without
5412 it, the exim user may also use -C and -D and retain privilege.
5414 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5415 command line option is disabled.
5417 6. Macro names set by the -D option must start with an upper case letter, just
5418 like macro names defined in the configuration file.
5420 7. Added "dereference=" facility to LDAP.
5422 8. Two instances of the typo "uknown" in the source files are fixed.
5424 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5425 the Configure-Makefile script screwed up while processing it.
5427 10. Incorporated PCRE 4.4.
5429 11. The SMTP synchronization check was not operating right at the start of an
5430 SMTP session. For example, it could not catch a HELO sent before the client
5431 waited for the greeting. There is now a check for outstanding input at the
5432 point when the greeting is written. Because of the duplex, asynchronous
5433 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5434 way, but not yet received, when the check is performed.
5436 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5437 on TCP/IP sockets, because this apparently causes some broken clients to
5440 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5441 unchanged) from the Cygwin maintainer.
5443 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5444 was supported instead of testing for mbx. Effectively a typo.
5446 15. The spa authenticator server code was not checking that the input it
5447 received was valid base64.
5449 16. The debug output line for the "set" modifier in ACLs was not showing the
5450 name of the variable that was being set.
5452 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5454 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5455 lookup on its own. Something else was needed to trigger the lookup. For
5456 example, a match in host_lookup or the need for a host name in a host list.
5457 Now, if $sender_host_name is referenced and the host name has not yet been
5458 looked up, a lookup is performed. If the lookup fails, the variable remains
5459 empty, and $host_lookup_failed is set to "1".
5461 19. Added "eqi" as a case-independent comparison operator.
5463 20. The saslauthd authentication condition could segfault if neither service
5464 nor realm was specified.
5466 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5467 error message that was logged was misleading, and incoming SMTP
5468 connections were dropped. The message is now more accurate, and temporary
5469 errors are given to SMTP connections.
5471 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5472 (except RSET) with a 421 error, until QUIT is received. However, it was
5473 failing to send a response to QUIT.
5475 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5476 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5477 ACL is now run after the helo_try_verify_hosts code.
5479 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5480 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5481 case-independent, so other case variants are also recognized). Apparently
5482 some systems use these upper case variants.
5484 25. If more than two messages were waiting for the same host, and a transport
5485 filter was specified for the transport, Exim sent two messages over the
5486 same TCP/IP connection, and then failed with "socket operation on non-
5487 socket" when it tried to send the third.
5489 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5491 27. The extern definition of crypt16() in expand.c was not being excluded when
5492 the OS had its own crypt16() function.
5494 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5495 as a preferred synonym for return_size_limit, both as an option and as an
5498 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5500 30. Changed the default configuration ACL to relax the local part checking rule
5501 for addresses that are not in any local domains. For these addresses,
5502 slashes and pipe symbols are allowed within local parts, but the sequence
5503 /../ is explicitly forbidden.
5505 31. SPA server authentication was not clearing the challenge buffer before
5508 32. log_message in a "warn" ACL statement was writing to the reject log as
5509 well as to the main log, which contradicts the documentation and doesn't
5510 seem right (because no rejection is happening). So I have stopped it.
5512 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5513 However, I am unable to do any testing of this.
5515 34. Fixed an infelicity in the appendfile transport. When checking directories
5516 for a mailbox, to see if any needed to be created, it was accidentally
5517 using path names with one or more superfluous leading slashes; tracing
5518 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5520 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5521 contents are added to the log line that is written for every discarded
5522 recipient. (Previously a log_message setting was ignored.)
5524 36. The ${quote: operator now quotes the string if it is empty.
5526 37. The install script runs exim in order to find its version number. If for
5527 some reason other than non-existence or emptiness, which it checks, it
5528 could not run './exim', it was installing it with an empty version number,
5529 i.e. as "exim-". This error state is now caught, and the installation is
5532 38. An argument was missing from the function that creates an error message
5533 when Exim fails to connect to the socket for saslauthd authentication.
5534 This could cause Exim to crash, or give a corrupted message.
5536 39. Added isip, isip4, and isip6 to ${if conditions.
5538 40. The ACL variables $acl_xx are now saved with the message, and can be
5539 accessed later in routers, transports, and filters.
5541 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5542 strings in the file are not string-expanded.
5544 42. If a MAIL command specified a SIZE value that was too large to fit into an
5545 int variable, the check against message_size_limit failed. Such values are
5546 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5547 day this will have to be increased, but I don't think I want to be around
5548 when emails are that large.
5555 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5556 iconv() is not standard in FreeBSD.
5558 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5559 IPv6 enabled. The observed symptom was a segmentation fault on return from
5560 the function os_common_find_running_interfaces() in src/os.c.
5562 3. In the check_special_case() function in daemon.c I had used "errno" as an
5563 argument name, which causes warnings on some systems. This was basically a
5564 typo, since it was named "eno" in the comments!
5566 4. The code that waits for the clock to tick (at a resolution of some fraction
5567 of a second) so as to ensure message-id uniqueness was always waiting for
5568 at least one whole tick, when it could have waited for less. [This is
5569 almost certainly not relevant at current processor speeds, where it is
5570 unlikely to ever wait at all. But we try to future-proof.]
5572 5. The function that sleeps for a time interval that includes fractions of a
5573 second contained a race. It did not block SIGALRM between setting the
5574 timer, and suspending (a couple of lines later). If the interval was short
5575 and the sigsuspend() was delayed until after it had expired, the suspension
5576 never ended. On busy systems this could lead to processes getting stuck for
5579 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5580 process, before it forks any delivery processes. The open lookup caching
5581 mechanism meant that the open file or database connection was passed into
5582 the delivery process. The problem was that delivery processes always tidy
5583 up cached lookup data. This could cause a problem for the next delivery
5584 process started by the queue runner, because the external queue runner
5585 process does not know about the closure. So the next delivery process
5586 still has data in the lookup cache. In the case of a file lookup, there was
5587 no problem because closing a file descriptor in a subprocess doesn't affect
5588 the parent. However, if the lookup was caching a connection to a database,
5589 the connection was closed, and the second delivery process was likely to
5590 see errors such as "PGSQL: query failed: server closed the connection
5591 unexpectedly". The problem has been fixed by closing all cached lookups
5592 in a queue runner before running a delivery process.
5594 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5595 seem to have the "const" qualifier which it has on other OS. I've
5598 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5599 *of the same type* (client or server) with the same public name that an
5600 error should be diagnosed.
5602 9. When Exim looked up a host name for an IP address, but failed to find the
5603 original IP address when looking up the host name (a safety check), it
5604 output the message "<ip address> does not match any IP for NULL", which was
5605 confusing, to say the least. The bug was that the host name should have
5606 appeared instead of "NULL".
5608 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5609 user that is built into the binary, and the -C or -D options is used, root
5610 privilege is dropped before the configuration file is read. In addition,
5611 logging is switched to stderr instead of the normal log files. If the
5612 configuration then re-defines the Exim user, the unprivileged environment
5613 is probably not what is expected, so Exim logs a panic warning message (but
5616 However, if deliver_drop_privilege is set, the unprivileged state may well
5617 be exactly what is intended, so the warning has been cut out in that case,
5618 and Exim is allowed to try to write to its normal log files.
5624 1. smtp_return_error_details was not giving details for temporary sender
5625 or receiver verification errors.
5627 2. Diagnose a configuration error if two authenticators have the same public
5630 3. Exim used not to create the message log file for a message until the first
5631 delivery attempt. This could be confusing when incoming messages were held
5632 for policy or load reasons. The message log file is now created at the time
5633 the message is received, and an initial "Received" line is written to it.
5635 4. The automatically generated man page for command line options had a minor
5636 bug that caused no ill effects; however, a more serious problem was that
5637 the procedure for building the man page automatically didn't always
5638 operate. Consequently, release 4.20 contains an out-of-date version. This
5639 shouldn't happen again.
5641 5. When building Exim with embedded Perl support, the script that builds the
5642 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5643 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5645 6. The freeze_tell option was not being used for messages that were frozen on
5646 arrival, either by an ACL or by local_scan().
5648 7. Added the smtp_incomplete_transaction log selector.
5650 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5651 was accepting AUTH without a new EHLO.
5653 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5654 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5657 10. Logging of TCP/IP connections (when configured) now happens in the main
5658 daemon process instead of the child process, so that the TCP/IP connection
5659 count is more accurate (but it can never be perfect).
5661 11. The use of "drop" in a nested ACL was not being handled correctly in the
5662 outer ACL. Now, if condition failure induced by the nested "drop" causes
5663 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5664 or "require"), the connection is dropped.
5666 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5667 that yield "discard" can now be used with an "accept" or a "discard" verb,
5668 but an error is generated for any others (because I can't see a useful way
5669 to define what should happen).
5671 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5672 are now processed in the same way as lines in the Exim configuration file.
5673 In particular, continuation lines are supported.
5675 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5677 15. Added -ti meaning -t -i.
5679 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5680 domains, and warn by logging if others are found.
5682 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5683 returned from GIFCONF when Exim is trying to find the list of interfaces on
5684 a host. The code in os.c has been modified to copy each ifreq to an aligned
5685 structure in all cases.
5687 Also, in some cases, the returned ifreq's were being copied to a 'struct
5688 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5689 means the last couple of bytes of an IPv6 address could be chopped if the
5690 ifreq contained only a normal sockaddr (14 bytes storage).
5692 18. Named domain lists were not supported in the hosts_treat_as_local option.
5693 An entry such as +xxxx was not recognized, and was treated as a literal
5696 19. Ensure that header lines added by a DATA ACL are included in the reject log
5697 if the ACL subsequently rejects the message.
5699 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5700 MD5 (which is deprecated).
5702 21. When testing a filter file using -bf, Exim was writing a message when it
5703 took the sender from a "From " line in the message, but it was not doing so
5704 when it took $return_path from a Return-Path: header line. It now does.
5706 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5707 with a valid header line field name (a series of printing characters
5708 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5710 23. Changed "disc" in the source to "disk" to conform to the documentation and
5711 the book and for uniformity.
5713 24. Ignore Sendmail's -Ooption=value command line item.
5715 25. When execve() failed while trying to run a command in a pipe transport,
5716 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5717 could be confused with a return value of 69 from the command itself. This
5718 has been changed to 127, the value the shell returns if it is asked to run
5719 a non-existent command. The wording for the related log line suggests a
5720 non-existent command as the problem.
5722 26. If received_header_text expands to an empty string, do not add a Received:
5723 header line to the message. (Well, it adds a token one on the spool, but
5724 marks it "old" so that it doesn't get used or transmitted.)
5726 27. Installed eximstats 1.28 (addition of -nt option).
5728 28. There was no check for failure on the call to getsockname() in the daemon
5729 code. This can fail if there is a shortage of resources on the system, with
5730 ENOMEM, for example. A temporary error is now given on failure.
5732 29. Contrary to the C standard, it seems that in some environments, the
5733 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5734 program. Exim now does this explicitly; it affects the formatting of
5735 timestamps using strftime().
5737 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5738 complaints. I've now initialized all the variables, to avoid this.
5740 32. Header lines added by a system filter were not being "seen" during
5741 transport-time rewrites.
5743 33. The info_callback() function passed to OpenSSL is set up with type void
5744 (*)(SSL *, int, int), as described somewhere. However, when calling the
5745 function (actually a macro) that sets it up, the type void(*)() is
5746 expected. I've put in a cast to prevent warnings from picky compilers.
5748 34. If a DNS black list lookup found a CNAME record, but there were no A
5749 records associated with the domain it pointed at, Exim crashed.
5751 35. If a DNS black list lookup returned more than one A record, Exim ignored
5752 all but the first. It now scans all returned addresses if a particular IP
5753 value is being sought. In this situation, the contents of the
5754 $dnslist_value variable are a list of all the addresses, separated by a
5757 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5758 to accept a host name and all its aliases if the forward lookup for any of
5759 them yielded the IP address of the incoming connection. Now it accepts only
5760 those names whose forward lookup yields the correct IP address. Any other
5761 names are discarded. This closes a loophole whereby a rogue DNS
5762 administrator could create reverse DNS records to break through a
5763 wildcarded host restriction in an ACL.
5765 37. If a user filter or a system filter that ran in a subprocess used any of
5766 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5767 the wrong values were passed to the pipe command ($thisaddress had the
5768 value of $0, $0 had the value of $1, etc). This bug was introduced by
5769 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5771 38. Improved the line breaking for long SMTP error messages from ACLs.
5772 Previously, if there was no break point between 40 and 75 characters, Exim
5773 left the rest of the message alone. Two changes have been made: (a) I've
5774 reduced the minimum length to 35 characters; (b) if it can't find a break
5775 point between 35 and 75 characters, it looks ahead and uses the first one
5776 that it finds. This may give the occasional overlong line, but at least the
5777 remaining text gets split now.
5779 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5780 file descriptors might be low, and that setting 1000 would always raise it.
5781 It turns out that in some environments, the limit is already over 1000 and
5782 that lowering it causes trouble. So now Exim takes care not to decrease it.
5784 40. When delivering a message, the value of $return_path is set to $sender_
5785 address at the start of routing (routers may change the value). By an
5786 oversight, this default was not being set up when an address was tested by
5787 -bt or -bv, which affected the outcome if any router or filter referred to
5790 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5791 log only when "message" or "log_message" are set. However, if one of the
5792 conditions was an address verification, or a call to a nested ACL, the
5793 messages generated by the underlying test were being passed through. This
5794 no longer happens. The underlying message is available in $acl_verify_
5795 message for both "message" and "log_message" expansions, so it can be
5796 passed through if needed.
5798 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5799 new expansion $bh_ to give the encoded byte string without charset
5800 translation. Translation happens only if iconv() is available; HAVE_ICONV
5801 indicates this at build time. HEADERS_CHARSET gives the charset to
5802 translate to; headers_charset can change it in the configuration, and
5803 "headers charset" can change it in an individual filter file.
5805 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5806 that creates RFC 2047 encoded "words" labels them as that charset instead
5807 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5808 expansion operator; (ii) when Exim creates a From: line for a local
5809 message; (iii) when a header line is rewritten to include a "phrase" part.
5811 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5812 buggy, causing it to skip the first lines of messages whose message ID
5813 ended in 'D'. This would not have bitten before Exim release 4.14, because
5814 message IDs were unlikely to end in 'D' before then. The effect was to have
5815 incorrect size information for certain domains.
5817 45. #include "config.h" was missing at the start of the crypt16.c module. This
5818 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5821 46. If there was a timeout during a "random" callout check, Exim treated it as
5822 a failure of the random address, and carried on sending RSET and the real
5823 address. If the delay was just some slowness somewhere, the response to the
5824 original RCPT would be taken as a response to RSET and so on, causing
5825 mayhem of various kinds.
5827 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5828 when I implemented it. It didn't allow for the fact that some option values
5829 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5830 the right test for positive values.
5832 48. Domain names in DNS records are case-independent. Exim always looks them up
5833 in lower case. Some resolvers return domain names in exactly the case they
5834 appear in the zone file, that is, they may contain uppercase letters. Not
5835 all resolvers do this - some return always lower case. Exim was treating a
5836 change of case by a resolver as a change of domain, similar to a widening
5837 of a domain abbreviation. This triggered its re-routing code and so it was
5838 trying to route what was effectively the same domain again. This normally
5839 caused routing to fail (because the router wouldn't handle the domain
5840 twice). Now Exim checks for this case specially, and just changes the
5841 casing of the domain that it ultimately uses when it transmits the message
5844 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5847 50. If a filter generated a file delivery with a non-absolute name (possible if
5848 no home directory exists for the router), the forbid_file option was not
5851 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5852 the existing equality matching.
5854 52. Exim was using ints instead of ino_t variables in some places where it was
5855 dealing with inode numbers.
5857 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5858 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5859 called TMPDIR, and if it finds it is different, it changes its value.
5861 54. The smtp_printf() function is now made available to local_scan() so
5862 additional output lines can be written before returning. There is also an
5863 smtp_fflush() function to enable the detection of a dropped connection.
5864 The variables smtp_input and smtp_batched_input are exported to
5867 55. Changed the default runtime configuration: the message "Unknown user"
5868 has been removed from the ACL, and instead placed on the localuser router,
5869 using the cannot_route_message feature. This means that any verification
5870 failures that generate their own messages won't get overridden. Similarly,
5871 the "Unrouteable address" message that was in the ACL for unverifiable
5872 relay addresses has also been removed.
5874 56. Added hosts_avoid_esmtp to the smtp transport.
5876 57. The exicyclog script was not checking for the esoteric option
5877 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5878 will work only if exicyclog is run under the appropriate euid.
5880 58. Following a discussion on the list, the rules by which Exim recognises line
5881 endings on incoming messages have been changed. The -dropcr and drop_cr
5882 options are now no-ops, retained only for backwards compatibility. The
5883 following line terminators are recognized: LF CRLF CR. However, special
5884 processing applies to CR:
5886 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5887 nor a local message in the state where . is a terminator.
5889 (ii) If a bare CR is encountered in a header line, an extra space is added
5890 after the line terminator so as not to end the header. The reasoning
5891 behind this is that bare CRs in header lines are most likely either
5892 to be mistakes, or people trying to play silly games.
5894 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5895 was being incorrectly given as 18 bytes larger than it should have been.
5896 This is a VOB (very old bug).
5898 60. This may never have affected anything current, but just in case it has:
5899 When the local host is found other than at the start of a list of hosts,
5900 the local host, those with the same MX, and any that follow, are discarded.
5901 When the list in question was part of a longer list of hosts, the following
5902 hosts (not currently being processed) were also being discarded. This no
5903 longer happens. I'm not sure if this situation could ever has previously
5906 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5909 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5910 header. This is something that is recommended in a new Internet Draft, and
5911 is something that is documented as being done by Sendmail. There are two
5912 possible values. For messages generated by the autoreply transport, Exim
5915 Auto-Submitted: auto-replied
5917 whereas for all other generated messages (e.g. bounces) it adds
5919 Auto-Submitted: auto-generated
5921 63. The "personal" condition in filters now includes a test for the
5922 Auto-Submitted: header. If it contains the string "auto-" the message it
5923 not considered personal.
5925 64. Added rcpt_include_affixes as a generic transport option.
5927 65. Added queue_only_override (default true).
5929 66. Added the syslog_duplication option.
5931 67. If what should have been the first header line of a message consisted of
5932 a space followed by a colon, Exim was mis-interpreting it as a header line.
5933 It isn't of course - it is syntactically invalid and should therefore be
5934 treated as the start of the message body. The misbehaviour could have
5935 caused a number of strange effects, including loss of data in subsequent
5936 header lines, and spool format errors.
5938 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5939 client host had authenticated. This control can now be exercised by an ACL
5940 for more flexibility.
5942 69. By default, callouts do not happen when testing with -bh. There is now a
5943 variant, -bhc, which does actually run the callout code, including
5944 consulting and updating the callout cache.
5946 70. Added support for saslauthd authentication, courtesy of Alexander
5949 71. If statvfs() failed on the spool or log directories while checking their
5950 size for availability, Exim confusingly gave the error "space shortage".
5951 Furthermore, in debugging mode it crashed with a floating point exception.
5952 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5953 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5954 serious problem, Exim now writes to the main and panic logs when this
5955 happens, with details of the failure. It then refuses to accept the
5956 incoming message, giving the message "spool directory problem" or "log
5957 directory problem" with a 421 code for SMTP messages.
5959 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5960 0, 1, and 2 exist, because some OS complain for execs without them (see
5961 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5962 descriptors. However, the code omitted to check that the open succeeded,
5963 causing mysterious errors if for some reason the permissions on /dev/null
5964 got screwed. Now Exim writes a message to the main and panic logs, and
5965 bombs out if it can't open /dev/null.
5967 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5968 interact so that it is all more flexible. It is supposed to remain
5969 backwards compatible. Also added extra_local_interfaces.
5971 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5972 to bomb out with an assertion failure - to the client this appears as a
5973 connection drop. This problem occurs in the part of the code that was taken
5974 from the Samba project. Fortunately, the assertion is in a very simple
5975 function, so I have fixed this by reproducing the function inline in the
5976 one place where it is called, and arranging for authentication to fail
5977 instead of killing the process with assert().
5979 75. The SPA client code was not working when the server requested OEM rather
5980 than Unicode encoding.
5982 76. Added code to make require_files with a specific uid setting more usable in
5983 the case where statting the file as root fails - usually a non-root-mounted
5984 NFS file system. When this happens and the failure is EACCES, Exim now
5985 forks a subprocess and does the per-uid checking as the relevant uid.
5987 77. Added process_log_path.
5989 78. If log_file_path was not explicitly set, a setting of check_log_space or
5990 check_log_inodes was ignored.
5992 79. If a space check for the spool or log partitions fails, the incident is now
5993 logged. Of course, in the latter case the data may get lost...
5995 80. Added the %p formatting code to string_format() so that it can be used to
5996 print addresses in debug_print(). Adjusted all the address printing in the
5997 debugging in store.c to use %p rather than %d.
5999 81. There was a concern that a line of code in smtp_in.c could overflow a
6000 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6001 initially expressed, the concern was not well-founded, because trailing
6002 spaces are removed early. However, if the trailing spaces were followed by
6003 a NULL, they did not get removed, so the overflow was possible. Two fixes
6006 (a) I re-wrote the offending code in a cleaner fashion.
6007 (b) If an incoming SMTP command contains a NULL character, it is rejected
6010 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6011 runs initgroups(), so that if the Exim user is in any additional groups,
6012 they will be used during message reception.
6018 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob